

small (250x250 max)
medium (500x500 max)
Large
Extra Large
large ( > 500x500)
Full Resolution


ASSESSING THE IMPACT OF VICARIOUS EXPERIENCES ON PRESERVICE ELEMENTARY SCIENCE TEACHER EFFICACY AND PRESERVICE ELEMENTARY TEACHER EFFICACY By RONALD ROBERT WAGLER Bachelor of Science in Biology Southern Illinois University Carbondale, Illinois 1990 Master of Science in Zoology Oklahoma State University Stillwater, Oklahoma 2003 Submitted to the Faculty of the Graduate College of the Oklahoma State University in partial fulfillment of the requirements for the Degree of DOCTOR OF PHILOSOPHY December, 2007 ii ASSESSING THE IMPACT OF VICARIOUS EXPERIENCES ON PRESERVICE ELEMENTARY SCIENCE TEACHER EFFICACY AND PRESERVICE ELEMENTARY TEACHER EFFICACY Dissertation Approved: Richard Bryant Dissertation Adviser Patricia LamphereJordan Caroline Beller James Shaw A. Gordon Emslie Dean of the Graduate College iii ACKNOWLEDGEMENT First and foremost, I would like to thank my wife Amy and my daughter Olive Elizabeth for their support, encouragement and love throughout my PhD. You are the best part of life! I would like to thank my committee members, Dr. Beller and Dr. Shaw, for the time and effort they put into this research project. I would also like to thank Dr. Jordan for all off her help throughout my PhD program. Last, but certainly not least, I would especially like to thank Dr. Bryant for his patience, understanding, friendship and assistance in bringing the current study to completion. iv TABLE OF CONTENTS Chapter Page I. INTRODUCTION.....................................................................................................1 Background..............................................................................................................1 Statement of the Problem.........................................................................................3 Purpose of the Study ................................................................................................3 Research Questions..................................................................................................3 Research Question 1 ................................................................................................3 Research Question 2 ................................................................................................3 Theoretical Perspective............................................................................................4 Significance of the Study.........................................................................................5 Definition of Terms..................................................................................................6 Composition of the Dissertation ..............................................................................9 II. REVIEW OF LITERATURE.................................................................................11 Introduction............................................................................................................11 Teacher Efficacy ....................................................................................................11 Teacher Efficacy ResearchInservice Teachers .....................................................12 Teacher Efficacy ResearchPreservice Teachers ...................................................20 Vicarious Experience .............................................................................................26 Preservice Teacher Early Field Experience ...........................................................28 III. METHODOLOGY ...............................................................................................31 Introduction............................................................................................................31 Research Question 1 ..............................................................................................31 Study Participants ..................................................................................................31 Testing Instruments and Procedure........................................................................32 Data Analysis .........................................................................................................33 Research Question 2 ..............................................................................................34 Study Participants ..................................................................................................34 Testing Instruments and Procedure........................................................................34 v Data Analysis .........................................................................................................35 Assumptions and Limitations of the Study............................................................35 Assumptions...........................................................................................................35 Limitations .............................................................................................................35 IV. RESULTS.............................................................................................................37 Introduction............................................................................................................37 Research Questions................................................................................................37 Research Question 1 ..............................................................................................37 Research Question 2 ..............................................................................................37 Testing Instrument Data Analysis..........................................................................38 Principal Component Analysis ..............................................................................38 Factor Analysis ......................................................................................................38 Cronbach’s Alpha Reliabilities ..............................................................................41 Summary Testing Instrument Statistics .................................................................43 Study Participants Demographic Data Analysis ....................................................44 Preservice Teachers ...............................................................................................44 Inservice Teachers .................................................................................................47 Introduction to Research Question 1 and 2 Analysis.............................................49 Research Question 1 Analysis ...............................................................................50 STEBIB Results....................................................................................................51 TES Results............................................................................................................51 Research Question 2 Analysis ...............................................................................51 STEBIB Results....................................................................................................52 TES Results............................................................................................................57 V. CONCLUSION AND RECOMMENDATIONS..................................................61 Introduction............................................................................................................61 Research Questions................................................................................................61 Research Question 1 ..............................................................................................61 Research Question 2 ..............................................................................................61 Testing Instrument Conclusions ............................................................................62 Principal Component Analysis ..............................................................................62 Factor Analysis ......................................................................................................63 Cronbach’s Alpha Reliabilities ..............................................................................64 Summary Testing Instrument Statistics .................................................................65 Study Participants Demographic Conclusions.......................................................67 Preservice Teachers ...............................................................................................67 Inservice Teachers .................................................................................................68 Research Question 1 ..............................................................................................69 Research Question 1 Conclusions..........................................................................69 STEBIB Factor 1 and 2 Results............................................................................69 vi TES Factor 1 and 2 Results....................................................................................69 Research Question 2 ..............................................................................................70 Research Question 2 Conclusions..........................................................................70 STEBIB Factor 1 Results .....................................................................................70 Inservice Teacher Ethnicity ...................................................................................70 Inservice Teacher Gender ......................................................................................72 Percentage of Students that Received Free and Reduced Lunch...........................73 STEBIB Factor 2 Results .....................................................................................74 Percentage of Asian/Pacific Islander Students ......................................................74 Percentage of Hispanic/Latino Students ................................................................75 Percentage of American Indian/or Alaskan Students ............................................77 ExCEL preservice teacher......................................................................................78 TES Factor 1 Results .............................................................................................80 Percentage of Hispanic/Latino Students ................................................................80 TES Factor 2 Results .............................................................................................81 The Problematic Nature of the Teacher Efficacy Scale.........................................82 The Specific Nature of Teaching Efficacy.............................................................83 Recommendations for Future Research.................................................................85 Implications for Practice ........................................................................................87 Concluding Remarks..............................................................................................88 REFERENCES ............................................................................................................90 APPENDICES .............................................................................................................97 APPENDIX AINFORMED CONSENT AND DEMOGRAPHIC QUESTIONS ......................................98 APPENDIX BTES ............................................................................................102 APPENDIX CSTEBIB ....................................................................................105 APPENDIX DINFORMED CONSENT AND CLASSROOM DEMOGRAPHIC AND COOPERATING TEACHER QUESTIONNAIRE .................108 APPENDIX EDEMOGRAPHIC QUESTIONS ...............................................113 APPENDIX FCATEGORIZATION OF VARIABLES FOR STATISTICAL ANALYSIS .....................116 APPENDIX GIRB.............................................................................................125 vii LIST OF TABLES Table Page 1 STEBIB Original Factor Analysis.......................................................................40 2 STEBIB Combined Factor Analysis ...................................................................41 3 TES Factor Analysis .............................................................................................42 4 Crombach’s Alpha Reliabilities............................................................................42 5 STEBIB Summary Testing Instrument Statistics ................................................43 6 TES Summary Testing Instrument Statistics ........................................................44 7 Preservice Teacher Question.................................................................................45 8 Preservice Teacher Question.................................................................................46 9 Preservice Teacher Question.................................................................................47 10 Select Inservice Teacher Questions .....................................................................48 11 Variables and Categories Associated with Research Question 1.........................50 12 Age of Preservice Teacher ANOVA....................................................................51 13 Age of Preservice Teacher Means .......................................................................52 14 Variables and Categories Associated with Research Question 2.........................53 15 ANCOVA Model Results for Research Question 2, Factor 1 of the STEBIB...54 16 Means for Factor 1 of the STEBIB Posttest Associated with the ANCOVA ....55 17 ANCOVA Model Results for Research Question 2, Factor 2 of the STEBIB...55 18 Means for Factor 2 of the STEBIB Posttest Associated with the ANCOVA.....56 viii 19 Tukey Simultaneous Confidence Intervals for Factor 2 of the STEBIB............57 20 ANOVA for TES Factor 1 ...................................................................................58 21 TES Factor 1 Means for Hispanic/Latino Students .............................................58 22 Tukey Simultaneous Confidence Intervals for Factor 1 of the TES....................59 23 ANOVA for TES Factor 2 ...................................................................................59 24 Means for TES Factor 2.......................................................................................60 1 CHAPTER 1 INTRODUCTION Background Teacher efficacy is a powerful idea with a long, sordid and controversial history. The first teacher efficacy study was influenced by Rotter’s (1966) social learning theory and focused on teacher’s beliefs about where control lies in student learning. Later studies would move away from this influence and would be more directly influenced by Bandura’s social cognitive theory (Bandura, 1977). More recently there has been a movement to develop teacher efficacy instruments that focused on more specific contexts of teaching such as science teaching efficacy. In this atmosphere of competing theories and competing contexts, a sense of confusion has arisen regarding the most appropriate way to understand and measure teacher efficacy. Researchers were intrigued by the need to address certain central issues that are fundamental in understanding the practical application of teacher efficacy to effective teaching and student learning in specific classroom contexts. Many questions arose regarding the study of teacher efficacy. How do specific classroom dynamics impact teacher efficacy? Does the Teacher Efficacy Scale (TES) (Gibson & Dembo, 1984), which has formed much of the basis for our understanding of teacher efficacy, even measure teacher efficacy? Does the idea of a general teacher efficacy with a low level of specificity even exist? Ultimately, does teacher efficacy need 2 to be captured within a specific classroom context and environment to have meaningful practical significance and application for effective student learning? The current study attempts to bring some clarity to these issues by looking at the impact that field experiences (Vicarious experiences) have on preservice elementary science teacher efficacy and preservice elementary teacher efficacy. Teacher efficacy has been defined as “the extent to which the teacher believes he or she has the capacity to affect student performance” (Berman, McLauglin, Bass, Pauly & Zellman, 1977, p. 137) or a “teachers’ belief or conviction that they can influence how well students learn, even those that may be difficult or unmotivated” (Guskey & Passaro, 1994, p. 4). According to TschannenMoran et al. (1998) “the research suggests that teachers’ sense of efficacy plays a powerful role in schooling” (p. 234). There has been extensive research over the last three decades to formulate a unified theory of efficacy and to develop valid, reliable instruments that could measure efficacy levels in teachers, especially elementary teachers. Vicarious experiences are one of the four main sources that influence the efficacy of the individual teacher (Bandura, 1997). Vicarious experiences are also a common component of teacher education programs. However, little research has been done to evaluate the impact of vicarious learning experiences in the context of perceived preservice teacher efficacy and perceived preservice science teacher efficacy. The vicarious experiences in this study occurred in elementary public school classrooms where the preservice elementary teachers conducted their field observations. The type of vicarious experience was dependant on the specific variables that existed in the specific 3 classroom where the individual preservice teachers observed. For the specific variables associated with the research questions see Table 11 and 14 in Chapter 4. Statement of the Problem Teacher efficacy has been positively correlated with the amount of effort a teacher will expend in a teaching environment and the level of persistence a teacher will show in the face of obstacles (TschannenMoran et al., 1998). However, there have been no studies that have looked at the impact that vicarious experiences in teacher preparation programs have on the construct of preservice elementary teacher efficacy and preservice elementary science teacher efficacy. Purpose of the Study The purpose of this study was to investigate the impact vicarious experiences had on perceived preservice elementary teacher efficacy and perceived preservice elementary science teacher efficacy. Research Questions The research questions that guided this study were: Research Question 1 What is the impact of a vicarious learning experience (preservice teacher field experiences) on perceived preservice elementary teacher efficacy and preservice elementary science teacher efficacy? Research Question 2 What is the impact of the characteristics of the field experience classroom, within the given school where the educational field experience occurred, on perceived preservice 4 elementary teacher efficacy and perceived preservice elementary science teacher efficacy? Theoretical Perspective The first formal efficacy research began over two decades ago when the RAND organization, influenced by Rotter’s (1966) social learning theory, added two items to an already existing questionnaire (Armor et al., 1976). With the findings of the two RAND organization items the construct of teacher efficacy was first formulated. In these early RAND studies, teachers were asked to designate their level of agreement with two efficacy item statements (Armor et al., 1976). The total of the scores on the two RAND items was called teacher efficacy (TE), a concept that professed to indicate the degree to which a teacher believed that the consequences of learning and student motivation were controlled by the teacher (TschannenMoran et al., 1998). In the late 1970’s a second line of efficacy thought developed directly from Bandura’s social cognitive theory and his construct of selfefficacy (Bandura, 1977). Bandura (1997) defined selfefficacy as “beliefs in one’s capabilities to organize and execute the courses of action required to produce given attainments” (p. 3). “Selfefficacy is a futureoriented belief about the level of competence a person expects he or she will display in a specific situation” (TschannenMoran et al., 1998, p. 207). He also proposed that “selfefficacy beliefs influence thought patterns and emotions that enable actions in which people expend substantial effort in pursuit of goals, persist in the face of adversity, rebound from temporary setbacks, and exercise some control over events that affect their lives” (TschannenMoran et al., 1998, p. 210). 5 Bandura’s theory and his construct of selfefficacy would later influence the development of such efficacy instruments as the Teacher Efficacy Scale (Gibson and Dembo, 1984), the Ashton Vignettes (Ashton, Buhr, & Crocker, 1984), the Science Teaching Efficacy Belief Instrument (STEBI) (Riggs & Enochs, 1990), the Ohio State Teacher Efficacy Scale (TschannenMoran et al., 1998) and many others. With the development of the Teacher Efficacy Scale (TES) by Gibson and Dembo (1984) two factors of teacher efficacy were identified: The first factor, General Teacher Efficacy (GTE) related to the teacher’s belief about the impact of external factors, such as hostility in the home or economic factors of gender, race or class, contrasted to the influence of the schools and teachers. The second factor, Personal Teacher Efficacy (PTE), related to the teachers’ belief in their ability to overcome the factors that make learning difficult for students. Because teacher efficacy is believed to be both subjectmatter and context specific (Tschannen et al., 1998), Riggs and Enochs (1990) developed the Science Teaching Efficacy Belief Instrument (STEBI) to measure efficacy of science teaching. The authors identified two uncorrelated factors within STEBI, which they named personal science teaching efficacy (PSTE) and science teaching outcome expectancy (STOE). Thus, the instrument measured both PSTE and STOE. The PSTE scale indicated teachers’ belief in their ability to perform a given behavior; the outcome STOE scale indicates the teachers’ belief that effective teaching can change behaviors (Riggs & Enochs, 1990). Significance of the Study The significance of this study concerns the impact of vicarious experiences on the construct of teacher efficacy. Teacher efficacy is an indicator of teacher performance and 6 success. This study will help to determine what role, if any; vicarious experiences play in impacting teacher efficacy. The results of this research will be significant for the elementary education faculty at the studied university in evaluating the effectiveness of vicarious experiences as a tool in preparing preservice elementary teachers to enter the teacher work force. The research findings could also benefit those attempting to incorporate vicarious events (e.g., elementary field experiences) into their preservice elementary teacher education curriculum. Furthermore, the findings of this study may be beneficial to those considering the degree and role that vicarious experiences should play in their preservice secondary teacher education programs. Evidence from this study could also be useful in furthering continued research related to elementary preservice teachers since a lack of research associated with efficacy and vicarious experiences in elementary preservice teachers exists. This study will also help to bring some understanding to the impact that specific classroom variables have on preservice elementary teacher efficacy and preservice elementary science teacher efficacy. Furthermore, the study will help to assess the reliability of the Teacher Efficacy Scale (TES) (Gibson & Dembo, 1984). Lastly, this study will attempt to bring some understanding to the specific level of teacher efficacy that is needed to have practical meaningful significance in the elementary classroom. Definition of Terms ExCEL The ExCEL (Excellence in Collaborative and Experiential Learning) Program is a collaborative partnership between the College of Education and Evergreen Public Schools. Open to elementary education majors in their final semester before student teaching, the program features a threeday a week full day internship in a public 7 school classroom in which teacher candidates are partnered with expert classroom teachers. Candidates return to the OSU campus one day a week for Semester Y methods classes. The ExCEL program is run by a team of OSU COE faculty members who work together to mentor students in their placements at the elementary school and also teach the ExCEL sections of methods classes, closely connecting course content with field experiences. The ExCEL program runs both fall and spring semesters, partnering with Driftwood Elementary in the fall and Norwood Elementary in the spring. Both schools are close to the OSU campus and each offers a richly diverse student population and talented, enthusiastic faculty and leadership. ExCEL program participants gain experience in almost every aspect of elementary school teaching and develop productive relationships with a variety of educational professionals. Many ExCEL students choose to stay in their placement classroom for student teaching so that they receive a full year of supported internship before starting their first job. Elementary Education majors interested in participating in ExCEL may apply for the program during the second semester of their junior year. A program representative will speak in classes about the program each semester, and invitations to apply will be posted around the college. General Teacher Efficacy (GTE): The teacher’s belief about the power of external factors, such as violence in the home or economic realities of gender, race or class, compared to the influence of teachers and schools (Ashton, Olejnik, Crocker, & McAuliffe, 1982); also known as teacher outcome expectancy. 8 Indefinable Factor 2 (IF2): Second TES factor from the study. Items associated with this factor were 2, 3, 5, 8, 12, 13, 15, 20 and 27. Based on the nonassociated nature of these items a common construct factor, such as TE, was deemed indefinable. Personal Science Teaching Efficacy (PSTE): teachers’ belief in their ability to perform a given behavior (Riggs & Enochs, 1990). Personal Teaching Efficacy (PTE): The teacher’s belief in their ability to overcome factors that could make learning difficult for students (TschannenMoran et al., 1998); in the context of the TES, PTE is referred to as selfefficacy (Gibson and Dembo, 1984). Preservice elementary teacher: University student majoring in elementary education who has not completed his/her educational certification. The student has not begun teaching formally and has not received his/her teaching certificate. Science Teaching Efficacy Belief Instrument (STEBI): Science teacher efficacy instrument designed to measure PTSE and STOE. The instrument consists of 25 items (Riggs & Enochs, 1990). Science Teaching Efficacy Belief Instrument B (STEBIB): Science teacher efficacy instrument designed to measure PTSE and STOE. The instrument consists of 23 items from the STEBI. The STEBIB is identical to the STEBI except items 20 and 25 have been removed and the verb tenses of some of the 23 items have been changed to accommodate preservice teachers. These changes were made so the instrument, which was originally designed for inservice teachers, could be used with preservice teachers (Enochs & Riggs, 1990). 9 Science Teaching Outcome Expectancy (STOE): the teachers’ belief that effective teaching can result in student learning (Riggs & Enochs, 1990). Teacher efficacy: “the extent to which the teacher believes he or she has the capacity to affect student performance” (Berman, McLauglin, Bass, Pauly & Zellman, 1977, p. 137) or “teachers’ belief or conviction that they can influence how well students learn, even those that may be difficult or unmotivated” (Guskey & Passaro, 1994, p. 4). Teaching Efficacy (TE): Teach efficacy (TE) in the context of the TES refers to outcome expectancy (Gibson and Dembo, 1984). Teacher Efficacy Scale (TES): Teacher efficacy instrument designed to measure personal teacher efficacy (PTE) and teacher efficacy (TE). The instrument consists of 30 items with a Likert scale ranging from 1 (Strongly disagree) to 6 (Strongly Agree). Gibson and Dembo (1984) found, after performing factor analysis, that sixteen of the original 30 items had acceptable reliability coefficients. Vicarious experiences: Within the context of teacher efficacy, a vicarious experience refers to observing another individual teaching. Within the context of this study it refers to the preservice teacher’s educational field experiences. Composition of the Dissertation The dissertation is composed of five chapters. Chapter 1 is the introduction which consists of a background introduction, statement of the problem, purpose of the study, research questions, theoretical perspective, significance of the study, definition of terms and composition of the dissertation. Chapter 2 describes the current literature that is essential to the study including an introduction and a description of the following: teacher 10 efficacy, teacher efficacy researchinservice teachers, teacher efficacy researchpreservice teachers, vicarious experience and preservice teacher early field experiences. Chapter 3 describes the methodology of the study. Chapter 4 describes the results of the study and Chapter 5 describes the summary conclusions and recommendations of the study. 11 CHAPTER II REVIEW OF LITERATURE Introduction This chapter describes the current literature that is essential to the study. The concept of teacher efficacy is described followed by current research on inservice teacher efficacy. Next, current preservice teacher efficacy research is discussed followed by Bandura’s concept of vicarious experience and, finally, a description of preservice teacher early field experiences. Teacher Efficacy According to TschannenMoran, WoolfolkHoy, and Hoy, W. (1998), teacher efficacy was first defined by the RAND organization “as the extent to which teachers believed they could control the reinforcement of their actions, that is, whether control of reinforcement lay within themselves or the environment” (p.202). Bandura’s (1977) social cognitive theory and his construct of selfefficacy, defined as “a cognitive process in which people construct beliefs about their capacity to perform at a given level of attainment” (TschannenMoran et al., 1998, p.203), provided a theoretical foundation for the construct of teacher efficacy as a specific type of selfefficacy. TschannenMoran et al. (1998) defined teacher efficacy as the “teacher’s belief in his or her capacity to organize and execute courses of action required to successfully accomplish a specific teaching task in a particular context” (TschannenMoran et al., 1998, p.233). 12 Gibson and Dembo (1984), equipped with the theories of the RAND researchers and the conceptual ideas of Bandura, developed the first reliable teacher efficacy instrument, the Teacher Efficacy Scale (TES). Since the development of the Teacher Efficacy Scale in the early 1980’s, researchers have developed a plethora (Gibson & Dembo, 1984; Ashton, Buhr, & Crocker, 1984; Riggs & Enochs, 1990; Tschannen Moran et al., 1998) of teacher efficacy instruments with the hope of understanding this powerful construct (TschannenMoran et al., 1998). Teacher Efficacy Research—Inservice Teachers Over the last 25 years there have been numerous studies, using many different efficacy instruments that have shown that a teacher’s sense of efficacy is a strong indicator of the teacher’s ability to be a productive, successful teacher. In this section some of the more historically important research findings concerning teacher efficacy will be addressed. The first formal efficacy research began over two decades ago when the RAND organization added two items to an already existing questionnaire (Armor et al., 1976). With the findings of the two RAND organization items the construct of teacher efficacy was first formulated. In these early RAND studies teachers were asked to designate their level of agreement with two efficacy statements (Armor et al., 1976). The total of the scores on the two RAND items was called teacher efficacy (TE), a concept that professed to indicate the degree to which a teacher believed that the consequences of learning and student motivation were controlled by the teacher (TschannenMoran et al., 1998). The first RAND item, “When it comes right down to it, a teacher really can’t do much because most of a student’s motivation and performance depends on his or her home 13 environment” (Armor et al., 1976), would be labeled general teaching efficacy (GTE) by future efficacy researchers. The second item, “If I really try hard, I can get through to even the most difficult or unmotivated students” (Armor et al., 1976), would be labeled as personal teaching efficacy (PTE) by future researchers (TschannenMoran et al., 1998). Armor et al (1976), using the two RAND items in the context of reading programs employed in Los Angeles schools, found teacher efficacy (TE) was strongly correlated to reading achievement variation among minority students. In the late 1970’s a second line of efficacy thought developed directly from Bandura’s social cognitive theory and his construct of selfefficacy (Bandura, 1977). Bandura (1997) defined selfefficacy as “beliefs in one’s capabilities to organize and execute the courses of action required to produce given attainments” (p. 3). “Selfefficacy is a futureoriented belief about the level of competence a person expects he or she will display in a specific situation” (TschannenMoran et al., 1998, p. 207). He also proposed that “selfefficacy beliefs influence thought patterns and emotions that enable actions in which people expend substantial effort in pursuit of goals, persist in the face of adversity, rebound from temporary setbacks, and exercise some control over events that affect their lives” (TschannenMoran et al., 1998, p. 210). Bandura’s theory and his construct of selfefficacy would later influence the development of such efficacy instruments as the Teacher Efficacy Scale (Gibson and Dembo, 1984), the Ashton Vignettes (Ashton, Buhr, & Crocker, 1984), the Science Teaching Efficacy Belief Instrument (STEBI) (Riggs & Enochs, 1990), the Ohio State Teacher Efficacy Scale (TschannenMoran et al., 1998) and many others. Bandura, after years of providing information for his everdeveloping theory, offered his own Teacher SelfEfficacy Scale 14 (Bandura, 1997). The instrument he developed is a 30item instrument with seven subscales: efficacy to enlist community involvement, efficacy to influence decision making, instructional efficacy, disciplinary efficacy, efficacy to enlist parental involvement, efficacy to influence school resources, and efficacy to create a positive school climate (Bandura, 1997). Bandura’s Teacher SelfEfficacy Scale uses a ninepoint Likerttype scale with response choices ranging from nothing (1) to a great deal (9) for each item. With the development of the Teacher Efficacy Scale (TES) by Gibson and Dembo (1984) two factors of teacher efficacy were identified: General Teacher Efficacy (GTE) or the teacher’s belief about the impact of external factors, such as hostility in the home or economic factors of gender, race or class, contrasted to the influence of the schools and teachers and Personal Teacher Efficacy (PTE), the teachers’ belief in their ability to overcome the factors that make learning difficult for students. Gibson and Dembo, using the TES, found that teachers who display a low sense of efficacy were more likely to instruct the class as a whole than to divide the class into small groups for instruction. They also found that low efficacy teachers were more likely to criticize students for an incorrect answer and were less likely to persist with a student in a difficult situation. Podell and Soodak (1993), again using the TES, found that there was a direct correlation between teacher efficacy and a teacher’s decision to refer a student to special education. They found teachers with higher levels of PTE were more willing to work with students who were experiencing problems rather than refer them to a special education program. Coladarci (1992) conducted a study to assess commitment to teaching. The subjects were composed of a random sample of 364 elementarylevel Maine teachers. 15 The TES (Gibson & Dembo, 1984) was administered to all subjects along with the teacher commitment question “Suppose you had it to do all over again: In view of your present knowledge, would you become a teacher?” (Coladarci, 1992, p. 328). After assessing the data it was found that the teachers who expressed a greater level of professional teaching commitment also tended to have higher levels of both PTE and GTE. Allinder (1994) administered the TES and the Teacher Characteristics Scale (Fuchs, Fuchs, & Bishop, 1992) to 200 randomly selected elementary special education teachers from four Midwest states. Teachers who displayed high levels PTE were more willing to try a variety of teaching approaches and materials, use new and innovative teaching methods and exhibited a desire to find better ways of teaching. Teachers who display elevated levels of PTE were also more apt to have higher scores in the areas of organization, enthusiasm, and fairness. Finally, enthusiasm and clarity in teaching were found to be related to elevated GTE. Research has also shown that just as a teacher’s efficacy level can influence a his or her behavior; a teacher’s level of efficacy can also influence students’ attitude toward the subject matter being taught and students’ attitudes toward school. Woolfolk, Rosoff, and Hoy (1990) concluded that students of teachers who exhibited high PTE tended toward greater interest in school, higher evaluations of their high PTE teachers, and showed a greater awareness that what they were being taught was important. Warren and Payne (1997) investigated middle school organizational patterns and their impact on teachers’ efficacy among 81 eighthgrade teachers. They concluded that teachers who were placed on interdisciplinary teams with the same planning times had 16 significantly higher PTE levels than teachers on interdisciplinary teams that did not have the same planning times. Teachers placed on interdisciplinary teams with the same planning times had higher PTE than teachers who were organized within their own department (Warren & Payne, 1997). In a similar study Reames and Spencer (1998) examined Georgia middle school teachers’ perceptions of their work environment, their perceived efficacy, and their organizational commitment. The study included 275 fulltime, certified teachers from 40 rural and metropolitan middle schools. Teachers completed a mailed survey that asked about demographics, organizational commitment, perceived efficacy, and the schoolwork culture (Reames & Spencer, 1998). Organizational structure and process variables were positively related to PTE. Process variables included collaboration, participatory decisionmaking, and supportive administrative leadership. Organizational structures included encouragement of innovation and risk taking, school goals and planning, and staff development to further goals (Reames & Spencer, 1998). Teacher efficacy has also been linked to family involvement practices. Garcia (2004) conducted a study that utilized the Teacher Efficacy Scale (Gibson & Dembo, 1984) and the Family Involvement Teacher Efficacy Scale (Garcia, 2000). Prior educational research has shown that positive benefits, for the child, are incurred as parents become involved in that child’s education. By utilizing these two instruments and a sample size of 110 urban elementary school teachers, Garcia concluded that elevated teacher efficacy was significantly correlated to and was also a predictor of five types of positive family involvement practices (Garcia, 2004). 17 The Ashton Vignettes were developed on the assumption that teacher efficacy can be context specific. Ashton, Buhr, and Crocker (1984) created a series of short literary sketches describing events a teacher might experience in the educational environment and asked teachers to make evaluations as to the possible causes involved in each written scenario known as the Ashton Vignettes (TschannenMoran et al., 1998). Two versions were developed and the second version, which asked teachers to compare themselves to other teachers, was significantly correlated with the two RAND items. Benz et al (1992) conducted a study in which they used the Ashton Vignettes (Ashton, Buhr, & Crocker, 1984) to assess perceptions of efficacy across a variety of educational situations with a variety of educators. They found college faculty, when compared to classroom teachers, student teacher supervisors and preservice teachers, had elavated selfefficacy for dealing successfully with a disobedient student, for selfmotivation, and for preparation. They also found both college faculty and preservice teachers were more confident about their efficacy in environments involving student socialization than were classroom teachers (Benz et al., 1992). In a related study Webb and Ashton (1987), the developers of the Ashton Vignettes, asked 42 middle and high school teachers to assess situational and environmental factors that they believed affected a teacher’s sense of efficacy. Six factors were identified: 1) inadequate salaries and low status, 2) excessive role demands, 3) lack of recognition and professional isolation, 4) uncertainty, 5) alienation and 6) low teacher morale. Because teacher efficacy is believed to be both subjectmatter and context specific (Tschannen et al., 1998), Riggs and Enochs (1990) developed the Science Teaching Efficacy Belief Instrument (STEBI) to measure efficacy of science teaching. The authors 18 identified two uncorrelated factors within STEBI, which they named personal science teaching efficacy (PSTE) and science teaching outcome expectancy (STOE). Thus, the instrument measured both PSTE and STOE. The PSTE scale indicated teachers’ belief in their ability to perform a given behavior; the outcome STOE scale indicates the teachers’ belief that effective teaching can change behaviors (Riggs & Enochs, 1990). Using the Science Teaching Efficacy Belief Instrument (STEBI), Riggs & Jesunathadas (1993) found elementary teachers with higher personal science teaching efficacy (PSTE) scores reported spending more time teaching science and were more apt to spend the needed time to develop the science concept being taught. Watters and Ginns (1995) found that teachers with a higher sense of PSTE tended to have a higher level of enjoyment associated with science activities. Elementary teachers who were involved in a oneyear science education training program who displayed low PSTE were less inclined to want to teach science and made smaller numbers of changes in their beliefs about how students could learn science. They also used less class time teaching science, were rated weaker by those who were observing them, and used a textbased teaching approach more often (Riggs, 1995). Czerniak (1999) used the STEBI to assess and compare the science teacher efficacy levels of science teachers who taught in either a middle school organizational structure or a junior high school organizational structure. After analysis Czerniak (1999) found that science teachers who were part of a middle school model versus a junior high model had significantly higher levels of science teaching outcome expectancy (STOE). Czerniak speculated that that these findings add merit to the use of a middle school model over a junior high model since middle school organizational structures “provide teachers 19 with the necessary support needed to remain committed to student learning” (Czerniak, 1999, p.36). Chun and Oliver (2000) conducted research into the quantitative examination of teacher selfefficacy and knowledge of the nature of science. They conducted a longitudinal study on 31 middle school science teachers in the southeastern part of the United States. All 31 participating teachers had been science teachers for over 5 years. Four sets of instruments were administered to the participants over a threeyear period including the STEBI. During the three years of the study the middle school science teachers participated in summer workshops. “The workshops were designed to enhance middle school teachers’ understanding about the nature and structure of science as well as pedagogical knowledge to teach science” (Chun & Oliver, 2000, p. 3). The STEBI was administered to the subjects five times during the study. A pretest and posttest were given during the first year workshop, at the second year workshop, and a posttest during the last year workshop (Chun & Oliver, 2000). Chun and Oliver (2000) concluded that scores of PSTE and STOE both increased and paralleled each other over the five test times. These findings add merit to the use of the type of workshops mentioned above to increase middle school science teachers’ selfefficacy and thereby make them more productive teachers. Rubeck and Enochs (1991) attempted to investigate an even more specific level of efficacy by distinguishing efficacy associated with chemistry teaching from efficacy associated with science teaching. Enochs, Smith, and Huinker (2000) further developed a similar instrument to measure efficacy of mathematics teaching while Coladarci and 20 Breton (1997) used a modified instrument to measure efficacy in the framework of special education (TschannenMoran et al., 1998). Teacher Efficacy Research—Preservice Teachers In this section some of the more important findings concerning preservice teacher efficacy are addressed. Evans and Tribble (1986), using the TES, compared the perceived teaching problems of 179 preservice elementary and secondary teachers with their level of efficacy. They found that preservice teachers who had elevated efficacy scores were more likely to have elevated levels of professional commitment. Czerniak (1989), using a revised Science Teaching Efficacy Belief Instrument (dubbed the STEBIB) (Enochs & Riggs, 1990), found that the level of personal science teaching efficacy (PSTE) could be positively correlated to confidence in teaching elementary science effectively and negatively correlated to science teaching anxiety. Czerniak (1989) found preservice elementary teachers with high selfefficacy “had less anxiety toward teaching science, were more likely to use openended inquiry and studentdirected teaching strategies, and were more confident about teaching elementary science effectively” (Czerniak & Schriver, 1994, p. 77). Czerniak and Schriver (1994), in a related study, examined elementary preservice science teachers’ beliefs and behavior related to selfefficacy. The 2year longitudinal study used the Science Teacher SelfEfficacy Instrument modified by Czerniak (1989) from the Gibson and Dembo (1984) Teacher Efficacy Scale. Using this modified instrument, Czerniak and Schriver found that “preservice teachers who were in the top 20% of the class on science teaching selfefficacy seemed to display greater conviction 21 that they could successfully help children learn science” (Czerniak & Schriver, 1994, p. 85). These preservice teachers, in order to become better educators, “analyzed their own strengths and weaknesses and sought to help all children learn” (Czerniak & Schriver, 1994, p. 85). Highlevel efficacy preservice teachers “selected strategies that they thought would help children learn science, and they used the educational theories they had learned in their methods class” (Czerniak & Schriver, 1994, p. 85). In contrast to the preservice teachers who were in the top 20% of the class on science teaching selfefficacy, the preservice teachers who were in the bottom 20% of the class on science teaching selfefficacy were unsure of their abilities to be successful teachers in the science classroom. They were overly concerned about noise in the teaching environment and regularly worried about student misconduct. They blamed others for their failures and avoided examining their own skills (Czerniak & Schriver, 1994). There is also research to suggest that student teaching can have an impact on overall teacher efficacy. Teaching experience gained during the student teaching time when evaluated by the TES has been shown to increase personal teaching efficacy (Hoy & Woolfolk, 1990) while general teaching efficacy has been shown to decrease during student teaching (Hoy & Woolfolk, 1990; Spector, 1990). This may be due to overoptimism that is challenged when the student teacher faces the difficulties of the teaching assignment (TschannenMoran et al., 1998). Student teachers with elevated PTE were also rated higher on classroom management, questioning behavior and lessonpresenting behavior by the teachers supervising them (Saklofske, Michaluk, & Randhawa, 1988). Emmer and Hickman (1990), using an adapted TES, found that preservice elementary and secondary teachers that show high teacher efficacy levels in all three 22 subscales (personal teaching efficacy, efficacy for classroom management and discipline and external influences) tend to use classroom management strategies that are aimed at increasing desirable responses through encouragement, praise, rewards, and attention. They also found that preservice teachers, with an elevated sense of personal teacher efficacy, when faced with student discipline problems, were more apt to ask for help. Efficacy has also been related to student control issues (Woolfolk & Hoy, 1990). Preservice teachers with low GTE and PTE or high GTE and low PTE, as measured by the TES tended to have a negative view of students’ motivation, relied on punishment to get students to study, and had a tendency to enforce stringent classroom rules. These findings are in contrast to preservice teachers who were high in both GTE and PTE. Student teachers with high GTE and PTE efficacy scores tended to be more humanistic in their manner of classroom control. Both PTE and GTE of preservice teachers are malleable and may be affected differently by experiences. Social persuasion and vicarious experiences, such as those encountered in college coursework, seem to have a greater effect on preservice teachers’ GTE (Watters & Ginns, 1995). In contrast, actual teaching experiences, such as those encountered in student teaching, seem to exert a greater influence on PTE (Housego, 1992; Hoy & Woolfolk, 1990), although GTE may also change (negatively) during student teaching (Hoy & Woolfolk, 1990; Spector, 1990). Schoon’s and Boone’s (1998) work with preservice elementary teachers using the STEBIB (Enochs & Riggs, 1990) has shown there is an association between elementary teachers’ low science efficacy beliefs and alternative science concepts. The study found that holding certain alternative concepts about science such as planets can only be seen with a telescope, dinosaurs lived the same time as cavemen, and that north is toward the 23 top of a map of Antarctica were linked to subjects with low science teacher efficacy. The study also found that preservice teachers that held fewer numbers of alternative concepts had significantly higher efficacy levels (Schoon & Boone, 1998). Current reform in teacher education has focused on the need for improvement of preservice training (National Research Council, 1996). With this in mind Wingfield and Ramsey (1999) conducted a study that examined the effect of a onesemester sitebased program where preservice teachers participated in authentic classroom and school experiences during their methods class. “The site experiences included: teaching experiences within the assigned classroom, teaching experiences during the methods classes, feedback from the university cluster coordinator, peer and sitebased teacher, observation of the sitebased teacher, and methods class assignments, text, instruction and instructor” (Wingfield & Ramsey, 1999, p. 2). The participants for the study consisted of 131 undergraduate elementary preservice teachers who completed the STEBIB (Enochs & Riggs, 1990) at the beginning and end of the fifteenweek sitebased program. A substantial increase in efficacy from pretest to posttest was noted. Wingfield and Ramsey (1999) concluded that the results indicated that the experiences of the onesemester program had a significant impact on the preservice teacher’s science teaching efficacy beliefs. They also speculated that the additional vicarious teaching experiences may have positively impacted the subjects’ science teaching efficacy. These vicarious teaching experiences specifically included observations of the methods instructor, other preservice teachers and classroom teachers (Wingfield & Ramsey, 1999). King and Wiseman (2001) conducted a study with the purpose of examining differences in science teaching efficacy beliefs among students enrolled in two versions 24 of a methods course in an elementary science teaching program. One group of preservice elementary teachers was enrolled in a semester long interdisciplinary methods class and another group of preservice elementary teachers was enrolled in a semester long more “traditional” noninterdisciplinary methods class. Both groups were given the STEBIB (Enochs & Riggs, 1990) after the methods class. When the results of the STEBI were compared between both groups, neither PSTE nor STOE were found to be significantly different. They concluded their study by stating that if the role of integrated instruction in the elementary curriculum is considered, “the findings of their study suggest that teaching in an integrated fashion and planning interdisciplinary units would seem to be no more effective than traditional teaching in terms of developing the science teaching efficacy of the students” (King & Wiseman, 2001, p. 149). Moseley, Reinke, and Bookout (2002) used Sia’s (1992) Environmental Education Efficacy Belief Instrument to evaluate the effect a 3day outdoor environmental education program would have on 72 participating preservice teachers. The Environmental Education Efficacy Belief Instrument (Sia, 1992), which is based on the STEBI (Enochs & Riggs, 1990), assesses outcome expectancy and selfefficacy in an environmental education teaching beliefs context. All items are based upon a 5point Likertscale response. Moseley, Reinke and Bookout (2002) found that the preservice teacher’s selfefficacy was high before the 3day program and remained unchanged immediately after the program. The preservice teacher’s efficacy was then checked approximately 7 weeks after the conclusion of the 3day program and it had dropped significantly. No change in the outcome expectancy of the subjects was observed over the complete length of the study. The authors accredited the lack of efficacy change 25 during the workshop to the positive characteristics of the 3day program. The drop in efficacy, approximately 7 weeks after the program, “was believed to have resulted from the preservice teachers reevaluation of their ability to teach as they learned more about teaching methodologies” (Moseley, Reinke & Bookout, 2002, p. 9). There are also data to suggest that the number of high school science subjects studied can have a long term effect on the science efficacy of preservice teachers. Mulholland, Dorman and Odgers (2004) used the STEBIB to assess the science efficacy of 314 elementary preservice teachers. They found that the preservice teachers’ PSTE scores were positively related to the number of science classes studied at the high school level but not to their STOE scores. Completing two science teaching classes with the preservice teacher training program also had a significant positive effect on the PSTE but not on the STOE of the subjects. Utley, Moseley and Bryant (2005) explored the impact an elementary methods course and student teaching had on both science and mathematics preservice teacher efficacy. Their study, which used both the STEBIB and the Mathematics Teacher Efficacy Beliefs Instrument (MTEBI) (Huinker & Enochs, 1995), found both a positive and negative relationship between science and mathematics teaching efficacy in their sample population of elementary preservice teachers. Specifically, as the preservice teachers progressed in their methods courses their mathematics and science teacher efficacy also increased significantly. Both science and mathematics efficacy showed a slight decrease after student teaching. Wagler and Moseley (2006) conducted a study to investigate the effects of a secondary contentspecific methods course and student teaching on preservice teacher 26 efficacy. The instrument used in the study was the “The Ohio State Teacher Efficacy Scale” (OSTES).The study employed a single group, pretestposttest Iposttest II design. The repeated measures ANOVA indicated no significant change in overall teacher efficacy from the beginning of the secondary methods course until the end of student teaching; however, overall efficacy did increase significantly after the secondary methods course but by the end of student teaching had returned to its original presecondary methods course level. Classroom management efficacy over all three test times – before and after methods course and after student teaching  was unchanged. Instructional strategies efficacy was shown to be statistically significant and positively affected by the secondary methods course, but no significant change in instructional strategies efficacy was detected after student teaching. No significant change in student engagement efficacy was found immediately following the methods course but student engagement efficacy significantly decreased after student teaching. Vicarious Experience Bandura’s (1997) construct of selfefficacy is influenced by four sources of information, which are (1) enactive mastery experience, (2) vicarious experience, (3) verbal persuasion, and (4) physiological and affective states. Mastery experience is considered the actual act of teaching by the individual. Physiological and affective states or physiological arousal are physiological effects an individual experiences during the teaching act. Vicarious experiences, within the context of teacher efficacy, refers to observing another individual teach. Verbal persuasion is the result of information about teaching conveyed to the preservice teacher or inservice teacher by someone perceived to be an authority. TschannenMoran et al. (1998) and other educational researchers had 27 utilized Bandura’s four sources of efficacy in their teacher efficacy models and instruments. For the purpose of this study we will focus on the source of vicarious experience and how it influences the construct of selfefficacy. Within the context of vicarious experience, modeling is an effective mode for enhancing selfefficacy. An example of this in the context of teacher efficacy would be a preservice teacher who observes, as a participating observer or as a passive observer, a teaching event. In this scenario, the teacher as the model in the context of the vicarious event would have the potential to influence the teaching efficacy of the observer (i.e., preservice teacher). Bandura (1997) points out that for many activities, such as swimming, proficiency and improvement can be measured. The criteria that denote when an individual is swimming are fairly welldefined. We can also quantify improvement by using a variable such as time. For many activities “there are no absolute measures of adequacy. Therefore, people must appraise their capabilities in relation to the attainments of others” (Bandura, 1997, p. 86). One of the ways this is done is by observing models performing tasks. Individuals seek out skilled models because these “competent models transmit knowledge and teach observers effective skills and strategies for managing environmental demands (Bandura, 1986). Acquisition of effective means raises beliefs of personal efficacy” (Bandura, 1997, p. 88). When a person observes another similar individual successfully model a given event, efficacy beliefs are typically raised. Conversely, when a person observes another similar individual fail at modeling a given event, individual efficacy beliefs typically decline (Bandura, 1997). This is especially true if the individual observed is deemed 28 competent by the observer. Competence at a given task, activity or event has been shown to be more effect at increasing efficacy than the age of the model, sex of the model or other personal characteristics (Bandura, 1997). “Model competence is an especially influential factor when observers have a lot to learn and models have much they can teach them by instructive demonstration of skills and strategies” (Bandura, 1997, p.101). Bandura (1977) also proposes that models that convey productive coping techniques can even raise the efficacy of subjects who have experienced many confirmatory personal inefficacious events. On the contrary, subjects who possess high levels of efficacy when performing a given task can have their efficacy raised even higher “if the models teach them even better ways of doing things” (Bandura, 1997, p. 87). “Models who express confidence in the face of difficulties instill a higher sense of efficacy and perseverance in others than do models who begin to doubt themselves as they encounter problems (Zimmerman & Ringle, 1981)” (Bandura, 1997, p. 88). Preservice Teacher Early Field Experiences The research associated with field experiences among preservice teachers is limited. Much of the research conducted in this area occurred in the 1980’s and 1990’s with a few studies occurring in the last six years. Much of the major research has been conducted within the context of physical education. For most preservice teachers, early field experiences involve assisting, in some capacity, in an offcampus school environment (LaMaster, 2001). In the majority of cases the preservice teacher is working in the public school. The situational nature of early field experiences can range from observing teaching to active involvement in the teaching process. Early field experiences occur prior to the preservice teacher’s student teaching assignment (Dodds, 1989) and 29 have been historically viewed as an important component in preservice teacher training programs (Paese, 1989). Because early field experiences are now seen as an essential component of preservice teacher training they have, over the past two decades, moved from a single early field experience to multiple early field experiences before student teaching. Dueck, Altmann, Haslett, and Latimer, (1984) believe these experiences “provide information to students so they can determine their suitability for the teaching profession, orient preservice teachers to schools, and begin the socialization process for potential teachers” (LaMaster, 2001, p. 28). Early field experiences have historically been looked upon as an essential part of a teacher’s socialization (Lasley, Applegate, & Ellison, 1986). Dodds (1989), in a related study on preservice teacher school socialization, stated “field experiences represent the closest juncture between formal teacher training in universities and onthejob training in schools” (p.81). Paese (1984) assessed Early field experiences in terms of their positive benefits in developing the skills of effective teaching and also found that graduates of teacher education programs found early field experiences to be a helpful factor in their teacher training. By providing “real world” experiences, early field experiences also have the possibility of influencing future career decisions (Paese, 1987). Paese (1989) lists seven teaching benefits that are achieved by incorporating EFE’s into preservice teacher training. Among them is the ability of EFE’s to help preservice teachers connect teaching theory to teaching practice, develop a more complete perception of students, gain a better understanding of their future inservice teaching responsibilities and have more of an opportunity to increase and improve their teaching skills. 30 A pilot study was conducted by the author during the spring semester of 2006. There were 50 participants (49 female, 1 male) who were preservice elementary education students enrolled in a course titled Early Lab and Clinical Experience in Elementary Education II at the university. The preservice teachers rated the teacher they observed during their educational field experience (see Appendix E minus questions 8 through 13) and completed the TES (see Appendix B). The results showed only a significant positive correlation between one undefined TES factor and item 1: Rate the quality of the lessons that your field experience teacher used. The undefined TES factor is associated with the teacher’s internal skills and techniques applied to the teaching process. These skills and techniques are learned through teacher training and teacher experiences. From the above literature review, it can be deduced that teacher efficacy has been positively correlated with many desirable teacher behaviors, but little research has been conducted to evaluate the impact of vicarious learning experiences in the context of perceived preservice teacher efficacy and perceived preservice science teacher efficacy. With this in mind, the purpose of this study was to investigate the effect of vicarious experiences on perceived preservice teacher efficacy and perceived preservice science teacher efficacy. The results should be most significant to the elementary education faculty at the studied university in evaluating the effectiveness of vicarious experiences as a tool in preparing preservice elementary teachers to enter the teacher work force. In a broader sense, the results of this study could benefit those attempting to incorporate vicarious events (e.g., elementary field experiences) into their teacher education curriculum. 31 CHAPTER III METHODOLOGY Introduction This chapter describes the way in which the study was conducted. Each research question consists of the study participants, the testing instruments, the procedure and the data analysis needed to answer that specific research question. The research methodology for the study was quantitative and is reflected in the way the data were collected and analyzed. Data were collected through the use of Likertscale instruments and questionnaires that were analyzed through quantitative statistical procedures. For the purposes of this study, a quantitative methodology was preferable to a qualitative approach because it permitted a larger sample size, thereby making the findings and conclusions more generalizable. Research Question 1 What is the impact of a vicarious learning experience (preservice teacher field experiences) on perceived preservice elementary teacher efficacy and perceived preservice elementary science teacher efficacy? Study Participants The participants for this part of the study consisted of 46 undergraduate elementary education students (preservice teachers) who were enrolled in a course titled 32 Early Lab and Clinical Experience in Elementary Education II at the university. Twelve of the 46 preservice teachers were also enrolled in the ExCEL program. The course involved direct observation and participation in classroom environments, kindergarten through eighth grade, and ran concurrent with seminars exploring multicultural education and integrated curricula. Testing Instruments and Procedure The Teacher Efficacy Scale (TES) is a teacher efficacy instrument designed to measure personal teacher efficacy (PTE) and teacher efficacy (TE). The instrument consists of 30 items with a Likert scale ranging from 1 (Strongly Disagree) to 6 (Strongly Agree). Gibson and Dembo (1984) found, after performing factor analysis, that sixteen of the original 30 items had acceptable reliability coefficients. For the current study all 30 items were used, and then factor analysis was conducted to evaluate what factors emerge within the specific test population. Science Teaching Efficacy Belief Instrument B (STEBIB): Science teacher efficacy instrument designed to measure PTSE and STOE. The instrument consists of 23 items from the STEBI. The STEBIB is identical to the STEBI accept items 20 and 25 have been removed and the verb tenses of some of the 23 items have been changed to accommodate preservice teachers. Items 20 and 25 were removed because both crossfactor loaded. These changes were made so the instrument, which was originally designed for inservice teachers, could be used with preservice teachers (Enochs & Riggs, 1990). After thorough analyses, Enochs and Riggs (1990) concluded that the STEBIB could be considered reliable and reasonably valid with a stable and unified factor structure. 33 The Educational Field Experience Teacher Rating and Preservice Teacher Questionnaire refers to the seven educational field experience inservice teacher rating items that have been validated as reflective of goals of the university teacher education program. The Preservice Teacher Questionnaire included items that helped assess the impact of the vicarious field experiences. Early in the spring 2007 semester, before field experiences (vicarious experiences) began, subjects signed informed consent forms and provided limited demographic data by completing a brief questionnaire (see Appendix A). They then completed the STEBIB (see Appendix C). The entire process took less than half an hour and occurred during a regularly scheduled class meeting on January 8, 2007. Near the end of the same semester, after field experiences were completed, subjects rated the teacher they observed during their educational field experiences and provided data about classroom events that occurred while doing their field experiences (see Appendix E), completed the TES (see Appendix B), and again completed the STEBIB (see Appendix C). The entire process took less than half an hour and occurred during a regularly scheduled class meeting on April 23, 2007. Data Analysis For the purposes of data analysis and interpretation, this part of the study was treated as observational (the assignment of subjects into a treated group is outside the control of the investigator), where teacher efficacy, science teacher efficacy, the rating of the observed teacher, the gender of the subjects, the age of the subjects, number of lessons taught by the subject, content area taught by the subjects, science lessons taught by the subjects, teaching rating of the lessons taught by the subjects and number of times 34 the subjects observed a science lesson being taught were the observed variables. The relationship of these variables was assessed using analysis of variance (ANOVA) and analysis of covariance (ANCOVA). Research Question 2 What is the impact of the characteristics of the field experience classroom, within the given school where the educational field experience occurred, on perceived preservice elementary teacher efficacy and perceived elementary preservice science teacher efficacy? Study Participants The participants for this part of the study consisted of the same 46 preservice elementary teachers described above along with twenty inservice teachers whose classrooms served as sites for the preservice teachers’ field experiences. Twelve of the 46 preservice teachers were also enrolled in the ExCEL program. Testing Instruments and Procedure The Classroom Demographic and Cooperating Teacher Questionnaire (see Appendix D) was a questionnaire that allowed the inservice teachers to record the ethnic demographics of the classes they teach, information about their school, and information about themselves. The preservice teachers delivered the questionnaire to their field experience inservice teachers midway through the spring 2007 semester. After granting informed consent, the inservice teachers completed the questionnaire, returned it to the preservice teacher that was observing them, who then returned it to the professor of their field experience course. The researcher collected the questionnaires from the professor. 35 Ethnic data collected from the preservice elementary teachers earlier (see Appendix A) were also used in this part of the study. Data Analysis For the purposes of data analysis and interpretation, this part of the study was treated as observational, where teacher efficacy, science teacher efficacy, the district in which the inservice teacher taught, the number of students the inservice teacher instructed each day, the grade instructed, the ethnicity of the inservice teacher’s classroom where the field observation occurred, the number of students in the classroom who received free and reduced lunch, the age of the inservice teacher, the gender of the inservice teacher, the ethnicity of the inservice teacher, the number of years the inservice teacher had taught, the number of years the inservice teacher had taught at their current school and the number of years the inservice teacher had taught at their current grade level were the observed variables. The relationship of these variables was assessed using analysis of variance (ANOVA) and analysis of covariance (ANCOVA). Assumptions and Limitations of the Study Assumptions The following assumptions were accepted: 1. The preservice elementary teachers understood the directions and the items on the testing instruments and responded to the items honestly, accurately, and to the best of their ability. Limitations The study was limited by the following: 36 1. Only one elementary field experience class with a specific set of components was studied. This prevents generalization to programs that do not resemble this course. 2. The study was conducted with preservice elementary teachers at a landgrant university. This prevents generalizations to other preservice teachers and to other types of elementary education programs at other universities. 3. The subjects were not randomly selected. Selection was determined by required enrollment in the methods courses. 4. The preservice teachers had many concurrent experiences during this phase of their teacher preparation, such as additional coursework, volunteering in public service work, substitute teaching, and/or parttime jobs. All of these experiences may have influenced the final results. 5. The possibility of confounding variables exists in the current study. With any preservice teacher, there were a number of variables that may have confounded the effects of each other. For example, the number of years of teaching experience of the inservice teacher may confound student ethnicity or school setting may confound the number of science lessons the preservice teacher taught. 37 CHAPTER IV RESULTS Introduction This chapter describes the results and statistical analysis associated with the current study. The two research questions that define the current study are presented followed by statistics for the two testing instruments used in the study. Next, some analysis of the demographic data for the study participants is presented. Lastly, the results and statistical analysis associated with the study’s two research questions are discussed. Research Questions Research Question 1 What is the impact of a vicarious learning experience (preservice teacher field experiences) on perceived preservice elementary teacher efficacy and preservice elementary science teacher efficacy? Research Question 2 What is the impact of the characteristics of the field experience classroom, within the given school where the educational field experience occurred, on perceived preservice elementary teacher efficacy and perceived preservice elementary science teacher efficacy? 38 Testing Instruments Data Analysis Principal Components Analysis Principal Components Analysis (PCA) transforms a set of correlated variables into a smaller set of uncorrelated variables (Johnson, 1998). This uncorrelated set of variables is called the principal components. Using the PCA is advisable in determining the number of factors to use in factor analysis (FA) (Johnson, 1998). PCA found two components for each of the Science Teaching Efficacy Belief Instrument B (STEBIB) pretest, STEBIB posttest and the Teacher Efficacy Scale (TES) responses. For the STEBIB pretest response variable, two principal components accounted for 40.3% of the variance. Two principal components accounted for 47.7% of the variance for the STEBIB posttest response variables. Lastly, two principal components accounted for 28.8% of the variance for the TES variables. Thus, since two principal components are sufficient for all three response variables, factor analysis for each instrument were run with only two factors. Factor Analysis A factor analysis (FA) model summarizes a large set of possibly related response variables with a smaller set of uncorrelated response variables (Johnson, 1998). This smaller set of uncorrelated variables explains the relationships that exist between the large set of original variables. The set of 23 items for the STEBIB (pretest and posttest) and the 30 items for the TES make up the three sets of possibly correlated variables. Once a smaller set of factors is determined, all other statistical analyses are performed on the underlying factors and not the original variables. Recall that PCA determined that two factors were identified for all three tests: the STEBIB pretest, the STEBIB posttest 39 and the TES. Table 1 presents the factor loading for both factors for the STEBIB pretest and STEBIB posttest. All STEBIB factor loadings (PSTE and STOE), as identified by Enochs & Riggs (1990), loaded on their correct factor. Using a cutoff factor loading value of 0.45, all items on the STEBIB loaded on either factor 1 (PSTE) or factor 2 (STOE). Values ranged from a low of 0.469 for item 7 to a high of 0.859 for item 18. There were some differences between the pretest and posttest regarding which items loaded on the two factors (PSTE and STOE) (See Table 1). To handle these differences, all PSTE items that loaded from the pretest and/or the posttest were combined and all STOE items that loaded from the pretest and/or the posttest were combined. Table 2 presents these combined factor loadings for the STEBIB pretest and posttest. Item 6 was used as PSTE based on the original instrument (Enochs & Riggs, 1990) and item 9 was omitted from the current study based on incorrect factor loading. Table 3 presents the factor loadings for the Teacher Efficacy Scale (TES). Note that for factor 1 three of the four items were Personal Teaching Efficacy (PTE) on the original Gibson and Dembo (1984) instrument. The fourth item, number 18: “If students are particularly disruptive one day, I ask myself what I have been doing differently,” did not load on the original instrument nor was this item defined as PTE or TE, although it appears to be a PTE item based on its emphasis on selfefficacy. 40 Table 1 STEBIB Original Factor Analysis Loadings Item # Original Instrument Factorc Factor 1 PSTE Factor 2 STOE Pretest 1 3 4 6 7 8 11 14 15 16 17 18 19 20 21 22 STOE PSTE STOE PSTEa STOE PSTE STOE STOE STOE STOE PSTE PSTE PSTE PSTE PSTE PSTE  0.709    0.707     0.676 0.751 0.646 0.562 0.543 0.588 0.625  0.494 0.594 0.494  0.761 0.619 0.654 0.628       Posttest 1 3 4 5 6 7 8 9 12 13 14 16 17 18 19 21 23 STOE PSTE STOE PSTE PSTEa STOE PSTE STOEb PSTE STOE STOE STOE PSTE PSTE PSTE PSTE PSTE  0.519  0.793 0.734  0.649 0.503 0.791    0.709 0.859 0.778 0.775 0.695 0.713  0.816   0.469    0.506 0.574 0.639      Note. Cut off for Factor Loading of >.45 aDouble Factor Loading. Item 6 was used as PSTE based on the original instrument (Enochs & Riggs, 1990). bIncorrect Factor Loading, Item 9 was omitted from the current study. cAs identified by Enochs & Riggs, 1990 41 Table 2 STEBIB Combined Factor Analysis Loadings Item # Positive/Negative Wording Original Instrument Factor Pretest Posttest Factor 1 PSTE 3 5 6 8 12 17 18 19 20 21 22 23 N P N N P N P N N N P N PSTE PSTE PSTE PSTE PSTE PSTE PSTE PSTE PSTE PSTE PSTE PSTE 0.709 0.326 0.375 0.707 0.367 0.676 0.751 0.646 0.562 0.543 0.588 0.378 0.519 0.793 0.734 0.649 0.791 0.709 0.859 0.778 0.437 0.775 0.270 0.695 Factor 2 STOE 1 4 7 11 13 14 15 16 P P P P N P P P STOE STOE STOE STOE STOE STOE STOE STOE 0.625 0.494 0.494 0.761 0.203 0.619 0.654 0.628 0.713 0.816 0.469 0.314 0.506 0.574 0.420 0.639 Note. Cut off for Factor Loading of >.45 Cronbach’s Alpha Reliabilities Cronbach’s alpha is a measure of the internal reliability of an instrument. Interpretation of Cronbach’s alpha presumes that an instrument consisting of K items is only a subset of all possible items that could be used to measure the topic of interest. Cronbach’s alpha is the correlation between the total scores of any two random samples from the entire population of all possible items (Johnson, 1998). Thus, just as any correlation, Cronbach’s alpha may range from 0 to 1, but generally any score above 0.7 is 42 considered to be an acceptable indicator of the instrument’s internal reliability (Johnson, 1998). Table 4 contains the Crombach’s alpha reliabilities for the STEBIB and TES. Table 3 TES Factor Analysis Item # Positive/Negative Wording Original Instrument Factor Loadings Factor 1 PTE 1 14 18 19 P P P P PTE PTE Undefineda PTE 0.688 0.791 0.489 0.928 Indefinable Factor 2 2 3 5 8 12 13 15 20 27 P P P P P N P P P TE Undefineda Undefineda Undefineda PTE Undefineda PTE Undefineda TE 0.550 0.462 0.675 0.534 0.458 0.640 0.554 0.689 0.656 Note. Cut off for Factor Loading of >.45 aNonFactor Loading on Original 30 Item Instrument (Gibson and Dembo, 1984). Table 4 Crombach’s Alpha Reliabilities STEBIB Factor 1(PSTE) Factor 2 (STEO) Pretest Posttest Pretest Posttest .846 .909 .727 .77 Factor 1 (PTE) Factor 2 (Indefinable Factor 2 ) TES .798 .909 43 Summary Testing Instrument Statistics Table 5 presents the means, standard deviations, ttest values and probabilities for the pretest and posttest scores on the Science Teaching Efficacy Belief Instrument B (STEBIB). Statistics for both factors personal science teacher efficacy (PSTE) and science teacher outcome expectancy (STOE) linked with the STEBIB are presented for both the pretest and posttest. A tTest (α = 0.10) revealed that the PSTE posttest mean of 3.955 was significantly greater than the pretest mean of 3.777 (t = 2.519, p = 0.015). Similarly, the STOE posttest mean of 3.685 was significantly greater than the pretest mean of 3.492 (t = 1.979, p = 0.054). Table 5 STEBIB Summary Testing Instrument Statistics PSTE pretest PSTE posttest STOE pretest STOE posttest Mean 3.777 3.955 3.492 3.685 S.D. 0.524 0.603 0.527 0.535 tTest 2.519 1.979 P(t) 0.015 0.054 Table 6 presents the means and standard deviations for the Teacher Efficacy Scale (TES). Both factors “Personal Teacher Efficacy (PTE)” and “Indefinable Factor 2 (IF2)” associated with the TES are presented for both the pretest and posttest. Indefinable Factor 2 (IF2) is the second TES factor identified in the present study. Items associated with this factor were 2, 3, 5, 8, 12, 13, 15, 20 and 27. Based on the nonassociative nature of these items, a common construct factor, such as TE, was deemed invalid. 44 Table 6 TES Summary Testing Instrument Statistics PTE IF2 Mean 4.446 4.592 S.D. 0.813 0.615 Study Participants’ Demographic Data Analysis Preservice Teachers Fortysix preservice teachers agreed to participate in the study. Fortyfive were female and one was male. Fortyone were White, four were American Indian/or Alaskan and one was Hispanic/Latino. Twelve of the 46 preservice teachers were also enrolled in the ExCEL program. The mean age of all 46 preservice teachers was 22. The minimum age was 20, the maximum age was 29, the median age was 22 and the standard deviation was 1.71. Table 7 presents all 46 preservice teacher responses to the question “How many times did you teach a lesson?” from the preservice post data collection event (See Appendix E). “T” denotes preservice teachers who were in the traditional observation program and “E” denotes preservice teachers who participated in the ExCEL program. The ExCEL (Excellence in Collaborative and Experiential Learning) Program is a collaborative partnership between the College of Education and the local public school district. Open to elementary education majors in their final semester before student teaching (Semester Y), the program features a threeday a week full day internship in a public school classroom in which teacher candidates are partnered with expert classroom teachers. By observing the range of lessons taught, thirtytwo of the thirtyfive 45 preservice elementary teachers who taught between 1 and 10 lessons were in the traditional observation program. Two traditional observation program preservice elementary teachers taught between 1120 lessons. All preservice teachers who taught more than 21 lessons were in the ExCEL program. Table 7 Preservice Teacher Question Mean Lessons Taught 10 Minimum Lessons Taught 1 Maximum Lessons Taught 50 SD of Lessons Taught 13.623 Range of Lessons Taught # of preservice teachers 1 to 10 Lessons Taught 32 T/3E 11 to 20 Lessons Taught 2 T 21 to 30 Lessons Taught 5 E 31 to 40 Lessons Taught 2 E 41 to 50 Lessons Taught 2 E Note. T = Traditional elementary education program, E = ExCEL elementary education program. Table 8 presents all 46 preservice teacher responses to the question “If you taught science, how many times did you teach a science lesson?” from the preservice post data collection event (See Appendix E). “T” denotes preservice teachers who were in the traditional observation program and “E” denotes preservice teachers who participated in 46 the ExCEL program. By observing the range of science lessons taught, fourteen traditional observation program preservice elementary teachers taught no lessons. Eighteen traditional observation program and five ExCEL program preservice elementary teachers taught only one or two science lessons. One traditional observation program and one ExCEL program preservice elementary teacher taught three or four science lessons. Lastly, all preservice teachers who taught more than five science lessons were in the ExCEL program. Table 8 Preservice Teacher Question Mean Science Lessons Taught 1.67 Minimum Science Lessons Taught 0 Maximum Science Lessons Taught 10 SD of Science Lessons Taught 2.35 Range of Science Lessons Taught # of preservice teachers 0 Science Lessons Taught 14 T 12 Science Lessons Taught 18 T/5 E 34 Science Lessons Taught 1 T/1 E 56 Science Lessons Taught 4 E 78 Science Lessons Taught 1 E 910 Science Lessons Taught 1 E Note. T = Traditional elementary education program, E = ExCEL elementary education program. 47 Table 9 presents all 46 preservice teacher responses to the question “How many times did you observe a science lesson being taught?” from the preservice post data collection event (See Appendix E). Again, “T” denotes preservice teachers who were in the traditional observation program and “E” denotes preservice teachers who participated in the ExCEL program. An examination of the range of science lessons observed revealed that all 34 of the preservice elementary teachers in the traditional observation program observed nine or fewer lessons, while ten of the twelve ExCEL preservice teachers observed more than ten science lessons. Table 9 Preservice Teacher Question Mean Observed Science Lessons 8.09 Minimum Observed Science Lessons 0 Maximum Observed Science Lessons 45 SD of Observed Science Lessons 13.133 Range of Observed Science Lessons # of preservice teachers 09 Observed Science Lessons 34 T/2 E 1018 Observed Science Lessons 3 E 1927 Observed Science Lessons 1 E 2836 Observed Science Lessons 2 E 3745 Observed Science Lessons 4 E Note. T = Traditional elementary education program, E = ExCEL elementary education program. 48 Inservice Teachers Twenty inservice teachers agreed to participate in the study. Nineteen were female and one was male. Eighteen were white and two were American Indian/or Alaskan. The mean age of all twenty inservice teachers was 44. The minimum age was 24, the maximum age was 59, the median was 40 and the standard deviation was 11.71 Table 10 presents all 20 inservice teacher responses to the questions (1) “How many years of teaching have you completed?”; (2) “How many years have you taught at your current school?”; and (3) “How many years have you taught at your current grade level?”. These questions are from the inservice data collection event (See Appendix D). Table 10 Select Inservice Teacher Questions Mean # of years teaching completed 15.3 Minimum # of years teaching completed 1.5 Maximum # of years teaching completed 37 SD of # of years teaching completed 9.99 Mean # of years taught at your current school 5.725 Minimum # of years taught at your current school 1 Maximum # of years taught at your current school 25 SD of # of years taught at your current school 5.63 Mean # of years taught at current grade level 6.7 Minimum # of years taught at current grade level 1 Maximum # of years taught at current grade level 23 SD of # of years taught at current grade level 6.764 49 Introduction to Research Question 1 and 2 Analysis In order to answer research questions 1 and 2 for the STEBIB test, analysis of covariance (ANCOVA) was utilized. ANCOVA is a statistical procedure that tests a set of factors for significance on the response variable while removing the variance for which the covariant accounts. For both research questions, the response variable is the posttest score for the STEBIB and the covariant is the pretest score for the STEBIB. The inclusion of the pretest score into the model as a covariant can increase power because it accounts for additional variability had the covariant been left out of the model. Forward stepwise selection was used as a variable selection method for the final ANCOVA linear model. This variable selection method selects the most parsimonious set of factors for the ANCOVA linear model. Research questions 1 and 2 for the TES were addressed using analysis of variance (ANOVA). ANOVA is a statistical procedure that relates a set of quantitative factors to a response variable. There is no covariant for the TES since it was given only one time. The categories associated with each variable were based on the distribution of the data for each specific category. Because six out of ten of the variables were discrete (discontinuous) and because the data of several of the continuous variables were not normally distributed, the responses were grouped into a manageable number of categories. There was an attempt to equalize the number of subjects in each category in order to pick up true existing differences between the categories. Appendix F contains tables detailing the makeup of each of these categories for variables related to preservice teachers, inservice teachers, and field experience classrooms. 50 Table 11 Variables and Categories Associated with Research Question 1 Variable Categories Age of preservice teacher < 22 years ≥ 22 years (N = 22) (N = 24) Gender of preservice teacher Male Female (N = 1) (N = 45) White Amer. Ind. or Alaskan Hispanic/ Ethnicity of preservice teacher Latino (N = 41) (N = 4) (N = 1) Rating of the inservice field experience teacher by the preservice teacher Likert Scale 15 (Poor to Excellent) Number of lessons the preservice teacher 1 24 510 1250 taught (N = 12) (N = 10) (N = 13) (N = 11) Number of science lessons the preservice 0 1 210 teacher taught (N = 14) (N = 22) (N = 10) Selfrating of the science lessons taught by the preservice teacher Likert Scale 15 (Poor to Excellent) Selfrating of all lessons taught by the preservice teacher Likert Scale 15 (Poor to Excellent) Number of science lessons the preservice 0 12 310 1545 teacher observed (N = 13) (N = 12) (N = 13) (N = 8) Was the preservice teacher part of the ExCEL program? ExCEL NonExCEL (N = 12) (N = 34) Research Question 1 Analysis Table 11 lists the variables associated with research question 1: “What is the impact of a vicarious learning experience (preservice teacher field experiences) on perceived preservice elementary teacher efficacy and preservice elementary science 51 teacher efficacy?” Each variable is presented along with the categories related to that variable. STEBIB Results For both factors 1 and 2, the ANCOVA model for research question 1 had no statistically significant independent variables (α = 0.10). TES Results The ANOVA model for factor 1 of the TES did not have any significant independent variables (α = 0.10). However, the ANOVA model for factor 2 of the TES did have a statistically significant independent variable, the age of the preservice teacher. Table 12 contains the ANOVA results. The means resulting from the ANOVA are given in Table 13. Research Question 2 Analysis Table 14 lists the variables associated with research question 2: “What is the impact of the characteristics of the field experience classroom, within the given school where the educational field experience occurred, on perceived preservice elementary teacher efficacy and perceived preservice elementary science teacher efficacy?” Each variable is presented along with the categories related to that variable. Table 12 Age of Preservice Teacher ANOVA Df Sum of Sq Mean Sq F Value Pr(F) Age of Preservice Teacher 1 1.09242 1.092424 3.017598 0.08936196* Error 44 15.92878 0.362018 *p< .10 52 Table 13 Age of Preservice Teacher Means Variable Mean of TES Age of Preservice Teacher (Below 22 years of age) 4.757 Age of Preservice Teacher (22 years of age or older) 4.449 STEBIB Results The final ANCOVA model for research question 2, factor 1 of the STEBIB test includes three independent variables as well as the covariate. Table 15 contains the ANCOVA model results. In the ANCOVA table, ‘Group’ refers to the collective effect of the independent variables in the model while ‘Covariate’ refers to the effect of the covariate (the pretest score on the STEBIB) on the response variable (the posttest score of the STEBIB). 53 Table 14 Variables and Categories Associated with Research Question 2 Variable Categories Name of the inservice teachers school Schools 1 through 7 Name of the inservice teachers district Districts 1 through 3 Number of students the inservice 1617 2225 1920 62147 teacher instructed per day (N = 5) (N = 6) (N = 5) (N = 4) Grade the inservice teacher 1st2nd 3rd 4th7th instructs (N = 6) (N = 6) (N = 8) Percentage of Hispanic/Latino 0% 3.1%11.8% 21%75% students the inservice teacher instructs each day (N = 8) (N = 7) (N = 5) 0% 4.5% 5.3% 8.1% 10.2% 12.5% 14.8% 18.2% 20% 77.3% Percentage of Black students the inservice teacher instructs each day (N = 4) (N = 4) (N = 4) (N = 5) (N = 3) 10.2%50% 52.9%73.7% 75%95% Percentage of White students the inservice teacher instructs each day (N = 7) (N = 7) (N = 6) Percentage of Asian/Pacific Islander students the inservice teacher instructs each day 0% 1.4%12% (N = 12) (N = 8) Percentage of American Indian/or Alaskan students the inservice teacher instructs each day 0%2.7% 4%57.9% (N = 10) (N = 10) Percentage of other ethnicity students the inservice teacher instructs each day 0% 4%18.8% (N = 13) (N = 7) Percentage of students the inservice teacher instructs that receive free and reduces lunch 9.736.8% 58.8100% (N = 5) (N = 15) Age of the inservice teacher ≤40 years >40 years (N = 10) (N = 10) Gender of the inservice teacher Male Female (N = 1) (N = 19) Ethnicity of the inservice teacher White American Indian/Alaskan (N = 18) (N = 2) Number of years teaching the inservice teacher had completed 2.5 yrs11 yrs 15 yrs37 yrs (N = 10) (N = 10) 54 Table 14 Cont’d. Variables and Categories Associated with Research Question 2 1 year 2 yrs3 yrs 4 yrs7 yrs 8 yrs25 yrs Number of years the inservice teacher had been at their current school (N = 5) (N = 5) (N = 5) (N = 5) Number of years the inservice teacher had taught at their current grade level 1 yr3 yrs 5 yrs23 yrs (N = 12) (N = 8) Was the preservice teacher part of the ExCEL program? ExCEL NonExCEL (N = 12) (N = 34) Table 15 ANCOVA Model Results for Research Question 2, Factor 1 of the STEBIB Df Sum of Sq Mean Sq F Value Pr(F) Group 3 1.023479 2.563234 0.09360104* Covariate 1 3.079642 3.079642 23.13828 0.00022924* Error 15 1.996460 0.133097 *p< .10 The test for the significance of the covariate in the model resulted in an F test statistic of F* = 23.13828 and an estimated pvalue 0.0002292445. The overall F test for equality of the means for each level of the independent variable was F* = 2.563234 and had an estimated significance level of 0.09360104. The significant predictors for the ANCOVA model included the inservice teacher ethnicity (p = 0.0832), the inservice teacher gender (p = 0.0065), and the number of free lunch students (p = 0.0197). Table 16 contains the means for factor 1 of the STEBIB posttest associated with the ANCOVA linear model. Adjusted means are computed for each category of each independent variable holding all other variables, including the covariate, constant using their respective mean values. 55 Table 16 Means for Factor 1 of the STEBIB Posttest Associated with the ANCOVA Variable Mean of Posttest STEBIB Ethnicity of the inservice teacher (White) 3.892 Ethnicity of the inservice teacher (American Indian/or Alaskan) 3.243 Gender of the inservice teacher (Male) 3.513 Gender of the inservice teacher (Female) 3.621 Percentage of students the inservice teacher instructs that receive free and reduces lunch (9.7%36.8%) 3.711 Percentage of students the inservice teacher instructs that receive free and reduces lunch (58.8%100%) 3.423 Table 17 ANCOVA Model Results for Research Question 2, Factor 2 of the STEBIB Df Sum of Sq Mean Sq F Value Pr(F) Group 5 2.76232 8.681305 0.00083516* Covariate 1 0.385161 0.3851610 6.05234 0.02866086* Error 13 0.827299 0.0636384 *p< 0.10 Factor 2 of the STEBIB yielded an ANCOVA model with four independent variables. Table 17 presents the ANCOVA results. Note that the test for the significance of the covariate yielded a significant F test (F* = 6.052, p = 0.0287). The test for overall equality of the group means was also significant (F* = 8.681, p = 0.00084). The significant predictors for the ANCOVA model included the percentage of Asian/Pacific Islander students (p = 0.0865), the percentage of Hispanic/Latino students (p = 0.0005), 56 the percentage of American Indian/or Alaskan students (p = 0.0293), and the indicator variable for ExCEL program participation (p = 0.0064). Table 18 contains the means for each level of each independent variable. Table 18 Means for Factor 2 of the STEBIB Posttest Associated with the ANCOVA Variable Mean of Posttest STEBIB Percentage of Asian/Pacific Islander students the inservice teacher instructs each day (0%) 3.727 Percentage of Asian/Pacific Islander students the inservice teacher instructs each day (1.4%12%) 3.556 Percentage of Hispanic/Latino students the inservice teacher instructs each day (0%) 3.817 Percentage of Hispanic/Latino students the inservice teacher instructs each day (3.1%11.8%) 3.301 Percentage of Hispanic/Latino students the inservice teacher instructs each day (21%75%) 3.807 Percentage of American Indian/or Alaskan students the inservice teacher instructs each day (0%2.7%) 3.512 Percentage of American Indian/or Alaskan students the inservice teacher instructs each day (4%57.9%) 3.771 ExCEL preservice teacher 3.456 NonExCEL preservice teacher) 3.827 Table 19 contains the Tukey simultaneous confidence intervals to determine where the means are significantly different. Tukey simultaneous confidence intervals are necessary when considering the independent variable for Hispanic/Latino students the inservice teacher instructs each day since there are three levels. Thus, we will compare all levels of Hispanic/Latino students the inservice teacher instructs each day with a 57 controlled experimentwise error rate of 0.10. Note that the Posttest STEBIB means for categoriesl 0% and 3.1%11.8% are different as are the means for categories 3.1%11.8% and 21%75%. However, the means for categories 0% and 21%75% are not statistically different. Additionally, point estimates for the mean differences appear in the “Estimate” column. For example, the mean for 0% Hispanic/Latino students is 3.817. The adjusted mean for 3.1%11.8% Hispanic/Latino students is 3.301. Their difference between these means is 0.516. Concerning the “Interval” column, if the interval does not contain 0 then the means are statistically significant. If the range is all positive numbers then the first mean is larger. If the range is all negative numbers, then the first mean is smaller. Table 19 Tukey Simultaneous Confidence Intervals for Factor 2 of the STEBIB Comparison level Estimate Interval 0% compared to 3.111.8% 0.516 (0.182, 0.850)* 0% compared to 2175% 0.009 (0.413, 0.432) 3.111.8% compared to 2175% 0.507 (0.873, 0.140)* *Significant interval TES Results Table 20 contains the results of the ANOVA for TES factor 1. Note that the percentage of Hispanic/Latino students the inservice teacher instructs each day was the only significant independent variable (F = 2.8977, p = 0.082621). Additionally, the means for TES for each category of Hispanic/Latino students the inservice teacher instructs each day are given in Table 21. 58 Table 20 ANOVA for TES Factor 1 Df Sum of Sq Mean Sq F Value Pr(F) Percentage of Hispanic/Latino students the inservice teacher instructs each day 2 3.22567 1.612835 2.8977 0.082621* Error 17 9.46183 0.556578 *p< 0.10 Table 21 TES Factor 1 Means for Hispanic/Latino Students Variable Mean of TES The percentage of Hispanic/Latino students the inservice teacher instructs each day (0%) 5.031 The percentage of Hispanic/Latino students the inservice teacher instructs each day (3.1%11.8%) 4.107 The percentage of Hispanic/Latino students the inservice teacher instructs each day (21%75%) 4.700 Table 22 contains the Tukey simultaneous confidence intervals to determine where the means are significantly different. Tukey simultaneous confidence intervals are necessary when considering the independent variable for Hispanic/Latino students the inservice teacher instructs each day since there are three levels. Thus, we will compare all levels of Hispanic/Latino students the inservice teacher instructs each day with a controlled experimentwise error rate of 0.10. Note that the means for TES are different for categories 0% and 3.1%11.8% Hispanic/Latino students the inservice teacher instructs each day, but not for categories 0% and 2175% or categories 3.1%11.8% and 21%75%. Additionally, the point estimate (Estimate) is the difference between the 59 means. For example, 5.0314.107 = 0.924 is the estimated difference between 0% and 3.111% Hispanic/Latino students. Concerning the “Interval” column, if the interval does not contain 0 then the means are statistically significant. If the range is all positive numbers then the first mean is larger. If the range is all negative numbers then the first mean is smaller. Table 22 Tukey Simultaneous Confidence Intervals for Factor 1 of the TES Comparison level Estimate Interval 0% compared to 3.111.8% 0.924 (0.075, 1.770)* 0% compared to 2175% 0.331 (0.604, 1.270) 3.111.8% compared to 2175% 0.593 (1.550, 0.368) *Significant interval Table 23 contains the results of the ANOVA for TES factor 2. Jefferson Middle School, one of seven schools where tie inservice teachers taught, was the only significant independent variable in this model (F = 4.641841, p = 0.04499545). The mean TES Factor 2 score at Jefferson Middle School was lower when compared to the mean score of the other schools (See Table 24). Table 23 ANOVA for TES Factor 2 Df Sum of Sq Mean Sq F Value Pr(F) Jefferson Middle School 1 1.141358 1.141358 4.641841 0.04499545 Error 18 4.425926 0.245885 60 Table 24 Means for TES Factor 2 Variable Mean of TES Jefferson Middle School 3.833 Schools other than Jefferson Middle School 4.630 61 CHAPTER V CONCLUSIONS AND RECOMMENDATIONS Introduction This chapter describes the conclusions and recommendations associated with the study. The two research questions that define the current study are presented followed by the testing instruments conclusions and study participants demographic conclusions. Next, the conclusions associated with the study’s two research questions are discussed. Lastly, recommendations for future research, implications for practice and concluding remarks will be presented. Research Questions Research Question 1 What is the impact of a vicarious learning experience (preservice teacher field experiences) on perceived preservice elementary teacher efficacy and preservice elementary science teacher efficacy? Research Question 2 What is the impact of the characteristics of the field experience classroom, within the given school where the educational field experience occurred, on perceived preservice elementary teacher efficacy and perceived preservice elementary science teacher efficacy? 62 Testing Instruments Conclusions Principal Components Analysis Principal Components Analysis (PCA) was performed on the Science Teaching Efficacy Belief Instrument B (STEBIB) pretest responses, STEBIB posttest responses and the Teacher Efficacy Scale (TES) responses. Johnson (1998) advises using PCA to determine the number of factors to use in factor analysis (FA). For the STEBIB pretest response variables, two principal components accounted for 40.3% of the variance. Two principal components accounted for 47.7% of the variance for the STEBIB posttest response variables. With both the STEBIB pretest and STEBIB posttest this percentage of variance further validated the use of a twofactor model. The STEBIB posttest variance level was 7.4% higher than the STEBIB pretest further strengthening the argument for two STEBIB factors for both the pretest and the posttest. When comparing the TES variance, which is 28.8% for the TES variables, to both the STEBIB pretest and posttest variance, the TES variables is 11.5% lower than the STEBIB pretest response variables and 18.9% lower than the variance for the STEBIB posttest response variables. Even though the amount of variance associated with the TES was lower than the STEBIB, the percentage of variance still suggests the use of a twofactor model. In conclusion, the PCA validated that two principal components are sufficient for both the STEBIB pretest and posttest and the TES although the amount of variance accounted for was lower for the TES than the STEBIB pretest and posttest. Based on the PCA results, the factor analysis was run with only two factors for both the STEBIB and the TES. 63 Factor Analysis Factor analysis was conducted using a twofactor model for both the STEBIB pretest and STEBIB posttest. Factor loading measures the contribution of the factor, for example PSTE, to the STEBIB response. Table 1 presents the factor loading for both factors for the STEBIB pretest and STEBIB posttest at a cut off for factor loading of >.45. Note that the items that loaded for factor 1 [Personal Science Teaching Efficacy (PSTE)] correspond to items concerning PSTE, as identified by Enochs and Riggs in their original instrument analysis (1990), with the exception of item 9 in the posttest. Conversely, the items that loaded for factor 2 [Science Teaching Outcome Expectancy (STOE)] correspond to items concerning STOE, as identified by Enochs and Riggs in their original instrument analysis (1990), with the exception of item 6 for the pretest. This leads to a clear interpretation for these two factors based on the PCA, the factor analysis and the commonality between this study’s factor loading and that of Enochs and Riggs’ (1990) original study. With confidence we can conclude that, in the current study, two factors, PSTE and STOE, were measured by the STEBIB. This is consistent with findings of other researchers have found (King & Wiseman, 2001; Mulholland, Dorman & Odgers, 2004). Factor analysis was also conducted using a twofactor model for the TES. Table 3 presents the factor loadings for the TES. Note that for factor 1 [Personal Teacher Efficacy (PTE)] three of the four items are PTE on the original Gibson and Dembo (1984) instrument. The fourth item, item 18: “If students are particularly disruptive one day, I ask myself what I have been doing differently,” did not load on the original instrument nor was defined on the original instrument as PTE or TE but is a PTE item 64 based on its emphasis on selfefficacy and locus of control. Note that for factor 2 [Teacher Efficacy (TE)] we have TE items, PTE items and items that did not load on either factors of the original 30Item Instrument (Gibson & Dembo, 1984). This leads to a very clear interpretation for these two factors based on the PCA, the factor analysis and the commonality and discrepancy between this study’s factor loading and Gibson and Dembo’s (1984) original study factor loading. With confidence we can conclude that, in the current study, one factor, PTE, was measured by the TES. The second factor deemed TE in the original Gibson and Dembo (1984) instrument was not identified in the current study. Rather a second factor was indefinable based on the nonassociated nature of the items that loaded on it. As pointed out previously, the items that loaded for this factor were TE items, PTE items, and items that did not load on any factor on the original 30 Item Instrument (Gibson and Dembo, 1984). This second factor in the current study associated with the TES has been, for this study, named Indefinable Factor 2 (IF2). The problematic nature of the TES is consistent with what other researchers have found and eliminates the ability to draw any conclusions associated with factor 2 of the TES. The problematic nature of the TES and the past research related to the instrument will be discussed in greater detail in the section labeled “The Problematic Nature of the Teacher Efficacy Scale.” Cronbach’s Alpha Reliabilities Cronbach’s alpha is a measurement that assesses the internal reliability of an instrument. Cronbach’s alpha was performed on the STEBIB for factor 1 (PSTE) and factor 2 (STOE) for both the pretest and the posttest data. Cronbach’s Alpha was also performed on the TES for factor 1 (PTE) and factor 2 (IF2) for the TES’s single data set 65 (See table 4). Cronbach’s alpha may range from 0 to 1, and generally, any score above 0.7 is considered an indicator of good internal reliability. All STEBIB Cronbach’s Alpha reliability results were above the 0.7 cut off that defines satisfactory internal reliability of an instrument. Concerning the specific STEBIB Cronbach’s Alpha reliability results it should be noted that the internal reliability of the instrument increased from the pretest to the posttest for both factor 1 (PSTE) and factor 2 (STOE). This trend further increases the validity of conclusions associated with the STEBIB. The TES Cronbach’s Alpha reliability results are also above the 0.7 cut off that defines satisfactory internal reliability of an instrument. In conclusion, all Crombach’s alpha values related to both instruments used in this study were above 0.7 indicating we have sufficient internal reliability to assess the internal factors of these instruments. Summary Testing Instrument Statistics Means, standard deviations, ttest values and probabilities for the pretest and posttest scores were calculated for both factors of the STEBIB (See Table 5). There was found a significant difference between the PSTE pretest and the PSTE posttest with the PSTE pretest score of 3.777 increasing to 3.955 for the PSTE posttest. There was also found a significant difference between the STOE pretest and the STOE posttest with the STOE pretest score of 3.492 increasing to 3.685 for the STOE posttest. For both PSTE and STOE all means for the pretest and the posttest were between the STEBIB Likert Scale categories of “Uncertain (3)” and “Agree (4).” Perhaps the overall increase that is observed in both PSTE and STOE is indicative of the positive efficacious events the preservice teachers experienced during CIED 3430 66 (Early Lab and Clinical Experience in Elementary Education II) and in their inservice teacher’s classroom over the testing time. Note that the pretest mean scores are moderately positive and the increase that we see over the study time is small. The case maybe that this tendency to be just above “Uncertain (3)” is due to the preservice teacher’s general anxieties, fears, and lack of selfconfidence associated with understanding science and science teaching during the study time. This relationship has been well documented in the responses of preservice elementary teachers (Czerniak, 1989; Enochs & Riggs, 1990; Czerniak & Schriver, 1994). Mean and standard deviation were calculated for factor 1 (PTE) of the TES (See Table 6). The mean for PTE was 4.446. The PTE mean was between the TES Likert Scale categories of “Agree slightly more than disagree (4)” and ‘Moderately agree (5).” Again, the level of PTE at the conclusion of the treatment was moderately positive. This tendency to be just above “Agree slightly more than disagree (4)” is maybe indicative of the preservice teachers’ experiences in their teacher training. In most cases they have very little, if any, teaching experiences or other experiences within a given school system to serve as a basis for their teacher selfefficacy beliefs. This inexperience is evident in their moderately positive responses associated with what they believe their teaching abilities are. The mean and standard deviation were also determined for factor 2 (IF2) of the TES (See Table 6). The mean for IF2 was 4.592. The IF2 mean was also between the TES Likert Scale categories of “Agree slightly more than disagree (4)” and ‘Moderately agree (5)” but, considering the indefinable nature of this factor, no conclusions can drawn. 67 Study Participants Demographic Conclusions Preservice Teachers Table 7 summarizes all 46 preservice teacher responses to the question “How many times did you teach a lesson?” from the preservice post data collection event. From the table it is apparent that there is a wide range in the number of lessons the preservice teachers taught during the treatment time. The great majority of the preservice teachers taught between 1and10 lessons with only 2 out of 35 of the preservice teachers in this category being in the ExCEL program. Two traditional preservice teachers taught between 11 and 20 lessons while the remaining preservice teachers who taught from 21 to 50 lessons were enrolled in the ExCEL program. Table 8 summarizes all 46 preservice teacher responses to the question “If you taught science, how many times did you teach a science lesson?” The number of science lessons the preservice teachers taught during their field experience ranged from 0 to 10. As with the previous question, the traditional preservice teachers were grouped on the lower end of the range while the ExCEL program preservice teachers were grouped on the higher end of the range. Table 9 summarizes all 46 preservice teacher responses to the question “How many times did you observe a science lesson being taught?” As with the previous two preservice questions, there is a similar trend where the traditional preservice teachers were grouped on the lower end of the range and the ExCEL program preservice teachers were grouped on the higher end of the range of science lessons observed. Thirtyfour out of 36 preservice teachers observed between 0 and 9 lessons. Only the ExCEL preservice teachers observed from 10 to 45 lessons. 68 In conclusion, the general trends we observed were the ExCEL preservice teachers are teaching more lessons, teaching more science lessons, and observing more science lessons. The traditional preservice teachers are teaching fewer lessons, teaching fewer science lessons, and observing fewer science lessons. The results are indicative of the constraints placed on the preservice teachers by their participation in either the Traditional or ExCEL program. Inservice Teachers Table 10 summarizes all 20 inservice teacher responses to the questions (1) “How many years of teaching have you completed?”; (2) “How many years have you taught at your current school?”; and (3) “How many years have you taught at your current grade level?”. With all three questions the responses disclosed a wide range of years. This is apparent by observing the “Minimum number of years,” “Maximum number of years” and “SD of number of years” of all three questions. The “Mean number of years teaching completed” is also much larger than the “Mean number of years taught at your current school” and the “Mean number of years taught at current grade level.” Obviously, this is to be expected since both “Mean number of years taught at your current school” and “Mean number of years taught at current grade level” are subcomponents of the “Mean number of years teaching completed.” Seventeen of the 20 inservice teachers have 7 or more years of teaching experience with a maximum of 25 years of teaching experience (see Table 10). This is not surprising considering the desire of the preservice teacher faculty to place preservice teachers in observation classrooms with inservice teachers who have many successful years of teaching experience. Eleven of the 20 inservice teachers have taught between 1 69 and 4 years at their current school. Thirteen of the 20 inservice teachers have taught between 1 and 5 years at their current grade level. Research Question 1 “What is the impact of a vicarious learning experience (preservice teacher field experiences) on perceived preservice elementary teacher efficacy and preservice elementary science teacher efficacy?” Research Question 1 Conclusions STEBIB Factor 1 and 2 Results Table 11 lists the variables and categories associated with research question 1. For both STEBIB factors PSTE and STOE, the ANCOVA model for research question 1 had no statistically significant independent variables (α = 0.10). TES Factor 1 and 2 Results Table 11 lists the variables and categories associated with research question 1. The ANOVA model for TES factor 1 (PTE) had no significant independent variables (α = 0.10). However, the ANOVA model for TES factor 2 (IF2) of the TES did have one statistically significant independent variable, the age of the preservice teacher (p = 0.089). Table 12 contains the ANOVA results. Unfortunately, the problematic nature of TES factor 2 (IF2) eliminates the ability to draw any conclusions associated with factor 2 of the TES. The problematic nature of the TES and the past research related to the instrument is discussed in greater detail below in the section labeled “The Problematic Nature of the Teacher Efficacy Scale.” 70 Research Question 2 “What is the impact of the characteristics of the field experience classroom, within the given school where the educational field experience occurred, on perceived preservice elementary teacher efficacy and perceived preservice elementary science teacher efficacy?” Research Question 2 Conclusions STEBIB Factor 1 Results Table 14 lists the variables and categories associated with research question 2. The significant predictors for the ANCOVA model for research question 2, factor 1 (PSTE) of the STEBIB test included the inservice teacher ethnicity (p = 0.0832), the inservice teacher gender (p = 0.0065), and the number of free lunch students in the classroom (p = 0.0197). The conclusions associated with these three statistically significant variables are addressed below in this order. Inservice Teacher Ethnicity. The mean of posttest STEBIB for factor 1 (PSTE) of the ethnicity of the inservice teacher for White and American Indian/or Alaskan is presented in Table 16. The mean posttest PSTE score of preservice teachers in classrooms with white inservice teachers was 0.649 higher than a preservice teacher who was in an observation classroom with an American Indian/or Alaskan inservice teacher. The posttest mean score of the preservice teachers in observation classrooms with white inservice teachers had a Likertscale score close to “Agree (4)” (Mean = 3.892), while the mean score of the preservice teachers in observation classrooms with American Indian/or Alaskan inservice teachers had a mean Likertscale score much closer to “Uncertain (3)” (Mean = 3.243). 71 The moderate increase (0.649) in the preservice teachers’ PSTE seen in White versus American Indian/or Alaskan inservice teachers may reflect the preservice teachers’ educational experiences associated with their teacher’s ethnicity during elementary, high school and college. The moderate increase (0.649) in the preservice teachers’ PSTE seen in White versus American Indian/or Alaskan inservice teachers may also reflect the high White ethnicity percentage (89.130%) of the preservice teachers during the study time. The moderate decrease we observed in the preservice teachers’ PSTE related to the inservice teacher’s ethnicity could be an issue of the White preservice teachers’ inability to relate to or find commonality with the American Indian/or Alaskan inservice teacher. Conversely, the moderate increase could be consistent with the White preserivce teacher’s ability to relate to or find commonality with the White inservice teacher or the White inservice teachers providing a more supportive teaching experience for the preservice teachers. Although the statistical techniques used are reliable, the 0.649 decrease in the PSTE score of preservice teachers who were in an observation classroom with an American Indian/or Alaskan inservice teacher could be an artifact of the small sample size of American Indian/or Alaskan inservice teachers (2 out of 20). This small sample size makes the probability of falsely concluding differences in the mean more likely. In conclusion, preservice teachers who were in classrooms with American Indian/or Alaskan inservice teachers had lower PSTE than preservice teachers who were in classrooms with white inservice teachers. 72 Inservice Teacher Gender. The mean of posttest STEBIB for factor 1 of the gender of the inservice teacher for male and female is presented in Table 16. The mean posttest PSTE scores of preservice teachers in observation classrooms with female inservice teachers were 0.108 higher than that of preservice teachers in classrooms with male inservice teachers. Preservice teachers in an observation classroom with female inservice teachers and preservice teachers in an observation classroom with male inservice teachers both had Likertscale scores almost directly between “Agree (4)” and “Uncertain (3).” The mean of preservice teachers’ posttest PSTE was 3.621 if they were in a classroom with a female inservice teacher and 3.513 if they were in a classroom with a male inservice teacher. The slight increase (0.108) we see in the preservice teachers’ PSTE in female versus male inservice teachers is consistent with the preservice teacher’s educational experiences during their own elementary, high school and college years. This is also consistent with the high female percentage (97.826%) of the preservice teachers during the study time. The small decrease we observe in the preservice teachers’ PSTE related to being placed with a male inservice teacher could be an issue of the female preservice teacher’s inability to relate to or find commonality with the male inservice teacher. Conversely, the slight increase could be consistent with the female preserivce teacher’s ability to relate to or find commonality with the female inservice teacher. Note that when comparing inservice teacher ethnicity versus inservice teacher gender, inservice teacher ethnicity had a larger negative impact (0.649) than inservice teacher gender (0.108). Although the statistical techniques used are reliable, the 0.108 decrease in the PSTE score of preservice teachers who were in an observation classroom with a male 73 inservice teacher could be an artifact of the small sample size of male inservice teachers (1 out of 20). This small sample size makes the probability of falsely concluding differences in the mean more likely and could explain the differences that are observed. In conclusion, preservice teachers who were in classrooms with male inservice teachers had lower PSTE than preservice teachers who were in classrooms with female teachers. Percentage of Students that Received Free and Reduced Lunch. The mean of posttest STEBIB for factor 1 of the percentage of students who received free and reduced lunch in the inservice teacher’s classroom be
Click tabs to swap between content that is broken into logical sections.
Rating  
Title  Assessing the Impact of Vicarious Experiences on Preservice Elementary Science Teacher Efficacy and Preservice Elementary Teacher Efficacy 
Date  20071201 
Author  Wagler, Ronald Robert 
Keywords  TES; STEBIB; preservice elementary science teacher efficacy; vicarious experiences; teacher efficacy; science; 
Department  Environmental Sciences 
Document Type  
Full Text Type  Open Access 
Abstract  The purpose of this study was to investigate the impact of vicarious experiences (preservice teacher field experiences) on perceived preservice science teacher efficacy and perceived preservice teacher efficacy. The participants for the study were 46 preservice elementary education students who were enrolled in CIED 3430 (Early Lab and Clinical Experience in Elementary Education II) at a large Midwestern state university and 20 classroom inservice teachers. A pretest was administered early in the spring 2007 semester, before the preservice teachers did their field experience and consisted of demographic questions and the STEBIB. A posttest was administered at the end of the spring 2007 semester, after the preservice teachers had completed their field experiences, and consisted of demographic questions, a rating of the teachers they observed during their educational field experience, the STEBIB and the TES. The field experience classroom inservice teachers provided personal, professional, and classroom data in the middle of the spring 2007 semester. All data were analyzed using analysis of variance (ANOVA) and analysis of covariance (ANCOVA). Factors of gender, ethnicity, socioeconomic status and preservice teacher program placement were found to be significant predictors of preservice teachers' efficacy scores. Even though, in some cases, these factors negatively impacted preservice teacher efficacy, preservice teachers should be placed in these environments when support is most available. The Teacher Efficacy Scale (Gibson & Dembo, 1984) is invalid. Even the construct of a general teacher efficacy is questionable. 
Note  Dissertation 
Rights  © Oklahoma Agricultural and Mechanical Board of Regents 
Transcript  ASSESSING THE IMPACT OF VICARIOUS EXPERIENCES ON PRESERVICE ELEMENTARY SCIENCE TEACHER EFFICACY AND PRESERVICE ELEMENTARY TEACHER EFFICACY By RONALD ROBERT WAGLER Bachelor of Science in Biology Southern Illinois University Carbondale, Illinois 1990 Master of Science in Zoology Oklahoma State University Stillwater, Oklahoma 2003 Submitted to the Faculty of the Graduate College of the Oklahoma State University in partial fulfillment of the requirements for the Degree of DOCTOR OF PHILOSOPHY December, 2007 ii ASSESSING THE IMPACT OF VICARIOUS EXPERIENCES ON PRESERVICE ELEMENTARY SCIENCE TEACHER EFFICACY AND PRESERVICE ELEMENTARY TEACHER EFFICACY Dissertation Approved: Richard Bryant Dissertation Adviser Patricia LamphereJordan Caroline Beller James Shaw A. Gordon Emslie Dean of the Graduate College iii ACKNOWLEDGEMENT First and foremost, I would like to thank my wife Amy and my daughter Olive Elizabeth for their support, encouragement and love throughout my PhD. You are the best part of life! I would like to thank my committee members, Dr. Beller and Dr. Shaw, for the time and effort they put into this research project. I would also like to thank Dr. Jordan for all off her help throughout my PhD program. Last, but certainly not least, I would especially like to thank Dr. Bryant for his patience, understanding, friendship and assistance in bringing the current study to completion. iv TABLE OF CONTENTS Chapter Page I. INTRODUCTION.....................................................................................................1 Background..............................................................................................................1 Statement of the Problem.........................................................................................3 Purpose of the Study ................................................................................................3 Research Questions..................................................................................................3 Research Question 1 ................................................................................................3 Research Question 2 ................................................................................................3 Theoretical Perspective............................................................................................4 Significance of the Study.........................................................................................5 Definition of Terms..................................................................................................6 Composition of the Dissertation ..............................................................................9 II. REVIEW OF LITERATURE.................................................................................11 Introduction............................................................................................................11 Teacher Efficacy ....................................................................................................11 Teacher Efficacy ResearchInservice Teachers .....................................................12 Teacher Efficacy ResearchPreservice Teachers ...................................................20 Vicarious Experience .............................................................................................26 Preservice Teacher Early Field Experience ...........................................................28 III. METHODOLOGY ...............................................................................................31 Introduction............................................................................................................31 Research Question 1 ..............................................................................................31 Study Participants ..................................................................................................31 Testing Instruments and Procedure........................................................................32 Data Analysis .........................................................................................................33 Research Question 2 ..............................................................................................34 Study Participants ..................................................................................................34 Testing Instruments and Procedure........................................................................34 v Data Analysis .........................................................................................................35 Assumptions and Limitations of the Study............................................................35 Assumptions...........................................................................................................35 Limitations .............................................................................................................35 IV. RESULTS.............................................................................................................37 Introduction............................................................................................................37 Research Questions................................................................................................37 Research Question 1 ..............................................................................................37 Research Question 2 ..............................................................................................37 Testing Instrument Data Analysis..........................................................................38 Principal Component Analysis ..............................................................................38 Factor Analysis ......................................................................................................38 Cronbach’s Alpha Reliabilities ..............................................................................41 Summary Testing Instrument Statistics .................................................................43 Study Participants Demographic Data Analysis ....................................................44 Preservice Teachers ...............................................................................................44 Inservice Teachers .................................................................................................47 Introduction to Research Question 1 and 2 Analysis.............................................49 Research Question 1 Analysis ...............................................................................50 STEBIB Results....................................................................................................51 TES Results............................................................................................................51 Research Question 2 Analysis ...............................................................................51 STEBIB Results....................................................................................................52 TES Results............................................................................................................57 V. CONCLUSION AND RECOMMENDATIONS..................................................61 Introduction............................................................................................................61 Research Questions................................................................................................61 Research Question 1 ..............................................................................................61 Research Question 2 ..............................................................................................61 Testing Instrument Conclusions ............................................................................62 Principal Component Analysis ..............................................................................62 Factor Analysis ......................................................................................................63 Cronbach’s Alpha Reliabilities ..............................................................................64 Summary Testing Instrument Statistics .................................................................65 Study Participants Demographic Conclusions.......................................................67 Preservice Teachers ...............................................................................................67 Inservice Teachers .................................................................................................68 Research Question 1 ..............................................................................................69 Research Question 1 Conclusions..........................................................................69 STEBIB Factor 1 and 2 Results............................................................................69 vi TES Factor 1 and 2 Results....................................................................................69 Research Question 2 ..............................................................................................70 Research Question 2 Conclusions..........................................................................70 STEBIB Factor 1 Results .....................................................................................70 Inservice Teacher Ethnicity ...................................................................................70 Inservice Teacher Gender ......................................................................................72 Percentage of Students that Received Free and Reduced Lunch...........................73 STEBIB Factor 2 Results .....................................................................................74 Percentage of Asian/Pacific Islander Students ......................................................74 Percentage of Hispanic/Latino Students ................................................................75 Percentage of American Indian/or Alaskan Students ............................................77 ExCEL preservice teacher......................................................................................78 TES Factor 1 Results .............................................................................................80 Percentage of Hispanic/Latino Students ................................................................80 TES Factor 2 Results .............................................................................................81 The Problematic Nature of the Teacher Efficacy Scale.........................................82 The Specific Nature of Teaching Efficacy.............................................................83 Recommendations for Future Research.................................................................85 Implications for Practice ........................................................................................87 Concluding Remarks..............................................................................................88 REFERENCES ............................................................................................................90 APPENDICES .............................................................................................................97 APPENDIX AINFORMED CONSENT AND DEMOGRAPHIC QUESTIONS ......................................98 APPENDIX BTES ............................................................................................102 APPENDIX CSTEBIB ....................................................................................105 APPENDIX DINFORMED CONSENT AND CLASSROOM DEMOGRAPHIC AND COOPERATING TEACHER QUESTIONNAIRE .................108 APPENDIX EDEMOGRAPHIC QUESTIONS ...............................................113 APPENDIX FCATEGORIZATION OF VARIABLES FOR STATISTICAL ANALYSIS .....................116 APPENDIX GIRB.............................................................................................125 vii LIST OF TABLES Table Page 1 STEBIB Original Factor Analysis.......................................................................40 2 STEBIB Combined Factor Analysis ...................................................................41 3 TES Factor Analysis .............................................................................................42 4 Crombach’s Alpha Reliabilities............................................................................42 5 STEBIB Summary Testing Instrument Statistics ................................................43 6 TES Summary Testing Instrument Statistics ........................................................44 7 Preservice Teacher Question.................................................................................45 8 Preservice Teacher Question.................................................................................46 9 Preservice Teacher Question.................................................................................47 10 Select Inservice Teacher Questions .....................................................................48 11 Variables and Categories Associated with Research Question 1.........................50 12 Age of Preservice Teacher ANOVA....................................................................51 13 Age of Preservice Teacher Means .......................................................................52 14 Variables and Categories Associated with Research Question 2.........................53 15 ANCOVA Model Results for Research Question 2, Factor 1 of the STEBIB...54 16 Means for Factor 1 of the STEBIB Posttest Associated with the ANCOVA ....55 17 ANCOVA Model Results for Research Question 2, Factor 2 of the STEBIB...55 18 Means for Factor 2 of the STEBIB Posttest Associated with the ANCOVA.....56 viii 19 Tukey Simultaneous Confidence Intervals for Factor 2 of the STEBIB............57 20 ANOVA for TES Factor 1 ...................................................................................58 21 TES Factor 1 Means for Hispanic/Latino Students .............................................58 22 Tukey Simultaneous Confidence Intervals for Factor 1 of the TES....................59 23 ANOVA for TES Factor 2 ...................................................................................59 24 Means for TES Factor 2.......................................................................................60 1 CHAPTER 1 INTRODUCTION Background Teacher efficacy is a powerful idea with a long, sordid and controversial history. The first teacher efficacy study was influenced by Rotter’s (1966) social learning theory and focused on teacher’s beliefs about where control lies in student learning. Later studies would move away from this influence and would be more directly influenced by Bandura’s social cognitive theory (Bandura, 1977). More recently there has been a movement to develop teacher efficacy instruments that focused on more specific contexts of teaching such as science teaching efficacy. In this atmosphere of competing theories and competing contexts, a sense of confusion has arisen regarding the most appropriate way to understand and measure teacher efficacy. Researchers were intrigued by the need to address certain central issues that are fundamental in understanding the practical application of teacher efficacy to effective teaching and student learning in specific classroom contexts. Many questions arose regarding the study of teacher efficacy. How do specific classroom dynamics impact teacher efficacy? Does the Teacher Efficacy Scale (TES) (Gibson & Dembo, 1984), which has formed much of the basis for our understanding of teacher efficacy, even measure teacher efficacy? Does the idea of a general teacher efficacy with a low level of specificity even exist? Ultimately, does teacher efficacy need 2 to be captured within a specific classroom context and environment to have meaningful practical significance and application for effective student learning? The current study attempts to bring some clarity to these issues by looking at the impact that field experiences (Vicarious experiences) have on preservice elementary science teacher efficacy and preservice elementary teacher efficacy. Teacher efficacy has been defined as “the extent to which the teacher believes he or she has the capacity to affect student performance” (Berman, McLauglin, Bass, Pauly & Zellman, 1977, p. 137) or a “teachers’ belief or conviction that they can influence how well students learn, even those that may be difficult or unmotivated” (Guskey & Passaro, 1994, p. 4). According to TschannenMoran et al. (1998) “the research suggests that teachers’ sense of efficacy plays a powerful role in schooling” (p. 234). There has been extensive research over the last three decades to formulate a unified theory of efficacy and to develop valid, reliable instruments that could measure efficacy levels in teachers, especially elementary teachers. Vicarious experiences are one of the four main sources that influence the efficacy of the individual teacher (Bandura, 1997). Vicarious experiences are also a common component of teacher education programs. However, little research has been done to evaluate the impact of vicarious learning experiences in the context of perceived preservice teacher efficacy and perceived preservice science teacher efficacy. The vicarious experiences in this study occurred in elementary public school classrooms where the preservice elementary teachers conducted their field observations. The type of vicarious experience was dependant on the specific variables that existed in the specific 3 classroom where the individual preservice teachers observed. For the specific variables associated with the research questions see Table 11 and 14 in Chapter 4. Statement of the Problem Teacher efficacy has been positively correlated with the amount of effort a teacher will expend in a teaching environment and the level of persistence a teacher will show in the face of obstacles (TschannenMoran et al., 1998). However, there have been no studies that have looked at the impact that vicarious experiences in teacher preparation programs have on the construct of preservice elementary teacher efficacy and preservice elementary science teacher efficacy. Purpose of the Study The purpose of this study was to investigate the impact vicarious experiences had on perceived preservice elementary teacher efficacy and perceived preservice elementary science teacher efficacy. Research Questions The research questions that guided this study were: Research Question 1 What is the impact of a vicarious learning experience (preservice teacher field experiences) on perceived preservice elementary teacher efficacy and preservice elementary science teacher efficacy? Research Question 2 What is the impact of the characteristics of the field experience classroom, within the given school where the educational field experience occurred, on perceived preservice 4 elementary teacher efficacy and perceived preservice elementary science teacher efficacy? Theoretical Perspective The first formal efficacy research began over two decades ago when the RAND organization, influenced by Rotter’s (1966) social learning theory, added two items to an already existing questionnaire (Armor et al., 1976). With the findings of the two RAND organization items the construct of teacher efficacy was first formulated. In these early RAND studies, teachers were asked to designate their level of agreement with two efficacy item statements (Armor et al., 1976). The total of the scores on the two RAND items was called teacher efficacy (TE), a concept that professed to indicate the degree to which a teacher believed that the consequences of learning and student motivation were controlled by the teacher (TschannenMoran et al., 1998). In the late 1970’s a second line of efficacy thought developed directly from Bandura’s social cognitive theory and his construct of selfefficacy (Bandura, 1977). Bandura (1997) defined selfefficacy as “beliefs in one’s capabilities to organize and execute the courses of action required to produce given attainments” (p. 3). “Selfefficacy is a futureoriented belief about the level of competence a person expects he or she will display in a specific situation” (TschannenMoran et al., 1998, p. 207). He also proposed that “selfefficacy beliefs influence thought patterns and emotions that enable actions in which people expend substantial effort in pursuit of goals, persist in the face of adversity, rebound from temporary setbacks, and exercise some control over events that affect their lives” (TschannenMoran et al., 1998, p. 210). 5 Bandura’s theory and his construct of selfefficacy would later influence the development of such efficacy instruments as the Teacher Efficacy Scale (Gibson and Dembo, 1984), the Ashton Vignettes (Ashton, Buhr, & Crocker, 1984), the Science Teaching Efficacy Belief Instrument (STEBI) (Riggs & Enochs, 1990), the Ohio State Teacher Efficacy Scale (TschannenMoran et al., 1998) and many others. With the development of the Teacher Efficacy Scale (TES) by Gibson and Dembo (1984) two factors of teacher efficacy were identified: The first factor, General Teacher Efficacy (GTE) related to the teacher’s belief about the impact of external factors, such as hostility in the home or economic factors of gender, race or class, contrasted to the influence of the schools and teachers. The second factor, Personal Teacher Efficacy (PTE), related to the teachers’ belief in their ability to overcome the factors that make learning difficult for students. Because teacher efficacy is believed to be both subjectmatter and context specific (Tschannen et al., 1998), Riggs and Enochs (1990) developed the Science Teaching Efficacy Belief Instrument (STEBI) to measure efficacy of science teaching. The authors identified two uncorrelated factors within STEBI, which they named personal science teaching efficacy (PSTE) and science teaching outcome expectancy (STOE). Thus, the instrument measured both PSTE and STOE. The PSTE scale indicated teachers’ belief in their ability to perform a given behavior; the outcome STOE scale indicates the teachers’ belief that effective teaching can change behaviors (Riggs & Enochs, 1990). Significance of the Study The significance of this study concerns the impact of vicarious experiences on the construct of teacher efficacy. Teacher efficacy is an indicator of teacher performance and 6 success. This study will help to determine what role, if any; vicarious experiences play in impacting teacher efficacy. The results of this research will be significant for the elementary education faculty at the studied university in evaluating the effectiveness of vicarious experiences as a tool in preparing preservice elementary teachers to enter the teacher work force. The research findings could also benefit those attempting to incorporate vicarious events (e.g., elementary field experiences) into their preservice elementary teacher education curriculum. Furthermore, the findings of this study may be beneficial to those considering the degree and role that vicarious experiences should play in their preservice secondary teacher education programs. Evidence from this study could also be useful in furthering continued research related to elementary preservice teachers since a lack of research associated with efficacy and vicarious experiences in elementary preservice teachers exists. This study will also help to bring some understanding to the impact that specific classroom variables have on preservice elementary teacher efficacy and preservice elementary science teacher efficacy. Furthermore, the study will help to assess the reliability of the Teacher Efficacy Scale (TES) (Gibson & Dembo, 1984). Lastly, this study will attempt to bring some understanding to the specific level of teacher efficacy that is needed to have practical meaningful significance in the elementary classroom. Definition of Terms ExCEL The ExCEL (Excellence in Collaborative and Experiential Learning) Program is a collaborative partnership between the College of Education and Evergreen Public Schools. Open to elementary education majors in their final semester before student teaching, the program features a threeday a week full day internship in a public 7 school classroom in which teacher candidates are partnered with expert classroom teachers. Candidates return to the OSU campus one day a week for Semester Y methods classes. The ExCEL program is run by a team of OSU COE faculty members who work together to mentor students in their placements at the elementary school and also teach the ExCEL sections of methods classes, closely connecting course content with field experiences. The ExCEL program runs both fall and spring semesters, partnering with Driftwood Elementary in the fall and Norwood Elementary in the spring. Both schools are close to the OSU campus and each offers a richly diverse student population and talented, enthusiastic faculty and leadership. ExCEL program participants gain experience in almost every aspect of elementary school teaching and develop productive relationships with a variety of educational professionals. Many ExCEL students choose to stay in their placement classroom for student teaching so that they receive a full year of supported internship before starting their first job. Elementary Education majors interested in participating in ExCEL may apply for the program during the second semester of their junior year. A program representative will speak in classes about the program each semester, and invitations to apply will be posted around the college. General Teacher Efficacy (GTE): The teacher’s belief about the power of external factors, such as violence in the home or economic realities of gender, race or class, compared to the influence of teachers and schools (Ashton, Olejnik, Crocker, & McAuliffe, 1982); also known as teacher outcome expectancy. 8 Indefinable Factor 2 (IF2): Second TES factor from the study. Items associated with this factor were 2, 3, 5, 8, 12, 13, 15, 20 and 27. Based on the nonassociated nature of these items a common construct factor, such as TE, was deemed indefinable. Personal Science Teaching Efficacy (PSTE): teachers’ belief in their ability to perform a given behavior (Riggs & Enochs, 1990). Personal Teaching Efficacy (PTE): The teacher’s belief in their ability to overcome factors that could make learning difficult for students (TschannenMoran et al., 1998); in the context of the TES, PTE is referred to as selfefficacy (Gibson and Dembo, 1984). Preservice elementary teacher: University student majoring in elementary education who has not completed his/her educational certification. The student has not begun teaching formally and has not received his/her teaching certificate. Science Teaching Efficacy Belief Instrument (STEBI): Science teacher efficacy instrument designed to measure PTSE and STOE. The instrument consists of 25 items (Riggs & Enochs, 1990). Science Teaching Efficacy Belief Instrument B (STEBIB): Science teacher efficacy instrument designed to measure PTSE and STOE. The instrument consists of 23 items from the STEBI. The STEBIB is identical to the STEBI except items 20 and 25 have been removed and the verb tenses of some of the 23 items have been changed to accommodate preservice teachers. These changes were made so the instrument, which was originally designed for inservice teachers, could be used with preservice teachers (Enochs & Riggs, 1990). 9 Science Teaching Outcome Expectancy (STOE): the teachers’ belief that effective teaching can result in student learning (Riggs & Enochs, 1990). Teacher efficacy: “the extent to which the teacher believes he or she has the capacity to affect student performance” (Berman, McLauglin, Bass, Pauly & Zellman, 1977, p. 137) or “teachers’ belief or conviction that they can influence how well students learn, even those that may be difficult or unmotivated” (Guskey & Passaro, 1994, p. 4). Teaching Efficacy (TE): Teach efficacy (TE) in the context of the TES refers to outcome expectancy (Gibson and Dembo, 1984). Teacher Efficacy Scale (TES): Teacher efficacy instrument designed to measure personal teacher efficacy (PTE) and teacher efficacy (TE). The instrument consists of 30 items with a Likert scale ranging from 1 (Strongly disagree) to 6 (Strongly Agree). Gibson and Dembo (1984) found, after performing factor analysis, that sixteen of the original 30 items had acceptable reliability coefficients. Vicarious experiences: Within the context of teacher efficacy, a vicarious experience refers to observing another individual teaching. Within the context of this study it refers to the preservice teacher’s educational field experiences. Composition of the Dissertation The dissertation is composed of five chapters. Chapter 1 is the introduction which consists of a background introduction, statement of the problem, purpose of the study, research questions, theoretical perspective, significance of the study, definition of terms and composition of the dissertation. Chapter 2 describes the current literature that is essential to the study including an introduction and a description of the following: teacher 10 efficacy, teacher efficacy researchinservice teachers, teacher efficacy researchpreservice teachers, vicarious experience and preservice teacher early field experiences. Chapter 3 describes the methodology of the study. Chapter 4 describes the results of the study and Chapter 5 describes the summary conclusions and recommendations of the study. 11 CHAPTER II REVIEW OF LITERATURE Introduction This chapter describes the current literature that is essential to the study. The concept of teacher efficacy is described followed by current research on inservice teacher efficacy. Next, current preservice teacher efficacy research is discussed followed by Bandura’s concept of vicarious experience and, finally, a description of preservice teacher early field experiences. Teacher Efficacy According to TschannenMoran, WoolfolkHoy, and Hoy, W. (1998), teacher efficacy was first defined by the RAND organization “as the extent to which teachers believed they could control the reinforcement of their actions, that is, whether control of reinforcement lay within themselves or the environment” (p.202). Bandura’s (1977) social cognitive theory and his construct of selfefficacy, defined as “a cognitive process in which people construct beliefs about their capacity to perform at a given level of attainment” (TschannenMoran et al., 1998, p.203), provided a theoretical foundation for the construct of teacher efficacy as a specific type of selfefficacy. TschannenMoran et al. (1998) defined teacher efficacy as the “teacher’s belief in his or her capacity to organize and execute courses of action required to successfully accomplish a specific teaching task in a particular context” (TschannenMoran et al., 1998, p.233). 12 Gibson and Dembo (1984), equipped with the theories of the RAND researchers and the conceptual ideas of Bandura, developed the first reliable teacher efficacy instrument, the Teacher Efficacy Scale (TES). Since the development of the Teacher Efficacy Scale in the early 1980’s, researchers have developed a plethora (Gibson & Dembo, 1984; Ashton, Buhr, & Crocker, 1984; Riggs & Enochs, 1990; Tschannen Moran et al., 1998) of teacher efficacy instruments with the hope of understanding this powerful construct (TschannenMoran et al., 1998). Teacher Efficacy Research—Inservice Teachers Over the last 25 years there have been numerous studies, using many different efficacy instruments that have shown that a teacher’s sense of efficacy is a strong indicator of the teacher’s ability to be a productive, successful teacher. In this section some of the more historically important research findings concerning teacher efficacy will be addressed. The first formal efficacy research began over two decades ago when the RAND organization added two items to an already existing questionnaire (Armor et al., 1976). With the findings of the two RAND organization items the construct of teacher efficacy was first formulated. In these early RAND studies teachers were asked to designate their level of agreement with two efficacy statements (Armor et al., 1976). The total of the scores on the two RAND items was called teacher efficacy (TE), a concept that professed to indicate the degree to which a teacher believed that the consequences of learning and student motivation were controlled by the teacher (TschannenMoran et al., 1998). The first RAND item, “When it comes right down to it, a teacher really can’t do much because most of a student’s motivation and performance depends on his or her home 13 environment” (Armor et al., 1976), would be labeled general teaching efficacy (GTE) by future efficacy researchers. The second item, “If I really try hard, I can get through to even the most difficult or unmotivated students” (Armor et al., 1976), would be labeled as personal teaching efficacy (PTE) by future researchers (TschannenMoran et al., 1998). Armor et al (1976), using the two RAND items in the context of reading programs employed in Los Angeles schools, found teacher efficacy (TE) was strongly correlated to reading achievement variation among minority students. In the late 1970’s a second line of efficacy thought developed directly from Bandura’s social cognitive theory and his construct of selfefficacy (Bandura, 1977). Bandura (1997) defined selfefficacy as “beliefs in one’s capabilities to organize and execute the courses of action required to produce given attainments” (p. 3). “Selfefficacy is a futureoriented belief about the level of competence a person expects he or she will display in a specific situation” (TschannenMoran et al., 1998, p. 207). He also proposed that “selfefficacy beliefs influence thought patterns and emotions that enable actions in which people expend substantial effort in pursuit of goals, persist in the face of adversity, rebound from temporary setbacks, and exercise some control over events that affect their lives” (TschannenMoran et al., 1998, p. 210). Bandura’s theory and his construct of selfefficacy would later influence the development of such efficacy instruments as the Teacher Efficacy Scale (Gibson and Dembo, 1984), the Ashton Vignettes (Ashton, Buhr, & Crocker, 1984), the Science Teaching Efficacy Belief Instrument (STEBI) (Riggs & Enochs, 1990), the Ohio State Teacher Efficacy Scale (TschannenMoran et al., 1998) and many others. Bandura, after years of providing information for his everdeveloping theory, offered his own Teacher SelfEfficacy Scale 14 (Bandura, 1997). The instrument he developed is a 30item instrument with seven subscales: efficacy to enlist community involvement, efficacy to influence decision making, instructional efficacy, disciplinary efficacy, efficacy to enlist parental involvement, efficacy to influence school resources, and efficacy to create a positive school climate (Bandura, 1997). Bandura’s Teacher SelfEfficacy Scale uses a ninepoint Likerttype scale with response choices ranging from nothing (1) to a great deal (9) for each item. With the development of the Teacher Efficacy Scale (TES) by Gibson and Dembo (1984) two factors of teacher efficacy were identified: General Teacher Efficacy (GTE) or the teacher’s belief about the impact of external factors, such as hostility in the home or economic factors of gender, race or class, contrasted to the influence of the schools and teachers and Personal Teacher Efficacy (PTE), the teachers’ belief in their ability to overcome the factors that make learning difficult for students. Gibson and Dembo, using the TES, found that teachers who display a low sense of efficacy were more likely to instruct the class as a whole than to divide the class into small groups for instruction. They also found that low efficacy teachers were more likely to criticize students for an incorrect answer and were less likely to persist with a student in a difficult situation. Podell and Soodak (1993), again using the TES, found that there was a direct correlation between teacher efficacy and a teacher’s decision to refer a student to special education. They found teachers with higher levels of PTE were more willing to work with students who were experiencing problems rather than refer them to a special education program. Coladarci (1992) conducted a study to assess commitment to teaching. The subjects were composed of a random sample of 364 elementarylevel Maine teachers. 15 The TES (Gibson & Dembo, 1984) was administered to all subjects along with the teacher commitment question “Suppose you had it to do all over again: In view of your present knowledge, would you become a teacher?” (Coladarci, 1992, p. 328). After assessing the data it was found that the teachers who expressed a greater level of professional teaching commitment also tended to have higher levels of both PTE and GTE. Allinder (1994) administered the TES and the Teacher Characteristics Scale (Fuchs, Fuchs, & Bishop, 1992) to 200 randomly selected elementary special education teachers from four Midwest states. Teachers who displayed high levels PTE were more willing to try a variety of teaching approaches and materials, use new and innovative teaching methods and exhibited a desire to find better ways of teaching. Teachers who display elevated levels of PTE were also more apt to have higher scores in the areas of organization, enthusiasm, and fairness. Finally, enthusiasm and clarity in teaching were found to be related to elevated GTE. Research has also shown that just as a teacher’s efficacy level can influence a his or her behavior; a teacher’s level of efficacy can also influence students’ attitude toward the subject matter being taught and students’ attitudes toward school. Woolfolk, Rosoff, and Hoy (1990) concluded that students of teachers who exhibited high PTE tended toward greater interest in school, higher evaluations of their high PTE teachers, and showed a greater awareness that what they were being taught was important. Warren and Payne (1997) investigated middle school organizational patterns and their impact on teachers’ efficacy among 81 eighthgrade teachers. They concluded that teachers who were placed on interdisciplinary teams with the same planning times had 16 significantly higher PTE levels than teachers on interdisciplinary teams that did not have the same planning times. Teachers placed on interdisciplinary teams with the same planning times had higher PTE than teachers who were organized within their own department (Warren & Payne, 1997). In a similar study Reames and Spencer (1998) examined Georgia middle school teachers’ perceptions of their work environment, their perceived efficacy, and their organizational commitment. The study included 275 fulltime, certified teachers from 40 rural and metropolitan middle schools. Teachers completed a mailed survey that asked about demographics, organizational commitment, perceived efficacy, and the schoolwork culture (Reames & Spencer, 1998). Organizational structure and process variables were positively related to PTE. Process variables included collaboration, participatory decisionmaking, and supportive administrative leadership. Organizational structures included encouragement of innovation and risk taking, school goals and planning, and staff development to further goals (Reames & Spencer, 1998). Teacher efficacy has also been linked to family involvement practices. Garcia (2004) conducted a study that utilized the Teacher Efficacy Scale (Gibson & Dembo, 1984) and the Family Involvement Teacher Efficacy Scale (Garcia, 2000). Prior educational research has shown that positive benefits, for the child, are incurred as parents become involved in that child’s education. By utilizing these two instruments and a sample size of 110 urban elementary school teachers, Garcia concluded that elevated teacher efficacy was significantly correlated to and was also a predictor of five types of positive family involvement practices (Garcia, 2004). 17 The Ashton Vignettes were developed on the assumption that teacher efficacy can be context specific. Ashton, Buhr, and Crocker (1984) created a series of short literary sketches describing events a teacher might experience in the educational environment and asked teachers to make evaluations as to the possible causes involved in each written scenario known as the Ashton Vignettes (TschannenMoran et al., 1998). Two versions were developed and the second version, which asked teachers to compare themselves to other teachers, was significantly correlated with the two RAND items. Benz et al (1992) conducted a study in which they used the Ashton Vignettes (Ashton, Buhr, & Crocker, 1984) to assess perceptions of efficacy across a variety of educational situations with a variety of educators. They found college faculty, when compared to classroom teachers, student teacher supervisors and preservice teachers, had elavated selfefficacy for dealing successfully with a disobedient student, for selfmotivation, and for preparation. They also found both college faculty and preservice teachers were more confident about their efficacy in environments involving student socialization than were classroom teachers (Benz et al., 1992). In a related study Webb and Ashton (1987), the developers of the Ashton Vignettes, asked 42 middle and high school teachers to assess situational and environmental factors that they believed affected a teacher’s sense of efficacy. Six factors were identified: 1) inadequate salaries and low status, 2) excessive role demands, 3) lack of recognition and professional isolation, 4) uncertainty, 5) alienation and 6) low teacher morale. Because teacher efficacy is believed to be both subjectmatter and context specific (Tschannen et al., 1998), Riggs and Enochs (1990) developed the Science Teaching Efficacy Belief Instrument (STEBI) to measure efficacy of science teaching. The authors 18 identified two uncorrelated factors within STEBI, which they named personal science teaching efficacy (PSTE) and science teaching outcome expectancy (STOE). Thus, the instrument measured both PSTE and STOE. The PSTE scale indicated teachers’ belief in their ability to perform a given behavior; the outcome STOE scale indicates the teachers’ belief that effective teaching can change behaviors (Riggs & Enochs, 1990). Using the Science Teaching Efficacy Belief Instrument (STEBI), Riggs & Jesunathadas (1993) found elementary teachers with higher personal science teaching efficacy (PSTE) scores reported spending more time teaching science and were more apt to spend the needed time to develop the science concept being taught. Watters and Ginns (1995) found that teachers with a higher sense of PSTE tended to have a higher level of enjoyment associated with science activities. Elementary teachers who were involved in a oneyear science education training program who displayed low PSTE were less inclined to want to teach science and made smaller numbers of changes in their beliefs about how students could learn science. They also used less class time teaching science, were rated weaker by those who were observing them, and used a textbased teaching approach more often (Riggs, 1995). Czerniak (1999) used the STEBI to assess and compare the science teacher efficacy levels of science teachers who taught in either a middle school organizational structure or a junior high school organizational structure. After analysis Czerniak (1999) found that science teachers who were part of a middle school model versus a junior high model had significantly higher levels of science teaching outcome expectancy (STOE). Czerniak speculated that that these findings add merit to the use of a middle school model over a junior high model since middle school organizational structures “provide teachers 19 with the necessary support needed to remain committed to student learning” (Czerniak, 1999, p.36). Chun and Oliver (2000) conducted research into the quantitative examination of teacher selfefficacy and knowledge of the nature of science. They conducted a longitudinal study on 31 middle school science teachers in the southeastern part of the United States. All 31 participating teachers had been science teachers for over 5 years. Four sets of instruments were administered to the participants over a threeyear period including the STEBI. During the three years of the study the middle school science teachers participated in summer workshops. “The workshops were designed to enhance middle school teachers’ understanding about the nature and structure of science as well as pedagogical knowledge to teach science” (Chun & Oliver, 2000, p. 3). The STEBI was administered to the subjects five times during the study. A pretest and posttest were given during the first year workshop, at the second year workshop, and a posttest during the last year workshop (Chun & Oliver, 2000). Chun and Oliver (2000) concluded that scores of PSTE and STOE both increased and paralleled each other over the five test times. These findings add merit to the use of the type of workshops mentioned above to increase middle school science teachers’ selfefficacy and thereby make them more productive teachers. Rubeck and Enochs (1991) attempted to investigate an even more specific level of efficacy by distinguishing efficacy associated with chemistry teaching from efficacy associated with science teaching. Enochs, Smith, and Huinker (2000) further developed a similar instrument to measure efficacy of mathematics teaching while Coladarci and 20 Breton (1997) used a modified instrument to measure efficacy in the framework of special education (TschannenMoran et al., 1998). Teacher Efficacy Research—Preservice Teachers In this section some of the more important findings concerning preservice teacher efficacy are addressed. Evans and Tribble (1986), using the TES, compared the perceived teaching problems of 179 preservice elementary and secondary teachers with their level of efficacy. They found that preservice teachers who had elevated efficacy scores were more likely to have elevated levels of professional commitment. Czerniak (1989), using a revised Science Teaching Efficacy Belief Instrument (dubbed the STEBIB) (Enochs & Riggs, 1990), found that the level of personal science teaching efficacy (PSTE) could be positively correlated to confidence in teaching elementary science effectively and negatively correlated to science teaching anxiety. Czerniak (1989) found preservice elementary teachers with high selfefficacy “had less anxiety toward teaching science, were more likely to use openended inquiry and studentdirected teaching strategies, and were more confident about teaching elementary science effectively” (Czerniak & Schriver, 1994, p. 77). Czerniak and Schriver (1994), in a related study, examined elementary preservice science teachers’ beliefs and behavior related to selfefficacy. The 2year longitudinal study used the Science Teacher SelfEfficacy Instrument modified by Czerniak (1989) from the Gibson and Dembo (1984) Teacher Efficacy Scale. Using this modified instrument, Czerniak and Schriver found that “preservice teachers who were in the top 20% of the class on science teaching selfefficacy seemed to display greater conviction 21 that they could successfully help children learn science” (Czerniak & Schriver, 1994, p. 85). These preservice teachers, in order to become better educators, “analyzed their own strengths and weaknesses and sought to help all children learn” (Czerniak & Schriver, 1994, p. 85). Highlevel efficacy preservice teachers “selected strategies that they thought would help children learn science, and they used the educational theories they had learned in their methods class” (Czerniak & Schriver, 1994, p. 85). In contrast to the preservice teachers who were in the top 20% of the class on science teaching selfefficacy, the preservice teachers who were in the bottom 20% of the class on science teaching selfefficacy were unsure of their abilities to be successful teachers in the science classroom. They were overly concerned about noise in the teaching environment and regularly worried about student misconduct. They blamed others for their failures and avoided examining their own skills (Czerniak & Schriver, 1994). There is also research to suggest that student teaching can have an impact on overall teacher efficacy. Teaching experience gained during the student teaching time when evaluated by the TES has been shown to increase personal teaching efficacy (Hoy & Woolfolk, 1990) while general teaching efficacy has been shown to decrease during student teaching (Hoy & Woolfolk, 1990; Spector, 1990). This may be due to overoptimism that is challenged when the student teacher faces the difficulties of the teaching assignment (TschannenMoran et al., 1998). Student teachers with elevated PTE were also rated higher on classroom management, questioning behavior and lessonpresenting behavior by the teachers supervising them (Saklofske, Michaluk, & Randhawa, 1988). Emmer and Hickman (1990), using an adapted TES, found that preservice elementary and secondary teachers that show high teacher efficacy levels in all three 22 subscales (personal teaching efficacy, efficacy for classroom management and discipline and external influences) tend to use classroom management strategies that are aimed at increasing desirable responses through encouragement, praise, rewards, and attention. They also found that preservice teachers, with an elevated sense of personal teacher efficacy, when faced with student discipline problems, were more apt to ask for help. Efficacy has also been related to student control issues (Woolfolk & Hoy, 1990). Preservice teachers with low GTE and PTE or high GTE and low PTE, as measured by the TES tended to have a negative view of students’ motivation, relied on punishment to get students to study, and had a tendency to enforce stringent classroom rules. These findings are in contrast to preservice teachers who were high in both GTE and PTE. Student teachers with high GTE and PTE efficacy scores tended to be more humanistic in their manner of classroom control. Both PTE and GTE of preservice teachers are malleable and may be affected differently by experiences. Social persuasion and vicarious experiences, such as those encountered in college coursework, seem to have a greater effect on preservice teachers’ GTE (Watters & Ginns, 1995). In contrast, actual teaching experiences, such as those encountered in student teaching, seem to exert a greater influence on PTE (Housego, 1992; Hoy & Woolfolk, 1990), although GTE may also change (negatively) during student teaching (Hoy & Woolfolk, 1990; Spector, 1990). Schoon’s and Boone’s (1998) work with preservice elementary teachers using the STEBIB (Enochs & Riggs, 1990) has shown there is an association between elementary teachers’ low science efficacy beliefs and alternative science concepts. The study found that holding certain alternative concepts about science such as planets can only be seen with a telescope, dinosaurs lived the same time as cavemen, and that north is toward the 23 top of a map of Antarctica were linked to subjects with low science teacher efficacy. The study also found that preservice teachers that held fewer numbers of alternative concepts had significantly higher efficacy levels (Schoon & Boone, 1998). Current reform in teacher education has focused on the need for improvement of preservice training (National Research Council, 1996). With this in mind Wingfield and Ramsey (1999) conducted a study that examined the effect of a onesemester sitebased program where preservice teachers participated in authentic classroom and school experiences during their methods class. “The site experiences included: teaching experiences within the assigned classroom, teaching experiences during the methods classes, feedback from the university cluster coordinator, peer and sitebased teacher, observation of the sitebased teacher, and methods class assignments, text, instruction and instructor” (Wingfield & Ramsey, 1999, p. 2). The participants for the study consisted of 131 undergraduate elementary preservice teachers who completed the STEBIB (Enochs & Riggs, 1990) at the beginning and end of the fifteenweek sitebased program. A substantial increase in efficacy from pretest to posttest was noted. Wingfield and Ramsey (1999) concluded that the results indicated that the experiences of the onesemester program had a significant impact on the preservice teacher’s science teaching efficacy beliefs. They also speculated that the additional vicarious teaching experiences may have positively impacted the subjects’ science teaching efficacy. These vicarious teaching experiences specifically included observations of the methods instructor, other preservice teachers and classroom teachers (Wingfield & Ramsey, 1999). King and Wiseman (2001) conducted a study with the purpose of examining differences in science teaching efficacy beliefs among students enrolled in two versions 24 of a methods course in an elementary science teaching program. One group of preservice elementary teachers was enrolled in a semester long interdisciplinary methods class and another group of preservice elementary teachers was enrolled in a semester long more “traditional” noninterdisciplinary methods class. Both groups were given the STEBIB (Enochs & Riggs, 1990) after the methods class. When the results of the STEBI were compared between both groups, neither PSTE nor STOE were found to be significantly different. They concluded their study by stating that if the role of integrated instruction in the elementary curriculum is considered, “the findings of their study suggest that teaching in an integrated fashion and planning interdisciplinary units would seem to be no more effective than traditional teaching in terms of developing the science teaching efficacy of the students” (King & Wiseman, 2001, p. 149). Moseley, Reinke, and Bookout (2002) used Sia’s (1992) Environmental Education Efficacy Belief Instrument to evaluate the effect a 3day outdoor environmental education program would have on 72 participating preservice teachers. The Environmental Education Efficacy Belief Instrument (Sia, 1992), which is based on the STEBI (Enochs & Riggs, 1990), assesses outcome expectancy and selfefficacy in an environmental education teaching beliefs context. All items are based upon a 5point Likertscale response. Moseley, Reinke and Bookout (2002) found that the preservice teacher’s selfefficacy was high before the 3day program and remained unchanged immediately after the program. The preservice teacher’s efficacy was then checked approximately 7 weeks after the conclusion of the 3day program and it had dropped significantly. No change in the outcome expectancy of the subjects was observed over the complete length of the study. The authors accredited the lack of efficacy change 25 during the workshop to the positive characteristics of the 3day program. The drop in efficacy, approximately 7 weeks after the program, “was believed to have resulted from the preservice teachers reevaluation of their ability to teach as they learned more about teaching methodologies” (Moseley, Reinke & Bookout, 2002, p. 9). There are also data to suggest that the number of high school science subjects studied can have a long term effect on the science efficacy of preservice teachers. Mulholland, Dorman and Odgers (2004) used the STEBIB to assess the science efficacy of 314 elementary preservice teachers. They found that the preservice teachers’ PSTE scores were positively related to the number of science classes studied at the high school level but not to their STOE scores. Completing two science teaching classes with the preservice teacher training program also had a significant positive effect on the PSTE but not on the STOE of the subjects. Utley, Moseley and Bryant (2005) explored the impact an elementary methods course and student teaching had on both science and mathematics preservice teacher efficacy. Their study, which used both the STEBIB and the Mathematics Teacher Efficacy Beliefs Instrument (MTEBI) (Huinker & Enochs, 1995), found both a positive and negative relationship between science and mathematics teaching efficacy in their sample population of elementary preservice teachers. Specifically, as the preservice teachers progressed in their methods courses their mathematics and science teacher efficacy also increased significantly. Both science and mathematics efficacy showed a slight decrease after student teaching. Wagler and Moseley (2006) conducted a study to investigate the effects of a secondary contentspecific methods course and student teaching on preservice teacher 26 efficacy. The instrument used in the study was the “The Ohio State Teacher Efficacy Scale” (OSTES).The study employed a single group, pretestposttest Iposttest II design. The repeated measures ANOVA indicated no significant change in overall teacher efficacy from the beginning of the secondary methods course until the end of student teaching; however, overall efficacy did increase significantly after the secondary methods course but by the end of student teaching had returned to its original presecondary methods course level. Classroom management efficacy over all three test times – before and after methods course and after student teaching  was unchanged. Instructional strategies efficacy was shown to be statistically significant and positively affected by the secondary methods course, but no significant change in instructional strategies efficacy was detected after student teaching. No significant change in student engagement efficacy was found immediately following the methods course but student engagement efficacy significantly decreased after student teaching. Vicarious Experience Bandura’s (1997) construct of selfefficacy is influenced by four sources of information, which are (1) enactive mastery experience, (2) vicarious experience, (3) verbal persuasion, and (4) physiological and affective states. Mastery experience is considered the actual act of teaching by the individual. Physiological and affective states or physiological arousal are physiological effects an individual experiences during the teaching act. Vicarious experiences, within the context of teacher efficacy, refers to observing another individual teach. Verbal persuasion is the result of information about teaching conveyed to the preservice teacher or inservice teacher by someone perceived to be an authority. TschannenMoran et al. (1998) and other educational researchers had 27 utilized Bandura’s four sources of efficacy in their teacher efficacy models and instruments. For the purpose of this study we will focus on the source of vicarious experience and how it influences the construct of selfefficacy. Within the context of vicarious experience, modeling is an effective mode for enhancing selfefficacy. An example of this in the context of teacher efficacy would be a preservice teacher who observes, as a participating observer or as a passive observer, a teaching event. In this scenario, the teacher as the model in the context of the vicarious event would have the potential to influence the teaching efficacy of the observer (i.e., preservice teacher). Bandura (1997) points out that for many activities, such as swimming, proficiency and improvement can be measured. The criteria that denote when an individual is swimming are fairly welldefined. We can also quantify improvement by using a variable such as time. For many activities “there are no absolute measures of adequacy. Therefore, people must appraise their capabilities in relation to the attainments of others” (Bandura, 1997, p. 86). One of the ways this is done is by observing models performing tasks. Individuals seek out skilled models because these “competent models transmit knowledge and teach observers effective skills and strategies for managing environmental demands (Bandura, 1986). Acquisition of effective means raises beliefs of personal efficacy” (Bandura, 1997, p. 88). When a person observes another similar individual successfully model a given event, efficacy beliefs are typically raised. Conversely, when a person observes another similar individual fail at modeling a given event, individual efficacy beliefs typically decline (Bandura, 1997). This is especially true if the individual observed is deemed 28 competent by the observer. Competence at a given task, activity or event has been shown to be more effect at increasing efficacy than the age of the model, sex of the model or other personal characteristics (Bandura, 1997). “Model competence is an especially influential factor when observers have a lot to learn and models have much they can teach them by instructive demonstration of skills and strategies” (Bandura, 1997, p.101). Bandura (1977) also proposes that models that convey productive coping techniques can even raise the efficacy of subjects who have experienced many confirmatory personal inefficacious events. On the contrary, subjects who possess high levels of efficacy when performing a given task can have their efficacy raised even higher “if the models teach them even better ways of doing things” (Bandura, 1997, p. 87). “Models who express confidence in the face of difficulties instill a higher sense of efficacy and perseverance in others than do models who begin to doubt themselves as they encounter problems (Zimmerman & Ringle, 1981)” (Bandura, 1997, p. 88). Preservice Teacher Early Field Experiences The research associated with field experiences among preservice teachers is limited. Much of the research conducted in this area occurred in the 1980’s and 1990’s with a few studies occurring in the last six years. Much of the major research has been conducted within the context of physical education. For most preservice teachers, early field experiences involve assisting, in some capacity, in an offcampus school environment (LaMaster, 2001). In the majority of cases the preservice teacher is working in the public school. The situational nature of early field experiences can range from observing teaching to active involvement in the teaching process. Early field experiences occur prior to the preservice teacher’s student teaching assignment (Dodds, 1989) and 29 have been historically viewed as an important component in preservice teacher training programs (Paese, 1989). Because early field experiences are now seen as an essential component of preservice teacher training they have, over the past two decades, moved from a single early field experience to multiple early field experiences before student teaching. Dueck, Altmann, Haslett, and Latimer, (1984) believe these experiences “provide information to students so they can determine their suitability for the teaching profession, orient preservice teachers to schools, and begin the socialization process for potential teachers” (LaMaster, 2001, p. 28). Early field experiences have historically been looked upon as an essential part of a teacher’s socialization (Lasley, Applegate, & Ellison, 1986). Dodds (1989), in a related study on preservice teacher school socialization, stated “field experiences represent the closest juncture between formal teacher training in universities and onthejob training in schools” (p.81). Paese (1984) assessed Early field experiences in terms of their positive benefits in developing the skills of effective teaching and also found that graduates of teacher education programs found early field experiences to be a helpful factor in their teacher training. By providing “real world” experiences, early field experiences also have the possibility of influencing future career decisions (Paese, 1987). Paese (1989) lists seven teaching benefits that are achieved by incorporating EFE’s into preservice teacher training. Among them is the ability of EFE’s to help preservice teachers connect teaching theory to teaching practice, develop a more complete perception of students, gain a better understanding of their future inservice teaching responsibilities and have more of an opportunity to increase and improve their teaching skills. 30 A pilot study was conducted by the author during the spring semester of 2006. There were 50 participants (49 female, 1 male) who were preservice elementary education students enrolled in a course titled Early Lab and Clinical Experience in Elementary Education II at the university. The preservice teachers rated the teacher they observed during their educational field experience (see Appendix E minus questions 8 through 13) and completed the TES (see Appendix B). The results showed only a significant positive correlation between one undefined TES factor and item 1: Rate the quality of the lessons that your field experience teacher used. The undefined TES factor is associated with the teacher’s internal skills and techniques applied to the teaching process. These skills and techniques are learned through teacher training and teacher experiences. From the above literature review, it can be deduced that teacher efficacy has been positively correlated with many desirable teacher behaviors, but little research has been conducted to evaluate the impact of vicarious learning experiences in the context of perceived preservice teacher efficacy and perceived preservice science teacher efficacy. With this in mind, the purpose of this study was to investigate the effect of vicarious experiences on perceived preservice teacher efficacy and perceived preservice science teacher efficacy. The results should be most significant to the elementary education faculty at the studied university in evaluating the effectiveness of vicarious experiences as a tool in preparing preservice elementary teachers to enter the teacher work force. In a broader sense, the results of this study could benefit those attempting to incorporate vicarious events (e.g., elementary field experiences) into their teacher education curriculum. 31 CHAPTER III METHODOLOGY Introduction This chapter describes the way in which the study was conducted. Each research question consists of the study participants, the testing instruments, the procedure and the data analysis needed to answer that specific research question. The research methodology for the study was quantitative and is reflected in the way the data were collected and analyzed. Data were collected through the use of Likertscale instruments and questionnaires that were analyzed through quantitative statistical procedures. For the purposes of this study, a quantitative methodology was preferable to a qualitative approach because it permitted a larger sample size, thereby making the findings and conclusions more generalizable. Research Question 1 What is the impact of a vicarious learning experience (preservice teacher field experiences) on perceived preservice elementary teacher efficacy and perceived preservice elementary science teacher efficacy? Study Participants The participants for this part of the study consisted of 46 undergraduate elementary education students (preservice teachers) who were enrolled in a course titled 32 Early Lab and Clinical Experience in Elementary Education II at the university. Twelve of the 46 preservice teachers were also enrolled in the ExCEL program. The course involved direct observation and participation in classroom environments, kindergarten through eighth grade, and ran concurrent with seminars exploring multicultural education and integrated curricula. Testing Instruments and Procedure The Teacher Efficacy Scale (TES) is a teacher efficacy instrument designed to measure personal teacher efficacy (PTE) and teacher efficacy (TE). The instrument consists of 30 items with a Likert scale ranging from 1 (Strongly Disagree) to 6 (Strongly Agree). Gibson and Dembo (1984) found, after performing factor analysis, that sixteen of the original 30 items had acceptable reliability coefficients. For the current study all 30 items were used, and then factor analysis was conducted to evaluate what factors emerge within the specific test population. Science Teaching Efficacy Belief Instrument B (STEBIB): Science teacher efficacy instrument designed to measure PTSE and STOE. The instrument consists of 23 items from the STEBI. The STEBIB is identical to the STEBI accept items 20 and 25 have been removed and the verb tenses of some of the 23 items have been changed to accommodate preservice teachers. Items 20 and 25 were removed because both crossfactor loaded. These changes were made so the instrument, which was originally designed for inservice teachers, could be used with preservice teachers (Enochs & Riggs, 1990). After thorough analyses, Enochs and Riggs (1990) concluded that the STEBIB could be considered reliable and reasonably valid with a stable and unified factor structure. 33 The Educational Field Experience Teacher Rating and Preservice Teacher Questionnaire refers to the seven educational field experience inservice teacher rating items that have been validated as reflective of goals of the university teacher education program. The Preservice Teacher Questionnaire included items that helped assess the impact of the vicarious field experiences. Early in the spring 2007 semester, before field experiences (vicarious experiences) began, subjects signed informed consent forms and provided limited demographic data by completing a brief questionnaire (see Appendix A). They then completed the STEBIB (see Appendix C). The entire process took less than half an hour and occurred during a regularly scheduled class meeting on January 8, 2007. Near the end of the same semester, after field experiences were completed, subjects rated the teacher they observed during their educational field experiences and provided data about classroom events that occurred while doing their field experiences (see Appendix E), completed the TES (see Appendix B), and again completed the STEBIB (see Appendix C). The entire process took less than half an hour and occurred during a regularly scheduled class meeting on April 23, 2007. Data Analysis For the purposes of data analysis and interpretation, this part of the study was treated as observational (the assignment of subjects into a treated group is outside the control of the investigator), where teacher efficacy, science teacher efficacy, the rating of the observed teacher, the gender of the subjects, the age of the subjects, number of lessons taught by the subject, content area taught by the subjects, science lessons taught by the subjects, teaching rating of the lessons taught by the subjects and number of times 34 the subjects observed a science lesson being taught were the observed variables. The relationship of these variables was assessed using analysis of variance (ANOVA) and analysis of covariance (ANCOVA). Research Question 2 What is the impact of the characteristics of the field experience classroom, within the given school where the educational field experience occurred, on perceived preservice elementary teacher efficacy and perceived elementary preservice science teacher efficacy? Study Participants The participants for this part of the study consisted of the same 46 preservice elementary teachers described above along with twenty inservice teachers whose classrooms served as sites for the preservice teachers’ field experiences. Twelve of the 46 preservice teachers were also enrolled in the ExCEL program. Testing Instruments and Procedure The Classroom Demographic and Cooperating Teacher Questionnaire (see Appendix D) was a questionnaire that allowed the inservice teachers to record the ethnic demographics of the classes they teach, information about their school, and information about themselves. The preservice teachers delivered the questionnaire to their field experience inservice teachers midway through the spring 2007 semester. After granting informed consent, the inservice teachers completed the questionnaire, returned it to the preservice teacher that was observing them, who then returned it to the professor of their field experience course. The researcher collected the questionnaires from the professor. 35 Ethnic data collected from the preservice elementary teachers earlier (see Appendix A) were also used in this part of the study. Data Analysis For the purposes of data analysis and interpretation, this part of the study was treated as observational, where teacher efficacy, science teacher efficacy, the district in which the inservice teacher taught, the number of students the inservice teacher instructed each day, the grade instructed, the ethnicity of the inservice teacher’s classroom where the field observation occurred, the number of students in the classroom who received free and reduced lunch, the age of the inservice teacher, the gender of the inservice teacher, the ethnicity of the inservice teacher, the number of years the inservice teacher had taught, the number of years the inservice teacher had taught at their current school and the number of years the inservice teacher had taught at their current grade level were the observed variables. The relationship of these variables was assessed using analysis of variance (ANOVA) and analysis of covariance (ANCOVA). Assumptions and Limitations of the Study Assumptions The following assumptions were accepted: 1. The preservice elementary teachers understood the directions and the items on the testing instruments and responded to the items honestly, accurately, and to the best of their ability. Limitations The study was limited by the following: 36 1. Only one elementary field experience class with a specific set of components was studied. This prevents generalization to programs that do not resemble this course. 2. The study was conducted with preservice elementary teachers at a landgrant university. This prevents generalizations to other preservice teachers and to other types of elementary education programs at other universities. 3. The subjects were not randomly selected. Selection was determined by required enrollment in the methods courses. 4. The preservice teachers had many concurrent experiences during this phase of their teacher preparation, such as additional coursework, volunteering in public service work, substitute teaching, and/or parttime jobs. All of these experiences may have influenced the final results. 5. The possibility of confounding variables exists in the current study. With any preservice teacher, there were a number of variables that may have confounded the effects of each other. For example, the number of years of teaching experience of the inservice teacher may confound student ethnicity or school setting may confound the number of science lessons the preservice teacher taught. 37 CHAPTER IV RESULTS Introduction This chapter describes the results and statistical analysis associated with the current study. The two research questions that define the current study are presented followed by statistics for the two testing instruments used in the study. Next, some analysis of the demographic data for the study participants is presented. Lastly, the results and statistical analysis associated with the study’s two research questions are discussed. Research Questions Research Question 1 What is the impact of a vicarious learning experience (preservice teacher field experiences) on perceived preservice elementary teacher efficacy and preservice elementary science teacher efficacy? Research Question 2 What is the impact of the characteristics of the field experience classroom, within the given school where the educational field experience occurred, on perceived preservice elementary teacher efficacy and perceived preservice elementary science teacher efficacy? 38 Testing Instruments Data Analysis Principal Components Analysis Principal Components Analysis (PCA) transforms a set of correlated variables into a smaller set of uncorrelated variables (Johnson, 1998). This uncorrelated set of variables is called the principal components. Using the PCA is advisable in determining the number of factors to use in factor analysis (FA) (Johnson, 1998). PCA found two components for each of the Science Teaching Efficacy Belief Instrument B (STEBIB) pretest, STEBIB posttest and the Teacher Efficacy Scale (TES) responses. For the STEBIB pretest response variable, two principal components accounted for 40.3% of the variance. Two principal components accounted for 47.7% of the variance for the STEBIB posttest response variables. Lastly, two principal components accounted for 28.8% of the variance for the TES variables. Thus, since two principal components are sufficient for all three response variables, factor analysis for each instrument were run with only two factors. Factor Analysis A factor analysis (FA) model summarizes a large set of possibly related response variables with a smaller set of uncorrelated response variables (Johnson, 1998). This smaller set of uncorrelated variables explains the relationships that exist between the large set of original variables. The set of 23 items for the STEBIB (pretest and posttest) and the 30 items for the TES make up the three sets of possibly correlated variables. Once a smaller set of factors is determined, all other statistical analyses are performed on the underlying factors and not the original variables. Recall that PCA determined that two factors were identified for all three tests: the STEBIB pretest, the STEBIB posttest 39 and the TES. Table 1 presents the factor loading for both factors for the STEBIB pretest and STEBIB posttest. All STEBIB factor loadings (PSTE and STOE), as identified by Enochs & Riggs (1990), loaded on their correct factor. Using a cutoff factor loading value of 0.45, all items on the STEBIB loaded on either factor 1 (PSTE) or factor 2 (STOE). Values ranged from a low of 0.469 for item 7 to a high of 0.859 for item 18. There were some differences between the pretest and posttest regarding which items loaded on the two factors (PSTE and STOE) (See Table 1). To handle these differences, all PSTE items that loaded from the pretest and/or the posttest were combined and all STOE items that loaded from the pretest and/or the posttest were combined. Table 2 presents these combined factor loadings for the STEBIB pretest and posttest. Item 6 was used as PSTE based on the original instrument (Enochs & Riggs, 1990) and item 9 was omitted from the current study based on incorrect factor loading. Table 3 presents the factor loadings for the Teacher Efficacy Scale (TES). Note that for factor 1 three of the four items were Personal Teaching Efficacy (PTE) on the original Gibson and Dembo (1984) instrument. The fourth item, number 18: “If students are particularly disruptive one day, I ask myself what I have been doing differently,” did not load on the original instrument nor was this item defined as PTE or TE, although it appears to be a PTE item based on its emphasis on selfefficacy. 40 Table 1 STEBIB Original Factor Analysis Loadings Item # Original Instrument Factorc Factor 1 PSTE Factor 2 STOE Pretest 1 3 4 6 7 8 11 14 15 16 17 18 19 20 21 22 STOE PSTE STOE PSTEa STOE PSTE STOE STOE STOE STOE PSTE PSTE PSTE PSTE PSTE PSTE  0.709    0.707     0.676 0.751 0.646 0.562 0.543 0.588 0.625  0.494 0.594 0.494  0.761 0.619 0.654 0.628       Posttest 1 3 4 5 6 7 8 9 12 13 14 16 17 18 19 21 23 STOE PSTE STOE PSTE PSTEa STOE PSTE STOEb PSTE STOE STOE STOE PSTE PSTE PSTE PSTE PSTE  0.519  0.793 0.734  0.649 0.503 0.791    0.709 0.859 0.778 0.775 0.695 0.713  0.816   0.469    0.506 0.574 0.639      Note. Cut off for Factor Loading of >.45 aDouble Factor Loading. Item 6 was used as PSTE based on the original instrument (Enochs & Riggs, 1990). bIncorrect Factor Loading, Item 9 was omitted from the current study. cAs identified by Enochs & Riggs, 1990 41 Table 2 STEBIB Combined Factor Analysis Loadings Item # Positive/Negative Wording Original Instrument Factor Pretest Posttest Factor 1 PSTE 3 5 6 8 12 17 18 19 20 21 22 23 N P N N P N P N N N P N PSTE PSTE PSTE PSTE PSTE PSTE PSTE PSTE PSTE PSTE PSTE PSTE 0.709 0.326 0.375 0.707 0.367 0.676 0.751 0.646 0.562 0.543 0.588 0.378 0.519 0.793 0.734 0.649 0.791 0.709 0.859 0.778 0.437 0.775 0.270 0.695 Factor 2 STOE 1 4 7 11 13 14 15 16 P P P P N P P P STOE STOE STOE STOE STOE STOE STOE STOE 0.625 0.494 0.494 0.761 0.203 0.619 0.654 0.628 0.713 0.816 0.469 0.314 0.506 0.574 0.420 0.639 Note. Cut off for Factor Loading of >.45 Cronbach’s Alpha Reliabilities Cronbach’s alpha is a measure of the internal reliability of an instrument. Interpretation of Cronbach’s alpha presumes that an instrument consisting of K items is only a subset of all possible items that could be used to measure the topic of interest. Cronbach’s alpha is the correlation between the total scores of any two random samples from the entire population of all possible items (Johnson, 1998). Thus, just as any correlation, Cronbach’s alpha may range from 0 to 1, but generally any score above 0.7 is 42 considered to be an acceptable indicator of the instrument’s internal reliability (Johnson, 1998). Table 4 contains the Crombach’s alpha reliabilities for the STEBIB and TES. Table 3 TES Factor Analysis Item # Positive/Negative Wording Original Instrument Factor Loadings Factor 1 PTE 1 14 18 19 P P P P PTE PTE Undefineda PTE 0.688 0.791 0.489 0.928 Indefinable Factor 2 2 3 5 8 12 13 15 20 27 P P P P P N P P P TE Undefineda Undefineda Undefineda PTE Undefineda PTE Undefineda TE 0.550 0.462 0.675 0.534 0.458 0.640 0.554 0.689 0.656 Note. Cut off for Factor Loading of >.45 aNonFactor Loading on Original 30 Item Instrument (Gibson and Dembo, 1984). Table 4 Crombach’s Alpha Reliabilities STEBIB Factor 1(PSTE) Factor 2 (STEO) Pretest Posttest Pretest Posttest .846 .909 .727 .77 Factor 1 (PTE) Factor 2 (Indefinable Factor 2 ) TES .798 .909 43 Summary Testing Instrument Statistics Table 5 presents the means, standard deviations, ttest values and probabilities for the pretest and posttest scores on the Science Teaching Efficacy Belief Instrument B (STEBIB). Statistics for both factors personal science teacher efficacy (PSTE) and science teacher outcome expectancy (STOE) linked with the STEBIB are presented for both the pretest and posttest. A tTest (α = 0.10) revealed that the PSTE posttest mean of 3.955 was significantly greater than the pretest mean of 3.777 (t = 2.519, p = 0.015). Similarly, the STOE posttest mean of 3.685 was significantly greater than the pretest mean of 3.492 (t = 1.979, p = 0.054). Table 5 STEBIB Summary Testing Instrument Statistics PSTE pretest PSTE posttest STOE pretest STOE posttest Mean 3.777 3.955 3.492 3.685 S.D. 0.524 0.603 0.527 0.535 tTest 2.519 1.979 P(t) 0.015 0.054 Table 6 presents the means and standard deviations for the Teacher Efficacy Scale (TES). Both factors “Personal Teacher Efficacy (PTE)” and “Indefinable Factor 2 (IF2)” associated with the TES are presented for both the pretest and posttest. Indefinable Factor 2 (IF2) is the second TES factor identified in the present study. Items associated with this factor were 2, 3, 5, 8, 12, 13, 15, 20 and 27. Based on the nonassociative nature of these items, a common construct factor, such as TE, was deemed invalid. 44 Table 6 TES Summary Testing Instrument Statistics PTE IF2 Mean 4.446 4.592 S.D. 0.813 0.615 Study Participants’ Demographic Data Analysis Preservice Teachers Fortysix preservice teachers agreed to participate in the study. Fortyfive were female and one was male. Fortyone were White, four were American Indian/or Alaskan and one was Hispanic/Latino. Twelve of the 46 preservice teachers were also enrolled in the ExCEL program. The mean age of all 46 preservice teachers was 22. The minimum age was 20, the maximum age was 29, the median age was 22 and the standard deviation was 1.71. Table 7 presents all 46 preservice teacher responses to the question “How many times did you teach a lesson?” from the preservice post data collection event (See Appendix E). “T” denotes preservice teachers who were in the traditional observation program and “E” denotes preservice teachers who participated in the ExCEL program. The ExCEL (Excellence in Collaborative and Experiential Learning) Program is a collaborative partnership between the College of Education and the local public school district. Open to elementary education majors in their final semester before student teaching (Semester Y), the program features a threeday a week full day internship in a public school classroom in which teacher candidates are partnered with expert classroom teachers. By observing the range of lessons taught, thirtytwo of the thirtyfive 45 preservice elementary teachers who taught between 1 and 10 lessons were in the traditional observation program. Two traditional observation program preservice elementary teachers taught between 1120 lessons. All preservice teachers who taught more than 21 lessons were in the ExCEL program. Table 7 Preservice Teacher Question Mean Lessons Taught 10 Minimum Lessons Taught 1 Maximum Lessons Taught 50 SD of Lessons Taught 13.623 Range of Lessons Taught # of preservice teachers 1 to 10 Lessons Taught 32 T/3E 11 to 20 Lessons Taught 2 T 21 to 30 Lessons Taught 5 E 31 to 40 Lessons Taught 2 E 41 to 50 Lessons Taught 2 E Note. T = Traditional elementary education program, E = ExCEL elementary education program. Table 8 presents all 46 preservice teacher responses to the question “If you taught science, how many times did you teach a science lesson?” from the preservice post data collection event (See Appendix E). “T” denotes preservice teachers who were in the traditional observation program and “E” denotes preservice teachers who participated in 46 the ExCEL program. By observing the range of science lessons taught, fourteen traditional observation program preservice elementary teachers taught no lessons. Eighteen traditional observation program and five ExCEL program preservice elementary teachers taught only one or two science lessons. One traditional observation program and one ExCEL program preservice elementary teacher taught three or four science lessons. Lastly, all preservice teachers who taught more than five science lessons were in the ExCEL program. Table 8 Preservice Teacher Question Mean Science Lessons Taught 1.67 Minimum Science Lessons Taught 0 Maximum Science Lessons Taught 10 SD of Science Lessons Taught 2.35 Range of Science Lessons Taught # of preservice teachers 0 Science Lessons Taught 14 T 12 Science Lessons Taught 18 T/5 E 34 Science Lessons Taught 1 T/1 E 56 Science Lessons Taught 4 E 78 Science Lessons Taught 1 E 910 Science Lessons Taught 1 E Note. T = Traditional elementary education program, E = ExCEL elementary education program. 47 Table 9 presents all 46 preservice teacher responses to the question “How many times did you observe a science lesson being taught?” from the preservice post data collection event (See Appendix E). Again, “T” denotes preservice teachers who were in the traditional observation program and “E” denotes preservice teachers who participated in the ExCEL program. An examination of the range of science lessons observed revealed that all 34 of the preservice elementary teachers in the traditional observation program observed nine or fewer lessons, while ten of the twelve ExCEL preservice teachers observed more than ten science lessons. Table 9 Preservice Teacher Question Mean Observed Science Lessons 8.09 Minimum Observed Science Lessons 0 Maximum Observed Science Lessons 45 SD of Observed Science Lessons 13.133 Range of Observed Science Lessons # of preservice teachers 09 Observed Science Lessons 34 T/2 E 1018 Observed Science Lessons 3 E 1927 Observed Science Lessons 1 E 2836 Observed Science Lessons 2 E 3745 Observed Science Lessons 4 E Note. T = Traditional elementary education program, E = ExCEL elementary education program. 48 Inservice Teachers Twenty inservice teachers agreed to participate in the study. Nineteen were female and one was male. Eighteen were white and two were American Indian/or Alaskan. The mean age of all twenty inservice teachers was 44. The minimum age was 24, the maximum age was 59, the median was 40 and the standard deviation was 11.71 Table 10 presents all 20 inservice teacher responses to the questions (1) “How many years of teaching have you completed?”; (2) “How many years have you taught at your current school?”; and (3) “How many years have you taught at your current grade level?”. These questions are from the inservice data collection event (See Appendix D). Table 10 Select Inservice Teacher Questions Mean # of years teaching completed 15.3 Minimum # of years teaching completed 1.5 Maximum # of years teaching completed 37 SD of # of years teaching completed 9.99 Mean # of years taught at your current school 5.725 Minimum # of years taught at your current school 1 Maximum # of years taught at your current school 25 SD of # of years taught at your current school 5.63 Mean # of years taught at current grade level 6.7 Minimum # of years taught at current grade level 1 Maximum # of years taught at current grade level 23 SD of # of years taught at current grade level 6.764 49 Introduction to Research Question 1 and 2 Analysis In order to answer research questions 1 and 2 for the STEBIB test, analysis of covariance (ANCOVA) was utilized. ANCOVA is a statistical procedure that tests a set of factors for significance on the response variable while removing the variance for which the covariant accounts. For both research questions, the response variable is the posttest score for the STEBIB and the covariant is the pretest score for the STEBIB. The inclusion of the pretest score into the model as a covariant can increase power because it accounts for additional variability had the covariant been left out of the model. Forward stepwise selection was used as a variable selection method for the final ANCOVA linear model. This variable selection method selects the most parsimonious set of factors for the ANCOVA linear model. Research questions 1 and 2 for the TES were addressed using analysis of variance (ANOVA). ANOVA is a statistical procedure that relates a set of quantitative factors to a response variable. There is no covariant for the TES since it was given only one time. The categories associated with each variable were based on the distribution of the data for each specific category. Because six out of ten of the variables were discrete (discontinuous) and because the data of several of the continuous variables were not normally distributed, the responses were grouped into a manageable number of categories. There was an attempt to equalize the number of subjects in each category in order to pick up true existing differences between the categories. Appendix F contains tables detailing the makeup of each of these categories for variables related to preservice teachers, inservice teachers, and field experience classrooms. 50 Table 11 Variables and Categories Associated with Research Question 1 Variable Categories Age of preservice teacher < 22 years ≥ 22 years (N = 22) (N = 24) Gender of preservice teacher Male Female (N = 1) (N = 45) White Amer. Ind. or Alaskan Hispanic/ Ethnicity of preservice teacher Latino (N = 41) (N = 4) (N = 1) Rating of the inservice field experience teacher by the preservice teacher Likert Scale 15 (Poor to Excellent) Number of lessons the preservice teacher 1 24 510 1250 taught (N = 12) (N = 10) (N = 13) (N = 11) Number of science lessons the preservice 0 1 210 teacher taught (N = 14) (N = 22) (N = 10) Selfrating of the science lessons taught by the preservice teacher Likert Scale 15 (Poor to Excellent) Selfrating of all lessons taught by the preservice teacher Likert Scale 15 (Poor to Excellent) Number of science lessons the preservice 0 12 310 1545 teacher observed (N = 13) (N = 12) (N = 13) (N = 8) Was the preservice teacher part of the ExCEL program? ExCEL NonExCEL (N = 12) (N = 34) Research Question 1 Analysis Table 11 lists the variables associated with research question 1: “What is the impact of a vicarious learning experience (preservice teacher field experiences) on perceived preservice elementary teacher efficacy and preservice elementary science 51 teacher efficacy?” Each variable is presented along with the categories related to that variable. STEBIB Results For both factors 1 and 2, the ANCOVA model for research question 1 had no statistically significant independent variables (α = 0.10). TES Results The ANOVA model for factor 1 of the TES did not have any significant independent variables (α = 0.10). However, the ANOVA model for factor 2 of the TES did have a statistically significant independent variable, the age of the preservice teacher. Table 12 contains the ANOVA results. The means resulting from the ANOVA are given in Table 13. Research Question 2 Analysis Table 14 lists the variables associated with research question 2: “What is the impact of the characteristics of the field experience classroom, within the given school where the educational field experience occurred, on perceived preservice elementary teacher efficacy and perceived preservice elementary science teacher efficacy?” Each variable is presented along with the categories related to that variable. Table 12 Age of Preservice Teacher ANOVA Df Sum of Sq Mean Sq F Value Pr(F) Age of Preservice Teacher 1 1.09242 1.092424 3.017598 0.08936196* Error 44 15.92878 0.362018 *p< .10 52 Table 13 Age of Preservice Teacher Means Variable Mean of TES Age of Preservice Teacher (Below 22 years of age) 4.757 Age of Preservice Teacher (22 years of age or older) 4.449 STEBIB Results The final ANCOVA model for research question 2, factor 1 of the STEBIB test includes three independent variables as well as the covariate. Table 15 contains the ANCOVA model results. In the ANCOVA table, ‘Group’ refers to the collective effect of the independent variables in the model while ‘Covariate’ refers to the effect of the covariate (the pretest score on the STEBIB) on the response variable (the posttest score of the STEBIB). 53 Table 14 Variables and Categories Associated with Research Question 2 Variable Categories Name of the inservice teachers school Schools 1 through 7 Name of the inservice teachers district Districts 1 through 3 Number of students the inservice 1617 2225 1920 62147 teacher instructed per day (N = 5) (N = 6) (N = 5) (N = 4) Grade the inservice teacher 1st2nd 3rd 4th7th instructs (N = 6) (N = 6) (N = 8) Percentage of Hispanic/Latino 0% 3.1%11.8% 21%75% students the inservice teacher instructs each day (N = 8) (N = 7) (N = 5) 0% 4.5% 5.3% 8.1% 10.2% 12.5% 14.8% 18.2% 20% 77.3% Percentage of Black students the inservice teacher instructs each day (N = 4) (N = 4) (N = 4) (N = 5) (N = 3) 10.2%50% 52.9%73.7% 75%95% Percentage of White students the inservice teacher instructs each day (N = 7) (N = 7) (N = 6) Percentage of Asian/Pacific Islander students the inservice teacher instructs each day 0% 1.4%12% (N = 12) (N = 8) Percentage of American Indian/or Alaskan students the inservice teacher instructs each day 0%2.7% 4%57.9% (N = 10) (N = 10) Percentage of other ethnicity students the inservice teacher instructs each day 0% 4%18.8% (N = 13) (N = 7) Percentage of students the inservice teacher instructs that receive free and reduces lunch 9.736.8% 58.8100% (N = 5) (N = 15) Age of the inservice teacher ≤40 years >40 years (N = 10) (N = 10) Gender of the inservice teacher Male Female (N = 1) (N = 19) Ethnicity of the inservice teacher White American Indian/Alaskan (N = 18) (N = 2) Number of years teaching the inservice teacher had completed 2.5 yrs11 yrs 15 yrs37 yrs (N = 10) (N = 10) 54 Table 14 Cont’d. Variables and Categories Associated with Research Question 2 1 year 2 yrs3 yrs 4 yrs7 yrs 8 yrs25 yrs Number of years the inservice teacher had been at their current school (N = 5) (N = 5) (N = 5) (N = 5) Number of years the inservice teacher had taught at their current grade level 1 yr3 yrs 5 yrs23 yrs (N = 12) (N = 8) Was the preservice teacher part of the ExCEL program? ExCEL NonExCEL (N = 12) (N = 34) Table 15 ANCOVA Model Results for Research Question 2, Factor 1 of the STEBIB Df Sum of Sq Mean Sq F Value Pr(F) Group 3 1.023479 2.563234 0.09360104* Covariate 1 3.079642 3.079642 23.13828 0.00022924* Error 15 1.996460 0.133097 *p< .10 The test for the significance of the covariate in the model resulted in an F test statistic of F* = 23.13828 and an estimated pvalue 0.0002292445. The overall F test for equality of the means for each level of the independent variable was F* = 2.563234 and had an estimated significance level of 0.09360104. The significant predictors for the ANCOVA model included the inservice teacher ethnicity (p = 0.0832), the inservice teacher gender (p = 0.0065), and the number of free lunch students (p = 0.0197). Table 16 contains the means for factor 1 of the STEBIB posttest associated with the ANCOVA linear model. Adjusted means are computed for each category of each independent variable holding all other variables, including the covariate, constant using their respective mean values. 55 Table 16 Means for Factor 1 of the STEBIB Posttest Associated with the ANCOVA Variable Mean of Posttest STEBIB Ethnicity of the inservice teacher (White) 3.892 Ethnicity of the inservice teacher (American Indian/or Alaskan) 3.243 Gender of the inservice teacher (Male) 3.513 Gender of the inservice teacher (Female) 3.621 Percentage of students the inservice teacher instructs that receive free and reduces lunch (9.7%36.8%) 3.711 Percentage of students the inservice teacher instructs that receive free and reduces lunch (58.8%100%) 3.423 Table 17 ANCOVA Model Results for Research Question 2, Factor 2 of the STEBIB Df Sum of Sq Mean Sq F Value Pr(F) Group 5 2.76232 8.681305 0.00083516* Covariate 1 0.385161 0.3851610 6.05234 0.02866086* Error 13 0.827299 0.0636384 *p< 0.10 Factor 2 of the STEBIB yielded an ANCOVA model with four independent variables. Table 17 presents the ANCOVA results. Note that the test for the significance of the covariate yielded a significant F test (F* = 6.052, p = 0.0287). The test for overall equality of the group means was also significant (F* = 8.681, p = 0.00084). The significant predictors for the ANCOVA model included the percentage of Asian/Pacific Islander students (p = 0.0865), the percentage of Hispanic/Latino students (p = 0.0005), 56 the percentage of American Indian/or Alaskan students (p = 0.0293), and the indicator variable for ExCEL program participation (p = 0.0064). Table 18 contains the means for each level of each independent variable. Table 18 Means for Factor 2 of the STEBIB Posttest Associated with the ANCOVA Variable Mean of Posttest STEBIB Percentage of Asian/Pacific Islander students the inservice teacher instructs each day (0%) 3.727 Percentage of Asian/Pacific Islander students the inservice teacher instructs each day (1.4%12%) 3.556 Percentage of Hispanic/Latino students the inservice teacher instructs each day (0%) 3.817 Percentage of Hispanic/Latino students the inservice teacher instructs each day (3.1%11.8%) 3.301 Percentage of Hispanic/Latino students the inservice teacher instructs each day (21%75%) 3.807 Percentage of American Indian/or Alaskan students the inservice teacher instructs each day (0%2.7%) 3.512 Percentage of American Indian/or Alaskan students the inservice teacher instructs each day (4%57.9%) 3.771 ExCEL preservice teacher 3.456 NonExCEL preservice teacher) 3.827 Table 19 contains the Tukey simultaneous confidence intervals to determine where the means are significantly different. Tukey simultaneous confidence intervals are necessary when considering the independent variable for Hispanic/Latino students the inservice teacher instructs each day since there are three levels. Thus, we will compare all levels of Hispanic/Latino students the inservice teacher instructs each day with a 57 controlled experimentwise error rate of 0.10. Note that the Posttest STEBIB means for categoriesl 0% and 3.1%11.8% are different as are the means for categories 3.1%11.8% and 21%75%. However, the means for categories 0% and 21%75% are not statistically different. Additionally, point estimates for the mean differences appear in the “Estimate” column. For example, the mean for 0% Hispanic/Latino students is 3.817. The adjusted mean for 3.1%11.8% Hispanic/Latino students is 3.301. Their difference between these means is 0.516. Concerning the “Interval” column, if the interval does not contain 0 then the means are statistically significant. If the range is all positive numbers then the first mean is larger. If the range is all negative numbers, then the first mean is smaller. Table 19 Tukey Simultaneous Confidence Intervals for Factor 2 of the STEBIB Comparison level Estimate Interval 0% compared to 3.111.8% 0.516 (0.182, 0.850)* 0% compared to 2175% 0.009 (0.413, 0.432) 3.111.8% compared to 2175% 0.507 (0.873, 0.140)* *Significant interval TES Results Table 20 contains the results of the ANOVA for TES factor 1. Note that the percentage of Hispanic/Latino students the inservice teacher instructs each day was the only significant independent variable (F = 2.8977, p = 0.082621). Additionally, the means for TES for each category of Hispanic/Latino students the inservice teacher instructs each day are given in Table 21. 58 Table 20 ANOVA for TES Factor 1 Df Sum of Sq Mean Sq F Value Pr(F) Percentage of Hispanic/Latino students the inservice teacher instructs each day 2 3.22567 1.612835 2.8977 0.082621* Error 17 9.46183 0.556578 *p< 0.10 Table 21 TES Factor 1 Means for Hispanic/Latino Students Variable Mean of TES The percentage of Hispanic/Latino students the inservice teacher instructs each day (0%) 5.031 The percentage of Hispanic/Latino students the inservice teacher instructs each day (3.1%11.8%) 4.107 The percentage of Hispanic/Latino students the inservice teacher instructs each day (21%75%) 4.700 Table 22 contains the Tukey simultaneous confidence intervals to determine where the means are significantly different. Tukey simultaneous confidence intervals are necessary when considering the independent variable for Hispanic/Latino students the inservice teacher instructs each day since there are three levels. Thus, we will compare all levels of Hispanic/Latino students the inservice teacher instructs each day with a controlled experimentwise error rate of 0.10. Note that the means for TES are different for categories 0% and 3.1%11.8% Hispanic/Latino students the inservice teacher instructs each day, but not for categories 0% and 2175% or categories 3.1%11.8% and 21%75%. Additionally, the point estimate (Estimate) is the difference between the 59 means. For example, 5.0314.107 = 0.924 is the estimated difference between 0% and 3.111% Hispanic/Latino students. Concerning the “Interval” column, if the interval does not contain 0 then the means are statistically significant. If the range is all positive numbers then the first mean is larger. If the range is all negative numbers then the first mean is smaller. Table 22 Tukey Simultaneous Confidence Intervals for Factor 1 of the TES Comparison level Estimate Interval 0% compared to 3.111.8% 0.924 (0.075, 1.770)* 0% compared to 2175% 0.331 (0.604, 1.270) 3.111.8% compared to 2175% 0.593 (1.550, 0.368) *Significant interval Table 23 contains the results of the ANOVA for TES factor 2. Jefferson Middle School, one of seven schools where tie inservice teachers taught, was the only significant independent variable in this model (F = 4.641841, p = 0.04499545). The mean TES Factor 2 score at Jefferson Middle School was lower when compared to the mean score of the other schools (See Table 24). Table 23 ANOVA for TES Factor 2 Df Sum of Sq Mean Sq F Value Pr(F) Jefferson Middle School 1 1.141358 1.141358 4.641841 0.04499545 Error 18 4.425926 0.245885 60 Table 24 Means for TES Factor 2 Variable Mean of TES Jefferson Middle School 3.833 Schools other than Jefferson Middle School 4.630 61 CHAPTER V CONCLUSIONS AND RECOMMENDATIONS Introduction This chapter describes the conclusions and recommendations associated with the study. The two research questions that define the current study are presented followed by the testing instruments conclusions and study participants demographic conclusions. Next, the conclusions associated with the study’s two research questions are discussed. Lastly, recommendations for future research, implications for practice and concluding remarks will be presented. Research Questions Research Question 1 What is the impact of a vicarious learning experience (preservice teacher field experiences) on perceived preservice elementary teacher efficacy and preservice elementary science teacher efficacy? Research Question 2 What is the impact of the characteristics of the field experience classroom, within the given school where the educational field experience occurred, on perceived preservice elementary teacher efficacy and perceived preservice elementary science teacher efficacy? 62 Testing Instruments Conclusions Principal Components Analysis Principal Components Analysis (PCA) was performed on the Science Teaching Efficacy Belief Instrument B (STEBIB) pretest responses, STEBIB posttest responses and the Teacher Efficacy Scale (TES) responses. Johnson (1998) advises using PCA to determine the number of factors to use in factor analysis (FA). For the STEBIB pretest response variables, two principal components accounted for 40.3% of the variance. Two principal components accounted for 47.7% of the variance for the STEBIB posttest response variables. With both the STEBIB pretest and STEBIB posttest this percentage of variance further validated the use of a twofactor model. The STEBIB posttest variance level was 7.4% higher than the STEBIB pretest further strengthening the argument for two STEBIB factors for both the pretest and the posttest. When comparing the TES variance, which is 28.8% for the TES variables, to both the STEBIB pretest and posttest variance, the TES variables is 11.5% lower than the STEBIB pretest response variables and 18.9% lower than the variance for the STEBIB posttest response variables. Even though the amount of variance associated with the TES was lower than the STEBIB, the percentage of variance still suggests the use of a twofactor model. In conclusion, the PCA validated that two principal components are sufficient for both the STEBIB pretest and posttest and the TES although the amount of variance accounted for was lower for the TES than the STEBIB pretest and posttest. Based on the PCA results, the factor analysis was run with only two factors for both the STEBIB and the TES. 63 Factor Analysis Factor analysis was conducted using a twofactor model for both the STEBIB pretest and STEBIB posttest. Factor loading measures the contribution of the factor, for example PSTE, to the STEBIB response. Table 1 presents the factor loading for both factors for the STEBIB pretest and STEBIB posttest at a cut off for factor loading of >.45. Note that the items that loaded for factor 1 [Personal Science Teaching Efficacy (PSTE)] correspond to items concerning PSTE, as identified by Enochs and Riggs in their original instrument analysis (1990), with the exception of item 9 in the posttest. Conversely, the items that loaded for factor 2 [Science Teaching Outcome Expectancy (STOE)] correspond to items concerning STOE, as identified by Enochs and Riggs in their original instrument analysis (1990), with the exception of item 6 for the pretest. This leads to a clear interpretation for these two factors based on the PCA, the factor analysis and the commonality between this study’s factor loading and that of Enochs and Riggs’ (1990) original study. With confidence we can conclude that, in the current study, two factors, PSTE and STOE, were measured by the STEBIB. This is consistent with findings of other researchers have found (King & Wiseman, 2001; Mulholland, Dorman & Odgers, 2004). Factor analysis was also conducted using a twofactor model for the TES. Table 3 presents the factor loadings for the TES. Note that for factor 1 [Personal Teacher Efficacy (PTE)] three of the four items are PTE on the original Gibson and Dembo (1984) instrument. The fourth item, item 18: “If students are particularly disruptive one day, I ask myself what I have been doing differently,” did not load on the original instrument nor was defined on the original instrument as PTE or TE but is a PTE item 64 based on its emphasis on selfefficacy and locus of control. Note that for factor 2 [Teacher Efficacy (TE)] we have TE items, PTE items and items that did not load on either factors of the original 30Item Instrument (Gibson & Dembo, 1984). This leads to a very clear interpretation for these two factors based on the PCA, the factor analysis and the commonality and discrepancy between this study’s factor loading and Gibson and Dembo’s (1984) original study factor loading. With confidence we can conclude that, in the current study, one factor, PTE, was measured by the TES. The second factor deemed TE in the original Gibson and Dembo (1984) instrument was not identified in the current study. Rather a second factor was indefinable based on the nonassociated nature of the items that loaded on it. As pointed out previously, the items that loaded for this factor were TE items, PTE items, and items that did not load on any factor on the original 30 Item Instrument (Gibson and Dembo, 1984). This second factor in the current study associated with the TES has been, for this study, named Indefinable Factor 2 (IF2). The problematic nature of the TES is consistent with what other researchers have found and eliminates the ability to draw any conclusions associated with factor 2 of the TES. The problematic nature of the TES and the past research related to the instrument will be discussed in greater detail in the section labeled “The Problematic Nature of the Teacher Efficacy Scale.” Cronbach’s Alpha Reliabilities Cronbach’s alpha is a measurement that assesses the internal reliability of an instrument. Cronbach’s alpha was performed on the STEBIB for factor 1 (PSTE) and factor 2 (STOE) for both the pretest and the posttest data. Cronbach’s Alpha was also performed on the TES for factor 1 (PTE) and factor 2 (IF2) for the TES’s single data set 65 (See table 4). Cronbach’s alpha may range from 0 to 1, and generally, any score above 0.7 is considered an indicator of good internal reliability. All STEBIB Cronbach’s Alpha reliability results were above the 0.7 cut off that defines satisfactory internal reliability of an instrument. Concerning the specific STEBIB Cronbach’s Alpha reliability results it should be noted that the internal reliability of the instrument increased from the pretest to the posttest for both factor 1 (PSTE) and factor 2 (STOE). This trend further increases the validity of conclusions associated with the STEBIB. The TES Cronbach’s Alpha reliability results are also above the 0.7 cut off that defines satisfactory internal reliability of an instrument. In conclusion, all Crombach’s alpha values related to both instruments used in this study were above 0.7 indicating we have sufficient internal reliability to assess the internal factors of these instruments. Summary Testing Instrument Statistics Means, standard deviations, ttest values and probabilities for the pretest and posttest scores were calculated for both factors of the STEBIB (See Table 5). There was found a significant difference between the PSTE pretest and the PSTE posttest with the PSTE pretest score of 3.777 increasing to 3.955 for the PSTE posttest. There was also found a significant difference between the STOE pretest and the STOE posttest with the STOE pretest score of 3.492 increasing to 3.685 for the STOE posttest. For both PSTE and STOE all means for the pretest and the posttest were between the STEBIB Likert Scale categories of “Uncertain (3)” and “Agree (4).” Perhaps the overall increase that is observed in both PSTE and STOE is indicative of the positive efficacious events the preservice teachers experienced during CIED 3430 66 (Early Lab and Clinical Experience in Elementary Education II) and in their inservice teacher’s classroom over the testing time. Note that the pretest mean scores are moderately positive and the increase that we see over the study time is small. The case maybe that this tendency to be just above “Uncertain (3)” is due to the preservice teacher’s general anxieties, fears, and lack of selfconfidence associated with understanding science and science teaching during the study time. This relationship has been well documented in the responses of preservice elementary teachers (Czerniak, 1989; Enochs & Riggs, 1990; Czerniak & Schriver, 1994). Mean and standard deviation were calculated for factor 1 (PTE) of the TES (See Table 6). The mean for PTE was 4.446. The PTE mean was between the TES Likert Scale categories of “Agree slightly more than disagree (4)” and ‘Moderately agree (5).” Again, the level of PTE at the conclusion of the treatment was moderately positive. This tendency to be just above “Agree slightly more than disagree (4)” is maybe indicative of the preservice teachers’ experiences in their teacher training. In most cases they have very little, if any, teaching experiences or other experiences within a given school system to serve as a basis for their teacher selfefficacy beliefs. This inexperience is evident in their moderately positive responses associated with what they believe their teaching abilities are. The mean and standard deviation were also determined for factor 2 (IF2) of the TES (See Table 6). The mean for IF2 was 4.592. The IF2 mean was also between the TES Likert Scale categories of “Agree slightly more than disagree (4)” and ‘Moderately agree (5)” but, considering the indefinable nature of this factor, no conclusions can drawn. 67 Study Participants Demographic Conclusions Preservice Teachers Table 7 summarizes all 46 preservice teacher responses to the question “How many times did you teach a lesson?” from the preservice post data collection event. From the table it is apparent that there is a wide range in the number of lessons the preservice teachers taught during the treatment time. The great majority of the preservice teachers taught between 1and10 lessons with only 2 out of 35 of the preservice teachers in this category being in the ExCEL program. Two traditional preservice teachers taught between 11 and 20 lessons while the remaining preservice teachers who taught from 21 to 50 lessons were enrolled in the ExCEL program. Table 8 summarizes all 46 preservice teacher responses to the question “If you taught science, how many times did you teach a science lesson?” The number of science lessons the preservice teachers taught during their field experience ranged from 0 to 10. As with the previous question, the traditional preservice teachers were grouped on the lower end of the range while the ExCEL program preservice teachers were grouped on the higher end of the range. Table 9 summarizes all 46 preservice teacher responses to the question “How many times did you observe a science lesson being taught?” As with the previous two preservice questions, there is a similar trend where the traditional preservice teachers were grouped on the lower end of the range and the ExCEL program preservice teachers were grouped on the higher end of the range of science lessons observed. Thirtyfour out of 36 preservice teachers observed between 0 and 9 lessons. Only the ExCEL preservice teachers observed from 10 to 45 lessons. 68 In conclusion, the general trends we observed were the ExCEL preservice teachers are teaching more lessons, teaching more science lessons, and observing more science lessons. The traditional preservice teachers are teaching fewer lessons, teaching fewer science lessons, and observing fewer science lessons. The results are indicative of the constraints placed on the preservice teachers by their participation in either the Traditional or ExCEL program. Inservice Teachers Table 10 summarizes all 20 inservice teacher responses to the questions (1) “How many years of teaching have you completed?”; (2) “How many years have you taught at your current school?”; and (3) “How many years have you taught at your current grade level?”. With all three questions the responses disclosed a wide range of years. This is apparent by observing the “Minimum number of years,” “Maximum number of years” and “SD of number of years” of all three questions. The “Mean number of years teaching completed” is also much larger than the “Mean number of years taught at your current school” and the “Mean number of years taught at current grade level.” Obviously, this is to be expected since both “Mean number of years taught at your current school” and “Mean number of years taught at current grade level” are subcomponents of the “Mean number of years teaching completed.” Seventeen of the 20 inservice teachers have 7 or more years of teaching experience with a maximum of 25 years of teaching experience (see Table 10). This is not surprising considering the desire of the preservice teacher faculty to place preservice teachers in observation classrooms with inservice teachers who have many successful years of teaching experience. Eleven of the 20 inservice teachers have taught between 1 69 and 4 years at their current school. Thirteen of the 20 inservice teachers have taught between 1 and 5 years at their current grade level. Research Question 1 “What is the impact of a vicarious learning experience (preservice teacher field experiences) on perceived preservice elementary teacher efficacy and preservice elementary science teacher efficacy?” Research Question 1 Conclusions STEBIB Factor 1 and 2 Results Table 11 lists the variables and categories associated with research question 1. For both STEBIB factors PSTE and STOE, the ANCOVA model for research question 1 had no statistically significant independent variables (α = 0.10). TES Factor 1 and 2 Results Table 11 lists the variables and categories associated with research question 1. The ANOVA model for TES factor 1 (PTE) had no significant independent variables (α = 0.10). However, the ANOVA model for TES factor 2 (IF2) of the TES did have one statistically significant independent variable, the age of the preservice teacher (p = 0.089). Table 12 contains the ANOVA results. Unfortunately, the problematic nature of TES factor 2 (IF2) eliminates the ability to draw any conclusions associated with factor 2 of the TES. The problematic nature of the TES and the past research related to the instrument is discussed in greater detail below in the section labeled “The Problematic Nature of the Teacher Efficacy Scale.” 70 Research Question 2 “What is the impact of the characteristics of the field experience classroom, within the given school where the educational field experience occurred, on perceived preservice elementary teacher efficacy and perceived preservice elementary science teacher efficacy?” Research Question 2 Conclusions STEBIB Factor 1 Results Table 14 lists the variables and categories associated with research question 2. The significant predictors for the ANCOVA model for research question 2, factor 1 (PSTE) of the STEBIB test included the inservice teacher ethnicity (p = 0.0832), the inservice teacher gender (p = 0.0065), and the number of free lunch students in the classroom (p = 0.0197). The conclusions associated with these three statistically significant variables are addressed below in this order. Inservice Teacher Ethnicity. The mean of posttest STEBIB for factor 1 (PSTE) of the ethnicity of the inservice teacher for White and American Indian/or Alaskan is presented in Table 16. The mean posttest PSTE score of preservice teachers in classrooms with white inservice teachers was 0.649 higher than a preservice teacher who was in an observation classroom with an American Indian/or Alaskan inservice teacher. The posttest mean score of the preservice teachers in observation classrooms with white inservice teachers had a Likertscale score close to “Agree (4)” (Mean = 3.892), while the mean score of the preservice teachers in observation classrooms with American Indian/or Alaskan inservice teachers had a mean Likertscale score much closer to “Uncertain (3)” (Mean = 3.243). 71 The moderate increase (0.649) in the preservice teachers’ PSTE seen in White versus American Indian/or Alaskan inservice teachers may reflect the preservice teachers’ educational experiences associated with their teacher’s ethnicity during elementary, high school and college. The moderate increase (0.649) in the preservice teachers’ PSTE seen in White versus American Indian/or Alaskan inservice teachers may also reflect the high White ethnicity percentage (89.130%) of the preservice teachers during the study time. The moderate decrease we observed in the preservice teachers’ PSTE related to the inservice teacher’s ethnicity could be an issue of the White preservice teachers’ inability to relate to or find commonality with the American Indian/or Alaskan inservice teacher. Conversely, the moderate increase could be consistent with the White preserivce teacher’s ability to relate to or find commonality with the White inservice teacher or the White inservice teachers providing a more supportive teaching experience for the preservice teachers. Although the statistical techniques used are reliable, the 0.649 decrease in the PSTE score of preservice teachers who were in an observation classroom with an American Indian/or Alaskan inservice teacher could be an artifact of the small sample size of American Indian/or Alaskan inservice teachers (2 out of 20). This small sample size makes the probability of falsely concluding differences in the mean more likely. In conclusion, preservice teachers who were in classrooms with American Indian/or Alaskan inservice teachers had lower PSTE than preservice teachers who were in classrooms with white inservice teachers. 72 Inservice Teacher Gender. The mean of posttest STEBIB for factor 1 of the gender of the inservice teacher for male and female is presented in Table 16. The mean posttest PSTE scores of preservice teachers in observation classrooms with female inservice teachers were 0.108 higher than that of preservice teachers in classrooms with male inservice teachers. Preservice teachers in an observation classroom with female inservice teachers and preservice teachers in an observation classroom with male inservice teachers both had Likertscale scores almost directly between “Agree (4)” and “Uncertain (3).” The mean of preservice teachers’ posttest PSTE was 3.621 if they were in a classroom with a female inservice teacher and 3.513 if they were in a classroom with a male inservice teacher. The slight increase (0.108) we see in the preservice teachers’ PSTE in female versus male inservice teachers is consistent with the preservice teacher’s educational experiences during their own elementary, high school and college years. This is also consistent with the high female percentage (97.826%) of the preservice teachers during the study time. The small decrease we observe in the preservice teachers’ PSTE related to being placed with a male inservice teacher could be an issue of the female preservice teacher’s inability to relate to or find commonality with the male inservice teacher. Conversely, the slight increase could be consistent with the female preserivce teacher’s ability to relate to or find commonality with the female inservice teacher. Note that when comparing inservice teacher ethnicity versus inservice teacher gender, inservice teacher ethnicity had a larger negative impact (0.649) than inservice teacher gender (0.108). Although the statistical techniques used are reliable, the 0.108 decrease in the PSTE score of preservice teachers who were in an observation classroom with a male 73 inservice teacher could be an artifact of the small sample size of male inservice teachers (1 out of 20). This small sample size makes the probability of falsely concluding differences in the mean more likely and could explain the differences that are observed. In conclusion, preservice teachers who were in classrooms with male inservice teachers had lower PSTE than preservice teachers who were in classrooms with female teachers. Percentage of Students that Received Free and Reduced Lunch. The mean of posttest STEBIB for factor 1 of the percentage of students who received free and reduced lunch in the inservice teacher’s classroom be 



A 

B 

C 

D 

E 

F 

I 

J 

K 

L 

O 

P 

R 

S 

T 

U 

V 

W 


