“ w»\. 3...... L . - .....n). . »,.. .. ... .I.u..... Emu} 1%“th pun»... .19, u “0‘. Vans” .. ... .. . , .. u. a ...L u- a! u. .kl. I... . ...... ... . .155... ...a . .. .. . ; ... ......m. . . . .. . ....aufiLL ......mi... figaxamw .. u... . . L .. . .... . 7. . .nvum..x...- mm. 4%.. .fifiwfimflfifikwi .. : _ .....cvwmflfi .... . .. . .1 . - , . r H tn! ‘13 .25 3?“. . Wm» I11... . . . . ......m ..wwvxfifi. ., ... . . Wanam? . .. tuWwT. . ... . . , . . . ...... ..M .. m1 ..AMWMWW . k. .. ..myfimw L4H. v.9..LU0rhwkmul. . . . . ... .3... ... . Iva“ rm Lh runny. «kr . v0 . flrn No firm... I? u. . . hwm"! Mun“... ...»...IM. . .. .... 1%. ...... m. . .J . .- {Infirmflfinfl . I... . . ......mflmwmu... .fimfigfih ...... ...- WW - . w I .. - I . g1 r. . .1 \ v . . H l . , o. LI .. . .... .1311." . . . . wrath...) mama»)... . I. . . I. . . Mmfsauwm .MMWRILEWJwE“ ...... -, E... Linuflwmflflmmfimra new.» ,L L... énflmmwfimanmmrt. ..Amwfimxu ...memmwwn. , . NI... RmeLIWW .IMW 1.»..meer ....»MWWMI. r .m.. %+¥wvlnm}vwnumm . . (n I.“ -... uIflLII‘ r .Vfit..fl§ku\»u7nmni . ...n. . vavmwntyrqlm M... h , - .. , fifmikrfiflus L . . . . 1.1. m. m... 1411.....mmfiu, t... . L .WWfiQfi... . . H . . .. .... .....-m H... mm. m. . .... .. km... .1... . . I - ...... ......mw....nw. m . ... .... _. -... ..wmwamwmd.flm{a - , , . .. , % --. .....x .uavnmuuxrwm..w.mw¢ux- L. 5mm? .wwxwfifgw.%mfimmk. L . ,. . . .- . «ILLMM .I_nu....mm..w..w-....éww ...... . 1...... .mg...fimw.w.mfl ,. . .. .-. .. .. .H- . - ...quflmfiuav..:wm.fifi.hw . .. .....nm....%.w..w1.. .... .. . . I .. . . . . . I . nil m - r. . . IN. ». Nun» . . . . .. . , . . ...... .-V4ma."q..p.rmm1wWM«mJumAmmfixwfi . .. .. . . ... *4...» km...» . . t I #3“ ... ...-WkaI....Hwhkahn. In. L. m% . I... MI. ..HanM—J I . . 1.... . . . um... unvgk ......“ Bhuufihfimfihfifi... .. 1 . ....uLVuL.umfim..m.w§mmWI§. I E": - .33; .121 J1” Eu I 213". 3.731 .31 . 2 I ! I .333 3! r. V: I 1 a.“ “It? .3: .13” ‘1 I L . ~. . a. . A . ..hfnuhhfin v... u . .. . )flhukfigflflmmiulplfl . . . o in-.. . . h. 153311.“. 11%.! f 5‘; -- :' iii .3. .1 .15 .th . } h? ’I I h: .35 A I In I . ”13:1: '19 R111. ...-t Iflm.flu..n.v . . .II» .I hut-I .énhlsvli... kl» l." . I11. . n .. . I§Lmflfi 3.15%.... II:1¥WVI.I“¥IL§KHIH§.VIII . - I ., IIleCo ...-1 "NAVAL. .\ .... » i 1 . . «a». . I . .. . .11 . . . I .. . wit... .. ”Lump waivinunuu fir. . .. i. 11% .5151”? .- I I hi1§fi+h¢llh§§ .mnmrnvfianf. , . . n «.113 . E1 E but - . uni. I 3‘: ..o. 3.1.5.(yohuhkauat§.5$unor.r&nrunl . . .II. I .1 UN... IIUWIII v I! . Iii}. I14. Judd." 119.1. . “figwfih. .... mWaWanflwmfifihs mummkhthtm h.” .....IN..4«<, .L. ... uhwm. . . In“... .. - I. hmnm2%.m1m u.» . ..v . . . . ...... .. .. émhfifiwfifim ...wsw... - . . ...»...anvddnuymehfl amkmxflmuufih . ......ufifiwsmmmmg. .. .. .. .... ...m «...»...m ...."an . -..». a . .. l.‘ 1%! H HLqun! .. ..L: 1 II. wul «Juan CELWI LEI}... . . 1L1: 61”...“ . . I l . .- vfiu ...éfiluIIIIII. “115% .11. .... ...IJII v .. .. 1.1.1.3111} F.11v1hw11grvhn III . . 591.1. I filflhfiraxlnhnpflkwmmlfil “.3131“: 1111bl111ol1|mfiu1§11¢14111 “In. .I. d. I... 1%1111L11Igg111huflk139 1.... III... r ..‘Uu I. .I L . . . 1%1.1I1 lung? . {mum-hr 3% ., ... ...mnmmmau. ITnmhILWV. I .. I... ... .13.... .... I. .......I....m...WII..I..¢I.I.n....I.nI...I.P.IIIiILY.Ilmmwalfruhmho... d.¢1”.r....mu..l.fluu. I 1.xuru9hwl, 1m... b..Y-uuh.u.LuLUI.1.h1.1h.hlomlmua3 menuk. Imw wlnulvrmlui :un1 . . . . . . , , .- . huh] I. ...":ande man! .NVI...In1 “Mitten... dfiuflvlttnwgnr . . . ..u . . . . . . “whexnoumeQflummlthiufll r Jun”)...mvfl. .3; 4.133)....» unlirhUNI. ., - 11 ., mafia... ... . . . . . u.%.t.m.urxmm..w1kmmmwnflmwmvflmsiILWIL. hm. 1.-...1fiéxfmm .tumfimflkfimma lflukh-.§1theoretica1 positions, attempts to answer 11m: questions should provide an incidental test of the theoretical positions. Limitations of the Study As the preceding analysis suggests, the fundamental questions that have been presented here have critical I_, _ i L ___‘- 8 implications for the professional education of teachers. They, therefore, demand precise consideration. This investi- gation sets out to test them in an exacting experiment. There are two salient features in the design of the investigation which were crafted to provide a severe test of the major research questions. First of all, in order to test the ex- tent to which self perceptions, formed as a result of self assessment, can be changed by external feedback, subjects had to be selected who had undergone or could undergo a prolonged and detailed process of self assessment as a preliminary stage to the experiment. In the second place, in order to compare in a one-on-one situation the strength of the in— fluence of self assessed feedback as against externally assessed feedback, a very exacting experimental procedure was followed that effectively resulted in a conservative test of the external feedback condition. To isolate and test the effect of external analysis and reporting, the role of the external evaluator was restricted to summarizing and re- porting the data according to a prescribed schedule. Thus, whereas self analysis was allowed full play, a careful effort was made so that only the defined function of external evalu- ation was operational in this experiment. If, under these stringent conditions, external feed— back is shown to have a strong influence on teachers' self perceptions, or externally evaluated feedback proves more potent than self analyzed feedback, then these findings would be persausive indeed. 9 Another limitation of this study should be noted. Re— search suggests three possible levels of response to feed- back which can be represented in the following model: Feedback —9 changes in self perceptions —§ intention to change behavior —9 actual changes in behavior. This study is limited to an investigation of the relation- ship between feedback and changes in self perceptions. It should be kept in mind that the broad concern underlying this study is the adequacy of self perceptions as a basis for making decisions about one's professional development. Research Questions A. Experimental Questions The basic purpose of this study is to provide defin— itive answers to the four specific experimental questions which follow. These questions are directly derived from the general questions which have been the focus of the discussion to this point. The four experimental questions are: 1. Does either of two forms of feedback——se1f— analyzed and reported (SAG), or externally analy— zed and reported (EAG) cause any change at all in a teacher's self perception of the relative levels of his or her abilities? 2. Which of two forms of feedback (SAG or EAG) will yield the greatest changes in a teacher's self perceptions? 10 Is the magnitude of change in self perceptions resulting from assessment feedback a function of empirically established levels of agreement (high, low) between feedback received and original self perceptions? Will the changes in self perceptions revealed in the posttest immediately following the experimen- tal intervention (seven to 10 day interval) differ from corresponding changes in perceptions revealed in a delayed posttest administered approximately six weeks after the intervention? B. Questions Concerned With Formative Evaluation A secondary aim of this investigation is to gather formative data on the program of evaluation which the teacher will be involved in during this experiment. Three questions concerned with instructional outcomes related to this pro— gram serve as the focus of this secondary purpose. Answers to these questions should in turn provide useful information for teachers and teacher educators involved in similar pro- grams of professional development. The three questions are as follows: 1. What elements in the two term evaluation program do the participants find most valuable or least valuable? How do they feel about participating in the program? ll 3. What are their perceptions of the feedback they receive? Overview of the Dissertation The remaining four chapters of this dissertation deal with (l) a review of the literature (2) the design of the investigation (3) the analysis of the data, and (4) conclus— ions and implications. The literature review is concerned with a theoretical investigation of the three central topics of this study. First, what is the importance of feedback in teacher educa— tion? Second, what is the importance of self perceptions in teacher education? Third, what does previous research say in relation to the four experimental questions? With regard to the first topic, a number of field related teacher training models that make use of self analyzed and externally analyzed feedback will be described and discussed. A number of issues will be explored, including the need for external support in evaluative feedback situations. The second section explores the history, nature and structure of self perceptions, how they are formed and changed, and how consistent and stable they are. These conjectures will be described from the per- spective of Combs' perceptual psychology and symbolic inter— actionist psychology. The concepts of the "looking-glass self" and "significant others” will be highlighted in the dis— cussion. The third section looks at empirical research in relation to the four experimental questions that are raised in this study. 12 Chapter Three begins by outlining the self analysis process which participants underwent in preparation for the experimental stage of the investigation. The two treatments, involving feedback that was self analyzed and reported by one group of participants (SAG), and externally analyzed and re— ported for the other group (EAG) are then described. Other sections of this Chapter deal with the selection and assign- ment of participants to experimental groups, and the develop- ment of the three measurement instruments used in the study—— including the Teacher Behavior Survey (TBS) used for gathering feedback and measuring participants' self perceptions. The nature of the dependent variables--inc1uding how they were determined, evidence of their internal and external validity, and the analyses used to determine if critical findings were statistically significant are also discussed. In Chapter Four a summary and analysis of the data are presented. The statistical analyses that are reported in— clude t-tests, analyses of variance tests, and correlational analyses. The data collected in response to the formative evaluation questions required only the computation of fre— quency scores and mean scores. Chapter Five briefly summarizes the purpose and de- sign of the investigation, the significant findings which resulted, and, finally, the implications of these results for teachers and teacher educators. Some far-reaching im— plications requiring further investigation are highlighted in this discussion. CHAPTER TWO: REVIEW OF LITERATURE INTRODUCTION The present investigation is directly concerned with two sets of variables. The treatment variables are two different ways of giving feedback to teachers. The depend- ent variables, which are being measured to determine the effects of the treatment variables, are derived from dif— ferences between self perceptions on three separate oc— casions. These variables are described operationally in Chapter Three. In this chapter, the literature will be re- viewed in relation to these variables. First, the importance of feedback in teacher education will be explored. Second, the literature in relation to the importance of self per— ceptions in teacher education will be reviewed. Third, answers to the four research questions regarding feedback and change in teacher self perceptions will be sought from the literature. IMPORTANCE OF FEEDBACK IN TEACHER EDUCATION Feedback is very important in teacher education, es- pecially in the area of field experiences. Where change and development of teacher behavior is concerned, many edu— cators would agree with Tuckman (1976) who writes that in 13 14 teacher education "the sine qua non of change is feed— back." The importance of feedback is attested to by the many ways in which feedback is utilized in various models of teacher field experience. A number of these uses will now be discussed. A. Clinical Supervision It would perhaps be difficult to think of any form of educational supervision in which some form of feedback is not given to the teacher. Whether in Cogan's eight stage model (Cogan, 1976), or in Goldhammer's five stage model (Goldhammer, 1969) or in any of the humanistic variations proposed by Sergiovanni (1975) or in any of Harris' models (1963), clinical supervision makes use of feeback, usually resulting from classroom observation, in a systematic cycle of supervision. Usually teacher and supervisor agree on goals and the specific focus of the observation, and the supervisor aims to give descriptive feedback within the limits set, after which plans for improvement and another cycle of supervision are made. An important aspect of clinical supervision is the democratic, supportive, optimistic ambience in which it op— erates. This is to be deliberately nurtured by the super— visor, who should operate as a colleague or professional helper, not as a superior. More generally, Sergiovanni (1976) 15 describes the practial philosophy of clinical supervision in these words: In clinical supervision it is assumed, for example, that operationally the cur— riculum of the school is manifested in what teachers do day by day; that changes in curriculum and in teaching formats re— quire changes in how teachers behave in classrooms; thattsupervisors are not teachers of teachers; that supervision is a process for which teachers and supervisors are both respon- sible; that the focus of supervision in on teacher strengths; that given the right cli— mate teachers are willing to improve; that teachers have large reservoirs of talent—— often unused; and that teachers want to in— crease competencies and to be successful for they seek and derive satisfaction for accom— plishing challenging and important work. This supportive element together with non-threatening, de— scriptive feedback are critical features in clinical super— vision. B. Peer Feedback Peer feedback based on classroom observation is recommended by a number of writers. Blumberg (1974) is often cited in support of peer evaluation. In answer to the ques- tion, ”Can teachers supervise teachers?" he reproduces two articles in support of the affirmative. The first is a re— search report by Amidon, Kies, and Palisi (1966) on the use of the Flanders System of interaction analysis in a peer group situation for giving feedback to teachers. He lists six useful rules for giving feedback that were devised by the faculty group in the study: 16 l. The person giving feedback describes rather than evaluates 2. Feedback is offered only in areas that are per— ceived as susceptible to change by the recipient 3. Feedback is given only upon request of the per— son whose teaching is being discussed. 4. Feedback is concerned with those aspects of teacher behavior that are characteristic of the teacher at the time that the discussion is taking place. 5. Feedback does not require a teacher to defend his personal opinion or feelings about the way in which he is teaching 6. Feedback is concerned with specific teaching acts, not with generalized interpretations. Amidon concludes that group supervision offers ad- vantages not found in the teacher-principal conference. Also, teachers do become more sensitized to verbal inter— action, and the group activity seems to influence faculty interpersonal relationships, communications, goal setting and behavioral norms in a positive way. The second article cited by Blumberg is by Abramson (1970) who provides a number of examples of the use of peer evaluation schools to point to its potential in staff im— provement. He does caution that evaluators need to be trained. Brophy (1979) recommends that teachers work in a group to observe one another's classrooms, and give feedback 17 and assistance to one another. Glassberg (1978) advocates peer supervision for student teachers, because it helps stu— dents to analyze their own and their colleagues' behavior, and in this way broadens their perspective from self to others with respect to the learning experience. Her experimental study gave evidence of signifcant gains in ego development. She stresses the need for a supportive environment to en— courage reflection, and the integration of experiences in order to promote the higher levels of ego development. Support for the efficacy of peer evaluation, or col— legialevaluation, as they call it, also comes from Roper et a1. (1976). They field tested a six step model involving goal setting, setting of criteria or standards, observing, ap— praising performance, communicating appraisals (providing feedback) and planning a program for improvement. They con— clude that it is a most useful and flexible approach. Bryant and Haack (1977) point to growing popu— larity and success of peer—centered systems of evaluation. They stress the need for training programs to define criteria for categorizing goals and behavior, to develop skills in data gathering and in giving feedback, and to gain expertise in planning improvement programs. Finally, Blumenthal (1977) reports favorably on the use of peer non evalu— 'ation feedback together with the use of video taped lessons. A highly organized feedback system that relies cen— trally on peer feedback is discussed separately in the next section. 18 C. Tuckman Feedback System The Teacher Feedback System developed by Tuckman (1976) is based on the Tuckman Teacher Feedback Form(TTFF). The form is a list of 28 adjectives, each describing a human element in teaching and paired with its opposite, for ex- ample, "original" is paired with “conventional,” and "patient" with "impatient". Four factors were derived from the data collected by this list: Creativity, Dynamism, Organized De— meanor, and Warmth and Acceptance. TheTHTT‘ and the four factors are the basis for describing and giving feedback on a teacher's performance. There are seven stages to the Feedback System: 1. Collect a team of volunteer teachers 2. Each teacher fills out the'FTFF describing the "good" teacher 3. Teachers observe one another 4. Each teacher is given a consensus summary of his or her ratings. Teachers meet as a group to discuss feedback. 5. Teachers engage in strength training. They learn their deficiencies and find out what they can do to improve by giving one another specific ideas. Role playing also takes place. 6. Leadership training is essential for group leaders. 7. They observe one another a second time to deter- mine whether there has been change, especially with reference to the four general factors. 19 Variations of this system have been tested in differ- ent environments. Spencer (1973) experimented with trainee teachers of vocational subjects and found that "warmth" and "acceptance" of the ideal teacher was rated much higher after the workshop, and that improvement in TTFF ratings was greater for the treatment group. Walencik (1973), substituting supervisors for peers, and high school students as the source of feedback using the TTFF, found results that supported the model. Student teachers who received TTFF feedback changed more than others. Kotula (1975) found that the group ap— proach led to greater increase in creativity in the experi— mental group, but that inexplicably, the control group had the greater gains in warmth and acceptance. Finally, Tuckman gives 12 specific rules for effec- tive feedback. These include the following, as summarized by Brophy (1979). Feedback should: 1. Involve specific, concrete behaviors or char— acteristics 2. Be credible and presented with good intentions and in understandable terms 3. Include specific guidelines for changes, and 4. Lead to a commitment to initiate specific changes D. Microteaching Cooper and Allen (1971) report that since its in— ception in 1963, microteaching has become an established teaching training procedure in many colleges, universities, 20 and school districts, to the extent that a national survey showed 44 percent of all teacher education programs to be using some form of microteaching. They define microteaching in the following way: Defined most succinctly, microteaching is a teaching situation which is scaled down in terms of time and numbers of students. Usually, this has meant a 4-20 minute lesson involving from three to ten students. The lesson is scaled down to reduce some of the complexities of the teaching act, thus allow- ing the teacher to focus on selected aspects of teaching. Frequently, one microteaching episode includes teaching a lesson and im- mediate feedback on the teacher's effective— ness. This feedback may come from videotape or audiotape recordings, supervisors, pupils, colleagues, or from the teacher's self- perceptions. Some of the variable aspects of microteaching include lesson length, number of reteaches, the amount and kind of supervision, the use of videotape or audiotape recordings, and number and types of pupils. Two points to note are the limited focus of micro— teaching, usually on one specific teaching skill, and the immediacy of feedback from a variety of sources. With re— spect to the feedback, microteaching does not require the supportive presence of a supervisor or colleague. Cooper and Allen describe the classic process as having eight steps: (1) trainees receive instruction in particular skill, (2) trainees see a videotaped or filmed model of a teacher demonstrating the skill, (3) the model is discussed until trainees are clear as to the skill they will be practicing, (4) trainees teach a short lesson to a small number of students; this is usually videotaped or audiotaped, (5) usually a supervisor helps trainees analyze 21 lesson and discuss improvements; frequently colleagues assist; sometimes student feedback is utilized, (6) trainee replans lesson, (7) trainee reteaches lesson to different group of students, (8) repeat feedback process as in step 5. They describe the basic model as having a teach—critique/reteach— critique format. Cooper and Allen give examples of the wide scale use of microteaching in a number of settings, naming the most comprehensive development of microteaching for inservice training as that conducted by the Far West Laboratory for Educational Research and Development. In addition to these descriptive studies, they also cite 10 experimental studies as evidence in support of the effectiveness of microteaching in changing and improving teaching behavior. In summarizing the research, they state: ”The feedback dimension of micro— teaching is probably the crucial one in terms of changing the trainee's behavior." Their endorsement of the efficacy of feedback echoes the remark by Tuckman quoted at the begin— ning of this Chapter: "The sine qua non of change is feed- back.” Many other studies have been done which support and many which question various aspects of the microteaching model. But these are not the concern of this investigation. The point to be made is the central importance of feedback to the process—and how widely used the process is. A secondary point is that, whereas clinical supervision and peer evaluation emphasized the importance of a support 22 system——whether supervisory or peer, microteaching found this to be optional. The mechanical device (audio- or video—tape) removed one need for human interaction, and presented the possibility of the individual confronting the objective data of his or her performance. The teacher could be assisted in the analysis of the performance by a period of training or the provision of a guidebook. The effective— ness of this last aspect of microteaching has been tested, but as discussed later in this chapter, the evidence is in— conclusive. Following from this discussion of microteaching, two other forms of giving feedback to teachers will be discussed: first, two "computer assisted" models, and then Flanders Interaction Analysis. Both of these involve ways of analy— zing teachers' behaviors and presenting feedback to teachers. E. Computer Assisted Models Two recent studies of the use of computer printouts as the source of feedback to teachers did not yield significant results (Froman and OWen, 1980; Trank, 1978). The first made use of student ratings, and the second made use Of the Student Perception of Instruction (SPOT) survey. But other ventures have been much more fruitful. As a result of computer program analysis and feedback using data collected by the Flanders Interaction Analysis teachers become more able to alter their behavior and more conscien— tious about planning (Hail, 1978). 23 A comprehensive, complex program, the Computer Assisted Teacher Training Program (CATTS) has been found effective with pre-service teachers of special education, where specific teaching skills were the aim (Semmel et al., 1976). The pro- gram is described in the System Document Manual (Semmel and Olson, 1977) which also presents extensive documentation for this syStem. CATTS is described ”as a system capable of pro— viding continuous, instantaneous, and/or delayed feedback of relevant teacher student interaction data in order to modify behavior through regulatory moves.” The component systems include CATTS stations, data flow, data collection, data analysis, feedback, storage and retrieval, and an observation system trainng subsystem. Chissom and Morris (1976) describe a system for the evaluation of student teachers employing automated data pro- cessing as an integral part of the system. It employed data gathered from four sources: public school pupils, student teachers, supervising teachers, and college supervisors. Feedback provided from the evaluation of the four sources was used to identify strengths and weaknesses of individual student teachers and cumulatively to evaluate the total stu— dent teaching program. Based on a Faculty—Course Questionnaire (FCQ) evalua— tion instrument, a computerized feedback system was developed at the University of Colorado (Whetsone, 1974). Its chief advantage is that each instructor receives detailed and 24 comprehensive information that is individually specific and at the same time allows comparisons with various norm groups within the university. Another computer assisted program makes use of a ques— tionnaire (Pohlman, 1976). The Instructional Improvement Questionnaire (IIQ) has four parts designed to collect evalu— ative feedback from students: (1) the Class Characteristics Section; (2) the Instructor Evaluation Section; (3) the Course Evaluation Section; and (4) the Optional Item Section, which consists of 60 "response positions" that the instructor may use to record student responses to locally supplied items. The answer sheets are optically scanned and responses are coded and written on magnetic tape. A computer program analy— zes the data and produces a printed report of the results. Again, it is to be noted that these computer assisted programs make it possible to bypass a human evaluator and reporter. The feedback can be transmitted without supportive human interaction. F. Flanders Interaction Analysis Flanders Interaction Analysis is a system for analy— zing teacher behavior originally designed by Flanders as a research tool, and later utilized by Amidon and Flanders (Amidon and Giamatteo, 1965; Amidon and Hunter, 1966; Amidon and.Hough, 1967; Flanders, 1965, 1968, 1970) to provide teachers and student teachers feedback about their verbal 25 behavior in the classroom. The analysis can be done on data collected by audiotape, videotape or peer observation, and can involve either self analysis or external evaluation by peer or supervisor. In itself, Flanders Verbal Interaction Analysis is not a process for giving feedback to a teacher; it is a sys— tem for analyzing teacher behavior. Both the procuring of the data and the reporting of results are not essentially determined by the Flanders analytical system. As reported in the section above, this system has even been used in com— bination with a computer program (Hail, 1978). There is a great deal of research evidence to support the validity of the Flanders system (Amidon and Flanders, 1961; Rosenhine, 1971). But the reason for including it in this discussion is the experiential one, that it is widely used as a device for giving feedback to teachers, with or without the involvement of an external human agent. To continue this section and to broaden its focus a little, two programs which make use of feedback and point to future directions in teacher education will be discussed. The first of these is concerned with the role of feedback in professional intervention; and the second is concerned with the development of specialist teachers in New York City and their use of feedback. G. Professional Intervention The work of Argyris (1970) and Argyris and Schon (1974) are concerned with the role and nature of intervention in 26 professional development and the need to respect and develop the client's professional autonomy. Many of their ideas are relevant to teacher education, and the place of feedback in the professional development of teachers. In Argyris' View, an "intervenor": ...assists a (client) system to become more effective in problem solving, decision making and decision implementation in such a way that the system can continue to be increasingly ef— fective in these activities and have decreasing need for the intervenor. The successful intervention depends on the three pri— mary tasks of the intervenor. He must ensure that the client is supplied with valid and useful information: First, it has been accepted as axiomatic that valid and useful information is the foundation for effective intervention. Valid information is that which describes the factors, plus their interrelationships, that create the problem for the client system. Next the intervenor must ensure that the client is able to make a free choice. But in order to have a free choice, the client must have a cognitive map of what he wishes to do: Free and informed choice entails what Simon has called "satisficing,” that is, selecting the alternative with the highest probability of succeeding, given some specified cost restraints. Free choice places the locus of decision making in the client system. Free choice makes it pos— sible for the clients to remain responsible for their destiny. Through free choice the Clients can maintain the autonomy of their sys— tem. But Argyris has a practial reason for insisting on free choice for the client: 27 Free choice is important because there are so many unknowns, and the interventionist wants the client to have as much willing— ness and motivation as possible to work on the problem. With high client motivation and commitment, several different methods for change can succeed. The final stage of the intervention leads to internal commit- ment on the part of the client: Internal commitment means the course of action or choice has been internalized by each member so that he experiences a high degree of ownership and has a feeling of responsibility about the choice and its implications. Internal commitment means that the individual has reached the point where he is acting on the choice because it fulfills his own needs and sense of re— sponsibility, as well as those of the sys- tem. These three tasks defined by Argyris were translated into practice in the Master of Arts in Classroom Teaching (MACT) program described by Cragun and Wilson (1980) and Bradley et a1. (1980)., first course in the MACT program called Classroom Analysis, incorporated the following three aims: (1) to help candidates collect and assess valid and useful information about their classroom behavior related to students“ needs and desired outcomes (2 V to help candidates who have constructed this "cognitive map” of their professional behavior make as free as possible decisions about courses, projects and learning experiences for the rest of their MACT program. (3) as a result of (1)and (2), to facilitate the development of a high degree of internal commitment 28 to their own professional development in the short term of the 2—year MACT program and in the long term thereafter as a lifelong goal. H. Individualized Professional Development Program Many professional development programs claim to be individualized and are based on a combination of self assess— ment and the evaluation of external feedback, frequently re— ferred to as a needs assessment. This is the basis, for ex— ample, of the needs assessment described by Craqun and Wilson, (1980) and Bradley et a1. (1980), as part of the Master of Arts in Teaching Program (MACT). Another sort of individualized professional develop— ment program was developed by the New York City Teachers Center (McDonald, 1980). The program has the following fea— tures: 1. It is an in—depth program which the teacher undertakes 2. It is developmental in character, that is, the activities of the program lead the teacher through a variety of progressive stages 3. It is built around the teacher's stated needs, so that it is not imposed by the Teacher Special— ist. 4. It is revised on the basis of evaluations and analyses by both teacher and Teacher Specialist. 29 The program is built around the stated needs of the teacher. But once these needs are stated, the teacher spec— ialist sets about collecting data from the teacher's class— room. Feedback is given in the context of detailed discussions of the data, with the goal that the teacher will make the diag— nosis. Thus, one of the important functions of the Teacher Specialist is the skillful gathering and evaluation of data. The Teacher Specialist is an innovation in this pro— gram. Each Teacher Specialist, defined as a teacher who specializes in teaching teachers, undergoes a carefully de— signed training program. The aims of the program are: 1. To provide Teacher Specialists with opportunities to gain further knowledge in curriculum and in— struction in the elementary school 2. To provide instruction for the Specialists in developing programs for individual teachers and groups of teachers at the Teacher Center sites 3. To provide training experience for the Specialists to learn to: (a) diagnose teachers' needs and problems, plan interventions for and evaluate instructional development (b) help teachers to diagnose their own needs and problems, plan their own strategies and make their own evaluations of instructional development 30 (c) plan workshops based on the stated needs of teachers and needs perceived by the Special- ists (d) select, organize and use appropriate instruc- tional materials with teachers in their Teacher Center programs. Two features stand out in this program. Specialists have to be able to help teachers to disgnose their needs, but the Specialists must also be able to diagnose these needs and give feedback to the teachers that will help them in their diagnoses. They also have to give evaluative feedback. The second feature is the supportive role of the Specialist which is fundamental to the program. Specialist intervention, based on feedback and developmental programs, is seen in this model as an essential ingredient. I. Change Agents in Education To go alittle further afield, it would be useful to look at feedback more broadly, in terms of the diffusion of knowledge and its use in schools in order to improve educational practice. Most schools apparently fail to utilize the fruits of much of the research that has been done, by not applying the knowledge or making use of the products that have been developed from it. From a synthesis of findings across five recent studies in educational dissemination and change (Emrick and Peterson, 31 1978), five major generalizations were derived. They stress once again that in order for this kind of feedback to bring about change, "directed personal intervention" and “contin— uous personal participatiOn" are essential ”to root and sus— tain utilization." In short, in this context, effective dissemination requires a support system. The five major generalizations are: 1. Meaningful change occurs as a process not an event Directed personal intervention is by far the most potent technical support resource and may be a necessary condition for many forms of utili— zation Continuous personal participation of the imple— menting staff is needed to firmly root and sus— tain utilization Administrators occupy a crucial role in supporting the utilization process Descriptive, instructional, and support materials are needed, particularly for utilization including organizational or instructional changes. J. Summary Feedback is used in a wide variety of teacher education models from clinical supervision to peer evaluation and from humanistically oriented to the mechanically or computer assisted forms of feedback fromndcroteaching and classroom observation. 32 There can be no doubt about the rich contribution feedback has made in teacher education. An interesting question is raised by the different levels of human support that characterize the models. On the one hand, clinical supervision in some forms cherishes an ideal of a helping relationship and peer supervision de— pends on the dynamic of group support and interaction. At the other extreme, some computer assisted systems have dis— placed the human support system, and microteaching does not require it. Instead, the dynamic is that of the self deter— mining individual confronting the objective data fed back from his or her own performance. IMPORTANCE OF SELF PERCEPTIONS IN TEACHER EDUCATION A. Introduction .\}7(( / Ryans and Teacher Charatteristics k Ryans (1960), in his monumental study of the char- acteristics of teachers, reveals that effective teachers differ in their perceptions of themselves, others, and their overall classroom behavior. Effective teachers, who were "high" with respect to overall classroom behavior, saw them— selves as ambitious and having initiative and were more sat- isfied with regard to their emotional adjustment. They liked other people and were willing to participate in school and college social groups. They were extremely generous in their perceptions of the behavior and motives of other people. The 33 "low" group of teachers were less satisfied with their emo— tional adjustment, and were restricted and critical in their appraisals of others. Ryans wasinterested only in describing these percep- tions and correlating them with different kinds of teachers. But his work draws attention to the importance of teachers' perceptions and self—perceptions as features that distinguish between teachers. Combs' Perceptual Psychology Combs (1965) states that the basic concept of percep- tual psychology is that the behavior of a person is the di- rect result of his field of perceptions at the moment of be— having. His behavior at any instant is the result of (1) how he sees himself, (2) how he sees the situation in which he is involved, and (3) the interrelations between the two. Thus, teacher behavior is a function of these three aspects of perception, the first of which is self perception. As Combs writes: of all the perceptions existing for an individual none are so important as those he has about himself.... It is the organ~ ization of seeing self that the modern psycho— logist calls the self concept. It represents the most important single influence affecting an individual's behavior. Just as Ryans did, Combs notes that there are certain kinds of self perceptions associated with effective teaching. Basically these are rooted in the fundamental need all people share to be adequate, and in a willingness to see others as _ +‘ 34 always motivated to be and become as adequate as they can be in all situations. Convinced of the importance of self and the self con— cept, Combs developed the idea of self as instrument. Thus, the effective teacher is "a unique human being who has learned to use himself effectively and efficiently to carry out his own and society's purposes in the education of others.” 8. History and Nature of Self Perceptions/Self Concept Webster (1974) gives a good account of the history of the self and self concept in modern times. It should be noted that ”self concept" and ”self perception” are being used interchangeably. There are two different ways in which the origin and development of the self have been viewed in sociological and psychological literature. On the one hand there is the view of the developmental self, which is roughly equivalent to a set of innate personal characteristics, or "personality," which develops by interacting with the possi- bilities inherent in one's environment. The crucial ele- mentixlthis school of thought is the "primacy of inborn needs and traits.” On the other hand, there is a view that focuses on the development of the social self. Those who share this view stress the importance of contact with others for the develop— ment of the personality, and are referred to as ”environ— mentalists.” There are at least two groups of environmen— talists: behaviorists, who stress the assumption that 35 "human behavior may be studied or analyzed as if it were determined only by specifiable external influences.” In this View, the individual is acted upon more than he acts. The second group, which this study will be most concerned with, employs the interactionist approach. The members of this group reject the idea of innate sources of behavior, and stress instead the effects of social interaction in shaping one's personality. The idea of "self concept“ plays a unique role intheir understanding of human behavior. The interactionist approach owes its origins to the writings of Cooley (1964) and Mead (1934). Cooley put for- ward the idea of the "looking—glass self," whichnmens as Webster (1974) says, that: a person’s self concept is considered to be dependent on observing the reactions and opinions of others toward the individ— ual. In other words, the personality is formed, not partially, but wholly through the experiences the individual has inter— acting with others. But the individual is not wholly at the mercy of these inter— actions. He is self aware and aware of others, and can con— trol his choices of action to some degree. Cooley also contributed the notion of the "internal- ized other." This is a mental image of others which an in- dividual develops as a result of social interaction. Mead modified this idea into the notion of the "generalized other" This concept described the organized community or social group which gives the individual his unified view of him- self, and is central to Mead's idea of the development of 36 the self. Mead emphasizes the importance of social context as a source of self concept. The individual thinks about himself in categories determined by his social groups, and probably he also applies to himself standards of comparison derived from the range of variation he sees in others. (Webster, 1974). Mead made another important contribution to the idea of the looking—glass self. He saw the self as having a semi- permanent structure, partially resistant to change-—not needing to change with every change in an individual's environment. Sullivan (1947) made a very important contribution to the development of the idea of the looking-glass self. He contributed the idea of the ”significant other," first ap— plied by him to parents, but since generalizedtx>others who are instrumental in forming the individual's self concept. Important here is the notion that not all people equally in- fluence the formation of one's self—concept. The individual judges some to be significant, and is much more influenced by them. In summary, the idea of the self that influences this investigation belongs to the tradition of the looking—glass self, developed by the school of interactionists, headed by Cooley, Mead and Sullivan. These ideas include the central perspective that one's self concept is directly dependent on the opinions and actions of others. But the individual in— terprets the significance of others in the light of what he knows of them, and is not blindly influenced. Also, the individual develops in a social context——within which he 37 operates a concept of the generalized other. His concept of himself is a semi—permanent structure, depending for change on his assessment of the significance of others and his own self image. This notion of the assessment of others is highlighted in Sullivan's concept of the significant other. C. Structurecfi Self Perceptions/Self Concept To this point, self concept and self perception have been used interchangeably. Combs, Richards and Richards (1976), however, do make a distinction. For them, the per— ceived self is the phenomenal self, and the core of this is the self concept: ...each person develops a large number of more or less discrete perceptions of self which he regards as characteristic of his being...these perceptions do not exist in the perceptual field as a s1mple enumeration of ways of seeing the self. Rather the concepts of self constitute an organization representing a person's own conception of himself in all his complexity.... This or— ganization of all the ways a person has of seeing himself We have called the phenomenal or perceived self. Combs et a1. (1976) distinguish between the phenomenal self (all perceptions of self irrespective of their significance) and the self concept (those perceptions about self which seem most vital to the person himself). Crucial Unboth of these conceptions is the notion of organization. Wylie (1961), in her survey of the research litera— ture concerned with self concept, pointed out that phenomeno— logical theorists like Combs are so called because they stress a, 38 the role of the conscious self concept in determining a person's behavior. A more elaborate structure of the self concept is de- scribed by Kash and Borich, (1978). Kash and Borich bring together both poles of self psychology, and include the "developmental self,” and the ”performing self” and the idea of the "significant" and ”salient" other in their theo— retical framework. Their description of the self concept is derived from Allport's categories of the self (1961). Self concept consists of five senses of the self: the senses of (l) bodily self (2) self identity (3) self extension (self— as—doer) (4) self esteem, and (5) self image. All component senses are interactive, interrelated and interdependent. Webster (1974) is not so interested in static pictures of how the self is constructed, but in the dynamic of how the structure of self is determined——how it is formed and maintained. D. How Self Perceptions/Self Concepts Are Formed And Maintained In discussing the history and nature of the self con— cept, the active role of the individual in reacting with significant others has been discussed in relation to the genesis of the self concept. Kash et a1. (1978) also sup— ports this point of view. Purkey (1978) also is of the opinion that no one is born with a self concept. For Purkey, the development and structure of self awareness is a "life— long research project.": 39 By experiencing the world through inter— actions with significant others, the de- veloping person develops a theory of per— sonal existence...we learn to identify ourselves both with categories (female, black, southerner, American) and with attributes (good, bad, valuable, able, unable, etc)... Gradually each person forges a self concept, complete with a complex hierarchy of attributes and cate— gories. Perhaps the main thing left in this discussion of the formation and maintenance of the self concept, is to indicate some empirical basis for these theoretical ideas. Webster (1974) cites three studies on the structure of the self concept by Miyamoto and Dornbusch (1956), Moore (1964) and Reeder, Donohue and Biblarz (1960), which agree that "self evaluation is a direct function of the (perceived) opinions of others, that such others may not be equally important to the individual, and that individuals may misperceive the others' opinions." Other studies that deal with change and stability will be discussed later in this chapter. Some further theoretical ideas about change and self perception will be discussed in the next section. B. How Self Perceptions Are Changed Combs (1965) has very interesting ideas about the ways in which changes in self perception are and are not brought about. An individual changes his self perceptions not by being told to do so; nor by the "objective analysis of self” which has been “vastly overrated as a device for 40 personality change," and can even by highly destructive. His position on perceptual change is this: Changes in behavior, including changes in one's personality, are most effectively brought about, not by introspection and analysis, but through slow changes in per— ceptions about outside events and their relation to the self. To produce a change in a person's self requires some new experience which helps him to perceive himself in a new way. Combs suggests that this sort of change can be brought about in three ways—-(l) through some direct provision of experience (2) as a consequence of perceiving an event in a new perspec— tive, and (3) through interaction following changed perception of others that is, a change in the perception of others causes them to behave in ways that change the self. For, as Combs concludes, "the self is learned from the looking glass held up for us by others.” The third way for changing self perceptions enunciated by Combs above, puts him firmly in the vanguard of those who see the need for a supportive presence, a significant other, perhaps, in the change process when feedback is given. Combs, as indicated at the beginning of this section, is skeptical of the virtues of solitary self analysis for the purpose of changing one's self perceptions. F. How Consistent Are Self Perceptions? Hamachek (1978) points out that perceptions need to be consistent for two reasons. First, there is less strain and anxiety if the social environment is not in a continual state of change. Second, consistency serves as the 41 foundation for stable human relationships. In fact, the world of social perceptions is very stable; once conceptual judgements are made, they tend to remain intact and unchanging. Hamachek then relates behavioral consistency to self concept theory, and reports that the values of “inner sameness“ has been stressed by many psychologists. For Hamachek, self concept theory strongly suggests that we will "act like" the sort of person we perceive ourselves to be, and that as we encounter new experiences in everyday life, we will tend to accept or reject them in terms of their compatability with our present concepts of ourselves. By behaving in this way, we reduce conflicts and maintain our individuality as persons. He then discusses Festinger's model of cognitive dissonance and the need for consistency. Festinger's model will be dis— cussed later in this paper. Combs et a1. (1976) support Lecky's position (1961) that the basic need of the organism is the maintenance of a unified organization. To achieve adequacy, one must develop a high degree of consistency within his phenomenal self. As did Hamachek, Combs then goes to Festinger's theory in order to explain the individual's need for consistency and the ef— fect of dissonance on his self concept. G. Festinger's Theory of Cognitive Dissonance Festinger (1957) sees a person as continually striving for cognitive consistency. His basic hypotheses are as follows: 1. The existence of dissonance, being psychologically uncomfortable, will motivate a person to try to reduce dissonance and achieve consonance. 42 2. When dissonance is present, in addition to trying to reduce it, the person will actively avoid sit— uations and information which are likely to in— crease the dissonance. The need for consistency or consonance drives the per— son to try to reduce dissonance or haavoid situations that increase it. Another of Festinger's maxims is that dissonance gives rise to pressures to eliminate the dissonance, and that the strength of these pressures to reduce the dissonance is a function of the magnitude of the dissonance. There is a limit to the amount of dissonance that can exist between any two elements, and this limit is set by the total resistance to change of the less resistant element. At the point of maximum possible dissonance, the less resistant element would change and the dissonance would be eliminated. Festinger describes three methods for reducing or eliminating dissonance stemming from social disagreement: 1. One person may change his opinion so that it corresponds more closely with one's knowledge of what the others believe. 2. One may try to get the others who disagree with him to change their opinions to conform more closely with his opinion. 3. One may attempt in some way to make the others not comparable to himself, either by attributing different characteristics, experiences or motives 43 to them, or by derogating or rejecting them out— right. One should mention that Festinger's theory rests on an abundance of research evidence. Hamachek (1978) reports that while not all the evidence is unequivocal, over 500 ex- perimental investigations do suggest that some such tendency as postulated by Festinger, does exist. In more recent times, Festinger's theory has continued to be fruitful ground for empirical research, although it has been modified and adapted to Suit particular contexts. For example, Rosenberg (1979) defines a particular adaptation of the theory in terms of ”contextual dissonance" and describes a number of empirical studies in this area. Contextual dissonance is based on the social similarity and dissimilarity of the individual to those around him, which affects his experience, and conse— quently his self concept. One of the questions being asked by this study is directly related to Festinger's theory. The question asks whether the magnitude of change in self perceptions re- sulting from assessment feedback is a function of levels of agreement between the feedback and one's original self per- ceptions. Festinger's theory would lead one to expect that this is so. H. How Stable Are Self PerceptionS/seLfConcepts Already the discussion of the literature concerning consistency suggests that self perceptions will tend to be 44 stable. Combs et a1 (1976) are quite emphatic about this. Stability, or resistance to change, is one of the character- istics of an organization. Once established, self perceptions have a high degree of stability. Combs et al. describe the perceived self as our "fundamental frame of reference, our anchor to reality," and claim that even an unsatisfactory self organization is likely to be highly stable and resis— tant to change. This stability has been demonstrated by a number of researches which they list (Balester, 1956; Bloom, 1964; Engel, 1959; Gollin, 1954; Kagan and Moss, 1962; Roth, 1959). I. Summary This section began with a discussion of the importance of self perceptions in the work of Ryans and in the percep— tual psychology of Combs. Combs is in the tradition of in— teractionist psychologists, who see the self concept, or one's self perceptions, as largely the result of social in— teraction especially with significant others. This notion of the self as a “looking—glass self“ was traced from Cooley through Mead to Sullivan. Two views of the structure of the self concept were reviewed, and then a number of its char— acteristics were discussed, including how it is formed and changed, and how consistent and stable it is thought:to be. 45 RESEARCH AND THE FOUR RESEARCH QUESTIONS l Throughout this literature review, research has been cited in support of a number of the theoretical positions discussed. In this section, the review of research studies will be focused on those studies that relate directly to the research questions. Question 1: Does either of two forms of feedback—-self analyzed and reported (SAG), or externally analyzed and reported (EAG)—- cause any change at all in a teacher's self perception of the relative levels of his or her abilities? This first question, precisely defined for the purpose of experimental investigation, can be broadly restated for the purpose of this literature review: Is there any evidence that feedback of any sort changes self perception? Theoreti— cal arguments have already been presented in the earlier sec— tions of this chapter. Webster (1974) in his review of self evaluation re— search, describes a number of experimental investigations that bear on the question. The prototype of change studies is the experiment by Israel (1956). Participants in the experiment were 107 students from four classes at a Swedish physical education college. Each class was divided into two experimental groups of 16 to 18 participants, who knew one another well. Each participant was asked (1) to rank all 46 members including himself on leadership ability (2) to es- timate the level of leadership others thought he possessed, and (3) to rank himself as he wished the others would. An attempt was then made to get individuals to change their self—ranking by reporting fictitious rankings from others. The significant findings are first, that the evidence sup— ports the basic idea of the looking—glass self, and conse- quently, the idea that the perception of others' opinions influences one's self perception. Second, the likelihood of change in one's self perceptions is directly related to the attractiveness of the group. Third, the effect of others in changing the self concept is inversely related to the in- dividual's accuracy of perception; so that, if the individ— ual did p93 accurately perceive the others' perceptions, he was more.likely‘U3change. This last finding is also rela- vant to the third question in this study to be discussed later. In a study by Baskman, Secord and Peirce (1963) college students rated themselves on personality items and told how close friends and relatives would rate them on the same items. Then they filled out personality measurement scales which were to be scored by an expert psychologist and this "objec— tive information” was fed back to them. But this ”objective information" included attempts to alter their self percep— tions on certain personality items. The significant findings are that participants do change their self ranking as a result of the manipulated reports, and that they change most 47 in those areas where they feel friends and relatives do not agree with them. Incidentally, this study also demonstrates the potency of expert evaluators in changing self perceptions. Videbeck (1960) experimented with 30 students from introductory speech classes, rated as superior by their in- structors. They rated themselves on a nine point scale on each of 24 items to do with adequacy of oral presentation. After this, they read six poems for a "visiting speech expert.” The major findings were that this procedure is effective in changing the individual's self rating. A second major find— ing has to do with the direction of the changes. These data show no greater tendency on the part of individuals to change in a positive than in a negative direction. In fact, the data suggests that those who received positive ratings changed less than those who received negative ratings. This finding is contrary to a finding by Moore (1964) who reported that individuals tend-agraise their self—perceptions more in re- sponsetx>feedback than to lower them. Maehr, Mensing and Nafzger (1962) replicated the Videbeck study using 31 boys enrolled from a high school physical education class,and confirmed Videbeck's finding except that regarding greater changes intimadisapproval condition. These studies strongly suggest that feedback does change self perceptions. Whether positive or negative feed— back is more potent is uncertain. But it does seem that the attraction of the assessing group is a strong influence, 48 as is the "expertness“ of the evaluator. This last point is more relevant to the second research question discussed below. Question 2: Which of two forms of feedback (SAG or EAG) will yield the great- est changes in a teacher's self perceptions? The evidence presented earlier, in the section dealing with the importance of feedback in teacher education, suggests a split verdict on this question. Practice and research findings from supervisory and peer evaluation seem to favor an external analysis of feedback, whereas experience and data from microteaching and computer assisted programs seem to promote self analysis of data. Two of the experiments reported in response'Uunestion one above, support the ef— ficacy of the ”expert" evaluator (Backman et al., 1963, and Videbeck, 1960). Hartman (1978) reports a number of very relevant findings from the Teacher Self Appraisal Research Project (Brooks, 1967). The major components of this project were (1) voluntary participation (2) leaders with the ability to provide an accepting, non-threatening climate (3) fre— quent videotaped feedback viewed only by the participants who were taught encoding and analytical skills, and (4) a year long inservice program of weekly meetings in which the principles of perceptual psychology were taught, in ad— dition to research findings about teaching and the analytical skills already mentioned. The assumption was that teachers 49 provided with these tools would be intrinsically motivated to improve. The findings clearly indicate that looking at videotape feedback is not enough; teachers need professional competencies in order to describe and evaluate their be- havior. Also there was evidence that group support was vital for the success of this sort of feedback, and that a long- itudinal rather than short range approach is necessary. This study is mentioned here to offer a caveat where the self assessment approach is concerned. Careful tutoring is necessary for it to be effective. This study also gives some support to the efficacy of self evaluation under these conditions, especially with the support of a peer group as part of the system. On the other hand, a study by Litwack (1974) which compared three types of feedback treatments: authority feedback, peer feedback and self feedback, using data from Flanders Interaction Analyses, suggests that participants feel more secure when they receive feedback from an author— ity figure, or peer group, than when they receive it from the videotape alone. The evidence cited in answer to this question is not' unequivocal, but does seem to weigh in favor of external analysis as a more potent force for change. 50 Question 3: Is the magnitude of change in self perceptions resulting from assess— ment feedback a function of empir- ically established levels of agree— ment (high, low) between feedback received and original self percep— tions? This question has also been anticipated by some of the research studies discussed in relation to question one. Israel's study (1956) indicated as one of its findings that if an individual had not accurately perceived others' per— ceptions of him, he was more likely to change, than if he had accurately perceived them. The degree of difference between one's original perception and the feedback received created dissonance which exerted pressure on the individual to change his self perceptions. Studies by Videbeck (1960), Moore (1964) and Maehr et.a1 (1962), though they disagree about relative potency, do agree that both positive and negative differences between self and other ratings exert pressure on the individual to change one's self perceptions. A few studies report on the effect of discrepancies between one's espoused platform of behavior and the platform one puts into use (Simon 1976), or discrepancies between one's ideal and actual behavior (Fages, 1978), or discrep— ancies between students' perceptions of their ideal and actual teacher fed back to the teacher (Gage, Runkel and Chatterjee, 1960). Two studies are concerned with the specific question of the effect of magnitude of discrepancy between self per— ceived and observed teaching behavior. Tuckman, McCall and 51 Hyman (1969) hypothesized that changes in teacher behavior, self perception, or the discrepancy between the two would be an increasing function of (a) the magnitude of the in- itial discrepancy between self perception and observed be- havior, and (b) the nature and specificity of the feedback. This hypothesis is based on Festinger's writing and research (1957) which show that the greater the dissonance between cognitive elements, the greater the pressure to change one or other of them. Twenty—four high school teachers were assigned to a High or Low Discrepancy group depending on dis- crepancies between their scores on the Self Perception Inven— tory and the scores given them by trained coders of audio- tapes of their behavior. Treatments were based on three dif— ferent types of feedback (Flanders, verbal feedback, and audiotape), and there was a control group. The findings were that the treatments had no differential effects on changes in self perception, but that initial discrepancy level did have a differential effect on changes in percep- tion. Teachers in the High Discrepancy level group changed their total perceptions across all treatments to a signifi—- cantly higher level than did Low Discrepancy teachers. Thus teachers with initially high discrepancies changed their view of their own teaching significantly to a greater extent than did teachers with low discrepancies. It was also found that initial discrepancy levels seemed to effect self perception change but not behavior change. 52 Doyle and McNally (1974) reported on the effect of intent—action discrepancy and student feedback on teacher behavior change. They surmised from the study of the Tuckman et a1. (1969) that in the absence of an external model, re— duction of discrepancy induced by feedback results in a modi- fication of personal intentions rather than teaching behavior. Their findings tended to confirm this surmise, for they con— Icluded that in the absence of an externally validated and supported model of approved behavior, verbal feedback appears to affect perceptions rather than behavior. The participants in their experiment were 36 junior high school teachers who volunteered to participate in an ex- periment in microteaching. Teachers were first introduced to the teaching task and asked to fill out an intent inventory based on the amount of time they expected to devote to various classroom behaviors during the teaching session. Students were tested on the material taught after the lab teaching session. Participants then received feedback on intent— action discrepancy and student test performance, and were given a similar teaching task to perform. The results showed that feedback concerning student learning outcomes did not have a significant impact on either the teachers' perceptions of how they would teach or how they actually taught afterwards. However, the data indicated that discrepancy conditions did have a significant effect on the amount of intent change in the areas of direct and indirect influence. These results indicate that when teachers are 53 asked to reteach the same content, they react to intent- action discrepancy by revising their intentions of how they will teach rather than by changing their actual classroom behavior. Thus, knowledge of student outcomes:m&mwsto have little impact on teaching behavior, and feedback regarding differences between teacher intention and teacher performance seems'UD lead teachers to change their perceptions but not their performance. As to the precise question of the differential ef- fects of magnitude of discrepancy, a comparison between high— discrepancy and low discrepancy participants showed that high discrepancy participants changed more on the two meas— ures of teacher intent (direct and indirect teaching be— havior). In summary, the evidence seems to indicate that mag— nitude of change in self perceptions as a result of assess— ment feedback is a function of high or low levels of agree— ment between feedback and original self perceptions. All the literature reviewed showed that dissonance is a potentmoti~ vator of change, and two studies suggested that_the greater the level of dissonance the greater the change in self perceptions. Perhaps we should note with caution a limitation reported by Glassberg (1978) in reference to Maves (1972). Glassberg writes, in the context of ego development and student teachers: Developmentalists emphasize the role of the environment in creating disequilibrium and point out at the same time that too much disequilibrium can become overwhelming re— sulting in fixation at a stage rather than progression to the next stages. ,—: _ 54 The studies reviewed here also suggest that there are three levels of response to feedback. Feedback can affect (a) self perceptions (b) intention to change, and (c) actual behavior. The relationship might be represented in the following way: Feedback ——) changes in self perceptions -> intention to change behavior -)~ actual changes in behavior. The present investigation is concerned with the relationship between feedback and self perceptions. As indicated by the studies reviewed, further research needs to be done on the relationship between feedback and intention to change behav— ior, and between feedback and actual changes in behavior. Question 4: Will the changes in self perceptions re— vealed in the posttest immediately following the experimental intervention (seven to ten day interval) differ from corresponding changes in perceptions revealed in a de— layed posttest administered approximately six weeks after the intervention? Psychologists seem to agree on the stability of self perceptions. Combs' arguments (1976) and the research he cited in favour of the stability of self perceptions have already been discussed in a previous section dealing with the importance of self perceptions in teacher education. Webster (1974) describes two studies that tested the stability of self perceptions over the exact time period (six weeks) used in this investigation. A series of studies reported by Haas and Maehr (1965) were designed to change various as— pects of self evaluation. Participants were eighth grade boys in physical education classes. The findings from the first experiment indicate that changes induced by the treatment 55 were great immediately following the evaluation, and that they persisted at the same level for the duration of the six week study. In a second experiment involving differences in the number of treatments, participants were found after the second “dosage” to have made a greater change in the pre— dicted direction, and this level of change persisted for six weeks. The conclusion is obviously that in the absence of other treatments, time alone, at least such a period of time as six weeks, will not affect self perceptions that have been changed by strong experimental treatments. Conclusion: Critical Issues Still Outstanding From Review of Research Although the research literature supports the view that feedback does influence self perceptions, a number of critical issues are left outstanding. The main question not addressed is whether self perceptions are altered by external feedback after they have been established by self assessment. This question is the main focus of the empirical section of this investigation. In the context of professional development, we need to know more about the specific attributes Of ”significant others“. Questions about the validity of their feedback need also to be pursued. What are the comparable effects and validity of feedback from ”experts”? With regard to the nature of feedback, levels of specificity, and alternative instruments and categories of 56 teacher behavior should be studied. The effects of differ— ent modes of analysis and transmission also need further com— parison. Questions about directionality are unresolved. Do positive or negative differences between feedback and self perceptions cause the greater changes in self perceptions? A number of demographic variables could influence judgement about the effect of feedback on self perceptions. These include sex, geographical location, level of school taught, age and years in teaching. They need to be investi- gated for their possible influence. In this investigation, an attempt will be made to control them. The question of whether an external support system is more effective than a self reliant one needs careful exam— ination. This question is a second major empirical focus of this investigation. An attempt will be made to distinguish between the function of the external agent as the analyzer and transmitter of feedback and other support functions some— times described as ”interaction”. Magnitude of dissonance as a motivator of change in self perception and behavior has not been sufficiently re— searched in teacher education. Is magnitude of self change always the result of magnitude of dissonance if one does not reduce dissonance by changing the opinion of others or one's opinion of them, as Festinger suggests. Perhaps it is im— portant to test also whether the expected result is the only result, and whether the ”maximum possible dissonance”, even 57 if it does cause the expected change, has harmful side ef— fects. Another important question is whether dissonance or intentions, changes only self perceptions or intentions, or also influences a teacher's classroom behavior, and if so, under what conditions. This is not a concern of the present empirical investigation, but it is a question that needs further experimental study in the context of teacher develop— ment. Finally, theory strongly indicates that self percep- tions are very stable, and the evidence cited here confirms that opinion, at least for periods up to six weeks. Whether stability is maintained beyond this time could be determined in other experiments. In this experiment, stability of any changes resulting from the treatment will be tested over the normative six week period. CHAPTER III DESIGN OF THE STUDY Introduction This investigation had two distinct stages: a pre- paratory stage that focused on instrumentation and instruc— tional methodology, and an experimental stage that focused on relations between feedback and self-perceptions. The preparatory stage represented an attempt to de- velop a process and a set of instruments for guiding teachers in their professional self assessment. These products were developed in the first of two consecutive graduate courses in a Master's Degree Program for Classroom Teachers (MACT), and were incorporated into the curriculum for that course. The main, or experimental, stage took place in the second courSe of the two term sequence. It consisted of two experiments which investigated the general question of whether externally mediated feedback is more likely to change a teacher's self perceptions than is self mediated feedback, or no feedback at all. The events which occurred in the first term had a major influence on the conduct of the two experiments. It is, there— fore, important to review these events prior to describing the experimental design. 58 59 Before undertaking this review, two reasons for the importance of the preparatory stage should be brought to mind. First, the basic question under investigation was not whether assessment feedback in general makes a difference to a teacher's self perceptions, but specifically whether it makes a differ- ence when the teacher has already undergone a careful process of professional self assessment. The kind of preparatory self assessment process the teachers in this investigation underwent is what is to be described here. In the second place, the Teacher Behavior Survey (TBS)—— the instrument used for collecting feedback data and for meas— uring changes in participants' self perceptions during the experimental investigation——was developed with the teachers as an integral part of their experience in the first term. Thus, the development of the TBS instrument is important both as a product which was used in the experiments, and as a part of the instructional process that guided the formation of the teachers' self perceptions in the first term. Description of Critical Events in the Preparatory Stage Throughout this description, then, two considerations should be kept in mind. The events were intended to contri- bute (a) to the participants' sophistication in self analysis, and (b) to the development of the Teacher Behavior Survey (TBS). As already mentioned, the preparatory stage of this investigation occurred during the first course in a two course 60 graduate level sequence for classroom teachers. Throughout this initial course, the teachers underwent an individual needs assessment of their professional strengths and deficiencies. These personal evaluations became the basis for consultation with the program adviser so that each teacher could frame his or her own individual plan for professional development within the overall resources of the program of the Master of Arts in Classroom Teaching (MACT). The investigator had a shared involvement in the de— sign and conduct of the course, drawing on a set of materials and procedures for a needs assessment developed by the pro— fessor, Dr. John Cragun, with whom he worked. The investi- gator's role was to assist in reorganizing the course and de— veloping new materials so that a more systematized and repli— cable model for conducting the needs assessment might emerge. The investigator had not previously met any of the teachers on the course. The needs assessment process which the participants experienced had six steps. These might be briefly described as follows: Step 1. Small Group Brainstorming on Teacher Abilities In the very first class session, the participants were divided into small groups and asked to discuss and list the characteristic strengths of the skilled teacher and the de— ficiencies of the unskilled teacher within each of the fol— lowing categories of teacher performance: 61 Planning Organization and Management Working with People Classroom Climate and Control Command of Subject Matter Teaching Methods Use of Audio-Visual and Other Materials Understanding Human Growth and Development of Children Understandingmgurriculumand Curriculum Development Evaluation Procedure Personal and Professional Characteristics Other The rough notes that described strengths and deficiencies were collected from the groups. They were read, and then sifted and elaborated by the investigator into a set of state— ments under each of the category headings. In this way two comprehensive lists were prepared. One entitled ”List of Teacher Abilities,” listed the strengths of the skilled teacher and the other, entitled "List of Teacher Deficiencies,” listed the deficiencies of the unskilled teacher, within each of the general categories. Step 2. Rating of Teacher Strengths and Deficiencies In the second session, the participants were asked to do two things. First, they were asked to rank from one to three the most important strengths in each category on the first list, and then to perform the same ranking of the most 62 characteristic deficiencies of the unskilled teacher in each category on the second list. Having in this way internalized some specific indi— cators of teacher behavior within a set of explicit categor- ies, the participants then began initially to examine their own performance. Step 3. Presentations on Areas of Teacher Behavior In the next four weeks, while the teachers were en— gaged in systematic self analysis of their professional be— havior, lecture,discussions and seminars were held on a num- ber of these categories of teacher behavior. The intention was to help the teachers to a deeper understanding of the particular category and to provide the opportunity for them to evaluate their current practices——their strengths and de— ficiencies—-in the light of this understanding. The presen— tations were descriptive of good practice and reviewed re- search on teaching in each area. The categories on which presentations were made were the following: Teaching Strategies Questioning Skills Verbal Interaction Analysis Helping Relationship Skills Classroom Management Planning Strategies The other categories were informally discussed from time to time during class sessions. 63 Step 4. Individual Needs Assessment Next, after they had been assisted by the presentation to evaluate their own behavior in a number of specific cate— gories, the teachers were now asked to prepare a written out- line of their needs assessment, using the lists of teacher abilities and deficiencies as a guide. They discussed these, first of all, in a small group session, and the following week they presented a full written account of their needs assess— ment to the instructor. These accounts were discussed in individual conferences. Step 5. Initial Program Planning Based on the results of the personal needs assessment, each teacher began to construct an individualized plan of courses, experiences and projects for the MACT program, with the help of the course adviser. Step 6. Individualized Mini—Projects Candidates planned and carried out a mini project of not more than three weeks duration to develop some aspects of their own classroom behavior suggested by their needs assessment. For example, a teacher might have discovered a limitation in questioning technique. The teacher would follow a set of project planning guidelines in mapping and executing a strategy for improving skills or a single skill, in this area. These guidelines required the teacher to describe his or her problem exactly, to state limited objectives as con— cretely as possible, to set out specific activities to be done 64 in the few weeks available and, finally to specify how pro— gress was to be evaluated.1 The intent of this summary of the six step needs assess— ment process has been to describe how the preparatory stage contributed to: a. the participants' sophistication in self analysis and the firm foundation of their perceptions at this point, and b. the development of the Teacher Behavior Survey (TBS). With respect to (a), it should be stressed that throughout the term, and particularly as a direct result of undergoing the first four steps of the process, the partici— pants had been concentrating on a specific set of categories of teacher behavior and had been examining their own behavior in the light of their understanding of these categories. By the end of the term, the teachers should have had quite spec— ific insights into their own behavior in each of these spec— ified areas. With respect to (b), it should be noted that the first of the six steps resulted in the development of a ”List of Teacher Abilities” and a ”List of Teacher Deficiencies.“ The two lists were similar in overall structure: both contained about 12 statements in each of the 11 categories of teacher behavior (for example: Planning, Organization and Management, 1This brief summary of the six step needs assessment process may prove inadequate for those who might wish to rep— licate the investigation. A more detailed account is provided in Cragun and Wilson (1980). 65 or Working with People). In step two, participants were asked to rate the dozen or so statements in each category on a scale of one to three in order of importance. From these weighted lists, the investigator selected the items with the highest ratings, and used them in constructing the two sections of the Teacher Behavior Survey (TBS). General Overview of the Design of the Experimental Investigation A. Introduction As the preceeding discussion suggests, the initial course in a two term graduate sequence is based on the as— sumption that teachers can gain detailed insights into their professional behavior quite economically and effectively through a process of self assessment. Evidence collected by others suggests that this assumption may be warranted (De Marte,Kelly, and Freeman, 1980). However, this evidence does not speak to the stability of self perceptions that are formed. This investigation therefore, focused on the general question of whether or not self perceptions formed as a result of a sys— tematic self assessment process will remain relatively un- changed when teachers receive feedback from others regarding their classroom performance. There are two distinct ways in which this feedback might be received after the data has been collected. Feedback data might be given directly to the in— dividual to be self—analyzed, or it might be externally analy- zed and then reported to the individual. Thus, the two experiments that represent the main focus of this investigation 66 also addressed the more specific question of whether feedback that is mediated by an external evaluator will alter a teacher's preformed self perceptions to a greater extent than is true for feedback in which the data is processed by the teacher himself or herself. B. Summary of Experimental Design The two experiments may be schematically summarized as follows: (R)0X00 Experiment I l 2 2 3 O1 O2 03 Experiment II 0 = Pretest (Teacher Behavior Survey, April 16-25, 1 1979) 02 = Posttest (Teacher Behavior Survey, May 28—June 6, 1979) 03 = Delayed Posttest (Teacher Behavior Survey, July 10—17, 1979) X1 = Data self analyzed and reported X2 = Data externally analyzed and reported This outline makes use of the Campbell and Stanley notation (1963), and represents the experimental design over time. The steps in conducting this experimental investiga- tion are summarized in the "Calendar of Events” presented in Appendix A. In Experiment I there is random assignment (R) 67 of pairs of participants selected from a total sub-population, two experimental treatments (X and X2), and the repeated use 1 of the same measure on three separate occasions as pretest, posttest and delayed posttest (01, O2 and 03). The design of Experiment II is related to that used in Experiment I in two ways. Both experiments took place over the identical time intervals, and used the same test measure repeatedly. The participants in Experiment II, however, were not randomly assigned, as in Experiment I. They also received no experi— mental treatment, as they were being considered in the role of a quasi control group. C. Narrative Outline of the Investigation Experiment I This experiment took place in the Spring term of 1979, and involved the MACT candidates who had taken the course already described as the preparatory stage to this experi- ment. They were now doing the second course in the profes— sional development sequence of the MACT program. Both these courses were offered in two middle sized cities in mid— Michigan, hereafter referred to as City A and City B. The teachers in these two classes were sorted into fourteen matched pairs according to geographical location (City A or City B), sex, and level of school (elementary or post elementary). Members of each pair were then randomly assigned to one or other of the two experimental groups. The first experimental group, the self assessed group (SAG), was 68 told that its members would be given evaluation data con- cerned with their teaching performance, which they would have to process and report on. The second experimental group, the externally assessed group (EAG), was told that its members would have the data processed for and reported to them. The members of both classes who could not be paired were dropped from the sample. The instrument used for pretest, posttest and delayed posttest, and for collecting the data, was a questionnaire called “Teacher Behavior Survey“ (TBS) with two sections: I. Questionnaire on Teacher's Strengths II. Questionnaire on Teacher's Possible Areas for Improvement. A copy of the TBS questionnaire is to be found in Appendix B and will be discussed later. It should, however, be noted that this questionnaire was developed from the two lists (”List of Teacher Abilities” and "List of Teacher Deficien— cies") which had come out of the small group brainstorming session in the previous term. The items on these lists had also been rated for each category of teacher behavior by these same teachers. The pretest was given to all the teachers at both 10— cations. Approximately six weeks later the posttest was given, and six weeks aftertfifltsthe delayed posttest was given. An attitude scale developed for this purpose was also distri- buted for the teachers to complete after the delayed post— test (see Appendix C). 69 During the period between pretest and posttest, the experimental treatments were done. Members of both experi- mental groups selected two peers and an administrator from their schools to observe them and provide assessment data on their performance. This the assessors were to do by filling out and returning the questionnaire to the investigator. A letter containing instructions for the assessors was attached to each questionnaire (see Appendix D). The assessors were also asked to complete and return a brief questionnaire giving information on how they performed as assessors (See Appendix E). The next stage involved two separate operations. First, the three assessment questionnaires for the teachers in the first experimental group (SAG) were copied out by hand (to avoid identification of evaluators by their handwriting), and these transcribed copies were given to the teachers to be analyzed and reported on according to a set of instructions read by the investigator (See Appendix G). For the second experimental group (EAG), the quesionnaires were processed and a report prepared by the investigator for each member of the group. (Appendix H provides a copy of the structured out— line the investigator followed in preparing reports for the members of the EAG group). Next, the members of both groups had individual in— terviews with the investigators. Those in the SAG group brought to the interview the written report on the feedback they received, and made an oral presentation. The 7O investigator's role was that of a carefully interested lis— tener who responded only by requests for clarification. The members of this group then wrote reports on the feedback they had received, following the same guidelines that were given to the SAG group. Approximately one week later, all teachers were re— quired to take the posttest, and six weeks later, they com- pleted the delayed posttest and the attitude scale. Experiment II Experiment II set out to investigate whether teachers who had previously experienced a self assessment process like that undergone by the teachers in Experiment I, but who did not receive any further treatment by way of external assessment feedback, would differ in the pattern of change or lack of change in professional self perceptions over the same period of time as in Experiment 1. Teachers in the second year of the MACT program who attended the City A center were the participants (EIIG) in this experiment. There was no second year program in City B. The City A second year candidates had done their needs assessment approximately a year previously under the direction of the professor with whom the investigator had been associa- ted in the preparatory stage of the investigation, and they had continued to take courses in the MACT program. First, the nature of the experiment being done with the first year MACT candidates was explained to these second year candidates. They were then told the purpose of their 71 involvement in Experiment II. Subsequently, they were treated as a control group, in that they received only the tests which the participants in Experiment I had received. They were given the pretest, posttest and delayed posttest during the same periods that were established for the first experiment. Description of Experimental Treatments A. Introduction The experimental treatments involved three operations: 1. Data Collection, which was identical for both of the groups in Experiment I (SAG and EAG). 2. Data Analysis,which was done py the members of the SAG, but was done for the members of the EAG by the investigator. 3. Reporting, which, for the SAG involved writing a report before the Interview, and making an oral presentation at the Interview. For the EAG, an oral presentation was made by the investigator to each member of the group, who then wrote a re— port after the Interview. A major concern where the data analysis and reporting were concerned was for the reliability of the experimental procedures. So, great care was taken to ensure that identical procedures were carried out within and between groups, where these were required for the integrity of the experiment. 72 B. Data Collection At the first class meeting of the Spring term, the members of the SAG and EAG groups were each given three copies of the Teacher Behavior Survey (TBS), and three large stamped, addressed envelopes, and asked to give these to two of their peers and one administrator at their schools. At- tached to each questionnaire was a letter to the assessor (See Appendix D), and a single sheet, headed "Description of Evaluator“ (See Appendix E). Participants were to choose the people who they felt confident would do the job conscientious— ly and give useful feedback on their teaching performance. The members of these experimental groups were themselves fam— iliar with the TBS since they had contributed to its develop— ment in the preparatory stage of the experiment. The letter attached to the questionnaire gave instruc— tions to the assessors as to the minimum required for doing the job properly. They were asked to observe the teacher for about one hour or one full class period. They were to look at the teacher's lessongfljulbook, teacher records and reports, teacher prepared materials, samples of students' work, etc. and they were to have informal talks with the teacher. They were also told to draw on their overall knowledge of the teacher over the period they had known him or her. Finally, they were given instructions about preserving their anonymity. The single sheet, entitled "Description of Evaluator,” was intended to gather descriptive data on the assessors, but also to provide a check on how they carried out their functions. 73 C. Data Analysis Self Assessed Group (SAG) When the three completed TBS forms were returned by the assessors for this group, the investigator copied them out by hand onto blank TBS forms. The summaries at the end of each section were also typed by him. These copies were given to each participant in the SAG group, and members of the group in City A and City B were given identical oral in— structions by the investigator. To ensure this, the instruc— tions had been written out beforehand, and virtually memor- ized by the investigator (See Appendix G). The instructions stressed that there was no unique way to analyze the data; so, SAG members should make the most careful and meaningful analysis of the data, keeping in mind that what they wanted from the data was help in making de— cisions about their future development as teachers. When the analysis was completed, they were to write a summary/ outline to be used as the basis of their oral report at the Interview with the investigator. Their summary should ade— quately cover the following points: 1. Show how they went about processing the data. 2. Give a summary of their findings—~what the main messages were. 3. Give an account of what they found that should be most helpful to them from the point of view of their professional development. 74 4. Explain what there was about the evaluation and feedback process that they found helpful, not helpful, or needing improvement. Externally Assessed Group (EAG) The investigator was responsible for analyzing the data for this group. The results of the three TBS surveys for each teacher were transcribed onto one blank TBS form, using three different colored pens——partly so as to be able to distinguish between the assessors, but more importantly, so that the researcher could quickly scan and read the re— sults to each participant in the later Interview. Then, the sectional summaries were retyped. The next step was to calculate the mean ratings of the subscales for each teacher, so that the top three strengths (or four, in the case of a tie), the middle strengths, and the bottom three (or four) strengths could be discerned and listed. The subscales were analyzed in terms of relative strengths, because a study of the questionnaires indicated that this was how the assessors rated the teachers. It was a relatively rare thing for an assessor to rate even a single item on the first section of the questionnaire as “Below Average”. On the second section of the TBS, which dealt with possible areas for improvement, assessors suggested very few areas in which the teachers seemed to them to need improve- ment. Where the assessors did indicate such areas, the in— vestigator reported the areas in the Interview. 75 Further analysis of the data had to be done in pre— paration for the presentation required in the Interview. The investigator had to be able to give a comprehensive review of the data to each participant on the spot in a live interview situation, and he had to do it in a way that was identical, or at least uniform, for all subjects. Reliability was a very important consideration here. How the investigator pre- pared himself and the data for this presentation will be dis— cussed in the next section. D. Reporting Self Assessed Group I (SAG) The teachers in this group had been told to prepare a written report which they should use as the basis of their oral presentation. When they came to the Interview, they were told that they should talk through their written reports feeling free to expand, explain, raise new points, and other— wise alter the written version in any way they wished. The problem for the investigator was to respond in such a way that he was perceived to be interested and en— couraging, and yet behave in such a manner that could be uni- formly replicated for each teacher. Simplicity seemed the best solution. With respect to verbal'interventions, he re— stricted these to requests for clarification. But overall, the main device was to explain to each teacher at the be— ginning of the Interview that in order to keep his attention 76 focused and also more easily remember afterwards the points the teacher raised, the investigator wished to take full notes. No one seemed put off by this request, nor did in seem to bother anyone in practice. The investigator at— tempted, then, to present to each teacher a uniform picture of absorbed attention as he sedulously took notes and oc- casionally interrupted to ask for clarification of a parti— cular point. At the end of the session, the teachers handed in their reports, were given a copy of the posttest to be filled out and handed in by the end of term, and were warmly thanked. Externally Assessed Group (BAG) (a) Investigator's Preparation It was of the greatest consequence in terms of re- liability, that the reporting to the individuals in this group should be carefully uniform. The first step was to make decisions about what form the data was to be presented. The next step was to develop an almost ritual format which the investigator could thoroughly practice before the Inter— view. The following decisions were made: 1. The teachers should be given the re-typed sum— maries of their strengths and needs for improve— ment. 2. They should know the order in which their abil— ities were rated, that is, those categories that were rated top, middle, or bottom on a scale of mean ratings. 77 3. Then they should be given an account of the ratings for each item in each of the categories in descending order. Categories and now, in turn, items within the categories were to be presented in descending order. 4. Any areas needing improvement suggested by the assessors should be reported. 5. They should be given a preliminary example of the procedure as part of the format. 6. Summaries should be given at the end of each section of the presentation. Following on these decisions, a detailed scenario was put to- gether, which incorporated the decision points. There fol— flowed a process of trying out the format in simulated In- terview situations and modifying the script until the in— vestigator felt that it was most likely to enable him to do as completely as possible what he wanted to do in the same manner for all teachers in this group. A copy of the final version of the script, which was used in the Interview, entitled ”Researcher's Schedule for Reporting to Group II (EAG),” is to be found in Appendix H. After a rehersal, in which he went completely through a few of the questionnaires, the investigator was ready. (b) Interviews After the candidates had been greeted, they were given a copy of the re—typed summaries which had appeared at the end of both sections of their assessors' questionnaires. 78 After a few minutes, when they seemed to have finished reading, they were given a blank TBS form on which to make notes. Then the interview scrupulously followed the pre— pared script, as described above. The investigator departed from the script only to answer requests for clarification. At the end of the reports, the investigator reminded the teachers that they had to write a summary/outline on the feedback they had received, which was to be turned in by the end of term. The report should cover the same four areas prescribed in the "Instructions to Group I" already cited under the description of data analysis for the SAG (see also Appendix G). They were then given a copy of the TBS for the posttest which was to be completed after they had written the report, and were asked to hand it in at the last class for the term. Then they were sincerely thanked. Participants in the Investigation A. Population The target population which this study has in mind is made up of teachers in any inservice training program that features a process of professional assessment for the purpose of professional development. The defined sub-populations which the investigation dealt with were, for Experiment I, the first year MACT candidates in two middle sized mid- Michigan cities (City A and City B), and, for Experiment II, second year MACT candidates in City A. 79 The fact that the defined sub—populations of both experiments consisted of teachers who had participated in a course that focused on self—analysis prior to the initia- tion of the treatments may severely restrict the general— izability of the findings. This issue is, therefore, dis— cussed in a later section of this chapter that describes the external validity of the study. b. Sample — Selection and Assignment For Experiment I, the sample was selected from 36 teachers enrolled in the first year of the MACT program in City A and City B in the Winter and Spring terms of 1979. The selection of participants and their assignment to the two experimental treatments went as follows. The names of the 16 teachers from City B and the 20 from City A were written on slips of paper and sorted into eight groups: City A City B Male post elementary Group 1 Group 2 Male elementary Group 3 Group 4 Female post elementary Group 5 Group 6 Female elementary Group 7 Group 8 Two containers were prepared, one labelled City A, the other City B. The slips from Group 1 were crumpled into balls and put in the City A container, and the slips from 80 Group 2 crumpled and put in the City B container. A flip of a coin decided which group would be selected first. Then one slip was blindly selected for the first experimental group, the Self Assessed Group (SAG), from the City A con— tainer, as the flip of the coin directed, and placed in a box marked SAG. A second slip was taken from the same container and put in the box marked EAG (Externally Assessed Group). These were matched by selecting two names from the City B container in the same manner. When the matched pairs in these two categories were exhausted, slips from Groups 3 and 4 were placed in the City A and City B containers. The pair- ing and matching process continued in the same way until all the possible matched pairs were assigned. Fourteen pairs were assigned-—seven pairs from City A matched by seven pairs from City B. Eight teachers, who could not be paired on these variables, were not included in the sample. It should be noted that in an attempt to control for certain potentially important confounding variables, the par— ticipants were blocked on sex, geographical location and level of school taught before the random assignment of pairs began. Late in the experiment, at the very end of the Spring term, one of the female elementary school teachers from the EAG group became tragically ill. During the Summer another female elementary teacher from the SAG group, from the same geographical location, had a prolonged illness and was unable to complete the delayed posttest. Both these teachers had to be dropped from the experiment. Nevertheless, the matching 81 of the groups was not seriously affected, as both subjects were female, elementary teachers, from the same geographical area, belonging to different experimental groups. Where it did affect balance, however, was that now each group had seven teachers from one geographical area and only six from the other. The experimental groups, therefore, continued to be identical in terms of the demographic criteria being used, but one variable (location) was not equally distri— buted in both. Finally, the total sample contained 26 participants, 13 in each of the two experimental groups. The demographic composition of each group was as follows: Two male post elementary teachers, one from each of the two locations. Two female post elementary teachers, one from each of the two locations. Five female elementary school teachers from one location Four female elementary school teaches from the other location. For Experiment II, the total group of 17 teachers in the second year MACT program in City A was initially in- cluded in the study. Eventually, two teachers who did not hand in one of the tests, were excluded from the sample. The final group of participants was made up as follows: Two male post elementary school teachers one male elementary school teacher 12 female elementary school teachers 82 A summary of the demographic characteristics of the two groups in Experiment I (SAG and EAG) and the group in Experiment II (EIIG) is presented in Table l. A description of the individual participants is included in Appendix I. As mentioned earlier, the defined sub—population from which the SAG and EAG were selected had been blocked on lo— cation, sex, and level of school. It had not been blocked on age and years taught for two reasons. First, the rel— atively small number of participants limited the amount of blocking that could be done. Second, it was felt that since age, and to a lesser extent, years taught, were more freely distributed among participantstlmniwere the blocked variables, these would be less vulnerable in a process of random assign— ment. However, as a study of Table 1 shows, leaving the assignment of these two variables to chance resulted in their uneven distribution among the SAG and EAG groups. The con- sequences in terms of internal validity will be discussed later in this chapter. Instrumentation A. Measures Taken In all, eight measures were taken. Six of these——the pretest, posttest, delayed posttest and feedback data col— lection by three assessors——made use of the same instrument, the questionnaire or Teacher Behavior Survey (TBS). The seventh measure was based on response to a one page 83 m H m N e m a H m N N mH NH ._ mHmH oHHm HH N N H a a m N N H m a a HH e a mH 04m HH 0 H N N N H N H m a a a HH e a NH omq ®>Onm Ucm om H m ®>0Qm Ucm ow n m 0H-mH n 4 am : mm n a «HIOH u N em . om n N m u m u N mN . mN n N a I o u H 4N . ON u H p£©5me mHmmw mm< HH oze H mezmszmaxm zH maaomo mo moHemHmmeoememo onaemoozmo ”H mamas 84 questionnaire entitled "Description of Evaluator." The eighth was based on an attitude scale entitled "Debriefing for Participants in the Two-Term MACT Professional Evalua- tion," and measured the teachers' attitudes to various ele— ments in the two stage evaluation process. The administra— tion and timing of the eight measures have been described earlier and are summarized in Appendix A. The number of respondents and rate of return was very high for all measures. Only two of the participants in Ex- periment II failed to return one of the tests and were drop— ped from the sample before the analysis of the data was un— dertaken. In Experiment I, only two female elementary school teachers from City B (one from the SAG and one from the EAG) did not participate in the posttest, due to serious illness in both cases. They were also omitted from the analysis. B. Instrumentation Characteristics Teacher Behavior Survey (a) Description The TBS is an original questionnaire developed by the investigator. It was derived from the two lists of statements describing general categories of teacher behavior ("List of Teacher Abilities,” and ”List of Teacher Deficiencies”) that were developed with the teachers in the preparatory stage of this experimental investigation. After the statements on the two lists had been rated by the teachers, the investi- gator took the five or so statements with the highest ratings 85 in each category and converted them into the items for the questionnaire. It should be noted that the two lists of statements describing teacher behavior were produced by the teachers in the first sessions of the initial course in the MACT pro— fessional sequence. Under these conditions the lists, and the TBS instrument based on them, were expressions of the teachers' entry level value systems rather than those shaped by participation in the MACT program. Thus, the value sys— tem implicit in the instrument should also be representative of the school systems from which the teachers came, rather than those of the instructors in the MACT program. When the assessors filled out the TBS instrument, they should then have found its implicit value system fairly congruent with their beliefs. The final version of the TBS contained 114 items, divided into two sections: Section I: Questionnaire on Teachers Strengths (58 items). Section II: Questionnaire on Teacher's Possible Areas for Improvement (56 items). Items in each section were organized in 11 categories of teacher behavior: Planning Organization and Management Classroom Climate and Control 86 Command of Subject Matter Teaching Method Use of Audio—Visual and Other Materials Understanding Human Growth and Development Understanding Curriculum and Curriculum Development Evaluation Procedure Working with People Personal and Professional Characteristics Responses to each item were on a five point scale: For Section I: 5 = Exceptional (This teacher is in the top 5% of teachersaatthis level or in this subject.) Strong (This teacher is in the top 15% of teachers at this level or in this subject). Above average Below average I have had no opportunity to observe or know this Section II: Strongly agree Agree Disagree Strongly Disagree I have had no opportunity to observe or know this At the end of Section I, respondents were asked to list in rank order the five most significant strengths of the teacher being assessed. At the end of Section II, they were asked to list, again in rank order, five specific areas 87 in which they thought the teacher might most need to improve. In both cases, they were requested to be as specific as pos- sible and not to restrict themselves to the categories or items on the TBS. A complete copy of the TBS questionnaire is provided in Appendix B. (b) Purpose The TBS was used for two purposes: (1) Self Assessment: On three different occasions, about six weeks apart, teachers filled out the TBS. The ad— ministrations represent the pretest, posttest and delayed posttest of the teachers' self perceptions. (ii) Feedback Data: Each teacher asked two peers and one administrator in his or her school to complete the TBS. Differences in how these data were interpreted and communica— ted to the teacher represent the two treatments in the experi— mental study. (c) Steps in Refining the TBS Two significant changes are made in the TBS instru- ment as a result of statistical analyses that focused on its 88 measurement properties. The most radical change prompted by this analysis was the deletion of the second section, "Questionnaire on Teachers' Possible Areas for Improvement." This section was derived from the "List of Teacher Defic- iencies" developed with participants in the preparatory stage of the experiment. Whereas the first section of the TBS required that the teacher be rated in positive terms, the second section required a negative rating of performance based on levels of deficiencies or ”possible areas for im— provement.” Furthermore, most of the items in the second section were negative versions of corresponding items in the first section. Thus, the participants themselves com- plained that the TBS was too long and repititious, especially since they had to complete it on three occasions as pretest, posttest, and delayed posttest. But more fundamental than this complaint about the tediousness of the task of completing the TBS was the lack of variance of response to the second section among the as— sessors. A few ignored the section altogether; but most simply rated all the items either (3) or (2)——indicating that they either disagreed or strongly disagreed that the teachers needed improvement in any of the suggested areas. A few wrote comments as to the confusion or resentment they felt as a result of the negative tenor of this section. It was decided, therefore, to drop this section from further consideration. Unfortunately, time had not allowed for a pretest of the TBS survey, and the author was genuinely 89 surprised by the assessors' response to the second section. Their reaction was not typical of the author's experience in his native country. The second major change in the TBS instrument was the decision to combine the "Organization and Management" and "Classroom Climate and Control" subscales into a single subscale labeled “Management”. This decision was based on the following considerations: (i) The interscale correlation for "organization" and ”climate” was higher than that for any other subscale pairing. (ii) The internal consistency (alpha coefficients) of these two subscales was relatively low. Cre— ating a single, longer subscale improved this reliability (See Table 2). (iii) The more general category, "Management,” suggested by the combination of the two subscales was con— ceptually meaningful. (d) Reliability Based on the responses from the 41 participants in both experiments, reliability analyses were done for the total scale and for each of the 10 subscales. Table 2 provides a summary of the internal consistency of the ten subscales. As these figures suggest, the subscale reliabilities were com- paratively high. As might be expected, the reliability of the total scale was even higher (coefficient alpha = .97). 90 TABLE 2: RELIABILITY ANALYSIS FOR INDIVIDUAL SCALES Subscale Coefficient Alpha Planning .77 Organization .74 Climate .76 gggggigggion/Climate} Combined .88 Subject Matter .81 Teaching Method .82 Use of A.V. and Other Materials .95 Understanding Human Growth .79 Understanding Curriculum Development .77 Evaluation Procedure .80 Working With People .92 Personal and Professional Characteristics .92 91 Item level analyses suggested that this figure would not be altered by the omission of any one item on the questionnaire (See Table in Appendix J). The item level analysis also re— vealed that item - total score correlations were consistently high across all items on the questionnaire. These correl— ations were greater than .60 for 37 of the 58 items and fell below .35 for only three items on the scale (6, 9 and 34). As a result of these analyses, item 16, which had an unusually low item - total correlation and a strong negative influence on the alpha level for the subscale to which it was assigned, was dropped from the scale. (e) Inter-Subscale Correlations As a final step in the statistical analysis of the TBS instrument, inter — subscale correlations were determined. These Pearson Product Moment correlations are summaries in the correlation matrix portrayed in Table 3. As the figures in Table 3 suggest, the inter—subscale correlations were typically quite high. There are at least two ways in which these data might be interpreted. Some would argue that the TBS instrument should be interpreted as a sin- gle global measure of teaching performance rather than as 10 independent sub—scales that measure distinct aspects of per- formance. Others would argue that high inter-subscale correla- tions do not necessarily suggest that all subscales are meas— uring essentially the same general phenomenon. In other words, ON. NO. mm. He. ON. Ne. ON. ON. HO. NO. mHmum HwHoe OO.H 4O. ms. O4. O». NO. mO. mm. mm. mm. OH OO.H as. mm. me. NO. HO. Om. Om. as. O OO.H OO. NO. mm. OO. NO. mm. OO. O OO.H mm. mm. me. as. ea. Nm. a OO.H am. ON. as. me. me. O OO.H He. mm. HO. mm. m OO.H me. OO. me. a OO.H O». HO. O OO.H NO. N OO.H H % OH O O A O m a m N H mmHmumnsm mmHmOmQDW moaumflumpomnmcu Hmcoflmmmmoum\amcowpmm H OH mHmHHmpmz HmCOHuUSHpmcH m mHaomm :HHB ocHxHoz n O eonumz mcHnomme a :OHHmsHm>m n O Hmuumz Humansm m DG®EQOHo>oQ ESHSUHHHSU n h upwemomcmz N npzouw amenm n m DuaccmHm H mmHMUmQDW AmezmHonmmoo onewhich they felt specific activities in the two term evaluation process influenced their perceptions of relative strengths and short— comings as a teacher. The next set of 17 items asked re— spondents to indicate their level of agreement with statements that described favorable or unfavorable attitudes toward specific or general features of the evaluative process. The third set of five items asked teachers to describe their re— sponse to the assessors and the feedback they received. The final set of five items invited respondents to complete a set of open—ended statements that provided an opportunity to elab— orate on their feelings about the program. Participants were asked to type or print their comments so as to preserve anonymity. This procedure was designed to encourage honest answers even when these were likely to be unfavorable. 97 A copy of this debriefing form is in Appendix C. Dependent Variables Analyses of responses mathe TBS questionnaire pro— vide measures of each of the following dependent variables: (a) Feedback Differential, (b) Change Scores I, (c) Change Scores II, and (d) Change Scores Differnetial. A. Feedback Differential Feedback differential is the difference between a teacher's original self perceptions and the perceptions of his or her three assessors. The teacher's original self per— ceptions were measured by the pretest. The perceptions of the assessors were taken as the average of their ratings. Feedback differential was, therefore, the difference between pretest and the mean feedback scores. In this investigation, feedback differential is analy— zed at two levels—-at the subscale and at the total scale levels. For each subscale, feedback differential is the dif- ference between the means of the pretest and the feedback scores. For the total scale, it is the difference between the weighted means of pretest and feedback scores. If feedback differential is large and positive, it suggests that assessors' ratings were considerably higher than the teacher's self ratings. Negative feedback differ- ential scores occurred when assessors' ratings were lower than the teacher's self rating. 98 B. Change Scores I A change score in this investigation is a measure of the degree of change in a teacher's self perceptions. Change Scores I refers to the difference between the teacher's ori— ginal self perceptions as measured by the pretest and his or her self perceptions six weeks later, as measured by the posttest. The means of subscales and the total scale for the pretest were subtracted from the corresponding subscale and total scale means for the pretest. These constituted Change Scores I. They represented an attempt to measure any changes that might have occurred in the teacher's self perceptions over the specified period. These changes could have been caused by the treatment received by the partici— pants in Experiment I, or by potential confounding variables, which were the only conditions affecting the participants in Experiment II. Large change scores suggest that a teacher may have altered his or her self perceptions to a consider— able extent. Small change scores suggest that the teacher's self perceptions may have remained unchanged or been only slightly altered. C. Change Scores II The second set of change scores attempted to measure the changes in teachers' self perceptions which might have occurred over the longer period of time from pretest to the delayed posttest, which was administered six weeks after the 99 posttest. These scores were derived in exactly the same way as for Change Score I, except that mean scores on the delayed posttest were substituted for mean posttest scores. D. Change Scores Differential One of the purposes of the investigation was to test the stability of any changes in self perceptions that took place during the time of the investigation as measured at six week intervals. Change Scores I measured such changes as might have taken place six weeks after the teachers' self perceptions were originally measured by the pretest. Change Scores II measured the changes six weeks later. Change Score Differential was the difference in corresponding mean scores at the subscale and total scale levels between Change Scores I and II. It should represent those changes in self percep— tions that took place between posttest and delayed posttest, and so, should give a measure of the relative stability or instability of any changes in self perceptions that occurred as a result of treatment or non treatment conditions. A large change score differential suggests a great deal of in— stability in a teacher's post—treatment self-perceptions, whereas a small change score differential suggests a greater degree of stability. E. Computational Formulas The computational formulas that guided the derivation of the dependent variables may be briefly summarized as follows: 100 (1) Feedback Differential; (a) For each scale: (b) For total scale: (2) Change Scores I (a) For each scae: (b) For total scale: (3) Change Scores II Subtract subscale for pretest from corresponding subscale mean for feedback. Subtract weighted total scale mean for pretest from weighted total scale mean for feedback. To get the total scale mean, multiply the mean score for each category by the number of responses in the category, Edd them, and divide by the total number of responses. Subtract subscale mean for pretest from corresponding subscale mean for posttest. Subtract weighted total scale mean for pretest from weighted total scale mean for posttest. Same as for Change Scores I, substituting delayed posttest means for posttest means. (4) Change Score Differential (a) For each scale: Subtract subscale score for Change Scores I from corres- ponding subscale score for Change Scores II. 101 (b) For total scale: Subtract total scale score for Change Scores I from total scale score for Change Scores 11- Internal and External Validity of the Study The following discussion of the validity of the design of the study is guided by the arguments of Porter (1980), and Campbell and Stanley (1963). According to these authors, the quality of an experimental design should be judged in terms of internal validity, precision and external validity. A. Internal Validity and Precision Porter defines internal validity in the following way: If the independent variable is the only rea- sonable explanation of the differences in the dependent variable, the study is said Unhave internal validity. Confounding variables, which are alternative explanations of the differences in dependent variables, are the main threats to internal validity. The best way to control confounding variables at the beginning of an experiment is to randomly assign the subjects from the initial pool to the experimental treatments. Porter cautions, however, that: The utility of random assignment for con— trolling variables must, of course, be tempered by the realization that the pro— cess is based on chance, and by chance alone, experimental groups will differ at least to some extent. 102 He goes on to point out that the smaller the pool of subjects initially identified for the experiment, the greater the chance that there might be "worrisome differences between experimental groups at the outset of the experiment," when random assignment only is used. To lessen this possibility, the largest pool of subjects should be sought. About precision, Porter writes: One way to think about precision is that the more precise an experiment, the less likely it is that the experiment will yield large chance differences between experimental groups. One way to improve precision is, again, “to increase the number of subjects assigned to each experimental condition." And, as pointed out above ...The larger the number of subjects to be randomly assigned, the less likely it is that random assignment will result in unusual groups of subjects. With respect to these two criteria of internal validity and precision, this study would seem to be caught in double jeopardy. The small number of subjects in the experimental groups in Experiment I (13 each) and in Experiment II (15) would seem to severely threaten the internal validity and precision of the investigation. Porter, however, does sug- gest a way out of this double predicament. Using the example of student aptitude as a variable in a particular experiment, he writes: 103 If you want to make sure that not all of the most able students are assigned to a single experimental condition, the most straightforward method is to first group students according to aptitude. Then for each aptitude group of students, randomly assign equal numbers to each experimental condition. This insures that each experi- mental condition has subjects with similar aptitude levels. The procedures is called blocking and is one way to improve precision. So, in designing the first experiment in this study, recognizing the problems in having a small number of sub- jects, the investigator decided to block the subjects on three variables: sex, location and school level taught. Then having done this, from each block he randomly assigned matched pairs of subjects to the experimental groups, as described earlier in this chapter. By this device, he hoped to avoid the "worrisome differences” between experimental groups on these three variables. It was also hoped in this way to increase precision in the experiment. There are at least two other variables that may con- found the results: age and years of teaching. However, prac— tical limitations in the ability to form matched pairs on several variables as well as anomalies in the distribution of participants across these variables prevented the investiga- tor from blocking on age and teaching experience. Thus, worrisome differences in age and teaching exper— ience did exist at the outset of Experiment I. These differ— ences are summarized in Table 4. As the data in this table suggest, members of the SAG group were younger and had fewer years of teaching experience than was true for the EAG group. 104 TABLE 4: IMPORTANT DEMOGRAPHIC DIFFERENCES ACROSS EXPERIMENTAL GROUPS Group n, Mean Age Mean Years Taught SAG 13 31.5 8.5 EAG 13 37 12.5 SAG & EAG 26 34 10.5 EIIG 15 37.5 12.5 Total 41 35.3 11.2 105 These differences must therefore be considered when inter— preting findings of this investigation. In short, age and teaching experience may serve as confounding variables in this study. As defined by Campbell and Stanley (1963), between the pretest and posttest many types of confounding variables can also develop during the course of a study even though random assignment may have ensured an unblemished beginning. The experimental designs in both experiments, where three tests were taken over a period of 12 weeks, would seem to make this study vulnerable in four respects: history, maturation, testing effects and mortality. With regard to history, even within the context of the course they were doing as part of the MACT program, the groups in both experiments completed major action research projects and interacted with one another and with the instructor in ways that could have altered their professional self percep— tions. Whether these effects were uniform or not, they still could have caused changes in the dependent variable not due to the influence of the experimental treatments. Maturation also could have been a factor affecting the dependent varia— ble in both experiments. So, too, could testing effects, where the subjects took the same test on three occasions. Responses on the second and third administration may have been influenced by the teachers' recall of how they had marked the items in the past. A negative attitude could also have developed as a result of too frequent exposure to the same 106 test. The effect of experimental mortality will be discussed a little later in this section. One of the main reasons for undertaking Experiment II was to ”control” for the effects of three of the four poten- tial confounding factors mentioned above. The participants in Experiment II (EIIG) shared the same sort of history, maturation and testing effects as experimental groups, but did not receive either of the experimental treatments. Thus it does not seem that differences between experimental and control groups could be attributed to systematic differences on these three variables. It might, however, be argued that the members of the control group could have been negatively affected by a sort of Hawthorne effect. For, they did not get the same level of attention which the members of the treatment groups experienced, particularly in the process of receiving or reporting on their feedback. The investigator could have controlled for this effect but felt that a half an hour visit with each of the participants would not have had an impact on their self perceptions in any of the areas under study. But there were two other confounding variables which Experiment II could not ”control". One was an aspect of history peculiar to the experimental groups, and the other was experimental mortality. The history of Experiment I in itself could have nurtured elements of the Hawthorne effect, or it may have had the opposite effect, namely, growing re- sentment at being asked to satisfy the demands of the experi— ment during a busy term. 107 Experimental mortality also had an effect in Experi- ment I. Two participants, because of serious illness, had to be dropped from the experiment. One withdrew in the last two weeks of the Spring Term, the other got 111 during the summer holidays when the delayed posttest was due. Both were female elementary school teachers from City B; so, on these two variables they constituted a pair. However, they left their counterparts from City A without a match, with the result that the variable, geographical location, was now unevenly distributed between the groups. Over the duration of Experiment II, the EIIG group also lost two members because they did not hand in one of the tests. But this is not seen to have had a serious effect on that experiment. The professional background of the 15 participants in Experiment II were very similar to those of the parti— cipants in Experiment I (See Appendix I). On the other hand, certain differences between these two groups suggest that participants in Experiment II should not be viewed as a “true" control group. The main differences were time spent in the MACT program, and to a lesser extent, restricted geo— graphical location. Nevertheless, results from the second year MACT teachers in the EIIG group should give some picture of what changes teachers in an inservice program of this sort, who have not been given the experimental treatments, are likely to experience in terms of their professional self 108 perceptions over a period of 12 weeks. To this extent, qualified comparisons with the experimental groups will be attempted. In summary, there is at least some reason to question the internal validity and precision of the experiments. At— tempts were made, however, to identify and neutralize the major effects of a number of confounding variables. Because blocking was used, the virtues of random assignment and pre— cision, though beleagured by the fact of a small pool of par— ticipants, should have been preserved. Nevertheless, age and years in teaching remain as potential confounding variables. B. External Validity External validity is concerned with the extent to which valid generalizations can be made from a study. In dealing with external validity, Porter quotes Bracht and Glass (1965) as follows: ...threats to external validity appear to fall into two broad classes: (1) those dealing with generalizations to populations of persons (what population of subjects can be expected to be— have in the same way as did the ...experimental subjects?), and (2) those dealing with the en— vironment of the experiment (under what condi- tions ... can the same results be expected?) Both of these threats to external validity will be considered. What Populations? To make generalizations to a population, ideally two conditions are necessary: a well defined population, and 109 random selection from that population. The problem is that random selection from a large well defined population is rarely the case in experiments. Porter pursues the conse— quent argument in this way: ...since random selection is rarely used in experiments, are the results of experiments typically limited to the subjects used? Most people would answer these questions with a decided "no". People who are willing to gen- eralize from an experiment, however, require a careful description of subjects. The de— scription defines a hypothetical population to be used for purposes of interpreting results. It is possible, then, for the results of an experi- ment to be generalized beyond its subjects to a hypothetical population based on a description of the subjects. In order to provide the reader with such a basis for the projection of an acceptable hypothetical population, a full description of subjects is provided in Appendix I. However, one critical characteristic of the sample must be considered. As described earlier, all teachers in the sample had engaged in an intensive self assessment exer— cise prior to their participation in the experiment. This characteristic tends to set them apart from other comparable groups of teachers and may therefore severely restrict the generalizability (external validity) of the study. This char- acteristic, however, also provides a unique opportunity to examine the impact of feedback on teacher self perceptions. If participants with this backgroundchbchange their self per— ceptions in accord with feedback they receive, it is logical to infer that other teachers will also alter their self rs... 110 perceptions under these conditions. In short, these unique background experiences of the participants may limit external validity, but they also provide a powerful test of the degree to which feedback may alter a teacher's perceptions of rela— tive strengths and shortcomings. Under What Conditions? Certain conditional features of this investigation should be noted. With regard to the experimental treatments, the need.tolimit the degree of spontaneous interaction be- tween investigator/instructor and participants/students during the reporting sessions for both experimental groups (SAG and EAG) was a distinguishing feature. It is less typical of evaluators, especially in the role of instructor or super— visor, not to interact more flexibly with their students or clients. This controlled interaction was a particularly troublesome feature of the EAG treatment. A positive feature of this study, where external val- idity is concerned, is the nature of the dependent variables. As Porter states: External validity can also be threatened to the extent that the experimenter limits his consideration of dependent variables. In this study, professional behavior, which is the object of the teachers' self perceptions, is widely represented in its various aspects in the subscales of the Teacher Behavior Sur— vey. 111 Porter also warns that: Differences on a variable observed immediately following the experiments may not perservere to later points in time. This study seeks to obviate that threat by a repeated meas— ures design that includes a delayed posttest as well as an immediate posttest. The concluding judgement might well be that the ex- ternal validity of this investigation, in so far as a gen— eral population of inservice teachers is concerned, is severely restricted by the self analysis backgrounds of the participants. The conditions of the experiment, especially the need for controlled interactions, also caution against making blind generalizations. But the value of this inves— tigation is in its attempt to explore a complex, practical question under conditions in which feedback might be least apt to influence self perceptions. In brief, the sharp focus of the experiment might discourage replication, but it should increase confidence in those inferences that are formed. Research Questions Two sets of questions are addressed by this investi— gation. The first set is concerned with changes in teacher self perception. The second set is concerned with the at— titudes of the participants in the experiment. 112 A. Questions Concerned With Changes in Self Perception Question l:Does either of two forms of feedback-—self analyzed and reported (SAG), or externally analyzed and re- ported (EAG)——cause any change at all in a teacher's self perception of the relative levels of his or her abilities? Question 2:Which of two forms of feedback (SAG or EAG) will yield the greatest changes in a teacher's self perceptions? Question 3:15 the magnitude of change in self perceptions resulting from assessment feedback a function of empirically established levels of agreement (high, low) between feedback received and original self perceptions? Question 4:Wi11 the changes in self perceptions revealed in the posttest immediately following the experi- mental intervention (seven to 10 day interval) differ from corresponding changes in perceptions revealed in a delayed posttest administered ap— proximately six weeks after the intervention? B. Questions Concerned With Attitudinal Outcomes Question 1:What elements in the two term evaluation program did the participants find most valuable or least valuable? Question 2:How did they feel about participating in the pro— gram? 113 Question 3: What were their perceptions about the feedback they perceived? Analytical Procedures Used A. Summary of Statistics Used 1. Questions Concerned With Changes in Self Perception. Data for the four research questions concerned with changes in self perceptions were analyzed using one of three statistical tests: E—tests, correlation coefficients, and analyses of variance. For all analyses, alpha has been fixed at .05. A simple summary of these tests is as follows: Question 1: Are there changes in self perceptions following feedback? (a)£—tests of means for Change Scores I and II at subscale and total scale levels for experimental groups (SAG and EAG) and for EIIG. (b)Tests to determine if correlations between Feed- back Differential and Change Scores I or II at subscale and total scale levels are significantly different from zero. Question 2: Are there differences between SAG and EAG groups? (a)ANOVA tests of means for Change Scores I and II at subscale and total scale levels for SAG, EAG and for EIIG. (b)Tests to determine the correlations between Feed— back Differential and Change Scores I or II at 114 subscale and total scale levels and to compare these coefficients by groups. Question 3: Is the level of change in self perceptions a function of the level of feedback differential? Tests to determine the correlations between Feedback Differential and Change Scores I or II at subscale and total scale levels for the combined experimental groups (SAG and EAG). Question 4: How stable are the changes in self perceptions? (a)ANOVA tests of means of Change Score Differentials at the total scale level by groups (SAG and EAG combined and EIIG). (b)Tests to determine if correlations between Change Scores I and II at subscale and total scale levels for SAG, EAG and EIIG are significantly different from zero. 2. Questions Concerned With Attitudinal Outcomes The questions concerned with attitudinal outcomes focus on three aspects of the participants' reactions to the program: first, what they found most or least valuable in the program; second, how they felt about participating in it; and third, what their perceptions were with respect to the feedback they received. For each of the three research questions in this sec— tion, mean scores of all items concerned with the question will be calculated and revised. For the open ended items a 115 frequency count will be made of the times a given activity or program feature is mentioned. B. Why Data Is To Be Analyzed In This Way The four research questions on self perceptions are motivated by the more general question of whether feedback which is different from one's initial self perceptions changes these self perceptions or not. The concern is with differences between feedback and original self percep— tions, and with changes that may have occurred later as a result of these differences. Thus, the nature of the in— vestigation requires that difference scores, or change scores, be the focus of analysis. Where it is necessary to see whether any change at all has taken place for a single or combined group, simple p-tests are appropriate. These test whether an apparent change is significantly different from zero, or should be attributed to change. Where it is necessary to distinguish between group differences, again p—tests (for 2 groups) or one way analy— ses of variance (for more than two groups) are appropriate for two reasons. First of all, only single dependent varia— bles are being tested; and, second, these tests discriminate between statistically significant and chance differences be— tween groups. A number of tests for significant correlations are to be performed. These are useful to test whether relationships 116 between feedback and original self perceptions are signifi- cantly related to later changes in self perceptions. The relationship between immediate and delayed change scores are also to be tested for correlational significance. A con— sistent pattern of significant correlations would suggest a high degree of stability in those changes that occur. The alpha level, or level of significance, has been set at .05, which indicates a five percent likelihood of re— jecting the null hypothesis when it is true. This is the most commonly chosen level, and seem suitable for this in- vestigation. The simple reporting of mean scores and frequency counts should provide adequate answers to the questions con— cerning attitudinal responses. CHAPTER IV PRESENTATION AND ANALYSIS OF DATA Part One: Changes in Self Perceptions Introduction The central concern of this investigation was to deter— mine to what extent the self perceptions of teachers who haveundergone a thorough process of self analysis may never— theless be altered by either of two types of external data feedback. If, after this training, teachers do not signifi— cantly alter their perceptions as a result of either form of external data feedback, then there is good reason to believe that they are in a good position to make decisions about their needs for professional development following an initial pro— cess of self analysis. If on the other hand, teachers do respond to additional feedback data from significant others, the provision of feedback may represent an essential step in needs assessment. Four specific research questions were addressed by this investigation. The data analysis that focus on each of these questions will be summarized in the sections that fol— low. While data relating to these questions were being analy— zed, an interesting question arose which could not be 117 118 answered within the context of the original four questions. The new question concerned individual responsiveness to ex— ternal data feedback. This was a more exacting question than any of the original four, and will therefore be con— sidered in the final section of this chapter. Question 1: Does either of two forms of feed— back——se1f analyzed and reported (SAG), or externally analyzed and reported (EAG)—-cause any change at all in a teacher's self per— ception of the relative levels of his or her abilities? This question asks whether feedback received by the teachers will cause them to change the perceptions of their professional behavior which they held at the conclusion of the self analysis course. The most direct answer to this question was to determine whether the changes in self per— ceptions that occurred for the combined treatment groups between pretest and posttest were significantly different from zero. These results had then to be compared with sim— ilar results for the control group (EIIG) in an attempt to determine whether differences in pre and post test perfor- mance should be attributed to treatment effects or to some other source. Means for the two sets of change scores (Posttest—Pretest and Delayed Posttest — Pretest), for the ten subscales and the total scale are presented in Tables 5 and 6. Results of E — tests which compare these means with a value of zero are also described in these tables. Table 5 summarizes the mo. V m... OH. v._m CH m manomnzm so NH H m OO.- ON.d. NO.._OO.- OH. NN. OO.H- NO. OO.- OO. OO.H _m OO. OO. NO. OO. HN. NO. HO. OO. NO. OO. OO. Im HO.- NHr. HO:.OH.- OO. OO. HN.- NO. HO.- OH. NH. m OH oHHm *ON.H ONO.N. *HO.NON. HO. HOO.H OO. OH.H HHO.H OOO.H HNO.H O OO. OO. ON. NN. HO. ON. ON. NN. OO. OO. OO. .mm OH. HO. OO. NH. OO. ON. OO. NH. NH. HN. NH. m O.ON OmcHneoo OOH O OOO OO. OO. OO. NO. OO. OO. OO. NO. OO. OO. OO. mm 9 I u NO. OO. OO. OO. ON. NO. OH. NO. NO. OO. OH. O OH OOH OO. HO. NO. OO. OO. HN. OO. HO. OO. OO. OO. mm HO. OH. OO. ON.- HH.I HH.I OH.- OO.- OO.- OO. OH. M OOH OOO Hmpoe OH O O N O O O O N H m Q50Ho wmamomnsm .Hmpomumno .monm\amcomumm H OH mHoomO nHHz ocHxHoz n O :HzOHo amaze n O Hmuumz HomOnsm n O . GOOumSHm>m H m mHMHHmpmz HmCOOuusnpmcH u m pcmewomcmz u N pcmEdoao>mQ EDHSUOHHDO n O pocumz chcommB n O ochcmHm n a anmomnzm memme I M.Ho Oeqammm QZO onemo HOOOZOHO .OZOmz "OOOOOO Om AHOOHOOOIHOHHHOOOO H Ommoow mozm n O vogue: OcHrumme n O mUOmepromHmno pcmEQon>mm ESHSUOHHSO n N prpmz pomhnsm n m HmcoOmmmmoum\HmcomHmm H OH Luzouo smear u o “cosmomcmz n m mHQoom npOz OCOxHoz n O meHHmumz HmCOHposuumcH n m @chcmHm u H mmHmomQDm mamme -.m Oo meaammm HZO OonHOH>mo OOOHZOHO .OZOmz "Omaomo Om lemmemmauemmeemom OOHOHOOO HH Ommoom OOZOOo OOHOH O24 OHOoOOOO HO mHmOe 121 results for Change Scores I (Posttest - Pretest). This table indicates, first of all, that the means for the combined treatment groups were in all cases higher than those for the control group. The corresponding p—tests suggest that the means for the treatment groups were significantly dif- ferent from zero for the total scale and for three subscales when the likelihood of Type I error was set at .05. Also, three other subscale means for the treatment groups approached statistical significance. On the other hand, none of the means for the control group were significantly different from zero under the same test conditions. These results suggest that there were greater changes in self perceptions for the treatment group than for the control group. A study of the data in Table 6 reinforces the in- sights already gained. Means for Change Scores II (Delayed Posttest _ Pretest) were always higher for the combined treatment groups than for the control group (EIIG). Further— more, : — tests of the means for the combined treatment groups were significantly different from zero for the total scale and for four subscales, when the likelihood of Type I error was set at .05. The means of two other subscales approached statistical significance. None of the means for the control groups were significantly different from zero, and only one approached statistical significance under these test conditions. The strong impression left from a study of Tables 5 and 6 is that the members of the treatment groups changed 122 their self perceptions to a greater extent than did the members of the control group. And these changes from the original pretest levels were greater for the treatment group members both as measured in the short and long term by the posttest and the delayed posttest. In general, there is a strong suggestion that feedback influenced the considerable changes in self perceptions that took place for the members of the treatment groups. A further study of the data in Table 5 reveals that the means and E - tests results for the combined treatment groups were all positive, while six of the eleven means for the control group were small and negative. (The latter was also the case for teachers who were members of the SAG treatment group, but this negative tendency was overcome by the strong positive scores for the EAG group). In Table 6, all of the combined treatment group means were positive, whereas two of the control group means, and three of the SAG means were negative. The significance of these relative positive or negative tendencies must be examined in the light of the data summarized in Table 7. The data summarized in Table 7 show that for every subscale and for the total scale, the means for feedback differential (Feedback—Pretest) were high and positive for the treatment groups, with most scores over .50. Feedback scores Were, therefore, typically higher than the teachers' original self perceptions as reflected in pretest scores. To return to the study of the data in Tables 5 and 6 the fact that the mean difference between posttest and I 123 OOO. HOO. OO. OO. OO. OO.H OO. OO. OO. NN. ON. O ON. HN. ON. NN. HN. OO. OO.H OO. HO. ON. HN. 1m OO. HO. OO. OO. OO. OO. ON. ON. OO. OO. OO. O OH OOO NO. NO. ON. OO. OO. NO. ON. OO. HO. OO. OO. 1m OO. OO. NO. OO. OO. ON. ON. OO. NO. HO. OO. M OH OOO Hmuoe OH O O N O O O O N H .m nacho mmHmUmQSm cOHumsHO>m n O scrum: OOHnomme n O mUOumOpromumso pcmeaoam>mm ESHSUOHHSU n b Hmppmz uownnzm n m HmCOOmmOMOHm\Hmcomem H OH cpzouo cmesm u m ucmemmwcmz n N manomm SHOZ ocflxuoz n m mHmOHmwmz HmCOOuosuumcH u m ocOccmHm n H mmamownsw omeozeem .mzmo "ABWmBmmm I¥O¢mmmmmv mmmOOm AdHezmmmmmHQ x0mHmQ pumpmumnemmuumom ummumnm I xomnpmmm wucmaflofluumm AH mamumnsmv AH mamomnsmv AH mamomnsmv I mmuoom HH mmuoom mmcmco H wououw mmcmco HmOucmHOMMOQ xumnpmmm QMEDQEOU mmmz mZOHEHmZOmmmm ¥Um n O Oorumz OcHnomme n O woOumOHmuomHmno unmemoam>wa ESHSUOHHDO u N Hmupmz pownnsm H m HmCOOmmmMOHm\HmcomHmm H OH cu30H0 cmezm H O pcmemmmcmz n N maaomm QuHB OCOxHoz n m mHmOHmvmz HmcoOuozuumsH u m ocOccmHm n H mmamomnsm ABmmBmmm I x0m n O Oonumz Ochumme n O mUOumOHmuomHmno pcmemoam>mm ESHSUOHMDU u N umupmz pumflnsm n m HmGOOmmmwoum\HmcomHmm H OH cuzouo smear n O unmewomcmz u N manomm bqu OCOxHoz u m HmOHmpmz HmcoflpusuumcH n m Ochcmam n H mmamomnsm . . moHeOm I m OZO Oonemo HOOOZOHO Ozm n O eoeemz Ochomms n O mUHumHHmpomumsu pcmEQon>mO ESHDUHHHSU H N Hespmz HUONQSO n m HMCOHmwmmoum\HmcomHmm H OH LHBOHO cwesm H O ucmEmomcmz u N mHQomm Lsz OCHxHoz n m mHmHHmpmz HOCOHHUSHpmcH n m OchcmHm n H mmHmomnsm moHHOO I M 02¢ mZOHBmQ Qmm n O eczema mcHnomms n O mUHumHHmpomHmnO pcmEQOH®>mQ EDHSUHHHSO n N Hmppmz uomNQSm n m HmGOHmmmmoum\HmcomHmm H OH QHBOHU cmssm n O Hamsmmmcmz u N mHQomm nqu ocHxHoz H O mHmHHmpmz HmCOHuosHpmcH n m OCHccmHm n H mmHmowQSm HemmBmmm I BmMBBmOm Omwm. so. 4*Na. *sms. mm. *xve. mm mm. ms. ssom. Hos. ow. slow. qu. so.a**om. swmm. xxms. mm. wam. HM mm mm em mm mm Hm om mH mH eH mH mH «H # aomsmsm mam HH.-*¥om. ow. ssos. He. so. 4mm. Has. mo.-*.ms. sow. mm. ssms. mm HH.! *vm. mm. *xow. mm. HHm. *Ho. *v0. *00. ssmh. sow. *mw. *xmm. HM MH NH HH OH m m e m m s m N H # Homemam lull Q m oHmzommmm JHozH "5H memes 149 individual responsiveness at these two levels. Another question suggests itself at this point. Is group membership a factor where individual responsiveness to feedback is concerned? To answer this question, the means for the two sets of correlations (g and r 1 2) were calculated for both treatment groups. The differences between the groups were tested for statistical significance. The results summarized in Table 18 show that where the first set of correlations (g ) is concerned, the means for the 1 two groups were almost identical (.54 for the SAG and .55 for the BAG), producing an insignificant difference. The differences between group means were only slightly higher for the second set of correlations (£2) and were also in— significant. Where individual responsiveness is concerned, group membership seems to make very little difference. In summary, the data suggest that individual re— sponsiveness to feedback was quite strong for the majority of participants. Twenty—two individuals significantly changed their self perceptions in line with the feedback they received. Individual responsiveness did not vary as a function of how the feedback was received (SAG or EAG).Further_ more the influence of feedback differential on individual responsiveness did not vary over time: the level of re— lationship was approximately the same for the delayed post- test as it was for the posttest- 150 TABLE 18: INDIVIDUAL RESPONSIVENESS: CORRELATIONS BETWEEN FEEDBACK DIFFERENTIAL SCORES (FEEDBACK — PRETEST) AND CHANGE SCORES I OR II (POSTTEST — PRETEST 2E DELAYED POSTTEST — PRETEST) FOR EACH PARTICIPANT IN EXPERIMENT I: MEANS, STANDARD DEVIATIONS AND RESULTS OF ANOVA TESTS £1 = Correlation Coefficients: Feedback — Pretest/Posttest - Pretest — Across 10 Subscales Analysis of Variance Group 9. X31 §2 Source g is E SAG 13 .54 .24 Main Effect 1 .00 .01 Error 24 .07 EAG 13 .55 .29 r = Correlation Coefficients: Feedback — Pretest/ Delayed Posttest — Pretest - Across lO Subscales Analysis of Variance Group E X £2 SE Source _gf _M§ _F SAG 13 .45 .33 Main Effect 1 .06 .65 Error 24 .09 EAG 13 .55 .29 151 Summary of Part One Part one of this chapter attempted to answer four questions about changes teachers make in their self percep— tions as a result of the feedback they receive from others. In response to the first question as to whether the feedback received causes any change in the teachers' self perceptions, analysis of the data strongly suggested that the teachers in this sample were clearly influenced by the feedback they re— ceived. Despite their training in self assessment, these teachers made significant and predictable changes in their self perceptions following feedback from others. Those who received feedback higher than their own estimation of them- selves raised their self perceptions in the direction of the feedback; those who received feedback lower than their own self evaluation, lowered their self perceptions. The second question asked which of the two forms of feedback (SAG or EAG) will yield the greatest changes in teachers' self perceptions. The analysis of mean change scores and correlation coefficients between change scores and feedback differential consistently suggested that changes in self perception were slightly higher for the teachers in the EAG group (who had the feedback analyzed for them and re- ported to them by an external evaluator), than for the teachers in the SAG group (who had analyzed and reported on their feed— back). The differences between the EAG group and the other two groups were most pronounced when posttest — pretest 152 scores served as the dependent variables. Also, the changes for the EAG group were strongest in the direction of the feedback received. The third question asked whether the magnitude of change in self perceptions resulting from assessment feed— back was a function of levels of agreement between feedback received and original self perceptions. The uniformly pos— itive and generally high correlation coefficients between feedback differential and change scores resulting from analy— sis of the data in response to this question, strongly in— dicate that high feedback differential scores were typically followed by large changes in self perception. Low feedback differential scores were followed by relatively small changes in self perception. Thus, magnitude of change in self per- ceptions could be said to be a function of levels of high or low agreement between feedback received and original self perceptions. The fourth question had to do with the stability of changes in self perception which result from feedback. Changes in self perception were found to be very stable, varying only slightly over the six week interval between the posttest and the delayed posttest. An additional set of analyses to determine each in— dividual's level of responsiveness to feedback was an exactig test of the central question in this study. Individual responsiveness to feedback was found to be strong for the majority of participants. Twenty—two of twenty—six teachers 153 changed their self perceptions in line with the feedback they received somewhere during the course of the study. How the feedback was provided (SAG or BAG) did not seem to affect the level of individual responsiveness. Part Two: Attitudinal Outcomes Introduction Following the delayed post test, the 26 participants who had experienced the two term evaluation program (guided self analysis in the first term, experimental treatments in the second) were asked to complete a debriefing form. As described in Chapter three, the purpose of this survey was to find out how these participants felt about the various aspects of the program, and in this way to collect formative data about the processes involved. Specifically, the de— briefing survey sought answers to three questions: 1. What elements in the two term evaluation program did the participants find most valuable, or least valuable? 2. How did they feel about participating in the pro— ‘gram? 3. What were their perceptions of the feedback they received? Summaries of responses that relate to these questions serve as the focus of the three sections which follow. 154 Question 1: What elements in the two term evalu- ation program did the participants find most valuable or least valuable? Table 19 presents a summary of responses to question- naire items that focused on this question. These data sug— gest that the self assessment done in the first term was felt by participants to be the most beneficial part of the two term process. Where the second term was concerned, feed— back from the first part of the questionnaire (TBS) was viewed as most valuable. According to the participants, the least valuable part of the experiment was the process of thinking through and writing up the summary outline at the end of the evaluation. However, even this feature was seen as having limited to moderate value. Five open ended items at the end of the attitude scale gathered further data on this question. Items 29 and 32 asked participants to describe the features of the program they most liked or valued. Their responses to these two items included the following (the numbers in brackets in— dicate the number of times the feature was mentioned): 1. Self evaluation (14) 2. Peer evaluation (11) 3. Interaction with colleagues (8) 4. Becoming aware of strengths and deficiencies (7) 5. Being given the opportunity to participate in an evaluation program (6) 155 TABLE 19 MEAN SCORES FOR ITEMS CONCERNED WITH VALUES PARTICIPANTS PLACED ON CHARACTERISTIC PROGRAM ELEMENTS. 4 = high value 2 w ll moderate value 1 limted value little or no value Item Number Items X l The process of deriving the lists of teacher strengths and needs in the first term 2 The self assessment done in the first term 3 Feedback from the first section of the questionnaire (”Teacher's Strengths") provided by your professional colleagues in the second term 4 Feedback from the lists of strengths cited by your colleagues at the end of the first section of the ques— tionnaire 5 Feedback from the second section of the questionnaire ("Possible Areas for Improvement") provided by your professional colleagues 6 Feedback from the lists of your possible needs for improvement cited by your colleagues at the end of the second section of the questionnaire 7 The process of thinking through and writing up the summary/outline at the end of the evaluation Grand Mean = 2.99 156 6. Becoming aware of the importance of evaluation (5) 7. Deriving lists through group discussion in the first term (5) 8. Being evaluated by administrators (5) These responses confirm that self evaluation was viewed as the most valuable feature in the total program. The process of deriving lists and interaction with colleagues which were also highly valued, were among the most character- istic features of the first term when the self evaluation was done. The participants also valued the two types of assessors, peer and administrator, who contributed feedback data during the experimental portion of the program. It is also important to note that several of the participants com- mented on the value of the evaluation process itself (see 4, 5 and 6 above). The features of the program that were least valued or liked, as measured by two open—ended items 30 and 33, were as follows (the numbers in brackets indicate the number of times each feature was mentioned.): 1. Repetitiousness (15) 2. Too much time required of participants (10) 3. Summary/Outline written at the end (7) 4. Too much time required of assessors (5) 5. Inefficient assessors.(4) 157 Items 1, 2 and 4 on this list indicate the participants had strong objections to the amount of time and repetition required by the program. They had spent a lot of time in the first term in activities that led to the production of the items from which the Teacher Behavior Survey (TBS) was de— rived, and then in the second term they were asked to fill out the TBS three times. In addition, the TBS they completed had 114 items in two sections, the second being a negative version of most of the items on the first section. As al— ready mentioned, this second section proved not very helpful in collecting data, and was eventually discarded. The summary/outline again proved to be unpopular, per- haps partly because it was also time consuming and came at the end of term when the participants were under pressure to hand in their other end—of—term assignments. But also, these summaries were never discussed with nor returned to the members of the externally assessed group (BAG), who had written their summaries after their Interview in which the feedback had been given them. On the other hand, the sum— mary was the basis of the report given by the self assessed group (SAG) in their interview; but even here the protocol of the experiment precluded discussion. Question 31 on the attitude scale asked the partici— pants morecommend changes that should be made before the program is used with other students. The main responses are listed below: 158 l. Shorten the process, decrease repetitions (l4) 2. Make assessors more efficient (10) 3. Give additional academic credit (6) 4. Condense the TBS (4) 5. Do earlier in the year, not in Spring term (3) The first and fourth items are consistent with the strong message already received in response to questions 31 and 33. The third item also indicates a degree of frustra— tion participants felt in having to undergo the experimental treatments in the second term in addition to other course requirements. Item 5 also refers to the time required of them and their assessors in what they described as the bus— iest term of the year. Where item 2 was concerned, ”make assessors more ef- ficient," the participants gave a number of more specific and useful suggestions: —clearer instructions should be given to assessors so their evaluations could be "more helpful and honest“. —more time should be spent on the in-classroom assess— ment by the assessors, and a ”more nearly equal amount of time” spent by all. —peer assessment should be done "over a longer period so as to be more valid". -peer assessors should be ”required to observe” 159 -peer assessors should be chosen from other schools —peer assessors should be members of the evaluation program —opportunity should be provided for participants "to go over the assessments with the assessors“ —assessors should not "guess“ In summary, participants seem to have valued most those activities done in the first term-—self evaluation, and deriving lists of teacher interaction with colleagues, strengths and deficiencies. Where the second term is con- w “v cerned, they also felt that they benefited greatly from the feedback they received from the first section of the TBS. They also valued the opportunity to receive feedback from peers and administrators. They felt that they achieved a growing awareness of their abilities, and were grateful for the opportunity to be involved in the evaluation process. They least appreciated the amount of time and repet— ition required by their participation, and strongly suggested that the program be condensed. They also did not favor the summary they were asked to write, although it was only men— tioned by one person as a number of suggestions or effectiveness of the Question 2: something to be changed. were given to improve the assessors . How did the teachers feel cipating in the program? Table 20 summarizes responses to questions on participants' attitudes toward the experience. Finally, efficiency about parti— that focused It is 160 TABLE 20: MEAN SCORES FOR ITEMS INDICATING HOW PARTICIPANTS FELT ABOUT TAKING PART IN THE TWO-TERM PROGRAM 4 = strong agreement 2 =-moderate disagreement 3 = moderate agreement 1 = strong disagreement No. on Scale Items Z 8 Because members of the control group had less to do, I initially felt some frustration/displeasure in being assigned to an experi— mental group 2.81 9 The benefits I received from being a member of an experimen- tal group adequately compensated for the extra work I had to do 2.31 10 The experiment procedure was too long and repetitive 3.35 11 I made considerable gains in my ability to assess my own perfor— mance as a result of this exper— ience 2.73 12 I made considerable gains in my ability to assess a colleague's performance as a result of this experience ‘ 2.42 13 I am more sensitive to short— comings in my classroom perfor— mance as a result of partici— pating in this project 3.23 14 I am more sensitive to strenghts in my classroom performance as a result of participating in this project 3.08 161 No. on Scale Items IXI 15 16 17 18 19 20 21 22 23 24 The benefits teachers might derive from some form of systematic evalu— ation in the MACT Assessment I course will depend in large part on the in— terpersonal/communication skills of the instructor who directs the ex— perience 2.88 The two—term evaluation procedure in which I participated should become a required part of the MACT Assessment I course 2.50 All teachers should undergo some form of professional assessment at least once every five years 3.65 I felt comfortable and relaxed during the classroom sessions in which I was being observed by my professional colleagues 3.13 The process of being observed and assessed by professional colleagues was a valuable experience-in and of itself 2.85 I would have taken the feedback I received more seriously if the as- sessors had spent more time in the classroom observing my performance 2.88 The assessors took this job ser— iously and did it conscientiously 2.92 There are few if any significant disagreements between my own ratings and those provided by my professional colleagues 2.96 The fact that the assessors tended to be very generous in their ratings reduced the value of the feedback experience 2.35 I have a higher regard for my own abilities as a teacher as a result of this experience 2.92 Grand Mean = 2.88 162 important to note that at the end of the two terms of evalu- ation-—a program which many must have felt to be arduous and sometimes irksome--the participants gave their strongest agreement rating (Y = 3.65) to the following item: 17. All teachers should undergo some form of pro- fessional assessment at least once every five years. This response confirms the finding that participants had a high positive attitude toward the main experience provided by the program. This positive feeling was also conveyed by their responses to items 13 and 14. Participants felt that they had benefited significantly from the program, for, as they reported, they had become more sensitive to both the strengths and shortcomings in their classroom performance. On balance, they also felt higher regard for their abilities as a result of their experience in the program (See item 24). Item 9 confirms the teachers' feelings that additional graduate credit should have been given for the second term of the course. Participants also reported some disagreement with the statement that the benefits from being a member of the experimental group adequately compensated for the extra work they had to do. Responses to items 18, 21, 22 and 23 reveal a high positive attitude toward the processes and results of the assessment they experienced. Participants felt comfortable and relaxed while being observed (item 18). Although they recognized shortcomings in their assessors, as already re— ported, they agreed that these individuals in general took 163 their job seriously and did it conscientiously (item 21). They saw few significant disagreements between their own ratings and those of their assessors (item 22), and denied that the generosity of their assessors' ratings reduced the value of the feedback experience (item 23). As already mentioned in discussing item 24, they reported having a higher regard for their abilities as teachers as a result of this experience. Briefly, then, participants seemed to have had favor— able attitudes towards the program and felt they had bene- fited from it. However, they also felt that in its present form it was too long and repetitive. Question 3: What were the participants' percep— tions of the feedback they received? Items 18, 21, 22 and 23 of the attitude scale pro— vide a partial answer to this question. The responses to these items have already been summarized in. the section above, where it was reported that participants responded favorably to the processes and results of the feedback assessment procedures. Some additional items on the scale were specifically concerned with participants' perceptions of the level and effects of the feedback they received. Responses to items 25 and 28 are summarized in Table 21. Only one partici- pant felt that the assessors' overall rating was lower than his or her own (see item 25). Ten participants felt that the assessors ratings generally coincided with their own. 164 TABLE 21: NUMBER OF RESPONSES FOR ITEMS RELATED TO PERCEPTIONS OF FEEDBACK RECEIVED No. on Scale Items Number 25 Overall, the assessors tended to rate me...I rate myself 1. higher than 15 2. at about the same level as 10 3. lower than 1 26. The extent to which the assessors agreed among themselves in rating my performance of specific skills was 1. very high 6 2. high 15 3. low 5 4. very low 0 27. I was most sensitive to these dif— ferences between my ratings and those of my assessors when they rated me 1. higher than I rated myself 6 2. lower than I rated myself 19 28. The most significant outcome of the feedback I received was 1. identification of some weak— nesses I did not know I had 9 2. identification of some strengths I did not know I had 2 3. confirmation of the percep- tions I had of my abilities prior to this experience 15 165 Fifteen felt the assessors' ratings were higher, but, as reported earlier (see item 23), this tendency of the assess— ors to rate generously did not reduce the value of the feed- back experience for most of them. Another positive attitude to the feedback data re- ceived is suggested by the responses to item 26. Only six participants felt that the extent1x>which the assessors agreed among themselves was low. With regard to item 27, the great majority of parti- cipants (19 out of 25) said they were most sensitive to these differences between their ratings and those of their assess— ors when the assessors rated them lower than they rated them- selves. It bears repeating here that, as reported in Part One of this chapter, the differences between feedback and participants' initial self perceptions were almost always positive, that is, feedback scores were most frequently higher than initial self perception scores. The second point is that although participants claim greater sensiti— vity to negative differentials, they nevertheless changed their perceptions significantly in the direction of the pos- itive feedback they received. Finally, the data collected in response to item 28 show that most participants (15 of 26) felt that the princi— pal outcome of the feedback they received was to confirm their original perceptions of their abilities. Slightly over one third (9 of 26) felt that the primary outcome was an identification of some weaknesses they did not know they 166 had; only two of the participants believed the main result was an identification of previously unrecognized strengths. Nevertheless, the teachers felt the feedback helped them to raise their estimations of their abilities. This sense of positive reinforcement might also explain why the partici— pant were generally positive toward the experience at the end of the demanding two term evaluation program. Summary of Part Two The principal outcome of the two-term experience in the eyes of most participants was a confirmation of the per— ceptions they already held regarding their own abilities. Although they judged that the level of feedback they re- ceived from colleagues and supervisors was generally higher than their initial self perceptions, they nevertheless, felt that this did not reduce its usefulness. In keeping with these perceptions, the participants generally expressed strong, positive feelings toward the program, even though it lasted two terms, and was seen-by some as being too long and repetitious. Where specific aspects of the program were concerned the opportunity to engage in self analysis in the first term was seen as the most valuable feature of the two term program. Peer groups interactions, which were a part of their analy— sis, were also seen as particularly valuable. The first half of the TBS questionnaire was felt to have provided profitable feedback. The second half, however, 167 was regarded as less useful. As suggested elsewhere, the, second section may have contributed to some feelings about repetitiousness and the amount of time required by the whole process. Another aspect of the program had a consistently negative rating. This was the final summary/outline which came at the end of a very busy term. The assessors who provided the feedback, were judged to have done their job conscientiously and well. This sen- sitive and positive attitude to the assessors resulted in participants giving some very useful suggestions for im— proving the assessors' performance. CHAPTER V: SUMMARY AND IMPLICATIONS Purpose and Design of the Investigation The main purpose of this study was to investigate whether external feedback causes teachers to change self perceptions of their professional behavior in those areas where feedback has been given. It was decided to test this question in very exacting circumstances. The investigation was therefore done with teachers who had recently undergone an extensive program of self analysis, and in this process had used and helped define the categories of teacher be- havior which were later used to provide them with feedback. Two practical considerations motivated this decision. First, from an experimental point of view, it seemed within the realm of common sense to predict that feedback would make some impact on the self perceptions of teachers who had not been asked to reflect on their own abilities in any systematic way. The literature gave strong indications that this would be so. On the other hand, it was not so certain that teachers would be affected by external feedback once they had firmly established their self perceptions through a systematic process of self assessment. The second practi— cal consideration was concerned with the fact that self assess- ment is becoming increasingly important in teacher education, 168 169 especially in inservice teacher education where teachers are increasingly being involved in determining their own needs and programs. In this context, self assessment by itself is being seen by some as providing teachers with adequate insights on which to base decisions about the course of their professional development. Putting these two practical considerations together, the question of interest became:after engaging in an exten— sive process of self assessment, does external feedback change one's self perceptions to any significant extent? Another question explored was concerned with which of two forms of feedback——externally analyzed and reported, or self analyzed and reported——caused the greatest changes in teachers' self perceptions. Also considered were questions about (a) the effect of the magnitude of differences be— tween self perceptions and the perceptions of others, and (b) the stability of changes in self perceptions. The design of the experiment took into account the need for all participants to have experienced a preparatory stage involving a rigorous term of self assessment. Next, since the two distinct modes of giving feedback were to be tested, three groups of participants were selected, one for each of the two feedback treatments, and one to act as a control group for the experiment. The groups were designated SAG (the self assessed group),hAG (the externally assessed group) and EIIG (the control group). The members of the SAG were given their assessment data and had Unanalyze them and 170 make an oral report. The members of the EAG had the data analyzed for them and were given the results in an evalua— tor's report. The assessment data took the form of perfor— mance ratings supplied by three assessors who were given precise written instructions on when and how to complete the feedback survey. The assessors were chosen by each participant. In order to provide a wider range of data sources, two were peers, and the third had to be an admin- istrator. The control group (EIIG) received no feedback at all. Neither this group nor its members were randomly selected; hence, the EIIG group was strictly a quasi—control group, chosen because its members had all undergone a one term self evaluation program similar to that experienced by the two treatment groups. The main instrument used in the investigation was the Teacher Behavior Survey (TBS), which the participants in the treatment groups had helped to develop. Responses to the TBS served as pretest, posttest, and delayed posttest for all the participants, and it was also the instrument used by the three assessors to rate teacher performance. A second instrument, an attitude scale, was given to the members of the experimental group after they had com— pleted the experiment. The scale was intended to find out how participants felt about the two term program and its constituent parts: what they valued and did not value,and what they felt ought to be altered. It was an attempt to collect formative data about this sort of evaluation program. 171 Feedback and Self Perceptions A. Major Findings A number of important conclusions were suggested by the analysis of the data focusing on the relation between feedback received and alterations in self perceptions. The three most significant findings were: 1. Teachers' self perceptions change as a rESUlt of feedback, even after they have previously rigorously assessed their own behavior. 2. Changes in self perceptions are a clear function of feedback differential. Thus, teachers change their self perceptions in the direction of the difference between their original self perceptions and the feedback they receive. Furthermore, the magnitude of changes in self perception are a function of the size of the difference between initial self perceptions and feedback received. 3. When feedback is externally analyzed and reported, it appears to be a slightly stronger motivator of change in self perception than when it is self analyzed and reported to an external evaluator. B. Discussion of Findings 1. Change in Self Perception as a Result of Feedback It is a very significant finding that teachers, even after they have analyzed themselves in specific areas and established fairly precise perceptions, are nonetheless 172 malleable to feedback that is given in those same areas. How- ever, once these self perceptions are changed as a result of the feedback received, they remain stable over time when no additional treatments are applied. How can this malleability be explained? At least three different explanations might be proposed. Common sense sug— gests one practical explanation. Teachers spend their pro— fessional lives locked away in their private classrooms, and professional politeness and the social structure of the school tend to reinforce this isolation. Principals do their required .1..." . evaluations reluctantly and with little in—class observation. Their feedback presents not only no challenge but little use- ful information to teachers. As a result of this situation, teachers crave helpful informative, professional feedback on their teaching performance. They desire the opportunity to get this type of feedback especially when it is not done for the sole purposes of retention, tenure and promotion. And they respond to it when they are given it. Interactionists would explain the malleability of self perceptions, where external feedback is concerned, in terms of the "looking glass self". They claim that an individual's self concept is very much determined by the way others view that person, and the extent to which their view is communi- cated to him or her. If the "others” are “significant” to the person, their views are more influential. Also, the clearer and more detailed the perception of these views are, the more they are likely to be influential. In this investi- gation, since the assessors were chosen by the participants, 173 it is very likely that they were "significant” to them. Next, the feedback was detailed and clear due to the fact that the participants had used the same categories as those in the TBS for their earlier self assessment, and the TBS itself made for a clear and detailed presentation of the data. In these circumstances, even after self assessment, the teachers were likely to be influenced by external feedback. In addition to sharing these views, Combs(1975) was very dubious about the efficacy of self assessment. He was quoted earlier as saying that "the objective analysis of self" is "vastly overrated as a device for personality change". Thus, Combs would explain the responsiveness to others' opinions even after detailed and prolonged self assessment as due to the inefficacy of self assessment. For him the explanation would be simple: the self assessment made little enduring impact. As already indicated in the review of the literature, however, there is experimental evidence supporting the ef- fectiveness of the use of self assessment as a valuable tool in behavior change. So, in brief, it does seem that the notion of the "looking-glass self“, in addition to the fact of their habit- ual isolation, is the best explanation for the malleability of teachers' self perceptions even after they have undergone long and detailed self assessment. 174 2. Change as a Function of Feedback Differential When teachers are given feedback that matches closely their own perceptions of their performance, there is little subsequent change in their self perceptions. This one would expect. On the other hand, when feedback is widely differ— ent from initial self perceptions, these self perceptions change in the direction of the feedback, and to a greater or less extent depending on the extent of the difference be— tween feedback and initial self perceptions. In other words, the magnitude of change in self perceptions appears to be a clear function of feedback differential. This behavior is consistent with Festinger's theory and experiments cited in the literature review. Elsewhere there was a concern expressed that too great a difference between a teacher's perceptions and those of the assessors might cause the teacher to become less responsive to the feedback or more impervious to change. This does not seem to have happened to any noticeable degree in the present investigation. Results of this experiment suggest that teachers do respond to cognitive dissonance caused by dif- ferences between their own perceptions and those of their chosen assessors by changing their self perceptions. The results also indicate that the greater the differences, the more the teachers seem to change. 175 3. Feedback Data: Externally Analyzed vs. Self Analyzed The results of this investigation suggest that data feedback which is externally analyzed and reported to teachers (EAG) influences their self perceptionsLto a somewhat greater extent than the same sort of data when it is analyzed and reported on by the teachers themselves (SAG). As pointed out earlier, both forms of data feedback have great in- fluence on teacher self perceptions, even after teachers have engaged in protracted self analysis. But when the ques— g tion of a comparison is raised as to which form of feedback ‘ (EAG or SAG) is the more potent, then the results show only slight differences between the two, with the differences consistently favoring the external mode. Studies discussed in the literature review in Chapter Two show that both modes of data feedback have been used with success. Those studies concerned with supervisor or peer evaluation support the value of externally analyzed feedback. The literature on "educational linking agents” also indi— cates that a ”support system” seems necessary to ensure the efficacy of feedback. On the other hand, research described in connection with microteaching and computer assisted pro— grams give some indications that self analyzed data feedback is a potent force for change. With regard to microteaching, studies reported by Cooper and Allen (1971) support the prac— tice of teachers confronting their videotaped performance with or without supervisor or peer assistance. Computer 176 printouts, self analyzed by teachers, also effect changes in teacher behavior, as reported by Pohlmann (1976), Semmel and Olson (1977) and Hail (1978). In addition, recent teacher education practice, promoted by many writers including Curwin and Fuhrmann (1975) and Rubin (1978), stress the importance of teacher self determination and the discovery of one's teaching self through processes of self analysis. The present investigation tested these two alterna— tive modes of feedback in a one—on—one experiment. Since previous evidence was that both forms of feedback do bring about change in teacher behavior, big differences between the two were not expected. It should be remembered, too, that the EAG treatment was restricted to one function of external evaluation, that of analysis and transmission of data. All the other functions of an external evaluator com- monly exercised in supervision, which are described as "helping,” ”support," or "interaction," were deliberately suppressed in the experiment so as to compare more exactly the two different modes of analysis and reporting. These experimental conditions constituted a very conservative test of the effects of external evaluation. Nevertheless, ex- ternal evaluation, so narrowly exercised in this experiment, proved to be more potent than self analysis. This seems to present a powerful argument in favor of the more normal uses of external evaluation found in teacher fieldebased experience, as compared to self analysis. For if an external evaluator I limited to one function can influence teacher self perceptions 177 consistently, even if only to a slightly greater extent than self analysis in the tests conducted in this investigation, then it seems likely that an external evaluator, freely interacting, helping and supporting the teacher, will be an even more powerful influence. , There are at least two possible explanations for the greater efficacy of external evaluation. First, there are arguments and research findings of interactionist psycholog— ists reported earlier in Chapter Two, which support the theory of the "looking-glass self". In this theory, the opinions of others—-particularly those of "significant others"—- communicated to an individual, present him with a mirror of his self which determines the development of that person's self concept to a very large extent. One's self concept, though it becomes fully developed, is never finally formed, and is therefore open to change. The most powerful agent for changing one's self perceptions is new information about one's performance from other people who one holds in esteem. In this experiment the significant others were, first of all, the assessors chosen by the teachers themselves to be the sources of their feedback. For those teachers who were given the EAG treatment, another significant person rein- forced the process of other-reflection. As co—instructor for the two—term program, the external evaluator can be ex- pected to have achieved a ”significant” status with these teachers. The effect of the feedback was, therefore, magni— fied by his intervention. 178 Another explanation for the greater strength of ex- ternal evaluation as revealed in this conservative test may be found in the deference customarily shown to people in authority. This seems the most straightforward explanation. Deference shown to persons in authority, including one's uni— versity instructor, is common human behavior. A set of opin— ions about one's professional behavior, analyzed and reported by one's instructor is likely to appear more authoritative and, therefore, persuasive than similar opinions which an individual weighs and sifts in the privacy and warmth of one's living room or study. The power of opinions when they are communicated by a significant other, to whom, in addition, one shows profession— al deference, would seem to go a long way in explaining the greater potency of external evaluation as compared with self assessment in giving feedback. Additudinal Outcomes General The main benefit participants said they derived from the two graduate courses that focused on the evaluation pro— cess was the confirmation of their perceptions of their strengths and weaknesses. In fact, the feedback they re- ceived was higher than their initial self perceptions, and most participants raised the level of their self perceptions as a result. Nevertheless, the feeling they were left with was one of support for their professional self perceptions. 179 It is also significant that participants expressed strong positive attitudes towards the program, even though it was long and demanding, continuing over two terms. They placed high value on the program in general, and were able to identify specific aspects from which they felt they benefited as well. Taking the program stage by stage, the outstanding characteristics identified by participants were as follows: Term 1: Guided Self Analysis The highest value of all was placed on the self analy— sis done in the first term. This process was characterized by a great deal of peer group interaction as exemplified by the derivation of lists of teacher strengths and deficiencies and review of one another's needs assessments. Term 2: The Experiment l. The Teacher Behavior Survey (TBS) Participants rated highly the feedback they received from the first half of the TBS ——the section concerned with the assessment of the teacher's strengths. Teachers pre— ferred this mode of receiving feedback, and it also proved an adequate basis for needs-based program planning. On the other hand, the second section of the TBS was not found to be particularly useful. Assessors communicated their antipathy to this section through messages or written notes. But, most of all, they restricted themselves to two 180 rating responses. They either disagreed or disagreed strongly with the items in this section, which all suggested specific areas in which the teacher might need to improve. In brief, the first half of the survey was an adequate instrument, framed as it was in positive terms requiring the assessment of relative levels of a teacher's strengths. The second half of the survey, which was framed in more negative terms requiring an assessment of what some perceived as the teacher's "weaknesses", was neither fruitful nor well re- ceived. 2. The Assessors The participants appreciated their assessors and the feedback they gave. They thought them to be conscientious, felt comfortable when observed by them, and valued their feed- back. They made a number of specific suggestions for im— proving their assessors' performance. Basically, they felt that in the circumstances, the assessors did a good job, but should be given training and supervision in future programs. 3. Feedback Participants perceived that the level of feedback was higher than that of their own initial self perceptions. How— ever, they did not think that this reduced its effectiveness. Generally, they felt that the effect of the feedback was to confirm perceptions they already had of their abilities. In fact, as already mentioned, feedback raised the levels of their 181 own assessment of their abilities. Nevertheless, they also reported that they responded most to these differences be— tween their ratings and their assessors' when the assessors rated them lower than they rated themselves. Implications of the Investigation A. Professional Development of Teachers This investigation has a number of implications re— lated to teacher education, and more specifically, to the professional development of teachers. The recent trend is increasingly to involve teachers in assessing their needs and in planning and even developing their own professional development programs. For teachers to be in the best position to do this they need to base their assessments on the best possible foundations. Objective self analysis is one factor, but by itself it may not be enough. Certainly this investi- gation strongly suggests that teachers' insights into their behavior are greatly influenced by the perceptions and opin- ions of others, even after these insights have been originally established by detailed and prolonged self assessment. Granted the validity of external data, this potent external influence, so strongly rooted in the interactionist psychological tra— dition, ought to be utilized in building the soundest founda— tions for professional development. Teachers, as evidenced by the participants in this study, are attracted to and feel rewarded by self assessment and peer interaction. But they 182 also respond very significantly to additional feedback from external sources, and the self perceptions so formed appear to be stable over time. Where the question of modes of feedback is concerned, this investigation suggests that it is slightly more effec- tive to transmit the feedback through an external agent. It should be noted that the person acting as the external agent in this investigation performed the barest functions usually associated with an evaluator in that interaction was reduced to a stark minimum. It is therefore possible that the dif— ference between external and self analysis of feedback data may be even more pronounced under "natural" conditions. How— ever, teacher educators will want to consider these options very carefully, since the evidence is not conclusive on either side of the question, and the influence of age and experience were not successfully excluded. In the professional development of teachers, various types of dissonance do seem to be potent motivators of teacher change. This investigation suggests that dissonance can be induced between self and other perceptions by using a teacher developed questionnaire, such as the first part of the Teacher Behavior Survey (TBS) developed for this study. Such an in- strument with a positive focus, allows for the collection and communication of clear and detailed perceptions of assessors. Feedback in this form is essential for the development of dis— sonance. Another factor affecting dissonance is the "attractive- ness" of the assessing group. It does appear from this study 183 that if teachers choose their assessors they are most likely to choose persons who for them are "significant others”. As a result, they will not easily derogate these assessors' opinions even when they differ widely from their own self perceptions. Another implication from this investigation having to do with dissonance is that it confirmed Festinger's idea that teachers are motivated to change their perceptions in proportion to the magnitude of the dissonance they ex— perience. The limits of teachers' tolerance of dissonance were not tested. In the absence of this evidence, sensitivity and prudence would counsel against excess. It also appears that assessors chosen by teachers will tend to rate them more generously than the teachers rate themselves. Even so, it does appear that teachers want and value this feedback as a strengthening experience. Teachers also value the opportunity both to engage in self evaluation and to receive external feedback. They seem, too, to wel— come the opportunity to be observed, certainly in an environ— ment where they do not feel threatened, but perceive it as part of an information gathering process contributing solely to their professional development. B. Suggestions for Further Research At the end of the literature review in Chapter Two, a number of outstanding research issues were defined. Some of these issues were specifically addressed by the primary research questions framed for this investigation. A number of issues, (1) (2) (3) 184 nevertheless, remain outstanding. These include: So many of the feedback differential scores were positive that the evidence concerning directionality was one—sided. Positive differential induces posi— tive change, but the effect of negative differential needs further investigation. The question of magnitude of dissonance needs to be further looked into. What are the effects of "maximum possible dissonance” on the perceptions and behavior of a much wider range and larger number of teachers? The attributes of significant others operating as asSessors in the context of professional develop- ment needs further definition. Also, granted that teachers given the choice might well choose signi- ficant others as their assessors,what,neverthe_ less are the criteria on which these choices are based? From the literature, expert assessors also seem very effective as motivators of change, and may be said to operate as "salient" others. What are the differential characteristics of these two groups and do they influence teachers in dif— ferent ways? And are expert assessors more re— liable and valid in their perceptions of teachers' behavior? The nature and effect of different forms of feed- back needs further investigation. Optimum levels (5) 185 of feedback specificity, for example, different categories of teacher behavior and other modes of transmission suggest themselves as fruitful areas for further research. In particular, since this study chose to limit the role of the external evaluator in transmitting the feedback, the ques— tion of the effect of greater interaction, which is usually associated with a supervisory or peer assessor, could serve as the subject of important additional research in this area. The first section of the TBS survey used in this study seems to have been an effective instrument. What would be the effect of using an instrument not developed by the participants, which might then not reflect their value system? Some variables that were dealt with as potential confounding variables in this investigation could usefully be explored as treatment variables in other experiments. What are the effects of sex, other geographical locations, and level of school taught on any of the research questions asked in this study? Also, the effects of age and years in teaching were not totally accounted for in this investigation. They are certainly variables that require further investigation. The stability of self perceptions seems well docu— mented in the literature and by the results of (8) 186 this investigation. The obvious area for further investigation is the effect of longer periods of time. The final question, however, that needs urgent exploration is the effect of changed self percep- tions on specific and general aspects of teacher behavior. It was suggested earlier that in the absence of a valid and supported model of teacher behavior, feedback differential affects perception but not behavior. Is this true only in the short term? If not, then efforts to change self per— ceptions, the assurances of Combs (1965) and the other perceptual psychologists notwithstanding would seem somewhat sterile or at least, limited exercises in improving teacher effectiveness. This is an area of research that needs extensive exploration. 1 Because the question addressed and the findings obtained seemed to have dramatic implications for teacher education, there is a clear need to repli— cate this study. Those who believe that inservice teacher education ought to be built on a firm foundation of teacher insights into their own professionalstrengths and needs should be parti- cularly sensitive tofflfififiiresults. The findings strongly suggest that Combs (1965) is right in arguing that self assessment by itself provides 187 an adequate base for personality change. If this is so, then the conclusion is inescapable that teachers should systematically process feed— back from others as an essential part of those ex— periences that constitute their professional de— velopment. Traditional methods of inservice training have requently included some process of external assess- ment. However the impact of external feedback on teachers may have been underrated by practitioners in this field. The findings of this study suggest that external feedback exerts a stronger and more dramatic influence on teachers than many who have worked in this area may have expected. The implications of these findings are far reach- ing not only for inservice teacher education but also for coordinators of student teaching and for in— school evaluators of teachers. They, too, hold a most potent instrument in their hands: the effect of their feedback on teachers and student teachers may also be more profound than has been imagined. Because of the critical nature of the implications of this study for teacher education, it is important that it be replicated. Other groups of teachers with different demographic features, and larger numbers of teachers ought to be used in order to ensure greater generalizability. APPENDICES APPENDIX A CALENDAR OF EVENTS 188 APPENDIX A CALENDAR OF EVENTS The steps in conducting the experimental investiga- tion may be briefly summarized as follows: I. Assigning Participations April 16, 1979 : Selection and assignment of participants II. Administering Pre—, Post—, and Delayed Posttest (These took place during three periods of approx— imately one week each, six weeks apart). April 16—25, 1979 : Pretest given to first year City A group on April 16, to the first year City B group on April 18, and to the second year City A group (EIIG) on April 24. May 28-June 6, 1979 : Posttest completed by participants in Experiment I between May 28 and June 6. Participants in Experi— ment II (EIIG) completed theirs on June 5. July 10—19, 1979 189 Delayed Posttest and Attitude Scale were handed out in the last class of the term during the week of June 3, with in- structions that they be kept until July 10 and completed before July 17. III. Experimental Treatments April 16, 1979 May May May May May 10 14 16 I I I 1979 1979 1979 24 and 29 30 and 31 Questionnaires were given to the participants in Experiment I to be distributed to their administrators and peers. Two and a half weeks was the time suggested for the data collection. All questionnaires had been re— turned by the assessors. Analysis of data began for CityA group Analysis of data began for City B group Interviews for City A group Interviews for’ City B group APPENDIX B TEACHER BEHAVIOR SURVEY (TBS) 190 APPENDIX B TEACHER BEHAVIOR SURVEY (TBS) Questionnaire on Teacher's Strengths PLEASE READ THE INSTRUCTIONS AND SUGGESTIONS IN THE ACCOMPANYING LETTER BEFORE FILLING OUT THIS QUESTIONNAIRE Teacher's name Grade Level or Subject Taught Please assess the teacher named above in terms of each of the following statements, and assign a mark from 5 to 1 in each case. 5 Exceptional (This teacher is in the top 5% of teachers at this level or in this subject) 4 = Strong (This teacher is in the top 15% of teachers at this level or in this subject). 3 2 Above Average 2 = Below Average 1 = I have had no opportunity to observe or know this. A. Planning This Teacher: 5 4 3 2 1 1. Makes plans that are flexible 2. Plans to meet the needs of individual students 3. Aims to involve all stu- dents in meaningful work 4. Can integrate different materials, methods and in— terests in a plan 5. Makes good short term and long range plans B. Organization 191 Teacher‘s Name: This teacher: 5 and Management 6. C. Classroom Climate and Control 10. 11. 12. 13. 14. 15. 16. Firmly establishes the rules of the classroom so that students know expectations and conse— quences Recognizes needs and problems early, and deals with them before they de— velop into major matters Attends to the special needs of indidivual stu— dents Is consistent in estab- lishing the day to.day patterns of the class Is flexible and adapt— able Helps students to build positive self— concepts Is fair and consis- tent in the classroom Communicates clearly and effectively with students Is flexible and adjusts to the needs of individ— ual students, and to the differences between groups and classes Is pleasant and relaxed in dealing with students Organizes classroom space in an attractive and stim- ulating manner, for example: bulletin boards, interest centers, materials available to students, etc. D. Command of E. Subject Matter Teaching Method 192 Teacher's Name: This Teacher: 5 4 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Adapts knowledge to the ability level of students Keeps up with current ideas and information in subject (3) Has personal interest in subject matter and communicates this to students Has inclusive knowledge of the subject matter including the content of last year's and next year's syllabuses Can admit ignorance and is willing to learn along with the class Uses a variety of methods and materials appropriate to different students, subjects and situations Includes all children, regardless of ability Communicates clearly and effectively with students in both speech and writing Communicates a sense of the purpose of what is being taught Keeps up with new develop- ments,and is willing from time to time to giVe a careful try to a new ap- proach F. Use of Audio- Visual and other materials G. Understanding Human Growth & Development 193 Teacher's Name: This Teacher: 27. 28. 29. 30. 31. 33. 34. 35. 36. Uses audio—visual and other suitable materials to enrich the curriculum, help students understand, or excite their interest Uses a variety of these materials Uses audio—visual mater— ials to engage more fully the students' sense in the learning process Has clear goals and sticks to them in using audio— visual and other instruc— tional aids and materials Prepares students before using audio-visual or other instructional mater— ial in class, and follows up on its use Understands and cares about human feelings Helps students to de- velop self-confidence Has expectations that neither exceed nor fail to challenge the stu- dents' levels of de— velopment Is familiar with the unique cognitive and social needs of students at the age he/she is dealing with and adjusts instruction accord— ingly Is aware of the home environment of students and its effects on them H. I. Understanding Curriculum & Curriculum Development Evaluation Procedure 194 Teacher's Name: This Teacher: 5 37. 38. 39. 40. 41. 42. 43. 44. 45. Knows in detail the objectives for the grade level above and below for each subject (or grade) he/she teaches Modifies curriculum guidelines to suit the needs of particular class (or classes) Uses pretesting to find where the students are Sets clear, reasonable objectives for each grade level (or sub- ject) Knows federal, state and local guidelines, and establishes prior- ities congruent with their goals and ob- jectives Uses a variety of eval— uation procedures ap- propriately and is not restricted, for example, to written tests Consistently evaluates own teaching Interprets results of student evaluation for the benefit of students, for example, by putting comments on papers and offering suggestions for improvement Makes sure students understand the evaluation procedures and what is expected of them J. K. Working With People Personal & Professional Characteristics 195 Teacher's Name: This Teacher: 46. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. Sees and uses tests as teaching instru— ments Is open and honest in his/her attitudes and dealings Avoids negative pre- conceived ideas about students or parents Is a good listener. Respects others-—stu- dents, staff, and parents Is sensitive and tactful Has a positive attitude toward teaching Relates to students, parents, community mem- bers and other teachers in a professional man- ner Respects professional confidentiality in dealing with both stu- dents and parents Is sincere and en— thusiastic Has a good sense of humor Is open to new ideas Seeks and/or accepts constructive criticism of his/her professional ideas, attitudes and hehavior 196 ON COMPLETING THIS QUESTIONNAIRE: List in rank order what you consider the 5 most significant strengths of this teacher. (Please be as specific as pos- sible, but you need not restrict your— self to the categories or statements used in this questionnaire). 1. 197 II. Questionnaire on Teacher's Possible Areas for Improvement PLEASE FILL OUT THIS QUESTIONNAIRE AFTER COMPLETING THE QUESTIONNAIRE ON TEACHER'S STRENGTHS Teacher's Name Grade Level or Subject taught Assess the teacher named above in terms of each of the following statements, and assign a mark from 5 to l in each case. strongly agree agree disagree strongly disagree = I have had no opportunity to observe or know this HPQOJbtfl ll This teacher needs to: 5 4 3 2 l A. Planning 1. Plan more adequately to meet the needs of the students 2. Have clearer objectives and priorities 3. Plan less rigidly and be more flexible 4. Plan more carefully the use of his/her own and the class's time 5. Keep an adequate lesson plan book for substitute teacher to use B. Organization & Management 6. Plan for and establish the sort of classroom routine that makes students feel more secure 7. Be more consistent in discipline and daily work patterns 8. Use fewer seat-work assignments and other types of busy work C. Classroom & Climate Control D. Command of Subject Matter 198 Teacher's Name: This teacher needs to: 9. 10. 11. 12. 13. 14. 15. 16. 17. Adapt materials and/or methods to the needs of individual students to a greater extent Have a more poised and relaxed manner of inter- acting with students Build up the dignity and self—worth of students to a greater degree Recognize and under- stand the needs of in- dividual students Be more consistent and strive to find the happy medium Relate to all stu- dents in a more positive manner Organize the classroom space in a more attrac— tive and stimulating manner, for example: bulletin boards, inter— est centers, materials more readily available for students, etc. Be better able to present subject mat— ter at students' level Have a more adequate background in subject matter B. F. Teaching Method Use of Audio— Visual and other materials 199 Teacher's Name: This teacher needs to: 5 18. 19. 20. 21. 22. 23. 24. 25. 26. Be willing to cover all aspects of the subject including areas of the curriculum where he/she might feel in— secure Be more curious about new methods and develop— ments in his/her area More adequately under— stand and model the ideas or skills being taught Be careful in practice not to expect the same from every child Be less rigid and in— flexible, and use a wider range of methods and materials Make learning more enjoyable Be better able to in- tegrate a variety of techniques into his/ her own style Focus more on teaching the students and not exclusively on teaching the books (or mater— ials) Be careful not to over- use audio-visual ma— terials, especially films G. Understanding Human Growth & Development 200 Teacher's Name: This teacher needs to: 5 27. 28. 29. 30. 32. 33. 34. 35. More carefully cor- relate the use of audio—visual, or other aids and materials, with the aims of the lesson Prepare the class be— forehand for the use of audio-visual or other instructional material, and follow up on its use Use more audio—visual or other aids and ma— terials, and/or a wider range of these materials Be more aware of the potential of audio—visual materials and other aids in the teaching process Be more aware of growth and develop- ment as important factors in the learn— ing process Be more aware of in— dividual needs and dif— ferences and treat children more as in— dividuals Adjust his/her teach— ing techniques to the level of the students Be able to cope more adequately with special learning or behavior problems in class Have more empathy for the emotional changes students are under— going I. Understanding of Curriculum and Curriculm Development Evaluation Procedure 201 Teacher's Name: This teagher needs to: 5 37. 38. 39. 40. 41. 42. 43. 44. Be more aware of and pay more atten— tion to the require— ments of what is being taught above and below the subject (or grade) level he/she teaches Pay more attention to curriculum objectives and/or more adequately integrade them with his/ her own perceptions of subject needs Have a better knowledge of relevant guidelines and objectives Set objectives or de- fine priorities more clearly More consistently eval- uate the curriculum and make modifications in it Make more adequate provisions to evalu— ate own teaching, and/ or be more responsive to evaluation feedback Know better how to in- terpret evaluation re— sults and how to give helpful feedback Evaluate students more frequently and com— pletely Be less affected by whether he/she likes a student or not J. K. Working with People Personal and Profess— ional Char- acteristics 202 Teacher's Name: This teacher needs to: 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. Be more familiar with a wider variety of evaluation procedures and instruments Show more tact or concern for the feel— ings of others Be less negative and apathetic Have a more open mind and be able to admit mistakes . Avoid showing favorit— ism and/or unfairly labelling students Be more reliable Be more careful not to reveal confidential information without proper need Show less concern for the paycheck and/or other benefits Be less narrow minded and lacking in concern for others Have less difficulty working with other teachers and/or parents Avoid coming down to the level of students, for example, by arguing With them or trying to be a "pal" 203 Teacher's Name: This teacher needs to: 5 4 3 2 56. Comply more completely with rules set by ad- ministration and/or staff ON COMPLETING THIS QUESTIONNARIE: List in rank order 5 specific areas in which you think this teacher might most need to improve. (Please be as specific as possible, but you need not restrict yourself to the categories or state- ments used in this questionnaire). THANK YOU VERY MUCH FOR YOUR HELP: APPENDIX C DEBRIEFING FOR PARTICIPANTS IN THE TWO-TERM MACT PROFESSIONAL EVALUATION 204 APPENDIX C DEBRIEFING FOR PARTICIPANTS IN THE TWO—TERM MACT PROFESSIONAL EVALUATION N.B: When completed please return separately in the stamped- self—addressed envelope providedll 1—7: To what extent have each of the following activities in the two—term evaluation program influenced your relative strengths and shortcomings as a classroom teacher? 1 = strong influence 3 = limited influence 2 = moderate influence 4 = little or no influence Please circle the appropriate number in the box on the right 1. The process of deriving the list of teacher strengths and needs in the first term 1 2 3 4 2. The self assessment done in the first term 1 2 3 4 3. Feedback from the first section of the questionnaire ("Teacher's Strengths”) provided by your professional colleagues in the second term 1 2 3 4 4. Feedback from the lists of strengths cited by your col- leagues at the end of the first section of the questionnaire 1 2 3 4 5. Feedback from the second section of the questionnaire ("Possible Areas for Improvement") provided by your professional colleagues 1 2 3 4 6. Feedback from the lists of your possible needs for improvement cited by your colleagues at the end of the second section of the questionnaire 1 2 3 4 205 The process of thinking through and writing up the summary/outline at the end of the evaluation 1 2 3 4 8-24 Please indicate your level of agreement with each of the following statements by circling the ap— propriate number in the box to the right. 1 2 strongly agree 3 agree 4 — — — — disagree strongly disagree Because members of the control group had less to do, I initially felt some frustration/displeasure in being assigned to an experi— mental group 1 The benefits I received from being a member of an experimental group adequately compensated for the extra work I had to do 1 10. The experimental procedure was too long and repetitive l 11. I made considerable gains in my ability to assess my own perfor— mance as a result of this exper— ience l 12. 13. 14. 15. I made considerable gains in my ability to assess a collegaue's performance as a result of this experience 1 I am more sensitive to short— comings in my classroom perfor— mance as a result of participating in this project 1 I am more sensitive to strengths in my classroom performance as a result of participating in this sproject l The benefits teachers might de— rive from some form of systematic evaluation in the MACT Assessment I course will depend in large part on the interpersonal/communication skills of the instructor who directs the experience 1 16. 206 The two-term evaluation procedure in which I participated should be- come a required part of the MACT Assessment I course .1 2 3 4 17. All teachers should undergo some form of professional assessment at least once every five years 1 2 3 4 I felt comfortable and relaxed during the classroom sessions in which I was being observed by my yprofessional colleagues 1 2 3 4 19. The process of being observed and assessed by professional colleagues was a valuable ex— perience in and of itself 1 2 3 4 20. I would have taken the feedback I received more seriously if the assessors had spent more time in the classroom observing my per- formance 1 2 3 4 21. The assessors took their job seriously and did it conscient— iously l 2 3 4 22. There were few if any significant disagreements between my own ratings and those provided by my professional _ colleagues 1 2 3 4 23. The fact that the assessors tended to be very generous in their ratings reduced the value of the feedback experience 1 2 3 4 24. I have a higher regard for my own abilities as a teacher as a result of this experience 1 2 3 4 25. 25-28 Please select the alternative that best describes your response to each of the following statements. In each case circle the number to the left of your answer Overall the assessors tended to rate me...I rated myself 1. higher than 2. at about the same level as 3. lower than 26. 207 The extent to which the assessors agreed among them- selves in rating my performance of specific skills was 1. very high 2. high 3. low 4. very low 27. I was most sensitive to those differences between my ratings and those of my assessors when they rated me 1. higher than I rated myself 2. lower than I rated myself 28. The most significant outcome of the feedback I received was 1. identification of some weaknesses I did not know I had 2. identification of some strengths I did not know I had 3. confirmation of the perceptions I had of my abilities prior to this experience 29—33 Please complete the following statements. PLEASE TYPE OR PRINT YOUR COMMENTS IF POSSIBLE. 29. The most valuable aspect of the two—term evaluation program was 30. The least valuable aspect of the evaluation program was 31. I would recommend that the two—term evaluation procedure should be required of all MACT candidates if, and only if, the following change(s) is (are) made: a) b) c) d) 32. I like 33. I did not like THANK YOU VERY MUCH!!! APPENDIX D LETTER TO ASSESSORS WITH INSTRUCTIONS FOR COMPLETING THE TEACHER BEHAVIOR SURVEY 208 C/o Dr. John Cragun Battle Creek Teacher Education Center Miller-Stone Building 77 Capital Avenue, N.E. Battle Creek, MI 49016 April 16, 1979 TO: School: For (Teacher's name) Thank you for agreeing to fill out the two attached questionnaires. We greatly appreciate your cooperation. You are participating in an exercise that has two purposes. First of all, you will be helping the teacher named above to complete the needs assessment of profes- sional abilities which is part of the Master of Arts in Classroom Teaching Program which this teacher is currently involved in. The teachers on this program rely on the re— sults of their needs assessment to help them to choose courses and to select their own special professional develop- ment projects. Secondly, you will be helping a research project which is being done by this researcher at the Battle Creek Area Teacher Education Center, School of Education, Michigan State University. The aim of this project is to compare the relative efficacy of two kinds of feedback—- self and external——in.aSSiSting teachers to evaluate their needs for the purposes of their own professional development. Please follow the following procedure in filling out the questionnaires: 1. Read the questionnaires carefully to see the sort of information required. Then, during the next two and a half weeks, which is the recom- mended length for the period of this assessment: 2. Visit this teacher and observe him/her teaching for about one hour, or, for high school teachers, one full class period. This observation may be done over two or three visits, but should total the time specified above. 3. Look at teacher's lesson plan book, samples of tests given, samples of students' work, teacher's records and reports, teacher prepared materials, and other teacher and student products which can -assist you. 209 4. Have informal talks with the teacher 5. When you have completed your periodwOf assess- ment, complete thesquestionnaires as quickly as possible. Please note that in answering the questionnaires you should draw an overall knowledge of this teacher over the period you have known him/her as a teacher. Please also note that the results of these question— naires are to be kept annonymous. Only the researcher will know who has filled out each set of questionnaires. To ensure this very important condition, please enclose the completed questionnaires in the stamped addressed envelope provided and return to me. I will then transcribe your re— sponses onto another questionnaire form which will then be seen by the subject of the questionnaires. You are one of three evaluators for this teacher whose questionnaires will be treated in this way. Finally, may I ask you to be certain to complete and return these questionnaires by: THANK YOU VERY MUCH INDEED!!! Yours Sincerely, Donald G. Wilson APPENDIX E DESCRIPTION OF ASSESSOR 210 APPENDIX E DESCRIPTION OF ASSESSOR~ Please fill out and return this sheet with the completed questionnaires. 1. Name School Position Have you done a formal evaluation of a colleague be— fore? If yes, about how many times? How long have you known the teacher being evaluated since he/she has been a teacher? Do you teach/work in the same building? If yes, for about how long? Have you had the opportunity to observe this teacher in the classroom before? If yes, please briefly de— scribe the nature and extent of the previous observa— tion: How many hours/class periods did you spend in the last three weeks observing this teacher prior to completing these questionnaires? How many times have you had discussions with this teacher for the purpose of the present exercise? About how long did these discussions last? 10. 11. 12. 211 What teacher and student products (e.g. lesson plans, tests, records, students' work, instructional material, etc.) did you look at in the course of the present evaluation? (Please list these) How long have you been in teaching? How long have you been in your present position? THANK YOU VERY MUCH!!! APPENDIX F SUMMARY OF DATA ON ASSESSORS +O ON 2H.zm.m.s.ms OHI O N O O mwscmzscmz HOQHUOHMO NH N NN OmOmHH mm HHO OHIOH O Oscmz m mm OOO HmeHucHHe NH N Om OoumHH mm HHO OH H N O O Oscmz .cHHO.pmmO HH A NN 2H.zm.s.es ONIOH O ..scmz OH OH OOO HmnHocHuO OH O OH zm.zH.m OH H H 36m ON Newscmz .cHHO .Hmme O O ON .m.es.zm.mm HmHnm N m awe N Nscmz HmaHocHHO O NH NN OmOmHH mm HHm OH H H +O O Oscmz HmnHocHHe s xHH OOH 2O.9.OO OH N H m H .Hscmz HmmHucHum O OH ON OopmHH mm HHO O O Oscmz N Oscmz HOQHUOHHO O O OH m.B.mq ON N O >HHm O O >cmz cmEHHmco IDCCO pcmspnmawo O m NH mm.zm .mpmms.ma ON H .Hscmz N m +OH pcmHHsmcoo CHOHCH m 2 H OH HN OmpmHH mm HHm ON m N s N N New: HmnHuaHHO N O OH m.zH.ea O H I O m m +OON HmmHUCHnO H . IA. . . . . . . . . . ddA 98 0.3...“ )T..I HGST..d IOSA MA 81 duv HL 013 BB n13 NUT. STOUGSGIESUBSI. OS A9 SSE 31 one I. w .151 JBSNBOEJM SS 99 T781 UIS 1.0.3 U09 pa A1.T..0881 M1 08 1.9 TO 1.98 T. 890. HT. nEUIO UI1 SUS 1S 1.S nu. TU UT. US 1.39 S 1.1.0.8A8 a T..S Be 01...... OH 31.1 83d S T..SS lent. 1m V. 00 1.1 u u a /, s n a I o PTyu a s u 1 a 8d (SH 0 UOS H I B S I P u 1 s 1 u QAO a D. 3 e S O I. S .8 I. I. U.. S _ O ..u 0 u a S ..u 1.1 b 1 a e P mpnwom chmHHDmumm sooummmHoumO .mpuouwm w.n®nomm9um .xsoz m_ucwosumnzm .mpmwsns .mHMHHmez HmcoHHUSHHmcHqu .mcmHm comma; Hm; mOBHmm HmHumeO mo HOHOCHUHOOO OH OH OH mmemm.zm.m.s OO N O mm» OH OH O Hmzomms OH OH OH OO.O.2O.2H.OO OHI OH O O mm» OH OH I Hanumms NH H mO .O.OO OO O O was O O I Ommm.uaoa HH H O 2H.2O.O.OO OH O +H oz H H I nmeomms OH HH HH 2H.2O.Oa OH N H mm» N O O Ommm.Hnmo O mH ON .m.9.mq I I I 02 N N O Hmzumme O O O I I H H oz O O I Hmaomms N O O H.OO OH O H mm» O\N H I Hmcomms O O O 2H.zm ONI OH O O mm» O +O I Hmeomms O O O s OHI O H O mm» O O N somoo O OH ON 2H.3O .OO ON O O mm» N N I Hmsumms O OH OH mm.zm.2H.mq Om O N mm> N N O Hmnomme N m O I OHI m N N mm> m m I Honmcsou H O O SdSvfl. )Ta. 0 a. O O O O 0 ad... .1... 9...; Nun mm OOOOO mm mm Om. OO. MO 018 88 80h? 1.9m 18 1 188 1.9 08 13w 88 Be 881 BE Uppo U1.8 PD A1.1.D8 ...-.1 M1 08 1.8 To 1.88 01 man. n n EU.OU.1 0.8 US 18 1.8 nu. «Os 8 HHO o u.O OMOO. HO 4.Inya A t. .O 11: e e Tru I. 141+; enud s Trshswle u_I 1 V oHO 4.1 OH+I. u?! s /, OOS.O I. o D. Run a s UHO e "u u sbu IIs u o "no she e s I p. 114 u wanw W D e s O 8 .U0 m U. 8 u «+1 9 a s wpumom :HumHHsmumm .EooummmHoumo .mcuuowm m.um£omuwum .xHoz .1 m m.uco©5umnzm .mpmmaue .mHmHHmum: HMCOHuUDHumcHHZH .mcmHm commmqumq umpospoum ucmpnpm\um£omme mmmU.<.2H.B.mq OO Ocmz >cmz new OH OH I Honomme ON O O OmomHH mm HHO OH H Ocmz mm» O O N Hmsumoe ON O NN empmHH mm HHm OO H I new O O +O pmom.uaoa NN OH OH 2H.s.Oa OH H H mmw OH N N nonomme HN O NH 2H.zm .OH ON N O oz N N I Honumms ON 5 NH OO .m.o.m.s.OO O H I mm» H H O nonomoe OH H O O 30.m.mm.e.mq + pcmuHsmcou OGHUmom .mvchSHO ONI OH O N mo» O\m O\m I .200 OcHUmom OH O O UmpmHH mm HH< ON m N no» O OH I Hosomoe OH OH OH pmpmHH mm HHO ON O N mo» O O O Hosomoe OH 6.0x“ 1.5 ,S.ac.£.1nrtim u.. s . . . . . . 0.19 a.O 9.1OOenwOeu_I a.m.evmmmmw%mwmfiu mmfl WWW WWW WNW sass. e e .Onvn errs. w 7.9 .1 1 efisr:mne o 2 15m sHs e e 11:: N51 unvplsu o 9 Own AulOyolOT.a.1 M.l oHO 1.3 Ola las Us neqnn n eqquP 3 us 1:: ls nq Tru I. «ou1e1.s s S «OI;P=.AOI e 7.8 p.e 0.4T: UT? 1+1;18 8 d S T:SOO1HOU Tr TO V OAu 111 u u ,bnu s [/9 I. a T: o O. Pay u a s u 1 a (.0 u o "uniwmw e s I P u 1 u q 1. HO 9 s s . uHu H s «Om a s mvnmom cHumHHsmnmm .EooummmHoumU .mpnooom m.uonomoaum .xuoz 1 W .mpmoane .mHmHHOHOE HOCOHuoanumcHqu .mcmHm commmquma "mposcoum pump:pm\uonomms m.pooO:uOuzO mom N O 1H5 mmmo.zH.e.Oa OI O O O\O mow N N I Hmcomos O OH ON 2H.zO.OH OO N N mow N N NH Hmcomms O O OH .zO.s.OO ONI OH O +H OOH O O O Omeomos O m m 3m.nq OH O N mm» H H I Hocomme O O O 2H.zm.m.s.en OH H O mm» O O I Hanomoe O H ON .m.9.ma OH O Bom mow H H Inomoo HHmQHoom O Honumos.m.e O O O 2H.2O.OH OH O H oz I N I .coo ocHOmom O 6 O O OOHOHH mm HHO ON O N mmw N N I Honomos N l 2 O O Oq.zm.e.zH O H 3mm mm» O O I HoumcHmemm H O HO: .ddfi TOK. SdSL )1.L. U.G.81. .QQ.8..A.X.A. HUI. dun. 31 OOO ... ...... ...... ......0.00000 OO OO 2 Trs 1 w m anunue I. m 7.3 .1 1 GAO 7.1 MOI o a 1.9 7.3 1.8Hs “no ”.mgwmw mmwe “BMW mmwmmeM ”we "us. 718 118 n q met. m. Osu.e 18.3 s .+7:Pnua "at. TO 9 V mwm ”Om u U 5m. m1” fimw ”wusswnrbu e s u: e (EOU O "OOHS L S w % Hm P col. u QAu 8 E q q. 1. 8 .8 B m 8 8 O U D 8 1 8 U 1. w. P moumom OHumHHsmnOO m . #GOUSflWHHSm .mumoBuB .EOOHOOMHUumU .meHHopmz HmcoHuosuumcHqu 1N1 Ommmoams mmmm .mpuouom m.umsomo9um .xHoz mm0wmmmm< ZO