:3: I -. -—“— -— no.— a v . c‘ v “0.... .. v». “—— -‘ , ~u— ‘1‘, n: P _ .q -." . 'r‘III-IIH : g: I I IIiIfII‘ II IIIIIIIIIZ'. I?!“ II II ‘ 'H. t’:' ‘,‘; 3.3.?“ III. IIIIIIIh ’IIII I II IIIJJ: II I '0‘ ”III II IIIIIII Mill} 1‘ IIIII '1' ”I" “WEI, .,“ IIII WI hf afalI 4, «III my [”WI III IIIIJIIII' I ‘I 'III' M “III!" ’1.) THESIg E. .. 1w... Wait-“WAS“ I1) I? in g I: £1.83 A It; ’2." C . “fiafifl 552;? ' Ufim I“ I}. If Thisistocertifythatthe dhnflummamfihd A PSYCHOLOGICAL STUDY OF TEACHERS‘ AND SCHOOL PSYCHOLOGISTS' PERCEPTIONS OF STUDENT REWARD PREFERENCES pmmmudby Jim Charles Van Treese lushnnqumwmmnnkfimmhmmt (Kmnundnmwmbfir Ph.D. in School Psychology man» ‘0 IQWI/fi” I 5 7"” fl Inu_§/12/82 ”Slit-WWW”- 042771 .— ” IIIIIIIIIIIIIIII 01 IIIIIIIHIIIIIIIIIWIIIIII " 3 1293 00992 6142 MSU RETURNING MATERIALS: Place in book drop to LIBRMuEs remove this checkout from ”In. your record. FINES will be charged if book is returned after the date stamped beIow. rga mom ‘ AUG 1 1 2000 A PSYCHOLOGICAL STUDY OF TEACHERS' AND SCHOOL PSYCHOLOGISTS' PERCEPTIONS OF STUDENT REWARD PREFERENCES BY Jim Charles Van Treese A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling, Educational Psychology and Special Education 1982 ABSTRACT A PSYCHOLOGICAL STUDY OF TEACHERS' AND SCHOOL PSYCHOLOGISTS' PERCEPTIONS OF STUDENT REWARD PREFERENCES BY Jim Charles Van Treese The study was designed to investigate female tea- chers', male school psychologists', and female school psychologists' perceptions of male school children's reward preferences by grade and behavioral characteris- tics of students. The main purpose of the study was to investigate possible age-related changes in male school children's reward preferences. The secondary purpose of the study was to compare the degree of congruence among the three groups of res- pondents' perceptions. A typology for reinforcers called Categories of Teacher Controlled Positive Reinforcers (CTCPR) was vali- daded. The CTCPR category names were: Concrete Edible Rewards; Concrete Non-Edible Rewards; Redeemable Symbolic Rewards; Non-Redeemable Symbolic Rewards: Oral Communication Rewards: Written Communication Rewards; Close Body Communication Rewards; Distant Body Communi- cation Rewards: Activity Rewards; Responsibility Rewards; Escape Rewards: and, Intrinsic Rewards. A pair comparison reward preference scale was devel- oped based on the CTCPR categories. The CTCPR pair compar- ison scale was presented to the three groups of respondents with grade levels (i.e., K, 3rd, 6th, and 9th) and student behavioral characteristics (i.e., average achievement, just barely passing achievement, overcontrolled behavior. and undercontrolled behavior) varied. The resultant category rankings were analyzed by multivariate analyses of vari- ance for the omnibus hypotheses and by univariate analyses of variance to examine the variation of each category within each omnibus test. The results were discussed in relation to previous research. Recommendations for future research were made. DEDICATION This dissertation is dedicated to my parents, the soil and Sun of my early life: and, to my wife and children. the joy of my present and future. ii ACKNOWLEDGMENTS The author wishes to thank the members of the author's committee Dr. Harvey Clarizio, Dr. Andrew Porter. Dr. Walter Hapkiewicz, and Dr. Ellen Strommen for their help with the formulation and completion of this project. Each gave invaluable assistance. The author wishes to credit doctoral candidate Gabriella Belli for her assistance with data analyses issues. The author also wishes to thank the many teachers and school psychologists who served as respondents. iii TABLE OF CONTENTS L181:OfTableSooeoonococococoon.00000000000ooooooooooViii LiSt OfFigureSOOOO0.0.0.000....0.000000000000000000IIOix CHAPTER ONE: TWO: INTRODUCTION. ........ . ..... . ........... .......1 Purpose of the Study..........................1 Statement of the Problem......................2 Research Objectives...........................9 Potential Contributions......................10 Chapter Summary..............................12 REVIEW OF THE LITERATURE.....................13 Section one: Measurement Theory.............13 Three Methods of Scaling Reward Preferences...............................14 The Problem of Scaling....................1# The Pair Comparison Method................15 The Method of Rank Order..................22 The Rational Zero Point Method............23 Example One: Researcher "A"..............25 Example Two: Researcher "B"..............26 Example Three: Researcher "C"............27 The "Solution"............................3O Section Summary...........................31 Section Two: Categorization Scheme Develop- ment and Instrument Development...........32 The PROS and its Predesessors.............32 .Reliability and Validity..................37 Other Systems of Categorization...........h0 iv THREE : The Development of a New Categorization Scheme: Categories of Teacher Controlled Positive Reinforcers (CTCPR)...............h1 The CTCPR's (In Preliminary Form) Validity...................................51 Reliability of the CTCPR Pair Comparison Scale......................................52 Section Three: Review of the Research on Age-Related Changes in Reward Preferences..55 Ware's Research............................55 Van Treese's Research......................57 Chapter Summary...............................59 RESULTS ..... .... ...... .................. ...... 60 Section One: The Construct Validity of the CTCPR Categories...........................6O Section Two: Teachers' and School Psycho- logists' Perceptions of Student Reward Preferences................................63 Design of the Study........................63 Purposes of the Study......................67 Hypotheses.................................67 Analysis...................................68 Respondents................ ..... ...........69 Procedure..................................7O Rationale for Including School Psycholo- gists as Respondents.......................71 Rationale for the Expansion of Grade Levels.....................................71 Rationale for the Selection of the Control Variables..........................71 Rationale for the Selection of Student Characteristics............................72 FOUR: FIVE: Results...................................73 Survey Return Rate......................73 Reliability Data........................7# Findings for Hypotheses.................76 Chapter Summary..............................83 DISCUSSION..... .............................. 84 Age-Related Effects.......................84 Intrinsic Rewards.......................8h Concrete/Symbolic and Immediate/Delay Dimensions..............................85 Praise..................................87 Other CTCPR Categories..................87 Type of Respondent and Student Characteristic Factors..................88 Chapter Summary...........................92 CONCLUSIONS AND RECOMMENDATIONS... .......... .9“ CTCPR Categorization Scheme...............9h CTCPR Pair Comparison Scale...............94 Age-Related Changes.......................95 Type of Respondent...;....................97 Student Characteristic....................97 Implication for Practice..................98 Limitations and Recommendations for Future Research...........................99 Chapter Summary..........................101 ReferenceSCCOOOOOO00.0.0000...OOOIIOIOOOOOOOOOOOCOOOO.102 Appendices A. B. C. Original PROS Pair Comparison Scale...........107 PROS Pair Comparison Scale as Reworded for Adolescent Delinquents....................112 CTCPR Category Validity Study.................116 vi D. Rater Reliabilities of Reinforcers...........128 E. surveyFomSOOOOOOOOOOOOOOIOO0.00.00.00.000001u2 vii TABLE 1. LIST OF TABLES Positive Reinforcement Observation Schedule (PROS): Definitions and Symbols...............34 Categories of Teacher Controlled Positive Reinforcers "" Preliminary Form. 0 o I o o o o o o o o o o o .44 Comparison of the CTCPR - Preliminary Form with other Categorization Schemes..............48 Revised CTCPR Category Symbols and Definitions.53 Validity Data on the CTCPR Categorization SChemeOOOO00.00.000.000...000.00.00.0000000000062 Experimental Variables.........................64 Control Variables..............................6# Design of Study................................66 Reliability Data "A"...........................75 Reliability Data "B"...........................75 Summary Table for Hypotheses...................76 viii LIST OF FIGURES Discriminal Dispersions of Ungrouped ReinforcerSOOOOO0......OOOOIIIOOIOOOOOOOOOO00.19 Discriminal Dispersions of Grouped Reinforcers...ICOOOOOIOOIOOOOOIIIIOOOOOOOIIOIOZO Categories by Grade over Type of Respondent and over Student Characteristic...............78 Categories'by Student Characteristic over Type of Respondent and over Grade.............80 Categories "A” and "R" Plotted by Type of Respondent by Student Characteristic over GradeOCO.0......0..O.OOOIOOIOOOOOOOOOOOOOOIO.082 Categories "WC", "DBC" & "A" Plotted by Student Characteristic over Type of Respon- dent and over Grade...........................91 ix CHAPTER ONE INTRODUCTION This chapter begins with a brief statement of the purposes of the dissertation. The statement of the pro- blem follows. The statement of the problem is carried out within the context of an overview of the literature. This chapter concludes with a statement of the research objectives and some of the study's potential contribu- tions. Purposes of the Study The study is designed to investigate female teach- ers', male school psychologists', and female school psy- chologists' perceptions of male school children's reward preferences by grade and by behavioral characteristics of the students. This design allows the establishment of three independent indicators of possible age-related changes in male school children's reward preferences. The use of the phrase "age-related changes" is not in- tended to imply ontological changes. The latter point is discussed in Chapter Three. 2 The main purpose of the study is to investigate possible age-related changes in male school children's reward preferences. These possible changes will be inferred from the three groups of respondents (i.e., the female teachers, male school psychologists, and female school psychologists). The secondary purpose of the study is to compare the degree of congruence among the three groups of res- pondents' perceptions to infer if they differ in their perceptions of the desirability of the specific types of rewards by grade levels and by student behavioral char- acteristics. Statement of_the Ppgblem The statement of the problem is carried out within the context of an overview of the literature. The liter- ature on age-related changes in reward preferences perti- nant to the main purpose of the dissertation is presented first. A brief mention of the research relevant to the secondary purpose follows. Research on age-related changes in reward prefer- ences is a relatively new research area which is gener- ating several distinct, but complementary, lines of re- search. One line of research has found evidence that reward preferences for specific reward objects may change with increased age, and that learning experiments which use 3 the same reward object over different age levels may be misinterpreting age-related changes in reward preference (i.e., incentive value) as developmental changes in learning ability (Haaf & Smith, 1976). This is an im- portant line of research. but since the focus is on developmental trends in learning and not on age-related changes in reward preferences per se, it is not likely to illuminate general age-related trends in reward pre- ferences. A second line of research has been the study of age- related changes in reward preferences for reinforcers which vary on dimensions such as "immediate/delay" and "concrete/symbolic" (e.g., Harter. 1967; Harter & Zigler, 197k: Mischel & Mitzner, 1962: Nisan, 197A; Walls, 1973: Weisz. 1978). This line has generally established both delay in gratification and preference for symbolic re- wards to be positively related to develOpment. This important line of research has yet. however. to address possible age-related changes in preferences for commonly used categories of reinforcers like social, edible, token, and so on. A third line of research is moving in that direction. The third line focuses on age-related changes in preferences for commonly used categories of reinforcers. A major conceptual and methodological problem in the third line concerns deficiencies in existing reward categoriza- tion systems. It is clear any research in the area of L; age-related changes in in categories of reinforcers is limited by the adequacy of the categorization system which is employed. To the degree a particular categor- ization system fails to assign all members of particular population of reinforcers to relatively homogeneous cat- egories in a reliable manner, any research based on it will be limited. Also. no nomological system can be evaluated apart from the purpose to which it is put. This, of course. is a specific reference to the general problem of developing any typology (Reynolds, 1971). The third line of research has been primarily re- stricted to investigating adult perceptions of school children's preferences for various categories of rewards commonly administered in the school setting. Bersoff and Moyer (1973. 1976) developed a categorization system which has been used as the basis for the development of an instrument called the Positive Reinforcement Obser- vation Schedule (PROS). The PROS has been used as an observation schedule and as a category preference scale (Bersoff & Moyer, 1973: Byalick & Bersoff, 197A: Dereven- sky & Rose. 1978). Bersoff and Moyer developed their cat- egorization scheme in response to two shortcomings which they identified in similar categorization schemes. The similar categorization schemes either failed to include categories of nonverbal reinforcement (e.g., Amidon & Flanders, 1967: BrOphy & Good, 1969) or were otherwise not inclusive enough to embrace all possible types of 5 reinforcement emissions in the schools (e.g., O'Leary & Becner, 1967). While Bersoff and Moyer's typology is a marked improvement over previous typologies, it suffers from at least three shortcomings: 1. Overlapping categories: 2. No empirical check to see if the typology can embrace all reinforcers in the specified pop- ulation of reinforcers: 3. Some categories of "reinforcers" represent mod- eling or prompting antecedents and not conse— quating events. These shortcomings, which are discussed in detail in Chapter Two, represent fundamental conceptual flaws in the typology. These flaws express themselves in the in- struments developed from this categorization scheme (the PROS observation schedule and the PROS pair comparison scale). This seriously limits the interpretability of any research which employs Bersoff and Moyer's typology or the instruments developed from it. Because of these problems, Van Treese (1980) began develOpment of a cate- gorization scheme for reinforcers used in the schools. This categorization system was intended to embrace all teacher controlled positive reinforcers recommended by authorities in the field for use in the schools (both regular and special education). This system, which is referred to as Categories of Teacher Controlled Positive Reinforcers (CTCPR), was evaluated both logically and empirically. The logical analysis consisted of a comparison of the CTCPR with existing categorization schemes for rein- forcers. It was concluded the initial work indicated the CTCPR was comprehensive in its ability to embrace all components of the other categorization schemes with the limitation that categories referring to non-teacher con- trolled reinforcers (e.g., peer reinforcement) and nega- tive consequating events were excluded. It was further concluded the CTCPR's subdivisions were sufficient to allow reasonable homogeneity within specific categories (Van Treese, 1980). The logical analysis was followed by an empirical analysis. The empirical analysis consisted of having experts assign a specified population of reinforcers to specific CTCPR categories. Rater reliabilities were computed for each item (i.e.. the raters' agreement on a specific rein- forcer's assigned category or categories). Rater relia- bility was found acceptably high. This provided initial evidence of the CTCPR categories' construct validity (Van Treese, 1980). Subsequently. the CTCPR categories were used as the basis for a pair comparison scale. This scale was used to assess teacher perceptions of reward preferences of kindergarten, 3rd. and 6th graders by sex and achieve- ment. The reliability study on the CTCPR pair comparison scale and the results of the preference study are 7 summarized in Chapter Two. The results indicated tenta- tive support for the CTCPR pair comparison scale's relia- bility and its sensitivity to age-related changes in children's reward preferences as inferred from teachers' perceptions. Thus far in this section, the research context in which the study is embedded has been reviewed in relation to the main purpose of the dissertation (which is to investigate possible age-related changes in male school children's reward preferences). Next, the research con- text for the secondary purpose is reviewed. The secondary purpose is to compare the degree of congruence among the three groups of respondents' percep- tions to infer if they differ in their perceptions of the social apprOpriateness or utility of the CTCPR categories. The principle comparison of interest is the school psy- chologists' (both male and female) perceptions with the teachers' perceptions. Three levels of student behav- ioral characteristics which are instances where school psychologists might, as part of an intervention plan, rec- ommend the use of a particular type of reinforcer for the teacher to employ in the classroom are included. They are included to make the comparisons more applicable to the practice of school psychology. A fourth level of student behavioral characteris- tics represents a "no problem" condition. The "no prob- lem" condition is included to obtain an indication of 8 "normal" children's reward preferences. Kazdin (1980) makes the point that in cases where several effective options for intervention are available. the acceptability (to the implementor or to the recipi- ent) of the different types of interventions should be considered when choosing the type of intervention to use. Kazdin also reported research which indicated the accep- tability of different interventions may vary with the particular presenting problem. The secondary purpose of the dissertation is to assess if there is any difference in the acceptability of the various CTCPR categories of rewards: and, whether the teachers (as implementors) see the various CTCPR categories' acceptability differently from the school psychologists (typically the recommenders or consultants). The study is related to Kazdin's (1980) effort to add to the knowledge of the social acceptabili- ty of intervention techniques, albeit in a restricted sense. The research objectives are stated in general form next. The order of presentation is based upon the chro- nology of the research steps. The first two relate to category development and the instrument development. The remaining three objectives are directed toward the two main purposes of the dissertation previously stated. Research Objectives 1. The study can be conceptualized in five major parts. The first part concerns a further empirical check of the construct validity of the CTCPR categories. Fol- lowing the initial work (Van Treese, 1980), slight changes in the wording of some categories and the con- soli ation of two related categories were done. These changes require a follow-up study to assess the possible changes in construct validity. That is the first research objective. 2. The second research objective concerns obtaining additional reliability estimates of the CTCPR pair com- parison scale. 3. The third research objective concerns the exten- sion of Van Treese's (1980) initial study of teacher per- ceptions of reward preferences of K, 3rd. and 6th graders by sex and by achievement. One additional level of grade, the 9th, is included. Specified student behavioral char- acteristics are also entered into the experimental design. Sex of student is fixed at male. The sampling of teachers is improved. ' h. The fourth research objective concerns obtaining equivalent data to that of the third research objective, but with the respondents being male and female school psychologists. This data provides two additional bases from which children's reward preferences can be inferred. 10 5. The fifth research objective concerns the compar- ison and the contrasting of the perceptions of the three groups of respondents. Potential Contributions The study has merit because it has the potential to make important contributions to both basic and applied research in the area. ‘ The refinement of the CTCPR categorization system (research objective one) not only provides the basis for the CTCPR preference scale, it may provide the basis for an observation schedule (i.e.. a method of coding observed reinforcement emissions by teacher or reinforcement selections by students). Bersoff and Moyer's work (1973, 1976) demonstrates how this can be done from a similar categorization scheme. Observational research conducted in naturalistic settings could be used as a check to see if teachers accurately report which rewards they, for example, believe are most effective with the actuality of which ones they use in their classrooms. The observa- tion schedule could also be used to establish base rates for the types of rewards which are used in classrooms. That would address, for example, issues such as whether or not the rate of teacher reinforcement emissions is high enough to promote learning or whether or not teach- ers tend to rely primarily on a given category. 11 The study is also designed to further basic research on age-related changes in children's reward preferences by providing three independent indirect measures of these. Furthermore. there is evidence (Van Treese. 1980) that by including additional student variables not controlled in previous studies. interpretation of the results will be less affected by problems of internal validity. The study is also designed to further applied re- search. For example. data concerning teachers' percep- tions of children's reward preferences could be discussed in classes for teachers in training. This data would also be of interest to school psychologists who often make recommendations to teachers concerning the use of incentives to help motivate children. Data concerning school psychologists' perceptions of children's reward preferences will be available to compare with teachers' perceptions. If great differences exist, insight might be gained on one possible reason that school pyscholo- gists' recommendations go unheeded in some cases. Some of the study's potential contributions to basic and applied research have just been reviewed. This study does not, however, provide direct measures of children's reward preferences. While this would be a most inter- esting aspect to investigate, restrictions in resources (e.g., both time and money) make it necessary to defer work on this issue. In the review of the literature, the complexities of obtaining direct measures of 12 children's reward preferences are discussed to further justify this ommission. Chapter Summary In this chapter, the purposes of the dissertation were stated. The statement of the problem was carried out within the context of an overview of the literature. The research objectives were stated in general form. The chapter concluded with a review of some of the poten- tial contributions to both basic and applied research. The next chapter is the review of the literature. CHAPTER TWO REVIEW OF THE LITERATURE This chapter is divided into three major sections. In the first section, the relevant measurement theory is reviewed. In the second section. the literature relating to categorization scheme develOpment and instrument development is reviewed. In the last section, the re- search on age—related changes in reward preferences is reviewed. SECTION ONE: MEASUREMENT THEORY In this section on measurement theory, the problem of establishing expressed preferences for various rein- forcers is introduced. Why the solution to the problem varies with the purposes of the investigator is discussed. Finally. the solution appropriate to this dissertation is presented. 13 14 THREE METHODS OF SCALING REWARD PREFERENCES The Problem of Scaling The purpose of this portion is to introduce the problem of establishing preferences for various reinfor- cers. This problem is complex in that its solution varies with and depends upon: 1. The population of children of interest; 2. Whether the children are directly assessed or whether adult perceptions of children's prefer- ences are assessed: 3. The population of reinforcers under considera- tions (st is the simplest case): 4. Whether or not a heterogeneous set of reinfor- cers is grouped into homogeneous categories when N is large; 5. Whether an idiographic or nomonthetic approach is taken. This dissertation focuses on school children in the K, 3rd, 6th. and the 9th grades (#1). Adult (i.e., teachers' and school psychologists') perceptions of chil- dren's reward preferences are assessed (#2). The popula- tion of reinforcers recommended for teacher use in the schools is considered (#3). Homogeneous categories are appropriate to this dissertation's purposes (#0). Final- ly, a nomonthetic approach is taken (#5). .With these points in mind, the problem of scaling either individual 15 reinforcers or categories of reinforcers will be ad- dressed. How this problem should or might be addressed assumes a basic knowledge of three statistical methods of ‘obtaining and scaling reward preferences. Therefore, this section begins with a survey of the pair comparison method, the rank order method and the rational zero point method. Later, three examples which illustrate the apprOpriate use of each of these methods are discussed and the method most appropriate for this study is presented. The Pair Comparison Method In 1927, Thurstone (Thurstone, 1927a; 1927b) devel- Oped the theoretical rationale and the statistical pro- cedures for the Law of Comparative Judgment which is the basis for this method. The writer will not recapitulate ~its development nor repeat its mathematical derivation. Excellent sources exist for the interested reader (Guil- ford, 195A: Thurstone, 1927a & 1927b). However, the central concepts are reviewed. In the method of pair comparisons, all stimuli (81's) are typically presented to an observer (Q) in all possible pairs of non-identical 81's. This results in N(N-1)/2 pairs and requires that the 9 pick one Si in each pair over the other one in the pair. The 9 com- pares one to another and judges which is "better" or "has more" of some defined quality or quantity. The same 9 may judge all pairs a large number of times on different 16 occasions or many similar Q's may judge all pairs only once . In both cases, the individual Q or group of 0's, there exists variation in how the pairs will be judged. The reason for this variation is a central concept to the Law of Comparative Judgement. It is called discrim- inal dispersion and is discussed next. Thurstone (1927a) developed the concept of discrim- inal dispersion via the following four propositions: 1. A series of stimuli S1, 32, 83...Sn can be ar- ranged in a continuum, with reference to any prescribed quantitative or qualitative stimulus attribute. These stimuli are differentiated by process of the organism of unknown nature, and they are designated R1. R2, R3...Rn respectively. Every stimulus Si is identified by the organism with the process R1. These processes may be either psychic or physiological or both. In this dis- cussion, they are referred to as the discriminal processes or qualities. ‘ When the discriminal processes R1...Rn are con- sidered in the same serial order as the corre- sponding stimulus series, they constitutewhat may be called the discriminal continuum or the psychological continuum. This continuum is the correlate of the already postulated stimulus continuum. It is assumed that the corresponding Sn -- Rn is subject to noticeable fluctualtion, so that Sn does not always produce the exact process Rn but sometimes nearly similar processes Rn+1 or Rn-1 and sometimes even Rn+2 or Rn-Z. It goes without saying that the numerical subscripts 17 are here used to denote qualitative similar- ity and that no quantitative attributes are thereby necessarily injected into the discrimi- nal processes. This fluctuation among the dis- criminal processes for a uniform repeated stimu- lus will be designated the discriminal disper- sion. (p. 22) Simply put. discriminal dispersion refers to the variation in judging, say, the weight of a two pound bar. Each judged weight is an R1. The Ri's are ex- pected to disperse around the actual value of the Si in question with fewer Ri's as one moves away from the actu- al value of the Si. In judging a pair of 81's, both of the 81's respec- tive discriminal dispersions must be considered. ~For example, if the discriminal dispersions of $1 and 82 do not overlap, the judgment between the two will be consistent. However, as the overlap between the discrim- inal dispersions increases, the judgement between the two becomes less consistent. In the extreme case of complete overlap, there is no consistency of judgment. Many times in psychology, the actual value of the Si is not known. It is necessary to "go in reverse" and estimate each Si from the Ri's which each Si elicits. Of course, the mean value of the Ri's elicited by a given Si equals the actual value of the Si with certainty only when an infinity of Ri's are elicited and averaged for it. One must be aware the estimation of a given Si can 18 vary greatly from its actual value when a more modest number of Ri's are elicited and averaged. The last point is critical to the understanding of why a set of heterogeneous reinforcers might be grouped into homogeneous categories in some cases. Consider Figure 1 which displays four Si's each with its own discriminal dispersion. Given this hypothetical example, it can be seen that S5 and 38 will be ordered consistently, since there is no overlap between their respective discriminal dispersions. That is, even if their estimated true values are consi- derably off (due to few observations), 58 will always be preferred over S5. This is also true for the pairs S5 and S7. 55 and S7. and S6 and 58' However. because of the overlapping discriminal dispersions, the pairs S5 and S6: and S7 and 88 cannot be reliably ordered from a Jmodest number of Ri's or observations on each. Assuming the discriminal distributions are fixed, the only way to reliably order S5 and S6: and S7 and $8 is to greatly increase the number of Ri's or observations on each. This solution has obvious drawbacks. Fortunately, a second solution exists and is especially applicable to the present purposes. Figure 2 diSplays an example in which the 81's are grouped into two categories. S5 and S6 have been placed in one category. S7 and S8 have been placed in another. If all of the 81's represented individual reinforcers: Figure 1. Discriminal Dispersions of Ungrouped Reinforcers 20 Ri's \ 866 ----------+w--Category 3758 60090069 ............. Category' S586 Figure 2. Discriminal Dispersions of Grouped Reinforcers 21 and if, specifically, S5 represented candy, 86 repre— sented a fruit, S7 represented a "pat on the back", and 58 represented an embrace, the two categories might be called: 1) concrete edible rewards: and, 2) close body communication rewards. Forming homogeneous categories may improve the ability to obtain reliable ordering from a limited number of Ri's or observations (now taken on . the categories and not on the individual Si's). What is not as clear from this example is that when the number of individual reinforcers is vary large (e.g., over 100), the forming of homogeneous categories lends considerable conceptual clarity to the picture. Also, the pair com- parison method applied to the 100 Si's requires (typi- cally) #950 pair comparisons [N(N-1)/2]. However, if the same 100 reinforcers could be divided into 10 homo- geneous categories, the number of pair comparisons I needed (between categories) drops to #5. Clearly, such categorization has desirability in some instances. How- ever, because the ability to discriminate within a cate- gory is lost and because some intracategoy variance exists, it is not always desirable as will be seen later. As should be clear by now, the ability to reliably order Si's depends on the "distance" between them and the variance of their respective discriminal distribution. As the overlap increases, the ordering becomes increas- ingly unreliable given the same number of Ri's on each 81. 22 One desirable property of the pair comparison method is that it yields an interval scale (Guilford, 1954)._The chief objection to the method is the number of pair com- parisons needed in the usual case. Guilford (1954) sug- gests that when Sn is greater than or equal to 16 the number of pair comparisons makes data collection diffi- cult. While some variations exist which allow a reduc- tion of the number of pairs (Guilford, 1954), they are not widely applicable and result in poorer estimates of the 81's. The Method oijank_Orgg; The method of rank order requires, in its simplest case, than an 9 order a set of 81's that are presented to the Q_simultaneously. It is preferred to the pair comparison method in some cases because it is easier to rank, for example 14 81's than to do 91 separate pair comparisons. It bears a fundamental relationship to the pair comparison method. In the simplest case, each 81 is, in principle, compared to all others. Also, 81's that have been ranked by a number of individuals (a slightly more complex case) can yield a "pooled" rank order which can result in interval scale properties. Furthermore, these "pooled" ranks can be used to approx- imate the pair comparison method. The interested reader is referred to Guilford (1954) for a detailed proof of these points and a discussion of modifications of the 23 rank order method. Even though there exists a fundamental relationship between the rank order method and the pair comparison method, the differences are important. While it may be easier to rank order, for example, 14 81's than to do 91 pair comparisons, the pair comparison procedure might be preferred. For example. if a six year old child is asked to rank order 14 81's, considerable unreliability may result because of the child's attentional and cognitive limitations. In short, the child might not compare each 81 with all others as the rank order method assumes. Therefore, it seems preferable to "force" the child to do that by presenting the 81's via the pair comparison for- mat. Generally speaking, the context in which one of the two methods is to be applied is probably the best basis of choice between them. The Rational Zero Point Method The rational zero point method is an elaboration of the pair comparison method which allows for the establish- ment of a rational zero point on preference scales. Horst (1932) initially formulated a rational zero point method, but he required that one list psychological var- iables of positive and negative values. This is often difficult and awkward in some contexts (e.g., if all var- iables are "good" it is awkward to assign some a negative value) (Thurstone & Jones, 1957). 24 Thurstone and Jones (1957) have developed a method of listing Si's which allows all to have positive values. Conceptually, it is rather simple. The following example illuminates the basic points. Via the pair comparison method, objects 51' 82, and each sum of pairs of Si (i.e., 81+SZ, 81+83, SZ+S3) are treated as separate stimuli and scale values for each are obtained. That is, each Si is compared with every other Si and every other combination of 81. The QL§ are asked to express preferences for each pair of stimuli such as $1 and 82 and they are asked to express their preferences for such choices as S1 and S1+Sz. The rational zero point is chosen so that the distance from the origin to S1+Sg is the sum of the dis- tances from the origin to S1 and 82. Every combination of Si determines the zero point in this manner. This assumes the anticipated satisfaction for the combination of two Si's (e.g., 81+82) is the sum of the anticipated satisfactions of owning each separate Si in a given com- bination. It should be noted that linearity is not assumed for composites of more than three 81's and that one Si cannot substitute for another Si. At any rate, through the assumption of additivity, several estimates of the rational zero point can be gotten. To the ex- tent that these various estimates of the zero point agree, the assumption of additivity is supported. The rational zero point is the average of these estimates. That, simply put, is how the rational zero point is 25 established conceptually. The mathematics are reported in Thurstone and Jones (1957) and are not repeated here. Being an extension of the pair comparison method, the rational zero point method shares the pair compar- ison's vices and virtues (which were noted earlier). However, the rational zero point method has one virtue that the pair comparisod method lacks, namely, it is a ratio scale. Therefore: it allows statements that one 81 is twice as valuable subjectively as another 81. An application of this virtue is discussed later. Unfor- tunately, the rational zero point method also exagger- ates a vice of the pair comparison method. Remember, the pair comparison method requires N(N-1)/2 separate pair comparisons. The rational zero point method in- creases this greatly. For example, when Sn=5 the num- ber of pair comparisons increases from 10 to 55. In the following examples, three applications of the scaling methods are considered. Example One: Researcher "A" Researcher "A" wishes to equate reward value in his experimental groups to control motivational incen- tives in a learning task. In order to insure that any differences in learning among the groups of, say, K, 3rd, 6th, and 9th graders are not wholly or partly due to changes in the "preference" for the reward used, it is necessary to insure that the reward objects used are 26 equally preferred by the children in the different grades. In order to obtain absolute preference values, it is necessary to establish a ratio scale of measurement. 0f the three scaling methods reviewed in this paper. the rational zero point method is the only applicable one. This example was adapted from a study by Haaf and Smith (1976) Example Two: Researcher "B" Researcher "B" wishes to explore teacher perceptions of possible age-related trends in preferences for the general set of reinforcers available in the schools. In this case, it is appropriate to employ some system of categorization to reduce the task of addressing hundreds of individual reinforcers to the more managable task of addressing a fewer number of relatively homogeneous cate- gories. Once researcher "B" has established ten or so cate- gories of reinforcers, the categories can be ordered by teachers via the pair comparison method with refer- ences to specific grade levels. The rank order method is not used in this case because some teachers might not compare each category with all others (as is implied in the method) because of the complexity of the task (Dunn- Rankin, 1965). This could result in considerable unre- liability. This possibility is avoided when the pair comparison method is used. The rational zero point 27 is not used since placing absolute values on the cate- gories is not necessary for Researcher "B's" purpose. Therefore, the additional work and expense needed to ob- tain the great many more pair comparisons required by the rational zero point method over the pair comparison method cannot be justified. Example Three: Researcher "C" Researcher "C" wishes to check teachers' perceptions of children's reward preferences against the "criterion" of the children's actual expressed reward preferences. This allows the investigation of age-related trends of expressed reward preferences and teachers' (or others) sensitivity to these possible changes. Like Researcher ”B", Researcher "C" has decided to use the pair compari- son method and would like to use categories but he is faced with the problem of assessing young children who cannot comprehend complex abstract categories. However, if he could sort the population of teacher controlled reinforcer into categories, he could ask the children to compare specific concrete exemplars of the categories. This would reduce the number of comparisons needed and lend cognitive clarity by being able to extrapolate from the specific reinforcers to the categories. 28 However, when one exemplar is chosen to represent an abstract category: 1. Its preference must equal the average preference of all reinforcers in the category it is intended to exemplify. ‘ 2. The preference of the exemplar must change over combinations of child variables in the manner the category as a whole changes in preference over combinations of child variables. Consider the first stipulation. It requires that the preferences of all reinforcers be determined so that the average preference can be computed. From what has been reviewed concerning the problem of scaling reward preferences, it is clear that the task of establishing reward preferences (either subjectively or by developing a situation where real reinforcers can be chosen by children) for all reinforcers in a category where the number of specific rewards is large (e.g., over 100) would require a great number of subjects. For example, if a category with 100 reinforcers is assumed, there are 4950 pair comparisons to make. This would require a partial set of 4950 comparisons to be administered to each subject. Fifty subjects would be required to obtain one complete ranking if each subject were administered a set of 99 comparisons. This must be multiplied by about 50 to get a reasonable estimate for one category and then multiplied again by the number of categories. In this 29 example, the total number of subjects would be 25,000 given 10 categories. Consider the second stipulation. Once the exemplar of average preference for each category has been deter- mined for a particular combination of child variables (e.g., grade level, sex), it must be determined that the selected exemplar's preferability changes across the other combinations in a manner representative of the way the category as a whole changes over types of chil- dren. If no one specific reinforcer meets that require- ment, then more than one reinforcer would best exemplify the category depending on which type of child was being considered. If the same exemplar was not the most ap- pr0priate for all combinations of child variables and yet only one exemplar was used, then the variance at- tributed to changes in preference between categories (inferred from experiments using concrete exemplars of abstract categories) would be confounded with within category changes in preferences for specific reinforcers over types of children. Because there is no reason to assume that such within category variance does not exist, (Half & Smith, 1976), even more subjects would be re- quired. For example, if three grade levels and two sexes were considered, the number of subjects need would in- crease from 25,000 (continuing from the previous para- graph) to 200,000. 30 It is clear that what might appear to be a simple task is quite complicated. The researcher is faced with selecting one of two subjective measures of children's reward preferences (both adult perceptions and the chil- dren's perceptions are subjective), or opting to do a naturalistic study where children select actual rein- forcers (the latter is the "ultimate" criterion). The difficulties just discussed relating to assessing chil— dren's subjective preferences also hold for naturalis- tic studies. In addition, naturalistic studies entail additional costs and logistical problems. The only practical choice is inferring children's reward prefer- ences from adult respondents. Furthermore, it is not unreasonable to expect adult perceptions to be valid indicators of children's reward preferences providing: care.is taken to assess adults who have observed children at the age level and in the physical setting to which inferences are made: and, care is taken to obtain a sam- ple of adults large and prOperly gathered so bias is not expected. The "Solution" Both researchers "B" and "C" would be building on the third line of research discussed in Chapter One. In both cases the pair comparison method is preferred and a categorization scheme for reinforcers is needed. While this dissertation is designed to be similar to Researcher 31 "B's" study, it would be desirable for the categorization scheme used be adaptable to Researcher"C's" purposes in- case the difficulties with doing such a study could be partially overcome (e.g., it might be possible to compli- ment a study of adult perceptions by adding a much nar- rower study of children's perceptions). Section Summary In this section, measurement theory as it related to the problem of scaling reward preferences was reviewed. It was concluded that the pair comparison method is the most applicable scaling method for the purposes of this dissertation. It was argued that it was not unreasonable to infer children's reward preferences from adult per- ceptions providing some cautions were taken. The need to either discover or develop a categorization scheme for the set of reinforcers used in the schools was iden- tified. SECTION TWO: CATEGORIZATION SCHEME DEVELOPMENT AND INSTRUMENT DEVELOPMENT The existing categorization schemes for positive reinforcers and the need for the development of an im- proved categorization scheme are presented in this sec- tion. The develOpment of the Categories of Teacher Con- trolled Positive Reinforcers (CTCPR) and the development of a pair comparison scale based on it are also presented in this section. The PROS and its Ppedegessors As noted in Chapter One, one line of research has been primarily restricted to investigating adult percep- tions of school children's reward preferences for vari- ous categories of rewards commonly administered in the school setting. Bersoff and Moyer (1973, 1976) developed a categorization system which is the basis for an obser- vation schedule and a pair comparison scale (Bersoff & Moyer, 1973: Byalick and Bersoff, 1974: Dervensky & Rose, 1978). Bersoff and Moyer developed their categories in response to two shortcomings which they identified in earlier categorization schemes which were used as the bases for scales. These earlier categorization schemes 32 33 either failed to include categories of nonverbal rein- forcement (e.g., Amidon & Flanders, 1967: Brophy & Good, 1969) or were otherwise not inclusive enough to embrace all possible types of reinforcement emissions in the schools (e.g., O'Leary & Becher, 1967). Because of the inadaquacies of the pre-PROS categorization schemes which have been used as the bases for scales, the present dis- cussion and criticism focuses on the PROS and the pair comparison scale developed from it. The categories of the PROS, their symbols and their definitions are found in Table 1. The 14 categories, when used as a preference scale, are presented to the respondents in a pair comparison format (Edwards, 1957) with each category being compared to each other. The original pair comparison scale and as reworded for ado- lescent delinquents are found in Appendices "A" and "B" re- spectively. While the PROS categories are an improve- ment over previous typologies, it has at least three shortcomings: 1. Overlapping categories: 2. No empirical check to see if the PROS categories embrace all reinforcers in the specified popula- tion of reinforcers: 3. Some PROS categories of "reinforcers" represent modeling or prompting antecedents and not conse- quating events. u. 3 “.mcma .noxoom a hummq.o Scum copmmo¢v .cwmmm cohoom an op know :mpdo mane Mom nonponm cam ouomfimm camped mac zoozvon omnmao szs Hm>hoPcH ocooom Io>wm m use hHoMmPMo mace CH mwfioaon omam passe m op mnwusoppm no Mawxooa copuoo . :aoo .wGHHflSm mafia; Hw>opnmm ho :ofivmowosa nonpo o>ww no .xawz .cmo: con names nopmop\uoncMoa .wzthm ma uafiso one was: no mason ma UHHno one pass op wcficsoppm no wcwawam ma nmpmwp\umgomop nos: Basso M ed mnflxoon .sowpnopp< Hawomm o>wpwmom +coa an Acmwsmnaoo now ma Avocadosv pampcoo Hmnho>no£ no Aooou .onm .sazaso .mowv pnmpCoo Hanuo> one ma mm commuwmnoo who mmmzommon ampmop honomoe .sofipmoassesoo Hmcomnmnsa cm haonwa so: pew adage on» op :oHvoMom Hmsomnm and: N no nowpmsam>o o>wpfimom m mopdowcsaaoo mm .oafiso on» wnfipsmawHQEoo haddpno> an xomnuomh m>fipfimom no nowpmahfiwmm canefim mo Hm>oa m.ampmmp\umsommp one czozmn ow sown; mCoHpowon m>fipmzam>m .omwmnm:99omawm mm .Aaeme .eeeeom s sheeq.o .mome .eeoo s seeohm Bosh cmpnmo< :.Hao3 manhospm on.zowe =.o:wm= =.coom m.pmse: mm moansmxo scam mo mpmfimzoo cam .pmom no wood on pogpfio has newsmanfiwmm Hmnno> .mpdwnmosnnw me how u>mson ma: Page no .oanmpmooom no voonnoo ohm noncommma ma: page adage a op cowme :zoesoo .Hm>onnmm mcwpwowcafi Homecoo Hanso> .now>mnmm opmfiamouam< mo sofipmsnfihw< m¢¢ .opm .mnoxowpm .mHMPm .mxowpm cohoaoo .moaaawp .mmwgo Hoxon ohm moansmxm nossoo .mSae chapsh msom pm mcnmzoh mvmhozoo poohfic pom coEoocoh on Haas scan: mummsop oaaonshm Ho wcfi>ww .Acoxoev mchmsmm oponoaoo Mo :owpmuwmfizfisc< emu .msam> nospo ho msxomp on o>mn sows; Rondo Page on nozmcm poohuoo Sons pcmmcwpnoo pawno a op mUAMo unmmam mzfl>wm mm sonny mcamzmp ofiaonshm pan opohocoo mo>wm hopmop\uonoMop on» non; moodemna mmonp mo mpmwmzoo omam ShowoPMo mane .oaap comm no .zozoa .hc:da mm scam . mcumzop opmhocoo Pomafio Mo wsfi>fiu .Apoohanv moamzom opmpozoo mo soapwhvmwawso< amo mmHmoomaHeHmom Immmnshm can m:o«p«:wwon .Amommv mascogom nowpd>nomno encamohomnwom o>fipwmom .a mummmw 35 .cafino one an cmpmwpncfi pow>mnon mo sowpmshwamm non amwo :Hpfiho hospwm: ma scans nopmop\uo£omop hp oncogmoh < .omconmom Haphm> onmapmoncoz m>z .nofipmosu on» ma Pammoua ma fiasco one mo zofipwsam>c o>fipmmmc no: o>apamon nonpwo: hhoMopdo was» sH .Aow>mnon page wafichmo :coc sow>mnon pcmvspm mzfisoaaom nopmov\umnomop an mcowpmosu wcfixm< .mcoapmosa mxm< a< mmHmooma¢o Hwmnommouzoc heaped ma cawco on» son: nopmov\umnommp an pow>mson opmHmmopmmm Ho Goapwhpmcoama .oaaamxm an mzwuw< .mzowpmowmsm Ho mownom commum m hoagao haddowpmaopmhm no .Azmco pmma one pow pmzw sowev pnosomMQSoocm m>am .wSammosw PmowwSm has hopmop\uogomop one .aoqum op mmmzmma cawgo map non: .sofipozconn omnommmu mom Paosowmhzooao mnwcfi>oag .mmmoozm and mocmcfiflcoo mcwposopm .wcwnmpmuom .wzprEopm .wcwwhb .huopmms psm>snu< A.uwma .maoosmHm d Cocaa< Eoum copnwc nopmop\uosomwp Cos: mm scam mcwcaflzp Psonpfis encasememon oHQaHm .mwocw wzwmoao>op use mafihmwumao mam mnofi>mnon nopmop\umsoMop he: one .oHMm mm: Psocspm map was: osHumEESm no .ham :sflm macs mood one mumpmoh .szsmpmpm m.»£oc:pm map omwpnmmumm has hovmop\umaomoa. .oaano one an oopmommsm memo“ mcwnooa>oc no .MGHUHHSD .mcwhawumao .mdooH mummoo< 73$ .meeeefim a edges son.“ 3335 69:38 we: no.“ edge 2:. cheese ea: :3 mg per» use .mwefieem «were 8. pzmfin m mm: on page .maomw caaco on» 3o: mandamhocc: on mhMm on son: mwnwammm magma now mopmop\uo:omop one .coozaocn ma mmswaooa wcaaawoon no weapoacohm .o>wvmmo: no o>pwmon on has mnoHPoSo pnovsvm no mmGAHoom .aonsms wzfinopwohspcoc a Ca pawgo one no ocop mafiaomm one mmwhanwao and mpaooom nopmop\uozomms .mMGHHoom memoooa A.uomfi .hmx :oom i hquH.o scam oopmmpommnm Ho swam a ma .ovo .cnmn wCAxMP .snm wawcao: m¢ £< H< m< .wcfiommnao .wsflppmn mm scam pompcoo Hmofimha Hospo< .pompsoo awowmhnm o>wpauom +om AU.p£oov H mflmsoz .noppzopp< Hmpomm m>ppmwmz n¢m .. .38; Peer? .. .emp weeds. on.zow= =.oz= mm scam .mpdppmopnnmsp no .oHDMpmooomnz .pomuuoonp ohm momsogmoa mp: pmnp capno w op mappMoppcp ampmop\uo£omop hp omnommoh Hmnam> .psosnmpco8p< n< mmHmooma¢o m>He¢umz AU.pfioov H mHmpppmom a snow pnwps honoMop map so .smmmm doom a so econ coow= no :pmcuw m.pmgp: spas: pswpa hmscmop one .oamsmxo pom .mosos sopppss m>pppmon Ham mGpdpaoo zaowopMo mane .mcnmzmm CoppMopcsasoo copppsz .epso some so .:no3m:m mafia: .=non coow: .:pmohm m.pm:p= mGHhMm an pamcspm m chases pzwps honoMop esp .oansmxc pom .mcoppmN :padpsm> hmnoMmp c>pppmom Ham medpsoc hAoMmpMo mans .mUhmsom :oHpMopczesoo Homo .nmxoppm comp amass m no .xnms xoono m .opmam coow m .nMpm m pcmospm map o>pw pzwpe nonoMop map .90 .mosmzoh umspo has peonppz hop>mgmn doom Ho madam a made pamps smnoMop map .magsmxo mom .hhomopMo nonposm Soup noosopspoa a you coacmuos on poGQMc scan: pan cosmsuopaom confined mo opaonamm was gowns macpfihm cam mpoonno Ham mfipMpcoo anomopmo maze .mcsmzom cpaonahm manmaoooomacoz .cmgomon ma gnaw map so papom copmswwmcc a son: amosopzpmu m pooamm ado pampSpm onp appp>on anommme wasp mp coGHMpaoo omam mum mam>oa pcmso>opncm mappMopccp magnum .mhomopmo Hospoem Soup hoosoHSHoh m pom owzmnoxm sopma zoo pnmczpm map sows: A.opo .xOppm cohoaoo .tha hHHMp .mwso noxom ..o.pv Smxop m psmUSpm map m>pm pnwpe honomop esp .mamamxo pom .hhowop :mo nonposm Soup poonopcwmp m pom cmwcmnoxm hapcoSUmmnzm cam possum on coo cops; mHonEhm was mpocwno Ham newspzoo haowmpmo mane .mcnmsom oppopfihm mandfimocmm .mzmxop can .mhmpm .pphos mo mmpwoppppsoo mam moamsmxo:Coz .unooos a so .hop Hanan m .xoon opsoo m .Hposon m .mcohmho msom no: no EH: wzp>pw an psmcspm m amazon psmws noncomp map .mamamxo pom .mo>amm8mnp csommn wcpnpmaom mspupaonshm no mappCommsnos mm copammmum zaps unoppsopcp pos mum hone .hoon amass cap opsp :oMMp zaamsm: pos was nope: and can lawnmp was cops; mpomnno Haw mGHMpCoo Showopmo mans .mcnwscm manpUMISoz oposocou .mouuasw a so .maoo m .pzcoo m .mp:Mo mo woman m so: so Sp: waw>pw an pcocspm a chosen pSMps sonoMop onp .maasmxo Mom .acon amass cap opcp Scamp hHHdSho: can scan; mmozmpmnsm Ham mnpwpnoo hHoMopmo maze .mcpmscm manpvm opononoo 03 oo mmlz mm 20 mo such mumspapaonm :: whooaomcpmm o>pppmom umaaoppcoo hosomoe mo mopaowopmo .m mumppom has on oasoz who: covspoap omp< .capzo vowmpsm>UMmpu a no mcpanpm a pop hop Hanan a spam no: so ape wcpppoa an pamuspm m cusses pawps pmnomop onp .oaasmxo pom .hmsm Co>pm no coamnm mp cmnpmpno Con: sows; passes has op whammy knowopMo mpna .mcpmzom pmhmsm m .oConn m wcpms so .hopoopunmpccp wzphmHm .oannmsom maphmam .mhmxoono maphsam .ozmpnp m op mszHMp mm scam macaw op po:¢do cam so on cops: can appapnmeogmos on mopppwo nope; hpp>ppom oaom cc so: 90 SH: mapppoa an pamuspm a chosen pamwe sonoMop esp .ngEMXw pom .haowopMo mcpmzom hppppnpmcommmm map opnp HHMH Mmmlmm scpsz was macaw snow on pounce cops; moppp>ppom Ham mcpmpzoo hAoMmme mpze .muawzom hpp>ppo< Hapoom ppom meow op so: so 8p: mapppoa an pnouspm m pauses pampa sonoMop map .mam :smxm pom .hhowmme mcpmzmm appapnpmzommom map opzp Hash Hmmlmo cops; can macaw snow was nepns moppp>ppom Ham mfipdpcoo zsommpMo owns .mupmzcm hpp>ppo<.Hmspp>po:H «H .mocmswpnpms aooummwao apps mspnao: cam .nopcapno hospo Madhopzp .noppnos mafia .Hopsoa Honpmn hpcpmm mm nosm amazon a ma pascapm m op sc>pw mopppapnpmzommcn :oxppupascm: Ham moosaocH onMopMo mane .mcnmzom appapnpmzommom m .cmpm =Q: mnssnp: s .so .xcpz m .coc m .mppEm m psopSpm map map>pm an pampzpm w phases pnwpa honoMop map .oamsmxm pom .podpsoo Hoopmmnm peozppz coopmsoucs on mac cops: :ommsmnma hoop: no mosspmmw Hoopmhsm Ham mSpMpCoo anomwpmo mpca .mupmzom coppmopGSEEoo zuom paupmpa 0mm .mmpx a no oxmnmunmn m .maxopp m .93: m .xomn onp no pen a pcopzpm map msp>pm an psoGSpm m phases pswpa pogoMop cap .oHQBmXo mom .pompcoo Hmopmhgm Hmspom o>Ho>np cows: :owmzmnda hoop: so megapmom Hmopmhng Ham mspmpcoo hmomopwompze .mpnmzom soppmopnzasoo hpom mmoao omo AU.pSoov N mqmm so noppmsppm m>pmsm>m cm upo>m op pcmuspm m soaps scan; mmwoap>psn mopaaocp anowmpmo mpne .mcpmzmm mnmomm m Ac.pnoov N mflmpee< rpmeeeem-empm .m Hepoem .H made .eepaaeee s mapper mpmpocoo .n empep>peee .N Hepeem .. meme .hmflaon phases :Bo ppmnp shame page meppp>ppoe refinemeee depep>ppo< . Hmaeom . mama .opepeeao ' . HNM: mmsmo .m meaee>peee .m redefine .H meme .emEhdeHm s emrsemam mnmxoa .m meppp>ppo< .m Hdpoom .H Heme ..Hm pd hmxeem mmHmowme<0 .mmzmmom 02HBmem mmamsom Soppmuppowopmo pmnpo apps shom hnmcpapaoum : mmoao map Ho zomphmmaoo .m mHmHpo< mmmzmmoHo mconmopnxm Hmonhnm mphoz 0mmH .Cmmvmz a Gmmums humpmms Ho mmcmm Hm>ohnm< HmHoom Amnmmnm ..w.mv mmmpmopm Ho mozmpp>m HmsmH> mxomno no mamxoe mpmxaphe so mace mohmzmm hhmanm mcmxoe Homncmmm m>HpmapoHcH mhoH>mnmm sepapnmpohm amp: mamohopsHmm HmHoom moHnmasmzoo nospo a pooh 0 m AN m N H mme .SMHHHEomE m .2 n N H 2 meme .speem mwcmHzosm Ho noHpHmpsd< passpoo wnHmm Hm>opnm¢ moocmemmxm hhomcmeszz mxmma mo :oHpmHnaoo mupmSom HmHoom refinememe .m .m .3 .m .N .H mode .ppesem Ae.pcoev m memes 50 m .Hpo< . mamxoa . moHnHmzma . mmHanm .H mmmH .hmzmz a Hoopmpo8om omszom oHsoz :oHnB aoHnoam moH>mnmm mooHMmm oz aoH>mcom oooc AHgmsoHpmHom HoomomnoSom smHmmosmo comm oHoon mopmpm OHEosooooHoom was: psoaspm so xsm moHanmm> Hompcoo .5 names moHaowmpmo mmoao .vo0 .oome A00 monowopmo poH>mnmm peanwoppo< .omHHompzooaoucD .moH>mnmm mmona0 Aomv oHpmHMopommmso .wsHmmmm hHommm pmsh .psmEo>oH:o< owmmo>< .AoQH .oome psoospm rec .spo .esm .m .aasH .eexam luv seesaw pmeoHonohmm onz .pmeoH nonohmm onsmm .anomoa :OHpMooom amstom I .aocH .oome r 390 psooGoAmom Ho maze moHomHmm> HmpsoaHmoaxm .0 MHmmmch pmopmmupmma a powwomm em: mesa sehHHssHHmm .os eHase .H H HNMd’WONQCfiO eoefiosm ea. some spHHassaHem .0 mqm¢s 76 Findings for Hypotheses In this section, a summary table of the findings concerning each hypotheses is displayed. A discussion of each hypothesis follows. TABLE 11. Summary Table for Hypotheses Main Effects p level Significance Hypothesis 1 (T) ‘<.0001 s Hypothesis 2 (G) ‘<.0001 s Hypothesis 3 (SC) ‘<.0001 s Interactions Hypothesis 4 (TxG) <.1420 ns Hypothesis 5 (TxSC) (.0005 s Hypothesis 6 (GxSC) '<.0805 ns Hypothesis 7 (TxGxSC) ( ‘<.2583 : ns 1 1. The first hypothesis, "There will be no main effect for Type of Respondent." was rejected Cp<.0001). However, Type of Respondent interacted significantly with Student Characteristic and Hypothesis 5 was rejected (refer to Table 11). Examining the interaction discloses that the differences suggested by the rejection of Hypo- thesis 1, by itself, are not particularly illuminating. Therefore, there is no detailed coverage of Hypothesis 1. 2. The second hypothesis, "There will be no main effect for Grade." was rejected (p<;0001). Examining the higher order interactions involving Grade (refer to '5 77 Table 11), one sees that Grade does not interact with Type of Respondent or with Student Characteristic. Therefore, the effect of Grade was independent of higher order inter- actions and can be considered separately. Figure 3 discloses the variation of the individual categories by Grade. In Figure 3 and in the figures which follow, lower numbers indicate mg§g_preferred cate- gories. Univariate analyses of the individual categories indicated significant effects for the following variables. The most dramatic change by Grade was the drop of the ‘ "Close Body Communication Rewards". "Escape Rewards" evidenced almost as large a change, but as an increase. "Concrete Non-Edible Rewards" evidenced the expected de- cline as Grade increased. _However, the other "concrete" category, "Concrete Edible Rewards", did not quite reach significance. "Non-Redeemable Symbolic Rewards" declined with increase in Grade which is at variance with expecta- tions based on other research. "Intrinsic Rewards", while not reaching significance, did not show a high level in Kindergarten. "Intrinsic Rewards" was the low- est category in Kindergarten and evidenced a weak non- significant trend to increase at each succeeding grade level. 3. The third hypothesis, "There will be no main effect for Student Characteristic." was rejected (p Category "X" 1B. Category "Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a NINTH GRADE BOY WITH UNDERCONTROLLED. ACTING OUT BEHAVIOR. That is. mark the one you think the student would mark if the student were doing this exercise. 158 Mr. Mrs. Ms.: PHONE: What grade are you presently teaching?................. ..... ........... How many years have you taught that grade?.. ................. .......... How many years have you worked as a teacher?........................... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a KINDERGARTEN BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher following student response for the purpose of strength- ening or acceleratingpappropriate behavior. The exercise is designed as a "paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category "X" to category ”Y". the item would be marked like this: Category "X" 1B. Category "Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a KINDERGARTEN BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. That is. mark the one you think the student would mark if the student were doing this exercise. 159 Mr. Mrs. Ms.: PHONE: What-grade are you presently teaching?................ .......... ....... How many years have you taught that grade?.............. ..... . ..... .... How many years have you worked as a teacher9....... ............ . ....... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a THIRD GRADE BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. I The student is a Caucasian with middle income level parents.w Home- school relations are good. The student is in a regular education program and has no serious behavior problems. In filling out the scale. use the following definition of reinforcement: Behavior 9y a teacher following student resppnse for the purpose of strengt_- enipg:or accelerating appropriate behavior. The exercise is designed as a "paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example, if you believe the student would prefer category "X" to category "Y". the item would be marked like this: Category "X" 1B. Category "Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a THIRD GRADE BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. That is. mark the one you think the student would mark if the student were doing this exercise. 160 Mr. Mrs. Ms.: PHONE: What grade are you presently teaching?................................. How many years have you taught that grade?........... .......... ........ How many years have you worked as a teacher?........................... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a SIXTH GRADE BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT.' . The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher following student response for the purpose of strength- ening or accelerating appropriate behavior. The exercise is designed as a ”paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category ”X“ to category "Y". the item would be marked like this: A. Category "X” 1B. Category "Y" Go rapidly_but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a SIXTH GRADE BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. That is. mark the one you think the student would mark if the student were doing this exercise. 161 Mr. Mrs. Ms.: PHONE: What grade are you presently teaching?......... ...... .... ...... ........ How many years have you taught that grade?............................. ‘How many years have you worked as a teacher? .................... .. ..... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a NINTH GRADE BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher following student response for the purpose of stren th- ening or acceleratipg appropriate behavior. The exercise is designed as a "paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category "X" to category "Y". the item would be marked like this: Category "X" 1B. Category "Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a NINTH GRADE BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. That is. mark the one you think the student would mark if the student were doing this exercise. 162 If. he he, PHONE: With which grade levels are you presently working?..................... How many years have you worked as a school psychologist?............... Have you ever worked as a teacher?................. ........ ............ o DIRECTIONS t Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a KINDERGARTEN BOY WITH AVERAGE ACADEMIC ACHIEVEMENT. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher following student rgsponse for the purpose of strength- ening or accelerating appropriate behavior. The exercise is designed as a "paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category ”X” to category "Y". the item would be marked like this: Category "X" 1B. Category "Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a KINDERGARTEN BOY WITH AVERAGE ACADEMIC ACHIEVEMENT. That is. mark the one you think the student would mark if the student were doing this exercise. 163 "to “1'3. "Be! PHONE: With which grade levels are you presently working?..................... How many years have you worked as a school psychologist?............... Have you ever worked as a teacher?.................... ..... ............ DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a THIRD GRADE BOY WITH AVERAGE ACADEMIC ACHIEVEMENT.- The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher following_§tudent response for the purpose of strengph- ening or accelerating appropriate behavior. The exercise is designed as a "paired comparisons” task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category "X" to category "Y". the item would be marked like this: Category "X" 1B. Category ”Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a THIRD GRADE BOY WITH AVERAGE ACADEMIC ACHIEVEMENT. That is. mark the one you think the student would mark if the student were doing this exercise. 161+ ”1'. me "8.! PHONE: With which grade levels are you presently working?..................... How many years have you worked as a school psychologist?............... Have you ever worked as a teacher?.......................... ..... ...... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a SIXTH GRADE BOY WITH AVERAGE-ACADEMIC ACHIEVEMENT. . The student is a Caucasian with middle income level parents. Home- school relations are good.’ The student is in a regular education program and has no serious behavior problems. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher following student response for the purpose of strength- eningTor accelerating appropriate behavior. The exercise is designed as a ”paired comparisons” task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category "X” to category "Y". the item would be marked like this: @ Category "X" 13. Category ”Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count . Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you eXpect would be most preferred by a SIXTH GRADE BOY WITH AVERAGE ACADEMIC ACHIEVEMENT. That is. mark the one you think the student would mark if the student were doing this exercise. 165 Mr. Mrs. "8.: PHONE: With which grade levels are you presently working?..................... How many years have you worked as a school psychologist?............... Have you ever worked as a teacher?................. ........ ............ DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a NINTH GRADE BOY WITH AVERAGE ACADEMIC ACHIEVEMENT. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious problems. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher following student response for the purpose of strength- enipg or accelerating_appropriate behavior. The exercise is designed as a "paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example, if you believe the student would prefer category ”X" to category ”Y". the item would be marked like this: Category "X" 13. Category ”Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a NINTH GRADE BOY WITH AVERAGE ACADEMIC ACHIEVEMENT. That is. mark the one you think the student would mark if the student were doing this exercise. 166 Mr. Mrs. Ms.: ' PHONE: With which grade levels are you presently working?...... ..... .......... How many years have you worked as a school psychologist?..... ...... .... Have you ever worked as a teacher?..................................... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a KINDERGARTEN BOY WITH OVERCONTROLLED. SHY-ANXIOUS BEHAVIOR. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. However. his tendency to take frustrations out on himself is worrisome to his teacher. In filling out the scale. use the following definition of reinforcement: Behayior by a teacher following student response for the purpose of stren th- ening or acceleratinggappropgiate behavior. The exercise is designed.as a "paired comparisons” task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category "X” to category "Y". the item would be marked like this: Category "X“ 1B. Category ”Y" . Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: rRead the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a KINDERGARTEN BOY WITH OVERCONTROLLED. SHY-ANXIOUS BEHAVIOR. That is. mark the one you think the student would mark if the student were doing this exercise. 167 Mr. "I": "8.: PHONES With which grade levels are you presently working?..................... How many years have you worked as a school psychologist?............... Have you ever worked as a teacher?..................................... — DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a THIRD GRADE BOY WITH OVERCONTROLLED. SHY-ANXIOUS BEHAVIOR. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. However. his tendency to take frustrations out on himself is worrisome to his teacher. In filling out the scale. use the following definition of reinforcement: Behavior byia teacher following student response for th§_pu§pose of strepgth- epipg or acceleratingpgppropriate behavior. The exercise is designed as a "paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category ”X” to category "Y". the item would be marked like this: ' C 1A9 Category "X" 1B. Category ”Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions.count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a THIRD GRADE BOY WITH OVERCONTROLLED. SHY—ANXIOUS BEHAVIOR. That is. mark the one you think the student would mark if the student were doing this exercise. 1A58 “to me "sci PHONE! With which grade levels are you presently working?...... ..... .......... How many years have you worked as a school psychologist?............... Have you ever worked as a teacher?..................................... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a SIXTH GRADE BOY WITH OVERCONTROLLED. SHY-ANXIOUS BEHAVIOR. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. However. his tendency to take frustrations out on himself is worrisome to his teacher. In filling out the scale. use the following definition of reinforcement: Behaviorppy a teacher followingpstudent response for thg_purpose of strenggh- ening or accelerating appropriate behavior. The exercise is designed as a “paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category ”X" to category “Y“. the item would be marked like this: Category "X" 18. Category ”Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a SIXTH GRADE BOY WITH OVERCONTROLLED. SHY—ANXIOUS BEHAVIOR. That is. mark the one you think the student would mark if the student were doing this exercise. 169 Mr. Ire. "8.8 PHONEz' With which grade levels are you presently working?..................... How many years have you worked as a school psychologist?............... Have you ever worked as a teacher?..................................... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a NINTH GRADE BOY WITH OVERCONTROLLED. SHY-ANXIOUS BEHAVIOR. The student is a Caucasian with middle income level parents. Home— school relations are good. The student is in a regular education program and has no serious behavior problems. However. his tendency to take frustrations out on himself is worrisome to his teacher. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher following student response for the purpose of strepgth- ening or accelerating appropriate behavior. The exercise is designed as a ”paired comparisons” task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you ’believe the student would prefer category ”X" to category "Y". the item would be marked like this: Category ”X" 1B. Category "Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. ' Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a NINTH GRADE BOY WITH OVERCONTROLLED. SHY-ANXIOUS BEHAVIOR. That is. mark the one you think the student would mark if the student were doing this exercise. 170 be I". he: - PHONE! With which grade levels are you presently working?..................... How many years have you worked as a school psychologist?............... Have you ever worked as a teacher?..................................... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a KINDERGARTEN BOY WITH UNDERCONTROLLED. ACTING OUT BEHAVIOR. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. However. his tendency to take frustrations out on others is worrisome to his teacher. In filling out the scale. use the following definition of reinforcement: Behavior byia teacher followipg student response for the purpose of strength- enipgjgr accelerating appropriate behavior. The exercise is designed as a ”paired comparisons” task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category ”X” to category ”Y”. the item would be marked like this: Category 'X" 1B. Category ”Y” Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a KINDERGARTEN BOY WITH UNDERCONTROLLED. ACTING OUT BEHAVIOR. That is. mark the one you think the student would mark if the student were doing this exercise. 171 “1‘. MP8- MBe’ PHONE: With which grade levels are you presently working?..................... How many years have you worked as a school psychologist?............... Have you ever worked as a teacher?................. .......... .......... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a THIRD GRADE BOY WITH UNDERCONTROLLED. ACTING OUT BEHAVIOR. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. However. his tendency to take frustrations out on others is worrisome to his teacher. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher following student response for the purpose of strength- enipg or accelerating appropriate behavior. The exercise is designed as a "paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category "X" to category "Y”. the item would be marked like this: I A. Category ”X" 1B. Category "Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a THIRD GRADE BOY WITH UNDERCONTROLLED. ACTING OUT BEHAVIOR. That is. mark the one you think the student would mark if the student were doing this exercise. 172 "re “1‘8. "8.! PHONE: With which grade levels are you presently working?...... ......... ...... How many years have you worked as a school psychologist? ............ ... Have you ever worked as a teacher?................ ...... . ........ ...... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a SIXTH GRADE BOY WITH UNDERCONTROLLED. ACTING OUT BEHAVIOR. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. However. his tendency to take frustrations out on others is worrisome to his teacher. In filling out the scale. use the following definition of reinforcement: Behavior bypa teacher following student response for the purpose of strength- ening or accelerating apprOpriate behavior. The exercise is designed as a "paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category ”X” to category "Y". the item would be marked like this: Category "X" 1B. Category "Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a SIXTH GRADE BOY WITH UNDERCONTROLLED. ACTING OUT BEHAVIOR. That is. mark the one you think the student would mark if the student were doing this exercise. 173 Mr. “"0 “8.: PHONE: With which grade levels are you presently working?..................... How many years have you worked as a school psychologist? ......... ...... Have you ever worked as a teacher?..................................... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a NINTH GRADE BOY WITH UNDERCONTROLLED. ACTING OUT BEHAVIOR. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. However. his tendency to take frustrations out on others is worrisome to his teacher. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher following student response for the puppose of strength— enipg or accelerating appropriate behavior. The exercise is designed as a "paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category "X" to category "Y". the item would be marked like this: (51A.> Category "X" 1B. Category ”Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a NINTH GRADE BOY WITH UNDERCONTROLLED. ACTING OUT BEHAVIOR. That is. mark the one you think the student would mark if the student were doing this exercise. 17L: Mr. Mrs. Ms.: ' PHONE: With which grade levels are you presently working?..................... How many years have you worked as a school psychologist?. ..... ......... Have you ever worked as a teacher?..................................... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a KINDERGARTEN BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher followipg student response for the purpose of strengzg- enipg or acceleratipg appropriate behavior. The exercise is designed as a ”paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category ”X” to category "Y". the item would be marked like this: Category "X" 18. Category "Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a KINDERGARTEN BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. That is. mark the one you think the student would mark if the student were doing this exercise. 175 Mr. Mrs. 318.: PHONE: With which grade levels are you presently working?...... ..... .......... How many years have you worked as a school psychologist?............... Have you ever worked as a teacher?..................................... : DIRECTIONS I Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a THIRD GRADE BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher following student response for the purpose of strengp ~ eningpor accelerating appropriate behavior. The exercise is designed as a ”paired comparisons" task. All of the categories have been paired with each other. In each case, choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category ”X” to category ”Y". the item would be marked like this: Category "X" 18 Category "Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a THIRD GRADE BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. That is. mark the one you think the student would mark if the student were doing this exercise. 176 Ire Me "Be! PHONE! With which grade levels are you presently working?.... ....... .......... How many years have you worked as a school psychologist?............... Have you ever worked as a teacher?...................... ....... ........ DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a SIXTH GRADE BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. In filling out the scale. use the following definition of reinforcement: Behavior by a teacher following student response for the purpose of strength- ening or accelerating appropriate behavior. The exercise is designed as a "paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category ”X" to category "Y". the item would be marked like this: ( EA.) Category "X” 1B. Category "Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a SIXTH GRADE BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. That is. mark the one you think the student would mark if the student were doing this exercise. 177 Mr. Mrs. Ms.: PHONE: With which grade levels are you presently working?...... ...... ......... How many years have you worked as a school psychologist? ........... .... Have you ever worked as a teacher?.................... ........... ...... DIRECTIONS Defined on the next page are thirteen categories of reinforcers or rewards which can be administered by a teacher. The aim of this scale is to determine which of these categories you expect would be most preferred by a NINTH GRADE BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. The student is a Caucasian with middle income level parents. Home- school relations are good. The student is in a regular education program and has no serious behavior problems. In filling out the scale. use the following definition of reinforcement: Behaviorppy a teacher following student response for the purpose of stren th— ening or acceleratingpappropriate behavior. The exercise is designed as a ”paired comparisons" task. All of the categories have been paired with each other. In each case. choose one of the two pairs. Circle the letter in front of the category you expect would be most preferred by the student. Thus. in the following example. if you believe the student would prefer category "X" to category "Y“. the item would be marked like this: Category "X" 1B. Category "Y" Go rapidly but carefully. Do not go back once you have marked an item. There are no repetitions of pairs. In cases of difficulty. let first impressions count. Remember: Read the category definitions carefully. Then. choose the one category in each of the pairs which you expect would be most preferred by a NINTH GRADE BOY WITH JUST BARELY PASSING ACADEMIC ACHIEVEMENT. That is. mark the one you think the student would mark if the student were doing this exercise. GRN W“ 1 TQTE S UNIV. LIBRARIES llll 1lNilllWM\IIHIHIVIWHWI 9 09926142 W 2 so nICHI $3