axiavz 1.v‘.... 3 1293 10113 3928 ' " _ Michigan State -' 7 University This is to certify that the thesis entitled THE EFFECT OF BEHAVIORAL TECHNIQUES WITHIN AN INTENSIVE DESIGN UPON ELEMENTARY SCHOOL CHILDREN'S ATTENDING BEHAVIOR presented by Elizabeth Ann Bridge has been accepted towarc’ of the requiremen L degree in _ Date 8'8‘73 0-7639 i9?" ABSTRACT THE EFFECT OF BEHAVIORAL TECHNIQUES WITHIN AN INTENSIVE DESIGN UPON ELEMENTARY SCHOOL CHILDREN'S ATTENDING BEHAVIOR BY Elizabeth Ann Bridge The purpose of this study was to measure the change in attending behavior exhibited by elementary age students through the application of a token economy and behavioral contracts through the use of an intensive design. The study was conducted over a 4vmonth, 73 school— day period of time. Initially 4 teachers were selected who in turn submitted the names of youngsters demonstrat— ing nonattending behavior in the classroom. From this list of referrals ll children were randomly selected to participate in the study. The process of the study followed an own control basic design format. Prior to the study 4 teachers were trained in the use of token economies and behavioral contracts. Con- currently a trained observer began a 20—minute daily observation of each child's attending and nonattending behavior, during phase one. This daily observation was Elizabeth Ann Bridge continued for the duration on the study. In the second phase each teacher initiated a treatment program with each individual child. Five of the 11 students received contracts, while the remaining 6 received tokens. In the third phase of the study the teachers withdrew the treat— ments while the observer continued to record the classroom behavior. The teachers again utilized a treatment tech— nique in the fourth phase of the study. The 5 children who had received contracts now used tokens, while those children who had received tokens initially now used con— tracts. In the fifth phase of the study the treatments were again withdrawn, while the children's behavior was still continuously observed. The teachers used whichever of the two treatments they felt most effective in increas— ing attending behavior in the sixth phase of the study. Due to the nature of the intensive design pro— cedure, the hypotheses were stated for both individual and group analysis. A number of criterion measures was used to evaluate change in attending behavior. These were: (a) the Behavior Analysis Form, (b) the Achievement Motivation Scale, (0) the Academic Achievement Scale, (d) the Assignment Completion Scale, and (e) the Indi— vidual Student Evaluation Scale. Two additional outcome measures were also used. These were the Teacher Research Evaluation Scale focusing on the teacher's assessment of the research, and secondly Elizabeth Ann Bridge the Reinforcement/Non—Reinforcement Ratings assessing habitual use or disuse of reinforcement regarding a stu- dent's classroom behavior. The Behavior Analysis Form data were analyzed using White‘s (1972) median statistics. The Achievement Scales (motivation and academic), the Assignment Com— pletion Scale, and the Individual Student Evaluation Scale were all analyzed using a summary of mean scores for each child across phases. The analysis of the Teacher Research Evaluation Scale was also done using comparative mean scores between phases. The Reinforcement/Non— Reinforcement Ratings were totalled for each child and the mean scores across all phases of the study were found. The results of the analyzed data indicate that the two treatments techniques were able to produce sig— nificant changes in each child's attending behavior on both a meaningful and a statistical level (p < .05). From a statistical standpoint there was no order effect in the presentation of treatments, nor was there a superior treatment across subjects. The Achievement (academic and motivation), Assignment Completion, and Individual Student Evaluation Scales data revealed some mean changes between phases, but these were unique to both teacher and student. The Teacher Research Evalu— ation analysis data revealed that teachers consistently rated the study on the positive end of the continuum Elizabeth Ann Bridge with little fluctuation between phases. The Reinforcement/ Noaneinforcement Ratings data indicated that while teachers progressively increased their reinforcement of attending behavior while simultaneously increasing their ignoring of the nonattending behavior, their overall percentage of reinforcement of any behavior, regardless of category, was low. THE EFFECT OF BEHAVIORAL TECHNIQUES WITHIN AN INTENSIVE DESIGN UPON ELEMENTARY SCHOOL CHILDREN'S ATTENDING BEHAVIOR BY Elizabeth Ann Bridge A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling, Personnel Services, and Educational Psychology 1973 ACKNOWLEDGMENTS Of all those to whom I wish to express my respect and gratefulness, the most important is Dr. Norm Stewart. From him I have learned a tremendous amount about people, my profession, and myself. He has helped me experience success in achieving a goal which will hopefully give me the impetus for further growth. His knowledge and skills, patience and caring have meant a great deal to me. I would also like to express my appreciation to Dr. Bob Craig, Dr. Richard Johnson, Dr. Dozier Thornton, and Dr. Bob Winborn for their help and direction in my doctoral study. The learning process I have been through would certainly have been incomplete without the friendships and help I experienced along the way. I would like to thank Kyle Euckert for his endless hours of observation, his work with the teaching staff, and his concern for the children in the study. His friendship and help I value highly. I am indebted to Connie Ripstra for her knowledge, ideas, and patience regarding this study. Her eXpertise in data analysis done for this study I admire. ii To my friend Dianne Singleton who worked on the pilot study and in observations I owe my thanks. Her friendship and encouragement I value and rely on so much. To Anne Baucon who helped in the observations I wish to express my appreciation. My thanks also go to Mr. James Swift and the staff at Reo Elementary School for their confidence in me and their cooperation in carrying out the research. To my family, particularly my father, who encouraged me to accept and follow through on another challenge, I am grateful. iii Chapter I. II. INTRODUCTION AND REVIEW OF THE LITERATURE TABLE OF CONTENTS Rationale for the Investigation . Attending Behavior. . . . . . Token Economy . . . . . . . Behavioral Contracts . . . . . Use of a Timer . . . . . Length of Attending Behavior . . Use of Observers . . . . . . Summary . . . . . . . . . OBJECTIVES AND OVERVIEW OF THE STUDY. Experimental Hypotheses . . . . Experimental Design . . . . . Schedule of Experimental Procedures I. II. III. IV. V. VI. Definit Descrip Description of Participating Profes- sion Coun Teac Obse Descrip kWNH 0 Prepare for the Training Program. Implement Program with Teachers Determine Treatment Components Teach Teacher Treatment . Conduct Treatment Program . Conclude the Study . . . ion of Terms . . . . . tion of the Sample . . . als O O O I O C O C Selor O O C O C O C O hers 0 O O O 9 O O O rver O O I C O O Q 0 tion of Criteria Measures . Behavior Analysis Form . . Academic Achievement Scale. Assignment Completion Scale Q \ O O Q Q Q Individual Student Evaluation Scale. iv Page 13 15 l6 17 18 19 21 23 25 27 28 28 30 31 37 39 4O 43 47 47 48 48 49 49 50 51 51 Chapter Page 1. Teacher Research Evaluation Scale . 52 2. R/N Rating. . . . . . . . . 53 Description of the Outcome Analysis. . . 54 Level Of Significance . . . . . . . 57 Summary . . . . . . . . . . . . 60 III . RESULTS 0 O O C Q Q Q 0 O O O Q Q 62 Analysis of the Outcome Data . . . . . 62 General Findings . . . . . . . . . 64 Hypotheses l and 2 . . . . . . . 64 Hypotheses 3 and 4 . . . . . . . 67 Hypothesis 5 . . . . . . . . . 7O Hypothesis 6 . . . . . . . . . 72 Hypotheses 7 and 8 . . . . . . . 74 Supplementary Findings . . . . . . . 77 Summary . . . . . . . . . . . . 87 IV. SUMMARY, DISCUSSION, AND IMPLICATIONS . . . 89 Summary . . . . . . . . . . . . 89 Discussion . . . . . . . . . . . 93 Implications of the Research . . . . . 110 Refinements of the Present Study. . . . 115 In Retrospect . . . . . . . . . . 122 REFERENCES . . . . . . . . . . . . . . 124 APPENDICES Appendix A. Figures . . . . . . . . . . . . . 133 B. Tables . . . . . . . . . . . . . 144 C. Forms Used in Treatment. . . . . . . . 150 LIST OF TABLES Research Schedule . . . . . . . . . Statistical SignificanceevProbabilities of Differences in Attending Behavior Between First Baseline (B1) and Treatment Phases (Tl/C2 or Cl/Tz), Last Baseline (B3) and Follow—Up (F) . . . . . . . . . . Probabilities of Differences in Attending Behavior Between Consecutive Phases for Each Subject . . . . . . . . . . Probabilities of Differences in Attending Behavior Between First Treatment Phase (T1 or C1) and FolIOWvup, and Between Second Treatment (C2 or T2) and Follow— up 0 o o 0 Q o 9 o O o g 0 I o Test for Statistical Differences Between the Two Treatment Positions, Significance Being Designated as p < .05 . . . . . Probabilities of Statistical Differences in Attending Behavior Between Phases 2 and 4 (T1, C1 and C2, T2) for Each Subject. . . Summary of Probabilities of Statistical Sig— nificance in Attending Behavior Between the Two Treatments for Subjects Who Demon— strated a Statistical Difference Between Treatments . . . . . . . . . . . Percentage of Increase in Median Amount of Attending Behavior Between Phases. . . . Summary of the Mean Scores Obtained for Each Treatment Group Across All Phases of the Study on the Individual Student Evaluation Scale (Behavior), The Academic Achievement Scale (Motivation and Achievement, and The Assignment Completion Scale Assignments) . Vi Page 32 65 68 71 73 75 76 78 81 Table Page 10. Summary of Treatment Technique, Total Number of Points Earned and Reinforcers Chosen by Each Student During the First and Second Treatment Phases as Well as the FOIlOW'UP o o o o o o c o o c O Q 83 11. Summary of Total Mean Scores Obtained Across Each Phase Regarding Teachers Reinforce— ment vs. Nonreinforcement of Attending Behavior . . . . . . . . . . . . 85 12. The Mean Rating on the Teacher Research Evalu— ation Scale for Each Phase; A 1—7 Scale was Used with 1 Being Negative and 7 Being Positive . . . . . . . . . . . . 86 3-1. Median Percentage of Attending Behavior for Each Subject Across Each Phase of the Study 9 O O 0 O I O O O O O O Q 144 3-2. Summary of Mean Percentages for Each Subject Across Phases for the Assignment Completion Scale and Comparisons Between Consecutive Phases . . . . . . . . . . . . . 145 8-3. Summary of Means for Each Subject Across Phases for the Individual Student Evaluation Scale and Comparison Between Consecutive Phases . . . . . . . . . . . . . 146 B—4. Summary of Means for Each Subject Across Phases for the Academic Achievement Scale (Achieve— ment) and Comparison Between Consecutive Phases . . . . . . . . . . . . . 147 8-5. Summary of Means for Each Student Across Phases for the Academic Achievement Scale (Moti- vation) and Comparison Between Consecutive Phases . . . . . . . . . . . . . 148 8-6. Summary of Mean Reinforcement or Nonreinforce— ment Scores of a Student's Attending and Nonattending Behavior Provided by Teachers for Each Phase of the Study . . . . . . 149 vii Figure 1. 2. 10. 11. LIST OF FIGURES Attending Behavior Observational Becky . . . . . . . . Attending Behavior Observational Jeremy. . . . . . . . Attending Behavior Observational Chris . . . . . . . . Attending Behavior Observational Michelle . . . . . . . Attending Behavior Observational Eric 0 O O O O O O O Attending Behavior Observational Jeff 0 o o o 0 0 Q 0 Attending Behavior Observational Todd . . . . . . . . Attending Behavior Observational Karen . . . . . . . . Attending Behavior Observational John O O O O O O O 0 Attending Behavior Observational Lisa 0 O O I O I O 0 Attending Behavior Observational Debby O O O 9 0 O 0 0 viii for for for for for for for for for for for Page 133 134 135 136 137 138 139 140 141 142 143 CHAPTER I INTRODUCTION AND REVIEW OF THE LITERATURE Rationale for the Investigation Counseling research has traditionally made use of comparative group designs in which the average behavior of one sample of individuals is compared at a given point in time to a control group. Through this approach coun— selors have been able to ascertain what techniques had a statistically significant effect on the behavior of a hypothetically sampled individual. Although the merits of such a group approach cannot be repudiated, it should be viewed as neither the only viable nor the most appro— priate method for researching the outcomes of counseling. Several current writers concur with this statement (Chas— san, 1967; Shontz, 1965; Strupp & Bergin, 1969; Yates, 1970; Lazarus & Davidson, 1971). These writers have seen deficits in traditional group research which might be remedied by an alternative, complementary approach to experimentation. Thoresen synthesized their challenges in citing three commonly found problems regarding the exclusive use of group designs. First, the group approach has been of limited value in understanding the causes of change in individual behavior. Second, an excessive reliance in group designs has been placed upon statis— tical significance rather than in replication in deter— mining the effectiveness of a technique. Third, there has been a need to develop experimental control through a research design, rather than by statistical operations as substitutes for control. The alternative approach which Thoresen and others have suggested is the use of an intensive or N = 1 design. This research method is the experimental study of an individual's behavior through continuous observation as various treatment techniques are applied. This approach emphasizes observation of a behavioral process as opposed to gathering data solely about an end result. Behavior, in an intensive design, is observed and measured on a day'to—day basis so that process changes can be recognized and analyzed. Thoresen feels that the N = I approach to research will begin to focus attention on specific and effective treatment techniques for individuals, rather than on general treatment categories which apply to hypothetical clients. If we accept the premise that counseling should result in concrete differences in client behavior, then we must devise research strategies which will provide this treatment information to the practitioner. This goal, however, requires a much more scientific approach to counseling than we have taken in the past. As Lazarus and Davidson (1971) have stated, "Eventually the practi— tioner may acquire a precise and clinical rule or pro— cedure with predictable success rates and clear lines of contraindication." The researcher then has to change his line of questioning from broad questions such as, "Are behavioral techniques effective?", to more precise and specific ones which ask, "What specific therapeutic inter— ventions produce specific changes in specific patients under specific conditions?" (Strupp & Bergin, 1969). Scientific inquiry focuses the researcher‘s attention on an individual's actions under certain con— trolled conditions. It moves away from inquiry which has provided data about a group's mean performance as it may apply to an idealized client. We begin, in the inten- sive design, to concentrate on answering very specific questions about the individual and his change pattern. The N = 1 design accepts the data provided by group research which indicates that a particular general tech— nique produces resultant change. The intensive design, however, moves beyond these insights to more detailed process observations. The design complements group research by capitalizing on its findings, and conse— quently trying to scientifically analyze these broad results in regard to specific change they might produce in individual behavior. From an historical standpoint the intensive design is soundly based. It has long been held that scientific inquiry is the most legitimate approach to uncovering facts. Scientific inquiry has emphasized careful observation, precise description and systematic control (Platt, 1966). In the past counseling research has relied on comparative group design rather than attempt— ing to pinpoint and capitalize on the three character— istics of scientific inquiry. While the physical sciences moved to more scientifically controlled inquiry in addition to classical research, the counseling profession has relied primarily on the latter (Thoresen, 1972). When researchers have used a more scientific approach our profession has made some of its most significant breakthroughs. When researchers have taken data provided by previous group procedures and used them to ascertain data about individuals, we as a profession have gained additional knowledge (watson & Rayner, 1920; Jones, 1924; Dukes, 1965; Lazarus & Davidson, 1971). One of the individuals who pioneered the use of carefully controlled data about individuals was B. F. Skinner. In his work, which changed traditional research, data were continuously gathered and evaluated as various interventions were introduced. In this vein, hypotheses which he had regarding change techniques were replicated over and over to achieve concrete data having generaliza— bility (Skinner, 1945, 1953). A counselor working with an individual needs data to confirm or regulate various techniques used to change specific problems in the individual. The intensive design offers this kind of scientific approach to research which facilitates the answering of questions about an indi- Vidual's problems with an increasing degree of certainty. Intensive design research is often criticized as being deficient in meeting several of the accepted stan— dards of experimental research. When analyzed, however, these criticisms can be challenged by well—planned, intensive design research. The first complaint usually lodged against this type of design is its lack of con— trol. It has been argued that traditional group research or extensive designs accounts for the individual varia— bility which occurs in research, supposedly making this a more tightly controlled design. This logic, however, demands that individual variability be seen as a chance occurrence. Those who View human behavior from a learnv ing theory perspective such as Thoresen (1972) alterna— tively have viewed individual variation as lawful and subject to the control of the environment. In the inten— sive design, "Individual variability is not looked upon as intrinsic or accidental, but as imposed or learned by the conditions under which it occurs" (Thoresen, 1972). Intensive design, then, recognizes such variability, and, therefore, uses experimental control to examine this variability. In addition to confronting individual fluctuation from a lawful perspective, the intensive design offers a statistical advantage to research. As Thoresen (1972) noted, "The intensive design offers excellent experimental control in having each subject serve as his own control for all kinds of past events prior to the experiment." Campbell and Stanley (1966) also pointed out the degree of control this design offers. The time series design controls for all the major internal variables that might confound experimental results during an investivation. A secbnd criticism lodged against the use of intensive designs in research is its lack of generaliza— bility. Chassan (1967) argues that in group designs, where significance and generalizability are thought to be straightforward, certain flaws exist for the pran titioner who intends to utilize the results. He states, "Even if such (group) comparisons are statistically sigv nificant, the observed mean difference between groups still does not clearly demonstrate that the difference was generally effective for all subjects." Thoresen (1972) adds that, "In fact the obtained differences can be the result of a few who changed dramatically while most changed little. The effect on the particular indi— vidual remains obscure." Issue can be taken then with the applicability of group results when they are applied to the change we are trying to produce in individual clients. From a theoretical standpoint the issue may be looked at as the difference in generalizability between statistical and meaningful differences. As Thoresen (1972) stated: The current obsession with a statistically signifi~ cant finding is not in itself a major finding. It is certainly not the basis for an investigator to claim that an important empirical fact has been established. The important facts in science are not established by statistical significance, per se, but by careful observation, precise control and systematic replication of findings. The intensive group design provides this type of control. Generalization is gradually established through a series of systematic replications with the same subject (intra— individual) or across subjects (interindividual). Repli- cation here means not only repetition for subjects but for situations and time limits involved in the experiment as well (Thoresen, 1972). Taken as a research strategy the intensive design offers an approach which is scientific, controlled, sta— tistically sound, and useful to both researcher and prac- titioner. Viewed as an alternative, complimentary approach to traditional research design, its advantages for counseling research are multiple. Thoresen (1972) summarizes these merits by listing eight attributes of intensive designs. 1. The specific actions of individuals are the unit of focus, rather than an average comparison between groups. 2. The frequency, magnitude and/or variability of the individual's actions can be examined con— tinuously during each phase and between phases of the investigation. 3. The individual subject serves as his own control in that the magnitude and duration of change is compared to his own baseline of actions. Past and individual differences are carefully controlled. 4. Experimental control of variables is greatly facilitated thereby reducing the need for sta- tistical control through complex inferential statistics. 5. The effects of treatments administered simul- taneously on one or more client behaviors (depen— dent variable) can be examined over time for a particular individual by use of multiple base- lining. 6. Causal relationships can be established by replication (reproductability) of specific results by means of certain intervention tech— niques across individuals. Evidence of generali— zation is systematically gathered without recourse to the untenable assumption of random sampling. 7. The clinician can determine the extent of rele— Vent changes in client action continuously during treatment and after treatment, if necessary, based on the data provided. 8. An intimate method for controlled inquiry of covert (internal) behavior of individuals is provided. This makes it well suited to humanists (Thoresen, 1972). Based on the merits delineated by Thoresen (1972) and others in discussing an intensive design approach to research, it seems to represent a viable alternative to traditional comparative group research designs. Apply— ing the N = 1 approach, data being uncovered through group comparison techniques could be refined to assess the specific change such findings have on the behavior of individuals. Intensive design research provides a means of scientifically verifying the mean results found in group studies through replication of treatments on individuals. Such an approach, if popularized, could provide the answers to treatment questions facing the practitioner. The intensive design should in fact further counseling research for the benefit of those who utilize the techniques as well as those who are the recipients of the treatment. Attending Behavior Since the intensive design demonstrates applica— bility for use in counseling research, its merits in dealing with common behavioral problems in the elementary classroom should provide practical results. One such prevalent concern facing the elementary counselor work- ing with children is to deal with inattention in the classroom. The practitioner is faced with the dilemma of finding effective techniques which produce demonstrable increases in the attending behavior of children in a school room. Counselors working in conjunction with teachers are concerned with the student as learner. According to Meacham and Wiesen (1970) we can classify this learning behavior into two broad categories: academic and nonacademic. They defined academic as any behavior related to the goals of the curriculum. How— ever we choose to categorize this learning, it divides superficially into those behaviors the teacher wishes to strengthen or maintain and those she wishes to diminish or eliminate. If our overall goal is to facilitate 10 learning, then, it is appropriate that the behaviors a teacher wishes to increase are those requiring a student to attend to the classroom activity. Specifically these behaviors are defined in numerous ways. They included "studying, reciting, asking questions, and attending to the teacher" (Meachan and Weisen, 1970). Others have described in even more detail the concept of attending and nonattending. Thomas, et a1. (1968), recognized five classes of disruptive behavior: (a) Gross motor (getting out of seat, hopping, skipping, standing up, walking, jumping, rocking the chair, kneeling in the chair, arm flailing etc.), (b) Noise making (tapping, clapping hands, tearing papers etc.), (o) Verbalizations (crying, screaming, singing, whistling, laughing, coughing, talking with other children etc.), (d) Orientating (turning head or body around towards another person, showing objects to another person, looking at another child etc., and (e) Aggressions (hitting, pushing, shoving, pinch— ing, slapping, striking, poking with objects, grabbing objects from another child, destroying objects etc.). He then described appropriate behaviors as "Looking at the teacher, raising hand and waiting for the teacher to respond, writing answers to workbook problems, reading assigned materials, etc." (Thomas, et al., 1970). Across definitions the component of increasing those behaviors which facilitate learning and decreasing those which impede it exist. Neither researcher attempted to define a standard learning style for the student, but rather examines those 11 observable behaviors which indicate that the classroom learning environment is being utilized by the student at that time. Teaching students to attend to their classroom tasks is a meaningful problem facing practicing counselors. The amount of literature dealing with attending techniques has given additional credibility to the need and concern for this behavioral problem. In attempting to deal with this concern, the literature revealed a variety of tech— niques which have been tried in order to increase attending behaviors or decrease disruptive, nonattending behaviors. These techniques fell into two basic cate— gories, reinforcement and/or punishment of a target behavior. The majority of studies produced at least a degree of the desired behavior when analyzed as the mean attending score for the group under study (Ferreira, 1969; Johnson, 1970; Mattor, Mattson, Walker, & Buckley, 1969; Broden, Vance, Dunlap, & Clark, 1970; Bearley, 1970; Wolf, et al., 1970; Hops, 1971). In these studies various reinforcement systems, such as verbal reward, physical attention, tokens, tangible rewards and modeling were used. Often coupled with these rewards was the simul— taneous use of aversive techniques such as time out or ignoring disruptive behavior (Hall, Lund, & Jackson, 1968; Thomas, Nieben, Kuypers, & Becker, 1968; O'Leary & 12 Becker, 1968-1969; Dickinson, 1968; Madsen, Becker, & Thomas, 1968; McAllister, Stachowaik, Baer, & Conderman, 1969). A legitimate question arises when the literature on attending techniques is reviewed. That is, of what utility is still another piece of research dealing with this tOpic? From a quantitative standpoint the area would seem to abound in conclusively effective techniques which could be applied to a given attending problem. Unfortunately, this type of scientific data does not exist for the practitioner's use. From a research standpoint there are several flaws in general research which may account, in part, for this lack of conclusive data. Gelfard and Hartman (1968) listed several sources of current criticism. Among them were: (a) The need for adequate baselines, (b) Gaps in data collection, (c) The desirability of multiple versus single behavior analyses, (d) A failure to employ contingency reversals. In still another effort to pinpoint the failure of various experiments, Kuypers, Becker, and O‘Leary (1968) mentioned several other research flaws which tended to weaken the resultant data. Among these were: (a) An absolute standard for reinforcable behavior, rather than a shaping procedure, (b) Teachers continued to attend to unwanted behaviors, and (c) The study took too much of the teacher's time. 13 An additional weakness which Thoresen (1972) and others suggested is the lack of specific, concrete data available on an individual which has been replicated sufficiently to produce generalizable results. One step towards producing this type of data, for the researcher as well as the practitioner, would be the use of an intensive design for carefully measuring the effects of various commonly tested techniques on the attending behavior of an individual. Then to produce concrete, conclusive results the same experiment should be repli— cated across numerous subjects (interindividual). Through this kind of research, which has been carefully planned so as to eliminate the common research flaws previously mentioned, the answers as to how to increase attending behavior should begin to emerge. At the very least it would give us more scientific data about the existing techniques which are being used in the field. In an effort to produce a research design which encompasses the desired characteristics mentioned prev Viously, two commonly utilized techniques were selected as treatment variables. Token Economy One technique employed in the counseling field to increase attending behavior was a token economy. Basically, this system involved rewarding a student for eliciting the desired behavior through the use of tokens, 14 marks, or chips. Generally these tokens were traded for rewards which the student selected. This system has been used by a number of experimenters with varying degrees of success (O'Leary, Becker, Evans, & Saudargas, 1969; Wagner & Guyer, 1971; Hops, 1971; Blaker, Bennett, 1970; O'Leary, Daniel, & Becker, 1967; Broden, Hall, Dunlap, & Clark, 1970). One variation used to supplement the token economy approach was the use of a work chart to record checkmarks or tokens which the student obtained (Ferreira, 1969; Martin, Schwyhart, & Wetzel, 1967). In both experiments this technique coupled with the teacher attention provided when the marks were administered increased the attending behaviors in the students. The advantage of this approach from the teacher‘s stand— point is that (a) the student is not tempted to play with the tokens or chips, and (b) the teacher is not hampered by having to carry around a number of chips to reinforce a student. The marks on the work charts were then exchangeable for primary rewards or preferred activities. As noted by Homme, C'deBaca, Deirre, Stein— horse, and Rickert (1963) this was an application of the Premack Principle (Premack, 1959). The token economy approach appears to not only increase attending behavior, but is simultaneously a simple method which a teacher can employ in a classroom setting. As stated by Kuypers, 15 Becker, and O‘Leary (1968), devising a token system which involves minimal teacher time, and which aids teachers in focusing on attending rather than nonattending, is essential to a successful reinforcement program. Behavioral Contracts A second technique reported in the literature is the use of behavioral contracts. Cantrell, Cantrell, and Huddleson (1969) used a behavioral contract approach to increase the academic achievement of a male student. Using the contract as a basic technique, Cantrell first specifically defined the behaviors he wished to increase in the boy's repertoire. He then determined potent r reinforcers which were made contingent upon the attain— ment of these academic behaviors. The student was rewarded on a point basis for each behavior obtained. When he had accumulated a sufficient number of points, the student was allowed to trade in his points on a reward of that "price." Cantrell, et al., reported the improvement of grades through this technique. In another study by Kiersey (1957) the written contract format was used in a number of cases within a particular California school district. Although no percentages of decreasing disruptive behavior were noted, Kiersey predicted that this technique would reduce disruptive classroom behavior to some extent 75—90% of the time. — JIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIll-IIIIII-IIIIl-I-I-I-I 16 As a change technique, then, behavioral contracts also offers a means of reducing disruptive behavior in the elementary school child. This method has several specific advantages. First, it is an approach which, once formulated and organized, requires only a minimal amount of teacher time and effort. Second, it necessi— tates that both teacher and consultant specifically list those behaviors which they feel are disruptive and those which are attending. This eliminates controversy and focuses teacher attention on particular behaviors. Third, it utilizes reinforcement techniques such as rewards to motivate attainment of attending behavior which is potent in producing change. Use of a Timer One potential drawback to the use of behavioral contracts has been their inherent delay in rewarding appropriate behavior. Renner's (1964) findings demon- strated that behavioral changes proceeded most effec— tively when reinforcement was made immediately contingent upon the behavior one wishes to foster. Generally, the degree of control provided by reinforcement over a par- ticular set of actions tends to decrease in effectiveness as the length of time between reward and behavior increases. In an attempt to maximize the probability of success during the use of behavioral contracts, a means 17 of shortening the time between behavior and reward must be found. A watch or timer used to measure the length of attending could be used for that purpose. Ferreira (1969) used this technique coupled with a token economy program to increase attending behavior. He reported that this was an aid in developing a time—work orientation. Wolf (1970) also used a timer as a means of identifying the time at which the subjects were to have demonstrated in—seat behavior. For purpose of this study the timer was coupled with the use of behavioral contracts. Such an approach could have (a) called both the teacher's and the stu— dent's attention to the attending behaviors themselves for a specified period of time, (b) made the use of successive approximation or shaping (Tharp & Wetzel, 1969) of attending behavior more systematic, and (c) given the student a sense of immediate feedback or reward for having attended for a specific amount of time. Length of Attending Behavior The approximate length of time we can expect a student to attend to a classroom task depends on many variables. The difficulty of the task, the age of the child, and the norm for attending in that particular classroom all come to bear on the ideal length of time spent attending. As a gauge, however, Hill and Buckley (1968) substantiated the ability of a student to move IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIllIl-----::__________ 18 from 30*600 seconds of attending. The data available indicate that setting a criterion level of attending is possible. A system which looks at the baseline behavior to identify the current level of attending and then successively approximates longer increments of attend— ing should prove effective. Thus by gradually increasing the segments of attending behavior, a long—term goal of at least 30 minutes seemed attainable. Use of Observers The need for accurate and continuous collection of behavioral data by trained observers has been attested to by several authors (Gelfand & Hartman, 1968; Thoresen, 1972; Ray, Shaw, & Cobb, 1972). Ray, et al., revealed that the verbal reports of parents and teachers, as well as a counselor's judgments about a client‘s behavior, has been unreliable. On the contrary what is needed, he states, is objective data from which to determine the specific degree of success obtained. For the purposes of this study the observer ful— filled this function of providing objective data. In addition he (a) provided Specific objective data for the experimenter and practitioner's utilization, (b) pro- vided the teacher and experimenter with frequent feed— back from which to assess a teacher‘s ability to carry out the techniques involved (Madsen, Becker, & Thomas, 19 1968), and (c) aided the teacher in attending to target Ibehaviors through a cueing method while the observation is being conducted. This last observer function, cueing, is a par— ticularly effective technique for alerting the teachers to attend to appropriate behavior. Several authors point to the efficacy of a cueing method in teacher training of reinforcement principles (Hall, Jackson, Lund, 1968; Allen, Henke, Harris, Baer, & Reynolds, 1967). Both Hall, et al., and Lund, et al., conducted studies in which the observer signaled the teacher when the attending or target behavior occurred. This helped teach discriminative skills and assured the ability of the teacher to at least initially reinforce appropriate behavior. The cueing method had merit for both treatment methods proposed, token economy, and contracts. Summary As an innovative and productive research strategy the intensive design offers the experimenter and praC* titioner alike a scientific method for studying behavior. It meets the criteria established for traditional research, yet provides unique and practical data about individual behavior which has not been sufficiently tapped in the past. From a research standpoint one area which could benefit from concrete data about effective 20 techniques is the problem of increasing attending behavior. The task of focusing a child's attention during class time remains troublesome to teachers and counselors despite accumulated data. This study was conducted in order to measure the change in attending behavior elicited by elementary age students through the application of a token economy and behavioral contracts, utilizing an intensive design approach. It was hoped that, through careful and continuous observations made on an individual, the specific effects of treatments on attending behavior could be more precisely defined. A further step, that of repeated replication of this pr0* cedure, could have helped clarify even further how indie viduals react to attending behavior change techniques. As research data are accumulated in this manner, scien- tific judgments about how techniques could be used across clearly defined groups will be ascertained. CHAPTER II OBJECTIVES AND OVERVIEW OF THE STUDY The overall purpose of this study was to measure the change in attending behavior elicited by elementary age students through the introduction of two different behavioral techniques. To accomplish this, 11 students were randomly selected from a group defined by teachers as nonattenders. Each child was initially observed for two weeks to determine his base rate of attending behavior. Simultaneously teachers of these 11 students were provided in-service training to familiarize them with two behavior change techniques they would be using. In the second phase of the study each student was given one of the two treatments. Six students received a token economy while five received behavioral contracts. This phase also lasted two weeks during which the students were observed daily to determine the extent of their attending behavior. In the third phase of the study a reversal was initiated. In this two—week segment the treatment techniques were not used. Each of the 11 students was again observed daily to assess what changes occurred in his attending 21 22 during this reversal. In the fourth phase, the same two treatments were again introduced. This time, however, the students who had previously received a token economy now received behavioral contracts, while thoSe who began with contracts now received tokens. This phase also lasted five weeks, and the students were again observed on a daily basis. In the fifth segment of the study ~another reversal was initiated. Treatment was withdrawn, while each student's behavior was still observed daily. In the sixth phase of the study the teachers were free to use either of the two previously tried techniques or a hybrid combination of the two.. Again in this two-week phase each student's attending behavior was observed on a day-to-day basis. Five outcome measures were used in this study. The first was a set of daily 20~minute observations recorded for each of the 11 students on the Behavioral Analysis Form. The second measure was an Academic Per— formance Form which the teachers filled out on the next to last day of each phase. This rating provided a record of each student‘s academic achievement and moti— vation level in each of his subjects over the preceding two—week period. The third outcome measure was an Assignment Completion Form also filled out once during each phase. This form recorded the percentage of com— pleted assignments in each class subject for each — ' -—:-~___.—_ — —_ 23 student over the previous two-week period. The fourth outcome rating was a Student Evaluation Form on which the teachers rated each student on a series of descrip~ tive phrases related to behavioral changes they might have exhibited. This was an attempt to get teacher input into the evaluation of a student's classroom behavior during each phase of the study. The fifth outcome measure was a Teacher Research Evaluation Scale. Through this rating form the teachers evaluated the study's effect on their own behavior as well as on the attending behavior of their students. This rating form was also completed on the next to last day of each research phase. Experimental Hypotheses A number of specific hypotheses regarding an increase in the attending behavior of students were evaluated on an individual basis in this study. The overall purpose of this N = 1 research was to determine if significant changes in attending behavior occurred in each student's case. Above and beyond this the 11 individuals were viewed as a whole to determine if any trends occurred across students. Two symbols used in the hypotheses might cause confusion and will therefore be clarified. The symbol TlCl refers to the first treatment and reflects the fact that some students initially participated in the token 24 economy treatment (Tl) while others received the behavioral contracting treatment (Cl) first. Likewise, the symbol C2T2 refers to the second treatment and reflects the fact that some students, on the other hand, received the behavioral contracting treatment (C2) and others received the token economy (T2) in the second treatment phase. In the C2T2 phase, those students who had originally received 2, while those who had received Cl now received T2 as their second treatment. For each individual it was hypothesized: T1 now received C Hypothesis 1: During the first treatment (T or Cl ) attending behavior will increase over the baseline observation level as indicated by the median slope trend. Hypothesis 2: During second treatment (C or T ) attending behavior will increase over the baseline observation level as indicated by the median slope trend. Hypothesis 3: During phase B (first return to baseline conditions) attending behavior will decrease from the first or C ) treatment level, as indicated by the gdian sIope trend. Hypothesis 4: During the phase B (second return to baseline con— ditions) attending behavior will decrease from the second (C or T ) treatment level, as indicated by the median slope trend. 25 Hypothesis 5: During the F phase (follow—up) attending behavior will be greater than during the B phase, but less than during either of the treatment phases (T1, C1 or C2, T2) c Hypothesis 6: No difference will be found in the attending behavior between the first treatment (T or Cl) and second treatment (C2 or T2), as indicated by the median slope trend. For the group analysis it was hypothesized: Hypothesis 7: No difference will be found among individuals receiv— ing the first combined treatment (T1, C2) and the second combined treatment (C1, T2). Hypothesis 8: No difference will be found between the group receiv- ing the first combined treatment (T1, C2) and the second combined treatment (C1, T2). Experimental Design The design chosen for this study was an intensive design which followed the general A, B, A, C, A, F format described by Thoresen (1972). The six phases are here described in the sequence used within the study: B1 = Initial baseline observation of attending behavior collected for ten days. 26 T or C = The first treatment phase lasting ten days in which six students received a token economy (T/C) and five students received contracts (C/T). B = The second baseline observation or reversal of attending behavior which lasted ten days. Treatment was withdrawn. C or T = The second treatment phase lasting ten days in which six students received con— tracts (C/T) and five students received a token economy (T/C). B = The third baseline observation or reversal of attending behavior which lasted 23 days. Treatment was withdrawn. F = The follow-up phase in which teachers used either of the two treatments, or a hybrid combination of them. This phase lasted ten days. The intensive design as a research approach has been discussed by several authors (Thoresen, 1972; Chassan, 1967; Shortz, 1965; Strupp & Bergin, 1969; Lazarus & Davidson, 1971; White, 1972a). The particular design selected for this study was discussed by Thoresen (1972) as an approach used in a study he conducted in conjunction with several associates. In this study 27 Thoresen was attempting to measure the performance of a class of students (serving as a single subject) regarding in-seat behavior as two change techniques were initiated by teachers. This general design format approximates Campbell and Stanley's (1963) time series design. Specifically, it resembles design eight, the equivilent time samples design. Campbell and Stanley noted that the most typical early use of this experimental design was in studies analyzing the efficiency of student's work under various conditions (Allport, 1920; Sorokin, 1930). Campbell and Stanley also noted (p. 45) that, "This approach could be applied to a sampling of occasions for a single indi— vidual." Campbell and Stanley indicated on a table (p. 40) that this quasi—experimental design meets all the criteria for an internally valid design. Schedule of Experimental Procedures The various phases of the study are presented here, as well as an account of the steps taken prior to the study. This is done to outline in a systematic fashion the research steps which were taken in carrying out this study. The underlined headings coincide with descriptors which might be used to transform these steps into a flow chart to visualize in capsule form the research process. 28 I. Prepare for the Training Program 1. Receive the teacher's referrals. Of students demonstrating nonattending behavior in the classroom. 2. Select students. Randomly select 11 subjects from the total number of referrals submitted. 3. Collect background data. Both the students and the teachers involved were asked to complete background information forms. The teacher's form was used to ascertain pertinent background data which would help in describing the student. The student completed forms were open-ended questions which were used to discover potential reinforcers unique to each student. II. Implement Program with Teachers 1. Conduct observer training program. Observational data were needed in order to draw conclusions about a student's attending behavior. At this point in the study an observer was trained to use the Behavior Analysis Form. The observer was given detailed information in the use of this form. Once familiar with the rating sheet, he was given observational guidelines described by Bijou and Peterson (1969). These included ‘IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII-----:______________ 29 information as to how to observe, where to sit, how to act when approached by children, etc. When the observer was ready he went into a classroom setting for several days to practice using the Behavior Analysis Form. On the fourth day of training, video-tapes of students not par— ticipating in the study were made. On the fifth day of training the observer, the experimenter, and two other individuals familiar with the rating form co-rated the video-taped behavior. A tape was used so that all co—raters could simultaneously observe the same behavior, and so that reliability scores between observers could be obtained. The proposed reliability score which would have allowed an observer to function on his own in a classroom was .75. The relia— bility obtained at the training time ran between .875 and 1.0 for all four co—raters. Begin observers collection of baseline data. At this point the observer moved into the class- room setting of each of the 11 students in the study. This step coincides with phase B1 of the design plan. Baseline attending behavior data were collected for 10 days on each student for 20 minutes each day. According to White (1972b) this amount of data gives the most stable behavior predictions. III. 30 Determine Treatment Components 1. Select the most powerful treatments. The review of the literature indicated that certain behavioral techniques proved effective in increasing attend— ing behavior. Those selected for this study were a token economy system using successive approxi— mation of behavior and cueing by an observer as supplemental techniques. The second, behavioral contracts, also utilized successive approximation, cueing by the observer and using a timer as supplemental techniques. List treatment components. The characteristics of each treatment are here defined. Token economy encompassed: (a) Work chart on desk (b) Successive approximation (c) Observer cues to teacher (d) Reinforcement from items selected by a stu— dent from a reinforcement menu, based on points obtained through increases in attending. Behavioral contracts encompassed: (a) Use of a structured behavioral contract which was discussed and completed by teacher and student simultaneously (b) Use of successive approximation 31 (c) Use of a pocket watch to cue the student to the attending behavior itself as well as to increase the length of attending (d) Reinforcement from items selected by the stu— dent from a reinforcement menu, based on points obtained through increases in attending behavior. IV. Teach Teacher Treatment 1. Schedule of the study. In order to familiarize the teachers involved with the overall process of the study as well as their particular inputs, a schedule of the research plan was presented (see Table l). A more detailed description of each research phase is presented earlier in the chapter. 5 10 10 10 10 23 10 Days Days Days DayS‘ Days Days Days Observer Training Bl T1 or C1 B2 C2 or T2 B3 F 2. Implement teacher training. The four teachers were given a packet of materials at their first in-service session. The training materials set included: directions for treatment implemen— tation, charts, contracts, pocket watches, 32 Table 1 Research Schedule Number of Date School Phase Teacher Function Observer Function Days Jan. 15-19 5 Observer 1. Select students 1. Use behavior Training 2. Fill out back— analysis form ground data in classrooms forms. and with co- raters using video-taped material until reliability is acceptable. Jan. 22- 10 B1 1. Attend in-service 1. Observe each stu— Feb. 2 sessions to learn dent in their the two change classroom 20 techniques. minutes each day. Feb. 5-16 10 T1 or Cl 1. Administer 1. Observe each behavioral con- student 20 tract treatments minutes each to six students. day. 2. Administer token economies to five students. Feb. 19- 10 B2 1. Review the two 1. Observe each March 2 treatment tech— student 20 niques in an minutes each in-service day. session. March 5- 10 C2 or T2 1. Administer 1. Observe each 16 token economy student 20 to six students minutes each who had received day. contracts. 2. Obtain a second 2. Administer inter—rater behavioral con- reliability tracts to five score on the students who had 'behavioral receiVed tokens. analysis form. March 19- 23 B3 1. Review the two 1. Observe each APril 27 treatment tech- student 20 niques in an in- minutes each service session. day. April 30- 10 F l. Administer either 1. Observe each May 11 tokens, contracts student 20 or combination to minutes each students without day. counselor super— vision. 33 reinforcers, successive approximation schedules, data accumulation forms, research schedules, and a reinforcement menu. Each piece in the packet was explained to the teachers so that they were familiar with its use. The teachers were also encouraged to ask any questions they had. They were also instructed as to the observer's role in collecting the 20—minute observations and in cueing them to attend to certain behaviors. Those teachers using a token approach first: (a) (b) Attached work charts to the desk of each student in the study. Explained to each student how the chart worked. That is, on the first day he was given a mark for each minute he attended up to 10 minutes. If for example on the first day he attended all 10 minutes he accumulated 10 points. When he had earned 10 points two days consecutively, he was then allowed to try and attend for 15 minutes on the third day. When he was able to earn 15 points two days consecutively, he could attempt to earn 20 on the third day, etc. 34 (c) Gave rewards to students at the end of each treatment day. Each reward from the rein- forcement menu carried a numerical value. That item or privilege could be "purchased“ for that number of points. The student was allowed to purchase items daily or he could save up his points to purchase a higher— priced item. (d V 3 days of the treatment phase, indicating to the teacher whether the student had been attending or nonattending each minute. The Received a cue from the observer for the first observer nodded to the teacher at each minute interval if the student had attended. At the end of each day's treatment time the observer told the teacher the number of minutes he observed the student attend so that the teacher could compare her observation with that of the observer. This procedure was used to help the teacher identify attending behavior, and to indicate when was the time to reward with marks on the work chart. This cueing technique was faded out after the 3—day period ended, and the teachers were left to independently observe and reward. 35 (e) Recorded accumulative data at the end of each day. Data included: (a) How many points could have been potentially earned, (b) the total number actually earned, and (c) the date. If the student decided to "spend" some of his points the teacher also indicated the number of points spent as well as a description of the reinforcer it was spent on. In this way the teacher kept a running tally on the stu— dent's points and purchases. It also indi- cated to the teachers when to move the cri— teria behavior up 5 minutes (when the student had earned the total number of attending points possible at a given level for two consecutive days). Those teachers using a contract approach first: (a) Took out a behavioral contract and discussed it with each student. The contracts stated that the students had to attend for a spe— cified period of time each day. This attend— ing behavior was also put under a successive approximation procedure. The students began by attempting to attend for 10 minutes at a time. When they were able to earn all 10 attending points two consecutive days, (b) (C) (d) (e) (f) 36 they were allowed to try and earn 15 points on the third day, etc. Helped the student associate to attending at a given time during the day by providing the student with a pocket watch to put on his desk. The watch remained on the student's desk during the period he was attempting to attend so that he was aware of the total number of successful minutes he had accumu— lated. This helped cue the students to attend as well as act as a source of imme— diate reinforcement for his attention. Verbally indicated to the student at the end of each.day the total number of points accumulated. Allowed the students to trade in his points on items from the reinforcement menu under the same conditions already outlined under token economy. Signed the behavioral contracts along with the student for an agreed upon time of one week. When the week ended a new contract was signed by both the teacher and student. Compiled data about points and reinforcers as described in Token Economy. 37 V. Conduct Treatment Program During this phase of the research the treatment was conducted. Each teacher participated using the tech— niques and materials they had studied during their in— service training. The students were given either a con— tract or a token treatment on an alternating basis. Between treatment phases reversals were used to check for changes in the levels of attending behavior. This phase of the research plan lasted 73 school days over a four—month segment of the school year. 1. Begin teacher implementation of treatment. On February 5 the teachers began using their assigned treatment programs. Two teachers used a token economy (T) with a total of six students, while the other two teachers used contracts with their five students (C). This T C phase lasted for 1’ l 10 days, or two school weeks. 2. Continue observation. During all phases of the study from the beginning of gathering baseline data to the end of the follow—up phase, the observer had been observing each of the 11 stu- dents for 20 minutes per day using the Behavior Analysis Form. All of these daily observations were then compiled nightly by the observer and put in a notebook for each student. 38 Begin phase B During this phase a reversal took 2' place. No treatment was given, but observation of attending behavior continued. During this phase the teachers attended another in—service session to discuss and review the treatments for the next phase. Conduct phase C2, T2. During this phase the two teachers who used tokens previously now switched and used contracts with their six students. The two teachers who had used contracts first, next used tokens with their five students. At this time also a second check on inter— rater reliability using the Behavior Analysis Form was done. As before, video—tapes of students not in the research were taken. The observer, the experimenter, and the same two co—raters used before viewed and rated the student‘s behavior. The proposed acceptable reliability score was set at .75. The actual reliability obtained ran between .90 and 1.0 for all inter—rater relia— bility scores. Conduct phase B3. In this phase another reversal, or return to baseline conditions took place. No treatment was administered, but the obser— vation of attending behavior continued. 39 During this phase the teachers attended a third in—service training session to discuss and review the treatments for the next phase. Conduct phase F. During this follow—up phase the teachers were free to use either of the two treatments previously used, or a combination of them in an attempt to increase the attending behavior of their students. They worked indepen— dently of the counselor during this phase. Daily observations of attending behavior continued as usual. VI. Conclude the Study l. Discontinue observation. The observer discon— tinued the collection of behavioral data at this point (May 12). Analyze data. The observational data were analyzed using median statistics as described by White (1972b). Evaluate data. The median statistics were evalu— ated to determine what behavioral changes occurred and what significance these changes demonstrated. Report the data. At this point the counselor gave specific feedback to the teachers concerning the effectiveness of the techniques used. 40 Definition of Terms Attending Behavior (as used on the Behavior Analysis Form) In—seat.——Seat contact. Assigned task.——Reading, writing, or discussing the assigned task. Raising hand.-—Raising hand as opposed to seeking attention through speaking or shouting. Answering.-—Responding when a question has been asked by the teacher for the whole class on a particular individual to answer. Nonattending Behavior Talking.——Talk not classified as answering. For example, blurting out answers when another was called on, or talking to peers when not asked to do so, etc. Verbal noise.——Whistling, humming, car or animal noises, etc. Physical noise.——Tapping feet, shuffling feet, kicking the chair, knocking, etc. Staring into space.—-Looking around the room, at the ceiling, out the window, or at classmates. Physically disturb peers.——Using hands, feet, or object to touch, hit, or disturb peer so that the peer responds verbally or physically. 41 Out of seat.——Lack of seat contact for more than 5 seconds (usually out of seat entirely, or wandering around the room). Squirming.——Wiggling in the chair, searching through the desk unnecessarily, turning around in the chair, etc. Handling objects.-—Playing with various objects such as toys, pencils, paper clips, etc. This would also include the throwing of these objects. Reinforcement menu.—-A reinforcement menu, as described by Krumboltz and Thoresen (1969), was "A kind of menu of activities that represent high probability responses." The reinforcement menu used in this study was a collection of potentially rewarding privileges and tangible items listed in a colorful folder called a Trade-in Book. Each item listed in the book had an accompanying numerical value. Students were allowed to leaf through the book and select rewards for which they had accumulated enough points to make a purchase. A student was only allowed to make a "purchase" at the point he had accumulated the total amount needed to buy a given item. Items in the reinforcement menu ranged in "price" from 10-200 points. Behavioral contracts.-—A reinforcement menu, as described by Krumboltz and Thoresen (1969), "Is an 42 agreement between two or more persons specifying what each will do for a stated period of time." The contracts used in this study were weekly paper agreements signed by both a student and his teacher. The contract described: (a) the behavior the student was to exhibit, (b) the total amount of time he was to attend each day, based on successive approximations, (c) the procedure involved in receiving points for attending, and (d) the reward available for accumulating points. Token Economy.—-A token economy, as described by Bandura (1969), is a system in which "Individuals earn tokens for performing desired behaviors which can later be used to obtain, among other things, access to pre- ferred activities." In this study the token economy system was run using work charts to record attending behavior. Each student was given a desk chart on which the day's date and the maximum number of points which could have been earned that day were listed. To the right of this information was a blank space where the teacher placed check marks for each minute the child attended while the behavior was going on. At the end of each treatment session these check marks were totaled for the student who received points equivilent to the total number of minutes he attended. The points could then be traded in for items listed in the Trade—in Book (reinforcement menu). 43 Description of the Sample The population of students considered for this study were elementary age children ranging from the first to sixth grades who had been categorized by their teachers as exhibiting nonattending behavior. The particular sample selected for this investi- gation were a group of 11 students within the population of students at Reo Elementary School in Lansing, Michigan. Each of these students possessed the following set of characteristics: 1. Each student was selected by his teacher for displaying nonattending behavior in the classroom. 2. Each student was referred by his teacher to the school counselor. 3. Each of the 11 students was randomly selected from the total group of nonattending referrals received by the counselor. 4. Each of the students was a member of a classroom whose teacher had agreed to participate in the research. In an attempt to further characterize the stu— dents selected, two background data forms were completed on each student prior to the commencement of the study. One form, the Student Background data sheet, was intended to elicit potential sources of reinforcement unique to 44 each child‘s taste. The second, a Teacher Background Data Form, was intended to assemble pertinent personal, family, and test data about each student. These two forms were used to describe each student thoroughly enough so that generalizations about treatments could be made following the research. Due to the nature of this research approach, an intensive design, it is important to describe each sub- ject briefly so that the student's individual and group qualities are known. 1. Jeff Age: 13 years, 2 months Grade: 6 Teacher: Ms. E. Siblings: 3 Parents: Living together Mother — 4 yrs. college Father — 2 yrs. college Test Scores: Stanford Achievement Test (1972) Total Reading - Stanine 3 Total Arithmetic — Stanine 2 Otis (1970) ~ Stanine 5 2. Todd Age: 12 years, 8 months Grade: 6 Teacher: Ms. E. Siblings: 1 Parents: Living together Mother - 12th grade education Father — 10th grade education Test Scores: Stanford Achievement Test (1970) Total Reading — Stanine 1 Total Arithmetic — Stanine 4 Otis (1970) - Stanine 2 45 3. Christopher Age: Grade: Teacher: Siblings: Parents: Test Scores: 4. Debby Age: Grade: Teacher: Siblings: Parents: Test Scores: 5. Karen Age: Grade: Teacher: Siblings: Parents: Test Scores: 6. Eric Age: Grade: Teacher: Siblings: Parents: Test Scores: 12 years, 6 months 5 Mr. R. 1 Living togehter , Stanford Achievement Test (1972) Total Reading ~ Stanine 4 Total Arithmetic — Stanine 4 11 years, 10 months 5 , Mr. R. ‘Living together Mother — 9th grade education Father - 12th grade education Stanford Achievement Test (1972) Total Reading - Stanine 3 Total Arithmetic - Stanine 2 8 years, 9 months 2 MS. K. 1 Living together Mother - 11th grade education Father - 10th grade education Stanford Achievement Test (1972) Total Reading — Stanine 4 Total Arithmetic — Stanine 5 9 years, 1 month 2 Ms. K. none Living together Mother — 12th grade education Father — 12th grade education Stanford Achievement Test (1972) 6. Eric 7. John Age: Grade: Teacher: Siblings: Parents: I Test Scores: 46 (continued) Grade Equivalent Scores Paragraph Meaning — 1.3 Spelling — 1.6 Language - 1.5 Arithmetic - 2.0 8 years, 6 months 2 Ms. K. 1 Living together Mother — 3 yrs. college Stanford Achievement Test (1972) Total Reading - Stanine 4 Total Arithmetic - Stanine 6 8. Michelle Age: .1 Grade: Teacher: Siblings: Parents: Test Scores: 9. Becky } Age: Grade: Teacher: Siblings: Parents: g‘ Test Scores: 11 years, 7 months 5 Mr. R. 3 Mother — 10th grade education Father — deceased Stanford Achievement Test (1969) Total Reading ~ Stanine 2 Total Arithmetic — Stanine 5 7 years, 10 months 1 Ms. C. Divorced or separated Mother — 12th grade education Stanford Achievement Test (1972) Total Reading — Stanine 1 Total Arithmetic — Stanine 2 47 10. Lisa Age: 6 years, 9 months Grade: 1 Teacher: Ms. C. Siblings: 1 Parents: Separated Mother — 12th grade education Test Scores: Metropolitan Reading Readiness (1972) Stanine 6 11. Jeremy Age: 6 years, 9 months Grade: 1 Teacher: Ms. C. Siblings: 1 Parents: Separated Mother - 12th grade education Test Scores: None available Description of Participating Professionals Counselor The counselor in this study was the existing coun— selor at Reo Elementary School, Elizabeth Bridge. She was also the experimenter in this study. It was felt that because she had access to the school's teacher and student population, this setting would provide a productive research climate. In this school the counselor could not only readily initiate a treatment program, but she was also able to use her consulting abilities as well. Her job was gply_that of a consultant with the teachers. She provided continuous planning and dis— cussion possibilities for the teachers. The execution of the treatment and all of the work done directly with 48 the students was accomplished by the teachers themselves. The counselor was a consultant only. Teachers The teachers for this study were chosen from the staff at Reo Elementary School. Four different teachers in all were selected, one from each of the sixth, fifth, second, and first grades. Each teacher contributed a list of potential nonattending students from which the subjects for this research were chosen. Ultimately each grade level contributed 2-3 students out of the 11 used in this study. Each of the teachers involved was selected on the basis of grade level, willingness to assist in the research, and the number of potential subjects in his classroom. Observer The observer in this was a male master's level student. He was selected on the basis of his background training in student personnel work and his availability for this long—term research project. He observed each of the 11 students in their classroom setting for 20 minutes each day, using the Behavior Analysis Form. The observations lasted from the first day of phase B to the last day of phase F. l 49 Description of Criteria Measures Four different measures were utilized to assess an individual student's behavior in the classroom. Each measure was aimed at evaluating a different facet of a child's classroom attending behavior. 1. Behavior Analysis Form A behavioral observation form was used in this study as a means of measuring concrete, observable attending behavior as it was exhibited by students on a day-to-day basis. The form can be basically broken into two sections: attending and nonattending behavior cate— gories. Four behaviors are listed as characteristic of attending: in—seat, assigned task, raising hand, and answering. Eight behaviors are listed under nonattending: talking, verbal noise, physical noise, staring into space, physically disturbing a peer, out of seat, squirm- ing, and handling objects. In using this instrument the observer, seated in the classroom, is asked to observe which behavior occurred for the longest period of time over a lS—second interval. In the next 15 seconds the observer recorded on the form which behavior he felt was exhibited. For the third 15 seconds the observer watched the students classroom behavior again. In the final 15 seconds the observer again decided which behavior occurred for the longest period of time and he then marked it on the 50 form. The observation during a onevminute period ran: 15 seconds observe, 15 seconds rate, 15 seconds observe, 15 seconds rate. Each student was observed in this manner for a total of 20 minutes each day for the entire duration of the study. In a given week, then, a student could have been observed a total of 1 hour and 40 minutes. Barring absences, the observation of a given child in the study totaled approximately 24 hours. 2. Academic Achievement Scale An assumption of learning is that a student's academic achievement will increase if he is provided an optimal learning environment and he can be taught to attend to the classroom task (Meacham & Weisen, 1969; Brooks & Snow, 1972; Hops, 1971; Masek, 1970; McKenzie, 1968). As a means of assessing the academic achievement of the subjects, an Academic Achievement Scale was used. This rating form was an attempt to rank a given indi— vidual's current grade or achievement level in all of his subjects. The rating form was set up on a 7—point scale. Each increment of the scale accounted for a higher letter grade. The possible scores were: F, D, D+: C-, C, C+: B—, B, A. To aid the teacher in making this grade judgment a descriptive phrase accompanied many of the letter grades. These were: F: little or 51 no achievement, D: minimal achievement, C: average achievement, B: above average achievement, and A: outstanding achievement. The teachers rated each of the children on this scale for every subject studied during the school day. The teachers attempted to approxi— mate each student's current academic achievement and motivation on the next to last day of each research phase. 3. Assignment Completion Scale A third measure of a student's attending behavior was the Assignment Completion Scale. The rationale behind such a scale was that, if in fact a student was attending to the academic tasks of the classroom, then the student's percentage of completed assignments should increase. The Assignment Completion Scale was devised to measure this behavior. The scale was a lO—point rating form. For each of the classroom subjects taken by a student, a teacher was to approxi~ mate the percentage of completed assignments each student finished. The percentages ran from O'to 100%. This rating form was also completed by the teachers on the I next to last day of each research phase. 4. Individual Student Evaluation Scale A fourth measure of a student's progress in attending was the Individual Student Evaluation Scale. 52 This was a 7—point scale on which a teacher rated each student's classroom behavior. On one end of each scale was a behavior category consistent with attending. On the other end was a behavior category which was incom— patible with attending. For example: industrious wastes time, or completes work——————rare1y completes work, etc. At the bottom of this rating form the teacher was asked to make an overall judgment about the student's classroom behavior. The poles of the scale were: behavior improving——————no improvement or disruptive behavior. This was also a 7—point scale. This rating form along with three of the four others mentioned was completed by the teacher on the next to last day of each research phase. Two different criteria measures were used to assess teachers in this study: 1. Teacher Research Evaluation Scale The teacher research evaluation scale was a means of assessing each participating teacher's feelings about the research and its effect on her students. This was another 7—point scale on which a teacher was to rank her feelings about a given stimulus phrase as it related to the study. Examples included: too time consuming worth the time commitment, or the results of this study 53 are obvious——————the results of this study are not obvious to me, etc. This rating form was an attempt to make more concrete a teacher's subjective feelings about the study, her commitment, and the on—going results. At the bottom of this form, space was provided for comments concerning any aspects of the study. 2. R/N Rating As a means of assessing a teacher's ability to reinforce or not reinforce a student's classroom behavior, the R/N Rating was devised. While the observer was rating the student's attending behavior, he simultane— ously rated the teacher's reinforcement or lack of rein— forcement of that behavior. Therefore, each time the observer rated the student's behavior, instead of using an X or check mark, he inserted the letters R or N to designate whether that previous behavior had been rein— forced or not reinforced by the teacher. Since behavior tends to increase when reinforced and decrease when ignored, it was believed that, if teachers responded in some way to the instances in which a student attended, the student's attending behavior would increase. This rating was done without the teacher's knowledge. Reinforcement of this kind was not taught specifically in any in—service training session. 54 The rating was done to assess a teacher's style in pro— moting or disregarding behavior as they naturally would have throughout the duration of the study. Description of the Outcome Analysis The analysis of the data gathered using the Behavior Analysis Form was analyzed using White's (1972b) median statistics. This type of outcome analysis is specifically geared to the kind of information gathered in an intensive design. According to White (1972b) it is "Basically a method for the description of progress in a single case (i.e., how the performance of a single individual changes over time), and for the prediction of future levels of performance based on that technique." This type of analysis has been used by White (1972b) in the past in order to test the use of median statistics on human behavior. The majority of work done in using median statistics was to gather data over an extended period of time. This information was then charted and finally analyzed to find the median slope, As White (1972b) points out, the researcher then predicts on the basis of the data available whether or not the individual will reach the established behavioral criterion within an acceptable time limit. If the slope indicates that the subject will not meet the desired criteria, then this signals the researcher that the program or inter— vention warrants revision. If, however, the slope 55 indicates probable attainment, or even better actual attainment, this then demonstrates the capability of the particular treatment to produce desired changes. Pre— diction can also serve to alert the practitioner of pre— mature attainment of criteria level so that preparation for the next phase can be made. The specific steps taken in the data analysis include (White, 1972b): 1. Construction of a median slope. 2. Determination of the step (immediate change in performance)and progress change (slope, that is, the difference between the slope of one line of progress and another). 3. Statistically test the data. (a) Establish that the slope is not a flat one. This is done by using Fisher's Exact Proba— bility to determine if there is any sig— nificance. (b) Determine whether the overall probability of change could have occurred by chance. This 51159! is done by using a bi—nomial test. It deter— (W‘b) mines whether general differences between slope trends exist. (c) Determine if the steps alone between phases are significant. This is also done using a bi—nomial test. (d) Determine if the slope alone between phases is significant. This is accomplished by using a bi—nomial test. Both (c) and (d) are done to determine whether the difference observed is in the step or slope. To utilize this type of analysis a specific amount of behavior is needed before a highly accurate (96—99%) prediction can be made about the behavior (White, 1972b). "Probabilities don't begin to approach satisfactory levels until seven data points are available for prediction. Typical research standards are not 56 attained until 9—11 data points are available (depending on the criteria used)" (White, 1972). For this study a slight variation of White's median slope analysis was employed in a few of the 11 cases under study. Following regular median analysis, a correction factor described by White (1972) was used. This was done in instances where the observational data gathered showed deviations. For example, the data may show a fluctuation from one data point above the median to one point below on successive dates. Continued fluc— tuation of this type could produce a deviation that would prevent any meaningful interpretation of the data. To cancel out these alterations, White (1972b) suggests the correction factor which averages the successive data points. This type of overall data analysis plan described here, using median statistics, fits the intensive design approach far better than traditional inferential statistics based on the mean. As White (1972b) points out, when using the mean, The greater the difference between any single score and the rest of the data, the greater the singular influence that score will have in the determination of the mean. The mean in a sense emphasizes the importance of deviant scores. On the other hand, the median is not effected by largely deviating scores, only by the fact that they are greater or less than the other scores in the distribution. As a second point, demonstrating the usefulness of the median, White (1972b) states that: 57 Using the median, we may be assured of meeting our expectation 50% of the time; using the mean on the other hand, we might base our estimates heavily on his (the subjects) best or worst days——resu1ting in either an over or under estimate of the majority of his performances. It seemed appropriate to use median statistics for the analysis of the data gathered in this study. The information in this research fits White's description of data which could be analyzed using median statistics. The resultant information should provide the researcher with descriptive as well as practical results. Level of Significance The level of significance achieved in an intensive design attempts to combine statistical significance with supplemental meaningful data concerning the individual. Typically in a comparative group design inferential sta— tistics are used to evaluate whether the average per— formance of one group differs significantly from that of comparison groups. Seldom is the individual's performance observed. When the performance of one person is the experimental, then a different set of criteria from that used in traditional group designs is needed. As Chassan (1967) points out in discussing inferential statistics, Even if such comparisons are statistically sig— nificant, the observed mean difference between groups still does not clearly demonstrate that the difference cannot be the result of a few who changed dramatically while most changed little. The effect of the particular treatment on the individual remains obscure. 58 In the intensive design significance and resultant generalization is focused on in a unique way. Control for generalization is not produced through sta— tistical maneuvers, but rather through a design which capitalizes on the variability. As Shontz (1965) states, The composite data of individuals in a group is con— sidered to represent one ideal subject, namely the overall performance of an ideal average individual. Yet there may be in fact no one who is average in the sense of having all extraneous variables elimi— nated. There may be, however, an individual who is "representative" of many others who can be studied individually. The intensive design, then, views individual variability not as a chance event, but rather as lawful and subject to the control of the environment. It focuses primarily on the individual's variability by using experimental control to examine the variability and altering certain conditions to see if change occurs (Thoresen, 1972). In this approach then: Generalization is based on the range of variations required to prove significant changes with each subject. Generalization is gradually established through systematic replication of results with the same subject (intraindividual) or across subjects (interindividual). Replication involves not only the individual, but situations as well as time (Thoresen, 1972). To maximize the usefulness of the data compiled in this study, both meaningful and statistical levels of significance were set. The statistical criteria are based on White's (1972b) work. Using his approach a 59 .05 level of significance is commonly accepted in research. These criteria were used to accept or reject the hypotheses dealing with individual change. The group trend analysis employed the same sta— tistical procedure and level of significance outlined for individual change. This was accomplished by combining the summary statistics for a group within a phase to produce a single line of progress. These lines of pro— gress were then tested for statistically significant differences between groups. In an absolute sense the practitioner could View any increase in attending behavior as significant. The actual increase might be large or extremely small. In order to sort out those changes which were miniScule from those that were meaningful, a criterion of a 20%- increase in attending behavior over that obtained as the median baseline behavior (phase B1) was arbitrarily set. That is, for example, a student whose baseline behavior had a median of 30 would need to show an increase of 20% and achieve a percentage of 36 or more for the increase to be considered meaningful. Likewise, a student with a median baseline of 50 would need to exceed 60 during the treatment phase to achieve a meaningful increase. The other research data analyzed in this study were evaluated separately. The information accumulated on the various 7—point scales was treated as follows: 6O 1. The Teacher Research Evaluation Scale, The Academic Achievement Scale, The Assignment Com— pletion Scale, and The Individual Student Evalu— ation Scale were all analyzed in the same manner. The data were totaled for every student on each of the 7—point alternatives. This information was used to note changes in response patterns of teachers over time. On a group basis the data for each 7—point alternative were totaled across students and converted into a single scale. The information was then compiled and compared for all individuals between the various phases. 2. The R/N ratings were tallied for each of the four teachers at the end of each phase of the research. This information provided some data about the amount of time teachers spent rein— forcing or ignoring attending behavior in their students. Summary This study was conducted in order to measure the change in attending behaviors exhibited by elementary age children through the application of behavioral tech— niques utilizing an intensive design approach to research. The research followed the A, B, A, C, A, F design outlined by Thoresen (1972). Two basic treatments were employed, behavioral contracts and token economies. 61 Following each treatment phase, the teachers returned to baseline conditions with their students. Observations were gathered on the students for 20 minutes each day over the course of the entire study. In addition, on a once—per—phase, data were gathered concerning each student's academic achievement, assign— ments completed, and overall classroom behavior. Teachers on a once—per—phase basis responded to their feelings about the study and its contribution to their students. Teachers were also watched by the observer to determine the number of times they reinforced or ignored the attending of their students. The data in this study were analyzed, in large part, using median statistics as described by White (1972b). Both group and individual analyses were made. CHAPTER III RESULTS Analysis of the Outcome Data This chapter presents an analysis of the data secured in this study. Due to the unique nature of the intensive design, two different types of hypotheses were examined. Six hypotheses were aimed at assessing changes in attending behavior as it occurred in each individual's case. Two additional hypotheses were focused on any group trends which may have occurred when individuals within a specific treatment type were combined. The analysis then attempts not only to assess what changes took place on an individual basis, but to assess group trends as well. Typically, research has been evaluated on the ~basis of statistical criteria. That is, change was said to have occurred if the differences which took place were statistically significant at a pre—determined numerical level. The intensive design approach to research recognizes the validity of such statistical criteria. The N = 1 design, however, supplements 62 63 statistical criteria with a meaningful standard as well. Thus, the intensive design attempts to satisfy the exper— imental researcher's standards for assessing change by acknowledging statistical levels of significance in reporting results. Additionally, however, the intensive design looks for meaningful results which are of practical significance for the counselor or practitioner in the field. Both types of assessment focus on the same ends, but taken separately they allow the researcher to approach two different aspects of the study's results. For the eight hypotheses formulated for this study, the .05 statistical level of significance was chosen. This conforms to typically accepted research standards (White, 1972b). From a meaningful standpoint, a criteria of a 20% improvement in the median percentage of attending behavior obtained by each subject compared with that score obtained by each individual for the median percentage baseline behavior observed in Bl was used. These two different criteria allow the experi— menter to assess the data using traditional research standards, as well as allowing an assessment which would satisfy questions dealing with meaningful dif— ferences found in the same set of data. The data collected for this analysis were obtained on two distinct time tables. The first 125-3 4w. :2. 3%“. ~--—- 64 collection schedule was that used for the observational data. Daily observations were collected for this rating throughout the course of the study. The second collection schedule was that used to gather info ation from teachers about their students on a series of Heti\g forms. This information was gathered on a once—periph‘ The analysis itself was accomplisheq_u51ng three different procedures. The available data wege analyzed in part using White's (1972b) median statistics. A. binomial test (Siegel, 1956) was also used in some instances to differentiate actual from chance occurrences in the outcomes. Teacher ratings were summarized and reported using mean scores across phases. General Findings For the purposes of analysis, each of the research hypotheses will be restated along with the accompanying outcome data to support or reject it. Hypgtheses 1 and 2 Hypothesis 1: During the first treatment (T1 or C1) attending behavior will increase over the baseline observation level (B1) as indicated by the median slope trend. Hypothesis 2: During the second treatment (C2 or T2) attending behavior will increase over the baseline observation level (Bl) as indicated by the median slope trend. “‘4‘. 65 Table 2 Statistical Significance——Probabilities of Differences in Attending Behavior Between First Baseline (B1) and Treatment Phases (Tl/C2 or Cl/Tz), Last Baseline (B3) and Follow—Up (F) T—C Subjects 1-2 1—4 1—5 1—6 Bl—Tl B -c2 31-33, B -F Michelle .0026 .0046' .000a .004a Debby .002a .0026 .0006 .0046 Chris .002a .0026 .0006 .0026 Jeremy .0026 .0316 .0006 .0316 Becky .0046 .0446 .154 .246 Lisa .0026 .0046 .0006 .0046 C-T Subjects Bl-Cl Bl-T2 Bl-B3 Bl-F John .002a .002a .000a .001a Karen .0186 .0016 .000a .001a Todd .0086 .0016 .0156 .001a Jeff .0016 .016 .0006 .0026 Eric .0186 .016 .0006 .0026 11 of 11 11 of 11 10 of 11 10 of 11 p < .0005a p < .0005a p < .005a p < .0056 aDenotes statistically significant difference (p <'.05) between phases. 66 As shown in the first column of Table 2, the experimenter found statistically significant (p < .05) difference between the baseline (Bl) phase and the two treatments offered during the first phase (T1 or Cl). As demonstrated by letter "a" in each individual sub— ject's case, attending behavior during the first treat— ment phase was significantly higher than it was during the baseline (Bl) phase. Hypothesis 1 was, therefore, accepted for all subjects. Similarly, the data presented in the second column of Table 2 are a comparison of attending behavior during the baseline phase and the second treatment (C2 and T2). Each individual subject increased significantly (p < .05) in attending behavior'and Hypothesis 2, there— fore, was also accepted for all subjects. The analysis in Table 2 was accomplished using White's (1972b) median statistics. The scores listed in this table represent the probability of a chance occurrence of statistical significance between base— line (B1) and all other phases (except BZ)° In addition to the data supporting Hypotheses l and 2, it is interesting to note thecomparison between phases B1 and B3, as shown in Table 2. The data represent the difference in attending behavior scores Obtained prior to any treatment administration and those scores obtained following the completion of both treatment phases. 67 With the exception of one subject, all students showed a statistically significant (p mne we» Eonm mafia pawns sconce concocono men can mafia econ» smocme m.mmm£m m cmmspmn Amo. v av oocmuommop unmoHMHcmHm maamowumoumum m monocmam oom. ioo moo. loo mao. loo mooo. Ame moao. m onus loo moao. loo woo. Loo oHo. loo mooo. Ame oaoo. o memo loo emoo. ioo moo. ioo oaoo. ooo. Ame mooo. o oooe mom. loo ooo. mom. oom. “so mono. o cones loo mooo. Ame ooo. so. see. loo omoo. m eeoo mnmm mmums msnmm mmuao Hclam muomnnom 910‘ ooa. mmo. Ame mamo. loo mooo. Ame omoo. H amen oom. man. so. Loo mooo. loo oooo. a sxoom loo oaoo. oHH. ooa. Loo maoo. loo mooo. a samumo oom. Ame ammo. msm. so. loo omoo. m mango Ame wooo. loo omoo. so. sHH. loo emoo. m shame mam. Hos. loo mooo. loo Mono. loo emoo. m managed: mumm mmumo Noumm mmuae Assam oum muo oum mum NIH wooed mucosnom one mousse pomflasm 30mm How mommem m>ousommcoo cocspmm Hoo>mcmm mcflcccuue so neoconmwmoo mo mmfluoaoncnoum m wanes 69 differences. Of these one demonstrated agreement with the hypothesis, and one contradicted the hypothesis showing an increase in attending behavior in the B3 phase. For the T2 and 33 comparisons there were five out of five statistical differences. Of these four were in agreement with the hypothesis that attending behavior would decrease in the B3 phase, while only one contra— dicted it. In summary, then, Hypothesis 3 was accepted for five out of the six statistically significant out— comes, while Hypothesis 4 was accepted for five out of the seven statistically significant outcomes demonstrated in Table 3. The individual scores for Table 3 were computed by using White's (1972b) test of significance for overall differences between two consecutive phases. Specifically, the test was a binomial test discussed by Siegel (1956) used in determining overall statistical differences. Having set the alpha level at .05, the data were then analyzed. When significance was found, it meant that the actual trend line data, found by extending a known trend line into the next consecutive phase, did not con- form to the predicted trend line. That is, the data points did not fall in equal numbers above and below the predicted trend line. Significance for the data in Table 3 is designated by the letter "a." In addition, listed on the table are the probabilities 70 of chance occurrences of differences between trend lines for two consecutive phases. As a means of describing this chance occurrence the table contains: (a) the probability of statistical differences occurring, (b) the letter "a" denoting a probability less than .05, and (c) a number in parentheses indicating, for those probabilities less than .05, the particular phase in which attending behavior was more frequent. Hypothesis 5 Hypothesis 5: During the F phase (follow—up) attending behavior will be greater than during the B1 phase, but less than during either of the treatment phases (Tl, C1 or C2, T2). As shown in Table 2 there was a statistical sig- nificance between the B1 and F phases in all but one instance. Therefore the first part of the hypothesis, stating that the level of attending for the F phase will be greater than that obtained in B is confirmed for 1' 10 out of 11 subjects. Table 4 contains the data needed to evaluate the second part of Hypothesis 5. According to the table, for comparisons between T1 and F there were three sta— tistical differences. Of these all three support the hypotheses that the T1 attending levels will be higher than the F levels. For the Cl and F comparisons, five 71 Table 4 Probabilities of Differences in Attending Behavior Between First Treatment Phase (T1 or C1) and Follow—up, and Between Second Treatment (C2 or T2) and Follow—up Phases T-C Subjects Grade 2-6 4—6 Tl-F C2-F Michelle 5 .2187 .2734 Debby 5 .03126 (T1) .0039a (c2) Chris 5 .00096 (T1) .0703 Jeremy 1 .0097a (T1) .2050 Becky l .2460 .1171 Lisa 1 .2187 .2187 C—T Subjects 2-6 4—6 C1-F Tz-F John 2 .00096 (F) .00976 (F) Karen 2 .0439a (F) .2050 Todd 6 .0019a (F) .0703 Jeff 6 .0439a (C ) .0097a (T2) Eric 2 .0175a (PI .2460 Treatment 4 of 11 Treatment 2 of 11 Follow—up 4 of 11 Follow—up l of 11 ‘ aDenotes a statistically significant difference (p < .05) between phases. 72 out of five comparisons were significant. Of these one supported the hypotheses that the Cl attending level would be higher than the F phase levels. Four out of the five contradicted the hypotheses. For the C2 and F comparisons there was one statistical difference. It confirmed the hypotheses. For the T2 and F comparisons there were two significant differences. Of these one confirmed the hypothesis and one contradicted it. Over— all then the second part of Hypothesis 5 was confirmed by 6 of the 11 subjects who demonstrated significant differences. The hypothesis was contradicted by 5 of the 11 subjects. Again, the analysis for Tables 2 and 4 was based upon White's (1972b) median statistics. The scores listed on the tables are the probability of a chance occurrence of the statistical differences between any two phases which were compared. The letter "a" denotes a significant difference between phases. Hyppthesis 6 Hypothesis 6: No difference will be found in the attending behavior between the first treatment (T1 or C1) and the second treatment (C2 or T2), as indicated by the median slope trend. Hypothesis 6 asks whether or not there is an order effect between the two treatment sequences. That 73 is, was the first treatment position always better at increasing attending behavior than the second treatment position, and vice versa, regardless of technique used. Table 5 contains the data needed to test that question. Table 5 Test for Statistical Differences Between the Two Treatment Positions, Significance Being Designated as p < .05 As demonstrated in the above table, the infor— mation obtained from the study indicates that there was no statistical difference (p < .05) between the two treatment positions. On the basis of these data, the hypothesis was accepted, confirming the fact that the T1 or Cl treatment position was not superior to the C2 or T2 position and vice versa. Specifically in computing significant differences between the two treat— ments, eight were significant. Of these eight, two were in the first treatment position and six were in 74 the second treatment position. Although this may appear to be meaningfully significant, the statistical proba— bility was .11. Hypotheses 7 and 8 Hypothesis 7: No difference will be found among individuals receiv— ing the first combined treatment (Tl C2) and the second combined treatment (C1 T2). Hypothesis 8: No difference will be found between the group receiv— ing the first combined treatment (T1 C2) and the second combined treatment (C1 T2) Both hypotheses 7 and 8 are asking if, in fact, either treatment combination was superior to the other treatment combination. Hypothesis 7 asks the question in regard to individual subjects grouped according to treatment type. An inspection of the data in Table 6 reveals that there was a statistical difference between treatments T1 Cl and C2 T2 for eight subjects. The phase column on Table 6 demonstrates that for the T1 Cl group four subjects showed a significant difference (p < .05) between the two treatment techniques. For the C2 T2 group four subjects also showed a statistical significance between treatments. The treatment column on Table 6 further clarifies these differences. For the T1 Cl group, two subjects showed contracts to be the superior Probabilities of Statistical Differences in Attending 75 Table 6 Behavior Between Phases 2 and 4 (T1, C1 and C2, T2) for Each Subject Phases Phase T-C Subject Grade —————— 2-4 2 4 Michelle 5 .2734 Debby 5 .0175a ca Chris 5 .1641 Jeremy l .0039a Ta Becky l .0019a Ca Lisa 1 .0039a Ta C—T Subject Grade 2-4 2 4 John 2 .0019a Ta Karen 2 .0439a Ta Todd 6 .0009a Ta Jeff 6 .2050 Eric 2 .0439a Ta Totals 2a 6a aDenotes a statistically significant difference at the p < .05 level. difference. T or C letters denote which treat— ment was superior for those who demonstrated a statistical 76 treatment, and two subjects showed tokens to be superior. For the C2 T2 group all four of the statistically sig— nificant differences occurred in the token treatment. Overall, then, Hypothesis 7 was rejected for eight subjects, while it was accepted for three subjects. Hypothesis 8 takes this part of the analysis a step further. The question posed here is whether or not for those subjects who showed a statistical difference between treatments there would be any overall statistical group differences (p < .05) between the two treatments. The data to assess that question are presented in Table 7. The information provided confirms that from a statistical standpoint there was no significant difference between the two treatment types (p o mmmp .mnma .mcsb “Hoummaumm>cH .mmpflum .¢ numnmNflHm .COHum>Hmmno cm mamsvm Unfiom Hoocom mm mo Esawxma m pmum>oo mQOHum>nwmno mmmse 133 camp £06m .Nmomm mom mmmmO mOH>¢mmm UZHQZMEBsH .mmpHHm .¢ nqumNflHm Hoonom mm mo ESEHxME m pmum>oo mCOHum>Hoon wmmca .pOHHmQ nucoe v m Hm>o what .coapm>uomno so mamsqm pcaom mwmp £06m .wzmme mom fifidm AfiZOHB<>mmmmO mOH>HCD mumpm 5mmflcoflz mafia pcouu cmflpmz lllll h F 0H . wmp WMO mom H0H>mnmn 1 mafipcou mo mmmucmouom Qllllb fl I ON a 1 m I On u 3 P In a I.ov o C T: . V 1. cm In... : . a! W A I! p " " To u 8 a u. P I on m. o 1 I om I A 1. I.om ” . II ooa o Hm .UOHHmm nusoE v m Hm>o mmmp .cOHum>meno cm mamsvm ucflom .mnma .mcsb “HopmmHUmm>cH .mmpfium .d nmemNHHm Hoonom mm mo ESEmeE m pmum>oo mGOHum>Homno mmmza :135 dump zoom .mHmmU mom defia AdZOHB¢>mmmmO m0H>¢mmm QZHQZMBB¢II.M mmDOHm mama mm on mm om mm om mv ow mm om mm om ma OH m — L P h P I P p I I O muflmumswcb mpmum cmmH£0Hz m .H pawn» swapmz IIIII OH wow 0 How Hofl>mzmn acflpcmu mo wmmucooumm QIIII‘ I ON a I. m o om a I m. I e I 6 I . a I 0v 0 a J / I W I Ia om m. I u / x . “w I ” om m I II I w I I H / I on m I I 3a m. I II I I I I I I r I om I I ‘1 a II. \ L .1. 1.... II I \ \ \K I om / ll. all I, I \ a / A I I i d \ a / 7. (III ’ II p - I I 4 ooa It .I mm 0 mm .I. am 136 .mnma .mcnn «Hoummflumm>cH .mmpflum .d numnmNHHm .UOHHmm aucofi v.m nm>o mwmp Hoonom mm mo Edsfixme m pmum>oo mGOHum>Hmmno muons .GOHum>Hmmno cm mamsvm.ucH0d mgmp zoom .mqqmmUHz mom «BIND .HINZOHBIQEmmmO MOH>§mm UzHszBB¢II.v MMDUHm mmmo mh on mm om mm om mv ow mm om mm om ma 0H m o — n h p P I P h P PI 0 mpflmum>flco mumbm cmmanowz on: @293 83me still . I OH amp one now Hougmzon I mcflpcmufim mo mwmucmonmm I I). II on I d I I m I I D I I .uomm s z 1. I I , w I : I S. a I m. I x I \ w I 1.0mm. I I \I u P s I. \ m. x I. \ II. 00 ID I x a I a I \ I... I \ m I _ .\ II ..\ . I om I \ II T om I x If I x. ._ / II - (on - ooa 137 .mnma .mcso “Hoummwbmm>aH .mmpfinm .< aumnmNHHm .UOHHom sucoa v m Hw>o mmmp Hoonom ms mo Edeflxma m poum>oo mGOAum>Hmmno wmmna .:0flum>ummno cm mamsvm Ucflom mumc comm .OHmm mom dado AdonB¢>mmmmo moH>mmmm ozHozmaB¢ll.m mmDon mama mm on mm om mm om mv ov 7 mm om mm on ma 0H m o I I _ I F I L I I I 0 man u 9.3 c332 I.II..I mafimumsflao mumum cwmflnoflz I.oa amp mso Ham mow>mnon msflpcwuum mo mmmucmouwmIQIIIlb om I c> Ln JorAeqea Burpusnnv JO abequeoxad filom \. .1 4 (4‘ ll ‘ OOH 138 .mnma ~masb “HoummfiummbcH ~mmwflnm .4 npmnmufiam .coaumm space v m Hm>o mmmc Hoonom mu mo Edfiflxmfi m Umum>oo meowum>nmmno mmmaa .GOHum>Hmmno cm mawdvm #Gwom mumv Somm .mmmb mom ¢BmmmmO mOH>¢mmm GZHQZHBBmnmn mcflvcmu mo mmmuamoumm 9|..llo IL 1 I c> In xogAeqag burpuanqv go abenueoxad <14 OOH 139 .mhma .masn “Hoummflumo>qH .wmoflum .m nquMNHHm .Uownmm Space v m Hm>o mwmw Hoonom mu mo EdEmeE m ©mum>oo mcowum>nmmno wmmza .cowum>nmmno cm mawavm unwom M#MU 50mm .0909 mom «Ema qmonB¢>mmmmo moH>mmmm wzHaszB mmeHm, mama . mm on mo om mm om mv 0v mm on mm om ma 0H m o — P b b h n n n b P o wpwmum>flaa mumum :mmflnowz mafia ucruu :maomz lllll 1 ca mac mac How How>mnmn maacqmuum mo wmmuamuumm Ollllb . 1.0m a \ \ I m .\ omnu q \ 2 x w v r Ofio 1.. 3 W : p T.- u l.owA. 8 a u. D. 1.05uw o 1 II .1. I om ..I \ L \ . \ .I om WI I ll 4 4’4 .1544 4 4 4 00.... on m mm a «m 9 am 140 .mhma sma5b “wepmmflumw>cH .mmownm .« numnmuflam .UOHHmm syncs v m Hw>o mama Hoonom mm mo Edfiflxmfi m wwnm>oo mGOHum>Hmmno mmmsa .cowpm>ummn0 cm mamsvm unwom mumv nomm .zmmmx mom «Baa AdonB¢>mmmmo moH>mmmm ozHszBBHGD mgmwm cmmflnoflz 70a mac man How How>mzmn maflvcwuum mo mmmucwoummollllb o I. m d a m rIOm.u u 3 e . a A 10¢ o 1: w . 10m 1. .v v \ w x m. \ v 1006 .\ H x: W. A . . .\ Ion. m T... o \ . x . \ \\ \ I. om \ . \ a A \ n“ \ \ ; . \ r.om \ . xx . . 4 x _ 4 03 a; m mm .H. mm o Hm 141 .mnma ~mcsn «Hoummflumw>cH .mmcflnm .m aumnmNHHm .UOHHmm nunofi v m um>o mmmc Hoonom mm mo Edswxms m Umnm>oo macaum>Hmmno mamas .cowum>ummno cm mawsvw ucflom mumu 50mm .zmon mom «Ema A¢onsm>mmmmo moH>¢mmm GZHQZmBB¢II.m mmstm mama mm on mm om mm om mv ow mm om mm om ma 0H m a — P — n b p . L b p P O muflmumbflcs mumum cmmHAOHz $5.." :93 cm.“ 02 lilll H 0 U r: 0H haw gqo How How>msmn mcflcqmuym mo ommucmoumm olllllo on d e I a a u 1 e ,b a c m. V H4- n4- ; m ¥ n. T.- u .b m I/ W. I / [1‘ J I A .1 4.1 II m. . : I . .I # I all I .I- I. I II. I nllu l \ v v/ I \\ I/ \ . . A4 4 4 44 00H 142 .mnma .GGSb “HoummHUmm>GH .mmvfium .¢ numnmNHHm .COflme suGOE v m Hw>o mama Hoozom mp mo ESEmeE m kum>oo mGOHHM>HmeO mmmne .cowum>ummno cm mandwm pawom MUMG 50mm .flmHA mom 4840 Qmmmm0 MOH>Emmm 02HQZMEB¢II.OH MMDOHW mama mm on mo om mm om mv ow mm on mm om ma OH m o h _ P P F p b b p p O mi: 98.3 nmflcmz lllll mnzmumtfica wwwgm GmmH£OHE OH mac mwo How How>mnmn mcaucmuufi mo mmmunmouom I om 3 .d a m r.om a u 1 D. ,n a o I. _. .fi w : 1. m n. T. u .n . 8 a u. D. A . v m. 1 I I I. i .\ //// II. I l// . 71/ I :\.\ VI . v A ooa Cu m mm o «m B 4 Hm .mnma .maSH “Hoummfipmm>cH .mmcwum .d numnmNHHm .mowumm Space v m Hm>o wand Hoonom mu mo ESEmeE m wmnm>oo mc0fl9m>uwmno mmmna .GOHum>ummno Gm mamnvm pcflom Murmv flOMW .Nmmma MO...“ amp—“4n.— .HdZOHBflxwmmmmO mOH>§mm UZHn—ZWBB/wllofld EDUHM 143 mama on we ow mm om mv 0v mm om mm om ma 0H m o — h — b F b k b p p O $8.3” Gamma 53."an l4444 mpflmumxfiwcb mpmum cmmHSOHZ 0H. Mme m 0 you HOfibmnon maflwamuu mo mmmuamonmmolullb om d e 1 rom m m / / .. w z 43. . , 0 z I. ,4 w I rIOm m. / m. r m. .7 m. x e / ION. A T. o O , 1 4 4 444 4 4 row . fi4 Iom 4 -4 4 I 4 ooa :5 m mm o Hm APPENDIX B TABLES APPENDIX B Table B-1 Median Percentage of Attending Behavior for Each Subject Across Each Phase of the Study Phases T-C Subjects 1 2 3 4 5 6 Bl T1 B2 C2 B3 Michelle 50. 52 72 95 81.25 93.75 Debby 53.75 90 95 100 80 63.75 Chris 72.5 97.5 92.5 95 90 82.5 Jeremy 57. 82.5 82 91 75 82.5 Becky 41.25 63.75 61.25 77 55 43.75 Lisa 55 79 81 86 82.5 82.5 C-T Subjects B1 C1 B2 T2 B3 F John 52 75 71.25 87 87.5 92.5 Karen 46 67.5 62.5 83.5 67.5 86.25 Todd 33.5 100 98.75 100 85 100 Jeff 50 91 85 97.5 71.25 87.5 Eric 66 83 88.5 95 80 95 144 145' Table B—2 Summary of Mean Percentages for Each Subject Across Phases for the Assignment Completion Scale and Comparisons Between Consecutive Phasesa Phases 1 2 3 4 5 6 1— Michelle 0 80 + 12 — 40 + 26 - 82 + + Debby 80 74 — 100 + 100 O 100 0 100 0 + Chris 58 94 + 72 — 86 86 O + Jeremy 72 73 + 83 + 100 + 100 0 90 - + Becky 38 33 - 23 - 50 + 67 + 90 + + Lisa 82 63 — 67 + 97 + 80 - 47 - - John 73 72 - 88 + 94 + 78 - 98 + + Karen 60 78 + 86 + 92 + 74 — 84 + + Todd 26 84 + 90 + 90 0 96 + 90 — + Jeff 92 82 — 95 + 100 + 100 0 84 — + Eric 85 62 — 92 + 94 + 82 — 96 + - Totals + 5 8 8 2 5 9 - 6 3 O 5 4 2 0 O 0 2 3 2 O aA plus (+) indicates an increased rating for a phase compared to the previous phase, a minus (-) indi- cates a decreased percentage mean rating, and a zero (0) indicates no difference between phases. 146 Table B—3 Summary of Means for Each Subject Across Phases for the Individual Student Evaluation Scale and Comparison Between Consecutive Phasesa Phases 1 2 3 4 5 6 — Michelle 2.00 4.40 + 3.00 - 3.30 0 3.10 5.40 + + Debby 1.90 3.90 + 2.50 - 3.90 + 3.90 3.40 — + Chris 3.60 5.40 + 5.10 — 4.90 3.75 — + Jeremy 3.25 4.90 + 4.90 0 5.50 + 4.60 5.25 + + Becky 2.50 3.60 + 1.70 — 2.25 + 2.60 2.50 - + Lisa 2.50 3.40 + 3.00 — 3.75 + 4.25 2.90 - + John 4.40 5.50 + 4.50 - 4.75 + 4.40 6.40 + 0 Karen 3.50 4.25 + 4.00 ~ 5.10 + 3.60 4.10 + + Todd 2.00 4.75 + 6.00 + 5.40 — 4.90 5.60 + + Jeff 3.90 6.10 + 7.00 + 6.00 - 7.00 4.10 — + Eric 4.60 4.10 — 4.50 + 4.10 — 4.00 5.00 + - ‘Totals + 10 3 6 4 6 9 - 1 7 3 5 5 1 0 0 l l l O 1 aA plus (+) indicates an increased rating for a phase compared to the previous phase, a minus (-) indi- cates a decreased mean rating, and a zero (0) indicates no difference between phases. 147 Table B—4 Summary of Means for Each Subject Across Phases for the Academic Achievement Scale (Achievement) and Comparison Between Consecutive Phasesa Phases 1 2 3 4 5 6 Michelle 3.00 2.00 — 1.40 — 2.40 + 2.20 - 3.40 + Debby 2.60 2.80 + 1.60 - 1.90 + 1.80 - 2.00 + — Chris 4.20 3.80 - 3.60 — 3.00 4.40 + — Jeremy 5.00 4.75 - 4.50 - 5.00 + 5.00 0 6.00 + 0 Becky 3.20 3.75 + 2.00 ~ 3.25 + 3.00 - 4.00 + 0 Lisa 4.70 5.25 + 4.00 — 4.75 + 4.00 - 4.33 + 0 John 4.40 4.60 + 5.8 + 5.60 - 5.80 + 6.00 + + Karen 4.60 4.60 0 5.8 + 6.40 + 5.80 - 5.60 — + Todd 3.60 5.60 + 5.8 + 5.60 - 6.40 + 4.50 + Jeff 5.60 5.60 0 6.4 + 6.40 0 7.00 + 6.20 + Eric 5.40 5.00 ' 6.2 + 5.80 - 5.80 0 5.80 0 + Totals + 5 5 6 3 7 5 - 4 6 3 5 3 3 0 2 0 l 2 1 3 aA plus (+) indicates an increased rating for a phase compared to the previous phase, a minus (—) indi- cates a decreased mean rating, and a zero (0) indicates no difference between phases. 148 Table B—5 Summary of Means for Each Student Across Phases for the Academic Achievement Scale (Motivation) and Comparison Between Consecutive Phasesa Phases 1 2 3 4 5 6 1-5 Michelle 1.00 2.00 + 1.20 - 2.80 + 2.00 — 2.60 + + Debby 2.80 3.60 + 2.20 - 7.00 + 1.80 - 2.00 + — Chris 4.40 3.20 - 3.20 0 2.40 4.20 + — Jeremy 3.00 4.25 + 3.75 — 4.00 + 3.75 — 5.33 + + Becky 2.33 2.75 + 1.25 — 2.75 + 3.00 + 3.00 0 + Lisa 2.50 5.00 + 2.75 — 5.50 + 3.25 - 4.00 + + John 5.80 5.20 — 6.00 + 6.00 0 5.60 - 6.00 + — Karen 4.80 4.20 - 6.40 + 6.40 0 5.80 - 5.80 0 - Todd 2.80 5.60 + 5.80 + 5.60 - 6.40 + 4.50 - + Jeff 5.00 5.60 + 6.40 + 6.40 0 7.00 + 6.20 - + Eric 6.40 4.80 - 5.20 + 5.40 + 6.20 + 6.20 O - Totals + 7 5 6 4 6 6 — 4 5 l 6 2 5 O O l 3 0 3 O aA plus (+) indicates an increased rating for a phase compared to the previous phase, a minus (—) indi- cates a decreased mean rating, and a zero (0) indicates no difference between phases. unwEwouowckucoz n mz “unmewouOMGHGm n m umcficcmuumqoz u z nonficcwuud n d 149 me m HeN OH OHN m OOO Hm om N HON eH HO O NmN mH mm m eVN mH mvH e veH NH m Hmuoe m4 H HHN NN mON e ONO mm cm H mHN 4N mm N ONN NN Ne m NvN ON NmH O ONH NH m an H eNN mm OHH e OHO Nm mv N OHM em NO N «Om HH mu m eON OH ONH v eNH m 0Hum mo N meN eH emN mH mmm Nm Hm H NNN mN we 4 me ON em v HON NN va ON mm MN mmmn N O mmm m mmH OH mem em v o oem mN OH H umm me HN H veH em OHN HH HO OH woos Ne O HHm eH mNm e mom HN OOH O NON eH HmH N NNN mH mHH N mNN ON OOH v em H coumx mm O Hmm on mmH v eme OH no N meN mH mmH m eON HH OOH H OVN e OmH H mOH OH snow muomensm 940 HOH m emN eH eHN m mom eN ev N NeN NH ee m veN OH mm m HmN OH OOH O mOH NH m mm e OmN NH eON m mam mm mm v veN NH mm m HNN eN mOH m NvN OH NvH eH meH HN mmHH meH HH HOH mm Omm OH mom em Om N mmN ON NNH e mHN m mNH N OHN mH mmH O ONH NH exoom me N eOm OH «ON OH mom ON ow O «4N ON mm vH emN ON mm e va ON eNH m NNH HH esmuon mm N NmN eH OOH N vmm m Nm O mmH o «v o mHm O on o mmN m He O mmH HH mHHso mHH m OOH NH NvH v mHO mm ON O eem m av O omm H mm O OON m HmH O HON O magma mm N ONN mH OOH H «Hm eH mm o eHm m VHH O HON e mVH O mHN N mOH H NVH HH 0HH020H2 muomnnsm U49 mz m «2 m «2 m m2 m mz m mz m mz m mz m mz m mz m mz m mz m .wcwoua02\.wcwom z 4 z < z a z a z a z a .006:02\.0u< O m v m N H mommnm modum ms» mo mwmnm comm uOM muwnommfi an cmvfl>oum HOH>mnom mcflccmuumcoz cam mGHccmuud w.ucmcsum a mo mwuoom usoEmUMOMCAGHcoz Ho ucmfimouOMCHom c002 uo mumfifidm wlm THQMB APPENDIX C FORMS USED IN TREATMENT Elizabeth Bridge Michigan State University January 1973 AGHIENDIDK<3 ASSIGNMENT COMPLETION SCALE TEACHERS NAME STUDENTS NAME DATE DIRECTIONS: PLEASE CIRCLE THE PERCENTAGE NUMBER WICH APPROXIMATES THE AVERAGE NUMBER OF COMPLETED ASSIGNMENTS THIS STUDENT TURNS IN FOR EACH SUBJECT. Percentages HEALTH 4 1.0 7.0 40 10510.40 7.0 0 910 400 r I F I I I I I I I I LANGUAGE ARTS L. 1 1 1 .1 1_ 1, 1 1. 1 ._1 I I I If I I T' I I I 1_1 MATH 11 1 1 1 4 1 1 4 1 1 4 I I I I I 4 T 4 1 1 SCIENCE 1 1 4 1 1 4 1 l .11 1 l I I j I I I I T f 1 1 . SOCIAL STUDIES 11. 1_. 1 1 4 4 -11. 1. 1 l __1 I I I I If I I I I I II MUSIC ART 1 1 4 1. 4 4 1 1. 1 1 111 P.E. T I I I I I I I I FI—fi 150 Elizabeth Bridge Michigan State University January 1973 151 ACADEMIC ACHIEVEMENT SCALE TEACHERS NAME STUDENTS NAME DATE DIRECTIONS: USE THE TOP LINE FOR ACHIEVEMENT USE THE BOTTOM LINE FOR MOTIVATION. PLEASE CIRCLE THE LINE ON THE SCALES WHICH YOU FEEL MOST CLOSELY CHARACTERIZES THIS STUDENTS ACHIEVEMENT AND MOTIVATION AT THIS TIME. O C 4U «U U U 00-“ C C: I: 4. 5 5 5 0 H 0) 0 5 5 E E 'U E 0 H :> 30 3 3’0 3 5 g 49-! 5 a Q) (U Q} 0) CO (I) U (I) .. E“ SE :42 32 J: '5 ~44 '5 > O .0 > O D 0 HEALTH '4 “’ 2‘ ‘I’ <1 <4! <: co «4 O m I I I TI I T’ i f I I ‘I 4 4 1' PPROXIMATE C C B B A RADE EQUIVALENTS F D m-C' +° ' LANGUAGE ‘ 1 . 1 L 1 .4 AR I l 7 I I I '1 T3 4 + 4e 4 4 J. 4 I‘ll-4A1"H ._ 1 1 1 1 1 J I I I I I I 'I 4 TE 4 4 4 1r 4. SCIENCE . , . 1 . 1 J I j I I l '17 I 4 a 4 1r 4 .L 4 STUDIES : 4 4 .r .r 4— A ART 4 + 4 4 4 1 4. ) ‘ I.h. Elizabeth Bridge Michigan State University January 1973 152 INDIVIDUAL STUDENT EVALUATION SCALE TEACHERS NAME STUDENTS NAME DATE DIRECTIONS: PLEASE CIRCLE THE NUMBER ON EACH SCALE WHICH YOU FEEL INDEPENDENT 1_ MOST CLOSELY CHARACTERIZES THIS STUDENT AT THIS TIME. NEEDS GUIDANCE, WORKER I 5 5 ti 5 SIS? CAN'T WORK ALONE DOESN'T FOLLOWS FOLLOW 1_ TL % I 1r 1f I DIRECT IONS DIRECTIONS {’ 2 3 4 5 6 7 DISTRUBINC HELPFUL T0 T0 OTHERS 1r I I 1r I I _I OTHERS 1 2 3 4 5 6 7 INDUSTRIOUS I: If I I I I I WASTES TIME 1 2 3 4 5 6 7 ICNORES RES PONDS To PRAISE ‘4 5 L 5 I I l PRAISE 1 2 3 4 5 6 7 COMPLETES RARELY COMPLETES WORK L2 s 4, 5 5 I _4 WORK 1 2 3 4 5 6 7 DOES NOT ATTENDS DURING ATTEND LA; I_ I L L 4 CLASS DURING CLASS 1 2 3 4 5 6 7 OVERALL EVALUATION OF CHILD: BEHAVIOR NO IMPROVEMENT OR IMPROVING 4— I II I I I I DISRUPTIVE BEHAVIOR 1 2 3 4 5 6 7 J. 153 Elizabeth Bridge Michigan State University January 1973 TEACHER RESEARCH EVALUATION SCALE TEACHERS NAME DATE DIRECTIONS: PLEASE CIRCLE THE NUMBER ON EACH SCALE WHICH YOU FEEL ‘MOST CLOSELY APPROXIMATES YOUR REACTION TO THIS STUDY AT THIS TIME. TOO TIME CONSUMING 1 1 1 1 1 1 J I l l ”I I‘ l I 1 2 3 4 5 6 7 DECREASES WORK LOAD 1 1 1 1 1 1 1 F I I I I— I I l 2 3 4 5 6 7 STUDY PERMITS ME TO GIVE ADEQUATE ATTENTION T0 IFII :I I l l I I I l OTHER.CHILDREN 1 2 3 4 5 6 7 TOO MUCH PAPER.WORK 1 1 1 1 1 1 1 I I I I F I I l 2 3 4 5 6 7 I FEEL POSITIVE TOWARDS THIS STUDY ’ 1 1 1 1 J 1 1 I I I wI I I I] l 2 3 4 5 6 7 I FEEL THIS STUDY ISN'T HELPFUL L 1 1 4 J 1 1 IV I I T 'I I fl 1 2 3 4 5 6 7 THE RESULTS OF THIS STUDY ARE OBVIOUS 1 1 11 1 1 1 1 I 7 I I I I I 1 2 3 4 5 6 7 I TEACH AS WELL 0R BETTER WITH OBSERVER IN 1 41 I1 1 1L 1 l ROOM r I I r I T’ *1 l 2 3 4 5 6 7 COMMENTS ON ANY ASPECTS OF STUDY WORTH THE TIME COMMITMENT INCREASES WORK LOAD STUDY PROHIBITS ME' GIVING ADEQUATE ATTENTION TO OTHER CHILDREN PAPER WORK.ACCEPTABLE I FEEL NEGATIVE TOWARD THIS STUDY ' I FEEL THIS STUDY IS‘ HELPFUL THE RESUETS OF THIS STUDY ARE NOT OBVIOUS TO ME I FEEL SELF§CONSCIOUS WITH OBSERVER IN ROOM 154 Elizabeth Bridge Michigan State University October 1972 TY‘f’IITFWD I"Trr :HjQ}; I...“ am new t L‘s... n It. a 1! I19 \' c'\."§' Directions / 1. Materials needed: pencil, ruler, hard writing surface, a.eoess to a clock or watch with a second hand. 2. Fi.ll in the child's name and the exact time you begin - your observation. (in the left hand corner of the page) 3. Your observation will last for 20 marking periods ea ach one minute in duration. You will record a behavior twigs during each one minute period. Ex mple- You are watching Eric in a third grade classroom. When the second hand reaches 12 you begin looking at Eric to observe ' ’ his behaviorc This lasts_15 seconds. For the next 15 seconds you will decide: a. Whether the predominant behavior for that pas- 35 seCOnd peri.od was attendi.ng or non-attending. (predominant: behavi.or consuming the most number 01 seconds) b. Then which snecifie behavior in that class ol"HenaV10rs was he engaged in. You may only select one. Once you have made these two decisions you record an X in the box directly under that specific behavior category. Now you begin looking at Eric for another 15 seconds. When the time elapses S, use the final 15 seconds in that minute to record a behavior again. Record a 1 in the minutes column and move on to the second rating periOd. Note: You may find it helpful to slide a ruler down the page as you pass each 15 second period so you won't lose your place. 155 Elizabeth Bridge Michigan State University Example" October 1972 Name: Eric Jones Time In: 10:10 Time Omt: 10:50 Minutes 1. 2. CNOIUC‘. 4 I’vvv . 1,“..L‘L, I III"; I .4 I ' , ‘ In." {she-A \O.\"tt. ‘ V. '61: mm 0’ w —— When you have completed the 20 rating periods ( a total of 40 X's ) stop. Fill in the time you completed the observation next to Time Out: . Next total the number of attending X's recorded in the first rating period and write the munber in the total box. Repeat for non~attending. Do this for each of the 20 rating periods. At the bottom of each page record the total number of attending and non- attending marks given during the ratings for that page. kfinatiop 2; Terms 1. in seat: seat contact 2. Egggpmedutgggz reading, writing, discussing assigned work. 3. raISIng hand: raising hand as opposed to shouting out a 4. answerigg= question or an answer. speaking when a question has been raised by the teacher for the whole class or his group to answer. 5. 6. 7. 8. .9. 10. 11. 12. 156 Elizabeth Bridge Michigan State University October 1972 talking: talk not classified as answering. For example, blurting out answers when another was called on, talking to peers when not asked to do so etc. verbal noifi e= whistling, humming, car or animal noises etc.' phlsicar D0180: tapping feet, shuffling feet, knocking etc, staring into space: looking around room, at ceiling, out window, at floor, or at classmates who are acting up. physically disturb peer: with hands, feet, or object - touch hit or disturb peer such that peer res ponds verbally or physically. 93h.gd.seat= lack of seat contact for more than 5 seconds ( usually out of seat entirely and we.ndering around the room). ’ squirm;ng= wiggling in chair, under desk top searching thru desk, turning around in chair etc. hgpgling gbflgglg; playing wi th various objects such as toys, pencils, paper, paper clips etc. This would also includ.e the throwing of these objects. 157 BEHAVIOR ANALYSIS FORM . Elizabeth Bridge NAME' - Michigan State Univ. TIME IN: ‘- October 1972 TIME OUT: DNQ-a Now-AU DING: ‘9 ‘5 R v 2. if 13 m . “g 3 3 T k “ a B Smeweus m $3” 15 OUT wt is; '3 "t ‘f l\ “ND 6 6‘ .._.L 1mm“ ‘YS k0 Mack ET ii .Lflml—__L___1j ‘MAL mow. MAM ‘ Tow. M-Aflauoms Mme . Ln 158 Elizabeth Bridge M.S.U./counseling Nov. 1972 TEACHER BACKGROUND DATA FORM TEACHER'S NAME: STUDENT'S NAME: I. GENERAL DATA A. AGE: B. GRADE: C. FAMILY BACKGROUND: 1. EMPLOYED? MOTHER FATHER WHERE? MOTHER FATHER HOURS? MOTHER FATHER 2. MARITAL STATUS: 3. SIBLINGS (NAME, AGE, SEX): 4. SOCIO-ECONOMIC STATUS: 5. LIST OTHERS LIVING IN HOME: 159 Elizabeth Bridge II. SCHOOL RECORDS M.S.U./counseling Nov. 1972 A . GRADES EARNQ 1. LAST YEAR (SPECIFY CLASS AND GRADE) 2. THIS YEAR (SPECIFY CLASS AND GRADE) B. TEST DATA l. ACHIEVEMENT: 2. READING LEVEL: 3. INTELLIGENCE TEST SCORE: 4. OTHER: 160 Elizabeth Bridge M.S.U./counseling c. PREVIOUS TEACHERS N°V~ 1972 1. LIST PREVIOUS TEACHERS THIS STUDENT HAD WHO ARE CURRENTLY IN SCHOOL HE IS ATTENDING: D. TEACHER COMMENTS (measurable and observable behaviors only) ? 1. STUDY HABITS 2. INTERACTION WITH OTHERS: 3. TEACHER/STUDENT INTERACTION 161 Elizabeth Bridge M.S.U./counseling Nov. 1972 E. PREVIOUS OR CURRENT SCHOOL SERVICES USED: 1. COUNSELOR (WHO, WHEN, WHY) 2. SOCIAL WORKER (WHO, WHEN, WHY) 3. OTHER (WHO, WHEN, WHY) III. SPECIFIC BEHAVIOR PROBLEMS A. WRITE A BRIEF BEHAVIORAL STATEMENT OF THE PROBLEM BEHAVIOR: 162 Elizabeth Bridge M.S.U./counseling Nov. 1972 B. LIST THE FREQUENCY AND/OR DURATION OF PROBLEM (AS PERCEIVED BY TEACHER) l. FREQUENCY: 2. DURATION: C. STATE THE APPROXIMATE DATE THE PROBLEM BEHAVIOR BEGAN: D. STATE THE CIRCUMSTANCES UNDER WHICH THE BEHAVIOR OCCURS: E. LIST PREVIOUS TECHNIQUES USED TO CHANGE CURRENT BEHAVIOR PROBLEM 163 Elizabeth Bridge M.S.U./counseling Nov. 1972 IV. POTENTIAL REINFORCERS A. LIST THOSE REINFORCERS THE TEACHER SEES AS REWARDING THE STUDENT 1. TANGIBLE FOOD: OBJECTS: PRIVILEGES: IN CLASSROOM: OUTSIDE CLASSROOM: ACTIVITIES: 164 Elizabeth Bridge M.S.U./Counseling Nov. 1972 TOUCHING (PATTING HEAD, SHOULDER, ETC.) 2. INTANGIBLE WORDS OF PRAISE (LIST SPECIFIC PHRASES) B. HOW MANY OTHER STUDENTS IN YOUR ROOM HAVE MORE DIFFICULTY ATTENDING THAN THIS CHILD? 165 Elizabeth Bridge M.S.U./Counseling Nov. 1972 STUDENT BACKGROUND DATA FORM STUDENT'S NAME: PLEASE ANSWER EACH OF THESE QUESTIONS AS BEST YOU CAN. 1. I THINK SCHOOL IS 2. IF I COULD DO ANYTHING TODAY I WOULD 3. I THINK MY TEACHER IS 4. IF I HAVE FREE TIME IN MY CLASSROOM I LIKE TO 5. IF I DON'T UNDERSTAND HOW TO DO MY SCHOOL WORK I WOULD 6. WHEN I AM AT HOME I LIKE TO 7. I LIKE TO COLLECT 8. MY FAVORITE SPORT IS 9. THE HARDEST THING FOR ME TO DO IN SCHOOL IS 10. I LIKE TO READ ll. 12. l3. 14. 15. l6. l7. l8. 19. 166 Elizabeth Bridge M.S.U./Counseling Nov. 1972 THE JOB I LIKE THE MOST TO DO IN OUR CLASSROOM IS WHEN I AM BORED IN SCHOOL I WHEN THE TEACHER TELLS ME I DID A GOOD JOB IT MAKES ME FEEL . DURING ART MY FAVORITE THING TO DO IS MY FAVORITE KIND OF CANDY IS ON SATURDAYS I LIKE TO IF I COULD BE THE TEACHER TOMORROW I WOULD MAKE SURE WE WOULD IF I WERE AT A STORE AND COULD BUY SOMETHING LIKE --CANDY --TOYS --COMICS --FOOTBALL CARDS --BASEBALL CARDS -—SMILE PINS -*MODELS MY FIRST CHOICE WOULD BE . MY SECOND CHOICE WOULD BE . ‘ MY THIRD CHOICE WOULD BE . THE THING I WOULD LIKE THAT ISN'T EVEN ON THE LIST IS ' . 167 Elizabeth Bridge M.S.U./counseling Nov. 1972 ANDRE CHART I W TEACHEL'S NMDE DATE POINTS EMUED TUES. MARCH b mlNUTES To ATTEND NED. mum-t» ‘1 FOMMHES TO (“THUD THURS. mnecp. ‘1 mwu‘I'ES To Wren!) __.__ 5mm: won: CHFHZT U321) Hon-H smowrs "0 TOW ECOmmY . meT m RECORD MINUTES OF WONG. 168 Elizabeth Bridge M.S.U./counseling Nov. 1972 CONTRACT I will sit in my chair and do my student's name work (reading, writing, or answering questions) each day until I am able to do it for a total of 30 minutes. I will start by doing my work for 10 minutes each day and work up slowly from there. DAY WORK MINUTES TIME OF DAY POINTS POSSIBLE Tues. Wed. Thurs. Fri. Each day the teacher will tell me how long I must work to earn my points. To help me she will put a watch on my desk so I can tell how long I have been working. If I am able to Sit and work for the number of minutes she says each day, I will earn points. I will get 1 point for each minute I am working. will tell me each day teacher . how many points I have earned. At the end of the day I can trade my points in on any items in the Trade-In Book. Each trade—in will have a price on it. I may onl trade for an item when I have enough points saved up E0 purchase it. will keep track of how many points I teacher earn and spend. If I can earn all my points two days in a row, I'll have a chance to earn more the next day. STUDENT TEACHER DATE DATA 169 Elizabeth Bridge M.S.U./counseling Nov. 1972 DATA ACCUMULATION FORM Teacher Student POINTS POSSIBLE POINTS EARNED POINTS "SPENT" ITEM TUES FEB. WED. FEB. THURS FEB. 8 *if maximum points earned 2 davs in a row--up time 5 min FRI. FEB. MON. FEB. 12 TUES. FEB. 13 DATE 170 Elizabeth Bridge M.S.U./counseling Nov. 1972 Teacher Student POINTS POSSIBLE POINTS EARNED POINTS"SPENT" ITEM WED. FEB. 14 THURS FEB. 15 FRI. FEB. 16 MON. FEB. 19 171 Elizabeth Bridge MSU/Counseling Dept. November, 1972 REINFORCEMENT MENU USED BY STUDENTS IN SELECTING REWARDS IN EXCHANGE FOR POINTS ACCUMULATED FOR ATTENDING BEHAVIOR 172 10 POINT TRADE—INS PASSING OUT PAPERS CHOOSE A STORY TO BE READ TO CLASS 1 STICK GUM WASH THE BOARDS READ A LIBRARY BOOK (10 MIN.) SWEEP FLOORS IN THE ROOM FREE TIME IN THE ROOM (5 MIN.) BRING ATTENDANCE CARDS TO OFFICE COLLECT PAPERS IN CLASSROOM HELP A STUDENT IN A YOUNGER GRADE CLEAN SINKS AND COUNTER HELP ANOTHER STUDENT IN THE ROOM 10 10 10 10 10 10 10 10 10 10 10 10 POINTS POINTS POINTS POINTS POINTS POINTS POINTS POINTS POINTS POINTS POINTS POINTS 173 20 AND 25 POINT TRADE-INS DOING ART WORK = 20 POINTS CANDY BAR = 20 POINTS SNOOPY STICKERS (1) = 20 POINTS SMALL SMILE :5 PINS = 20 POINTS HELP WITH A BULLETIN BOARD = 20 POINTS RUN ERRANDS FOR TEACHER = 25 POINTS USE PUZZLES IN THE ROOM = 20 POINTS PLAY A GAME IN THE ROOM = 20 POINTS 20 POINTS USE THE LISTENING CENTER 30 POINTS COMIC BOOK LARGE SMILE PIN MIN. IN THE GYM FELT TIP MARKER PEN SMALL TOY CAR CAR OR MOTORCYCLE MODEL POSTER 174 AND UP TRADE*INS = 30 POINTS = 30 POINTS = 30 POINTS = 50 POINTS 100 POINTS 150 POINTS 200 POINTS "‘IWWW