AC . ._ Rigor z'vi C . n-“ .E E ‘ ‘p ‘ 0‘ ~ . ". "M-K Jdfiusqf 'U p C‘- u. .. , . ’ H U‘ul. v- ABSTRACT A STUDY OF THE USE OF HIGHER FUNCTIONING RETARDATES AS LANGUAGE ACQUISITION TRAINERS OF LOWER FUNCTIONING RETARDATES IN ATTENDANT SUPERVISED TRAINING SESSIONS ON INSTITUTIONAL NARDS By Martha Evelyn Snell Rowland Within most institutions for the retarded are barriers which prevent adequate intellectual stimulation of its residents. These bar- riers include the low ratio of attendants to residents, the low ratio of education and speech professionals to residents, the inadequate preparation and motivation of attendants to train residents, and the administrative priorities which in practice generally do not result in adequate educational programming for residents. As a consequence of these factors, institutions for the retarded have been shown to retard even further the social, cognitive, and language skills of their resi- dents. The purpose of this study was to demonstrate and compare the effectiveness of a strategy for utilizing the higher functioning resi- dents of an institution to train the lower functioning residents under the guidance of the attendant staff. Specifically this study sought to determine whether significant gains in language functioning and in the ability to respond to a standard measure of intelligence could be T i'fqued by re‘. *ertally ordere fn‘2rcezent an: t'lecess of the ,__ .. I >w:Wi51ng atte :Iged where tra ra‘r:. b J i rce“ent oh 6' . 7'! a . Ll. Tn‘ - fr‘m‘qn‘h‘ " '~b on. Y . ICERt7( Martha Evelyn Snell Rowland. produced by retarded trainers who were trained to administer a develop- mentally ordered language program which incorporated principles of re- inforcement and modeling (Group E). In order to determine the effec- tiveness of the instruction given to these retarded trainers and their supervising attendants in Group E, a control group (Group C) was em- ployed where trainers were encouraged to provide their students with reinforcement on a non-contingent basis, verbal and tactual stimula- tion, training based on their own or their attendant's ideas, and play interaction. Identical in both groups, the organization consisted of basic training units. Each of these units was composed of one supervising attendant who directed two trainable mentally retarded trainers. Each trainer in turn, was assigned to two profoundly retarded students, al- though all training was done with one student at a time. Each group included three of these training units or a total of three attendants, six trainers and twelve students. Trainers and students were paired together first and then randomly divided into the two groups. Atten- dants were randomly divided into two groups and then paired with two trainer-student triads. The experiment consisted of three phases. Phase I, lasting two weeks, was an initial training period, which varied for Groups E and C. Phase II lasting three weeks, involved the training of atten- dants in the supervision of one triad at a time, while Phase III, last- ing three weeks, involved each attendant supervising her two triads, vvith maintenance levels of supervision and instruction from experimen- tal staff. r! . ‘I r IIU “IV ote fra‘fied t 22 n to tr vv .u vs“ C s t a" .n b“** las‘pd eig‘v . .9'6" I; ’ 'I'Iyr g i“r- 1‘ P \ . - fianrn Martha Evelyn Snell Rowland During Phase I and II Group E trainers and attendants were trained to teach toward specific language goals by applying condition— ing principles as embedded in the Modified Language Acquisition Program (MLAP). During this same time, Group C trainers were also given lan- guage training goals for their students, but without training in the use of this program or behavioral principles. Insofar as possible, all other training variables were made equivalent for the two groups. Thus, the independent variable was the application of a developmentally or- dered language training program which incorporated the principles of re- inforcement and modeling. In both groups the daily training sessions between trainer and student lasted 20 minutes. Each trainer worked with one of his assigned students for a single session and then with the second student for an- other single session. Throughout the entire 8-week study each student received a mean of 13 hours of training. Except for brief demonstra- tions given to the trainers by attendants and experimental staff and for some extreme cases of behavior control, all the direct interaction with the students during these training sessions was done by the trainers. The main dependent variables were measures of language ability (The Modified Language Acquisition Program Final Tests, MLAPFT) and in- telligence (Cattell Infant Intelligence Scale). The design analyzed the pre-test, posttest scores on both dependent variables of all pairs of students trained by a single trainer, by a one-way analysis of variance of the index of response. The scores of the two students in a triad were combined in the analysis, since they were considered not to be in- «dependent, but to be, in part, a function of the skill of the trainer in 'the triad. .tiiln Experi“ bu- AA' .: .: new to tr; It was "AA . .- u ? ~‘ ‘“ J: fF-rs - , . c 'V‘A " H -s . SEEN-S On " “:3 ”I n ' "on: C ,~:a.tmt&] ‘ . i ‘ *‘5 3f the a, ENC." ‘ can In N .:‘§ 3: Martha Evelyn Snell Rowland A less formal dependent variable, the Behavioral Observational Record, BOR, was utilized to gather data in both groups regarding the in- structional interactions between trainers and students in regular train- ing sessions and between trainers and students and attendants and stu- dents in experimenter structured training sessions which utilized stu- dents new to trainers and attendants. It was the main hypothesis of this study that the intensive training of Group E at the attendant and trainer levels in the use of re- inforcement and modeling techniques and the Modified Language Acquisition Program would yield greater growth in language skills by the Group E stu- dents than by the Group C students who were paired with untrained re- tarded trainers and supervising attendants. This hypothesis was sup- ported since the mean gain in language skills in Group E students was significantly greater than the gain in Group C. It was further hypothesized that Group E students, being dif- ferentially reinforced for attending skills and early language skills commonly tested by infant intelligence tests, would demonstrate greater mean scores on the Cattell Infant Intelligence Scale than would the stu- dents in Group C. Although this hypothesis was not supported by a sig- nificant mental age gain difference in favor of Group E students, the results of the analysis did reveal that both groups of students made significant MA gains as compared to the expected zero gain for the pop- ulation of older, profoundly retarded students from which they were drawn. The series of questions posed to investigate the actual inter- actions between an instructor and a student was answered informally by ar exac‘nation were collected *‘35 trainer Car 3‘ re study. TIMES in lac-3f S'CES StFUC‘uv—e: h PIA-:14 ”if“ v Hug-d "U‘. Cn‘ly :rfl- ‘ 'J' 353‘ 9'0.;s "in ‘f‘ «.e r: Hi“ i»... I ' all Ubt". Cr .u Q T " .n Cl trfi‘r‘;n- ('1: 3' Ev (n T J 'C" T7 1’ 1.. ('3 ‘3 "I Ill 3 CO '4‘! [ll Martha Evelyn Snell Rowland an examination of the data gathered with the BOR. Some of these data were collected from three-minute observations made of each student and his trainer during regular training sessions at the end of each phase of the study. The remainder of the observations, also each three- minutes in length, was taken before and after the study in training ses- sions structured by the experimenter. These latter observations in- cluded not only trainers and students but also attendants and students from both groups. All students trained during these observations were new to the attendant and the trainer. The results of these observations of attendants and of train- ers in both groups are summarized: A. In comparison to trainers using unstructured play methods of training with untrained attendant supervision, retarded train- ers using operant and developmental language training techniques and supervised by attendants trained in the same skills, were ob- served to consistently and more frequently l. present complete trials to students 2. have students attending to stimulus/presentations 3. evoke student responses 4. demonstrate appropriate reinforcement of student's re- sponses 5. couple tangible reinforcers with praise 6. demonstrate appropriate prompting of students 7. provide consistent stimulus presentations and with appro- priate repetition 8. fade prompts appropriately 9. ts l0. Sol deN 8. Ir. in the a;;' training pr 1. pm EY‘S Martha Evelyn Snell Rowland 9. teach tasks appropriate to the students l0. successfully generalize their training skills to new stu- dents in experimenter structured training situations. B. In comparison to untrained attendants, attendants skilled in the application of an operant, developmentally ordered language training program, were observed to consistently and more frequently l. provide reinforcement and instruction to the retarded train- ers they supervised 2. and to demonstrate a high quality of training skills when asked to instruct a student in a particular early language task (these training skills are described in A. 1. through A. 9). 43—31 in A STUDY OF THE USE OF HIGHER FUNCTIONING RETARDATES AS LANGUAGE ACQUISITION TRAINERS OF LOWER FUNCTIONING RETARDATES IN ATTENDANT SUPERVISED TRAINING SESSIONS ON INSTITUTIONAL NARDS By Martha Evelyn Snell Rowland A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Elementary and Special Education 1973 Coypyright by Martha Evelyn Snell Rowland l973 Ann-"I. ¢ re“ec.s the to: :e acxiowledged Host 9 iiie‘Ter, COTitte siegecent ard ACKNOWLEDGMENTS The development, implementation and completion of this study reflects the cooperation and support of many people; some of whom will be acknowledged at this time. Most grateful appreciation must be given to Dr. Edwin J. Keller, committee chairman, for his valuable guidance, continued en- couragement and personal support. Gratitude is also strongly expressed toward Dr. Donald Burke, committee member, for his unfailing inspira- tion and for his assistance in testing and videotaping during the study. Appreciation is extended also to Dr. M. Ray Denny and Dr. Stephen Yelon, committee members. Deep appreciation is extended to Diane Klein, past speech pathologist at Fort Custer State Home and Training School, for her con- tinued support and most valuable assistance in the development of the Modified Language Acquisition Program, and the selection and testing of participants in the study. Gratitude is also extended to Dr. Louise Kent, author of the Language Acquisition Program, for her approval and support of the Modified Language Acquisition Program. Thankfulness is extended to Linda Glendening and Andrew Porter for their continued assistance in the statistical analysis of the re- sults of this study. Most deep gratitude and appreciation must be extended to the many staff members of Howell State Hospital whose interest, support and ii is extended to 33:3 and invest It is staff without wt tear possible. ‘3." their daily 1’6 study and ft 5730 be deeply ' "Elite“.t and u Final Fl: " ~32tora] wor participation made this study possible. Specifically this appreciation is extended to Dr. Detterbeck, Superintendent, for his interest in and approval to have the study take place at Howell; Dr. Robert Gutshall, Program Director, Mrs. Doris Holley, Social Worker, and Carol Lappin, Speech Therapist, for their assistance throughout the pilot work and study; the six attendants and twelve trainers who participated in this study and invested their energy, talents and enthusiasm. It is important also to extend thanks to the experimental staff without whose invaluable assistance this study would not have been possible. Susan Miller and Nancy Williams deserve great thanks for their daily guidance of the Control Group during the eight weeks of the study and for their continual personal support. Penny Rogers must also be deeply thanked for her interest and participation in the de- velopment and use of the Behavioral Observational Record. Finally deep and sincere appreciation is extended to my fam- ily and friends who offered personal encouragement throughout the years of doctoral work. TABLE OF CONTENTS LIST OF TABLES ........................ LIST OF FIGURES ....................... Chapter PROBLEM AND REVIEW OF RELATED RESEARCH ........ Difficulties in the Implementation of InStitutional Resident Training Programs ...... Attendant-Resident Ratio .............. Professional Staff-Resident Ratio ......... Preparation of Attendants as Resident Educators . . Motivating Attendants to Train Residents ...... Methods of Inservice Training for Institutional Attendants .................... The Nonstimulating Environment of the Institution . . Language as Verbal Behavior ............. Language Acquisition Training Based on Operant Conditioning Procedures for Lower Functioning Retardates ..................... Research on Imitative Training as an Important Aspect of Operant Training in Retardates ..... Research on Control of Disruptive Behaviors During Language Training ............. Research on Specific Aspects of Language Training ..................... Language Training Programs for Lower Functioning Retardates ................... Using Peers as Helpers, Behavioral Engineers, and Tutorsin Classrooms and Institutions ....... Utilization of Normal Peers as Change Agents for Special Education Students .......... iv Page viii dVO‘N N l2 15 l8 23 24 28 32 41 53 r ., .(‘d Page Utilization of Retarded Persons as Helpers, Substitute "Mothers" Trainers, and Behavioral Engineers of Other Retardates in Classrooms and Institutions ................. 55 Summary Relevant to the Purposes of the Study . . . . 64 II. METHODOLOGY ....................... 66 Purpose ........................ 66 Design ........................ Overview of Design ................. 66 Research Questions ................. 69 Modifications in Design from Related Research . . . . 73 Subjects ....................... 74 Students ...................... 74 Trainers ...................... 77 Attendants ..................... 8l Assignment of Subjects to Treatment Groups ...... 84 Treatment Procedures ................. 86 The Modified Language Acquisition Program ...... 86 Phase Structure ................... 92 General Controls Common to Both Groups ....... 94 Phase Ia: Initial Instruction Procedures Common to Trainers in Both Groups ............ 96 Phase Ia: Initial Instructional Procedures for Group E Trainers ................. 97 Phase Ia: Initial Instructional Procedures for Group C Trainers ................. 99 Phase Ib and c: Initial Training Procedures . Common to Supervising Attendants in Both Groups . . lOO Phase lb: Initial Training of Group E Supervising Attendants .............. lOl Phase Ic: Initial Training of Group C Supervising Attendants .................... lOZ Phase II: Procedures Common to Both Groups ..... 104 Phase II: Training Group E Attendants to Supervise One Trainer at a Time ............... lO6 Phase II: Training Group C Attendants to Supervise One Trainer at a Time . . . ............ lO9 Phase III: Procedures Common to Both Groups . . . . lO9 Description of Measures and Testing Procedures . . . . llZ Overview of Testing ................. llZ IN (‘7 (.17 J'- . . . "TI ['7 (' J I J I Treatne Hyaot RESULTS . HIDCthes Infarmal ”L7 (.1) ._4 (,1! I J (7 a: £3. (D III. IV. Cattell Infant Intelligence Scale .......... Behavioral Observation Record (BOR) ......... Hypotheses ...................... Treatment of the Data Relating to the Major Hypotheses ..................... RESULTS ......................... Hypothesis .................. Informal Questions . . . . . . . . . . . . . . : Z . . Questions and BOR Results for Trainers in Regular Training Sessions ................. Questions and BOR Results for Trainers in Experimenter-Structured Training Sessions ..... Questions and BOR Results for Attendants in Experimenter-Structured Training Sessions ..... Informal Accounts of Training Sessions by Attendants and Experimental Staff ............... Examples Taken from Group E Accounts ........ Examples Taken from Group C Accounts ........ Summary ........................ SUMMARY AND CONCLUSIONS ................. Summary of the Study ................. Discussion .................... . . Implications for Research ............... BIBLIOGRAPHY ......................... APPENDICES .......................... Appendix (D ‘T'lfT'lU OW) Letter to Parents and Guardians of Students ....... Letter to Parents and Guardians of Trainers ....... Students' Placement in MLAP Based Upon Initial or Final Test Results .................. Sample Training Observation ............... Picture Training Card Examples ............. Guidelines for Experimental Assistants' Work with Trainers During Phase I ................ General Trainer Goals for Attendants in Control Group ...................... vi Page 121 124 139 140 142 142 153 154 172 177 181 181 182 184 187 187 194 201 204 215 215 216 217 220 221 222 225 (DZE‘ Exarples Provice General Perfon'a" lhe Modi‘ to Acc: Prograr Exaerimer f0r Ad? ACJulSl Sefiavlora 367316 Be. Directions Attendar Students Pretest ar Totals 1 7 ,/// 2 R D 2 ng \R3e— sR SD-—)R3 Figure 3. The Infant's Vocal Repertoire Comes Under the Control of Both Antecedent and Consequent Stimuli (p. 130). the child as an organism, (2) variation within the initial setting or stimulation conditions for the child, and (3) variation in the rein- forcement variables attendant on his behavior. The source of varia- tion occurring within the child which may result in language retarda- tion includes such disabilities or major pathologies as profound hear- ing loss or severe paralysis of the muscles in the speech mechanisms. "Though these are obviously 'organic' conditions," Siegel notes that, "it is helpful to think of them as conditions which affect the child's capabilities for interacting with his environment rather than as static attributes of the organism (p. 131.)" Deaf children with the help of hearing aides or with exposure to gestural systems of language, can still learn language by the same system of behavioral principles. Like- wise, a person, unable to develop speech due to muscle paralysis, can learn a variant form of expressive language, perhaps manual signing. Quality and quantity of verbal stimulation in a child's en- vironment may be a source of retardation in that child's language de- velopment. To a great extent, speech appears to be modeled after the signif- icant adults in the child's environment. It is not unusual to find that a child with a severe articulation disorder has a parent with a similar problem. If the opportunities to model speech are scarce, or if V 0r 1t "'9 ment iS_ prOL‘iens‘I Rei ‘n; aTJDFO-Dr'at tle exoerlfiert iatsfiee» one ' m parents 0‘ ::*:ed to have early years. 5 that the child 1 was not differei :te environrent rsspeech deve' element as nor ‘erced child's e :erl y developmer aficeand there haviors that do AnothE 'rgege occurS “Til behaviors =-Ctrced in an 5'8: all 111$: s'fcnn “Wis with a . 22 or if the model is itself defective, disordered speech may develop. Or it may be that the kind of stimulation provided by the environ- ment is somehow not properly matched to the child's behavioral propensities (p. 132). Reinforcement is the primary means of strengthening and shap- ing appropriate language skills. Siegel notes that since there is lit- tle experimental evidence to specify the factors leading to delayed language, one often resorts to anecdotal evidence. Often in interviews with parents of nonverbal speech-retarded children the child is re- ported to have been quiet and without any "need for speech" during his early years. Siegel says that "in motivational terms this suggests that the child produced little verbal behavior, and that verbalization was not differentially reinforced over other methods for manipulating the environment (p. 131)." The mentally retarded child is delayed in his speech development but appears to follow the same sequence of de- velopment as normals (Blount, 1968). However, if the adults in the re- tarded child's environment are unaware of this predictably slow but or- derly development they may "give up" on the child's speech and fail to notice and therefore to differentially reinforce the early language be- haviors that do occur. Another aspect of improper reinforcement which acts to retard language occurs when one portion of a child's environment reinforces verbal behaviors that are incompatible with the verbal behaviors being reinforced in another portion of that same child's environment. For example, an institutionalized child may be reinforced on the ward for silence or for echoic verbal responses while during language training sessions with a speech therapist he may be reinforced for gross vocal responses or tending upon hiher envirc P Lar. Fifi-fillies of according to 5 . . . it i for respo. Presence D61 repem DEDdVlorS exprESSlor I0 bring E rect reini ltative re and gnawé If a Ch11( prCPFIdte the antECe mentaHY < Pattern j, (D. 132.3 23 responses or for nonimitative expressive responses to questions. De- pending upon the reinforcers and the length of the child's exposure to either environment, language retardation is apt to result. Language training according to learning theory utilizes the principles of operant conditioning or behavior modification in that, according to Siegel: . it consists of the systematic application of reinforcement for responses that increasingly approximate some norm, and in the presence of appropriate cues. If the child has little or no ver- bal repertoire, the task is to develop the necessary prerequisite behaviors that will move the child toward at least minimal verbal expression. Initially, this will involve a concentrated effort to bring about any sort of reasonable vocal behavior through di- rect reinforcement, shaping, and perhaps the nurturing of an im- itative repertoire along the lines described by Baer, Peterson, and Sherman (1967). If a child has the necessary responses, but emits them in inap- propriate contexts, the speech program may have to concentrate on the antecedent control. A great deal of speech therapy is funda- mentally concerned with the stimulus shift--bringing a response gattern igto the context of appropriate stimulus conditions p. 132-3 . Language Agguisition Training Based on Operant Conditioning Procedures for» Lower Functioning Retardates Research in the application of operant conditioning or behav- ior modification procedures with institutionalized and noninstitutional- ized, moderately, severely, and profoundly mentally retarded persons are numerous and often demonstrate successful results (Gardner and Watson, 1969; Watson, 1967). Recent examples of these studies describe experimental modification of behaviors in areas of self help (Azrin and Foxx, 1971; O'Brien and Azrin, 1972; O'Brien, Bugle and Azrin, 1972), self abusiveness (Corte, Wold, and Locke, 1971; Smolev, 1971), language hawence, 1 die prograr cihd's entir 192;Lent, l Mos be reviewed (it erating princi - ~ . the Should be in nomai Sequence '1 MinEuistic i339 in the n “3:49 consul t I "'3“; Willie til 24 (Lawrence, 1971; MacCubrey, 1971; Kent, 1972) as well as in comprehen- sive programs which apply behavioral principles throughout the retarded child's entire day (Browning and Stover, 1971; Kazdin and Bootzin, 1972; Lent, 1967; Pizzat, 1973; Thompson and Grabowski, 1972). Most of the behaviorally based language training programs to be reviewed next, including the one used in this study, follow the op- erating principle described by Yoder and Miller (1972): . . . the content for language training for retarded children should be taken from the data available on language development in normal children, and this content should be taught in the same sequence that it is acquired by the normal child (p. 11). The linguistic studies that trace the developmental acquisition of lan- guage in the normal child (Bellugi, 1967, 1972; Bloom, 1970, 1972; etc.) must be consulted to provide the content and the sequence of instruc- tion; while the behavioristic principles of operant conditioning are used to specify the instructional procedures which most efficiently teach that content (Gray and Ryan, 1973). Research on Imitative Training as an Important Aspect of Operant Training in Retardates Risley, Hart and Doke (1972) summarized operant language mod- ification research and found an overwhelming emphasis placed on the im- portance of imitation in these studies. The reasoning for the impor- tance of imitative training is explained as follows: Most behavior modification efforts are premised upon the effective- ness of the therapist's 'modelling' of the desired behavior in pro- ducing similar behavior from children. When such demonstrations are not effective-~that is, when the child does not imitate--there “l can be new bet ferenti. desired conside' slow an: establi: part of Har‘ C'tiOhing of { 593;.ence of si Chfldren Wlth hum—- r+ “i363 alread se:.er.:e the f. 3- 1| 4- ll ti 5. ,. BaEr .3300 deVe 2:“ an alte 25 can be no hope for rapidly producing significant acquisition of new behavior. The alternative, shaping each new behavior by dif- ferentially reinforcing successively closer approximations of the desired behavior, while often necessary, requires a therapist (of considerable skill). Even then the shaping process is frequently slow and arduous. Consequently, the development of procedures to establish or to improve imitative responding has been a necessary part of operant behavior modification research (p. l07). Hartung (l970), in his review of the research on verbal con- ditioning of nonspeaking children through imitation training, notes the sequence of skills often necessary to develop meaningful speech in children with no rudimentary speech and little or no attending skills: . the control and limitation of disruptive behavior; the conditioning of attention and eye contact; the development of motor imitation; . the transition from motor imitation to gross vocal imita- tion; systematic shaping of vocalizations into vowels, consonant- vowel combinations, and/or words; and . the transition from imitation to naming. 0’ 01 thd Other research reviewed by Hartung and by Risley et_al, (l972) revealed an alternate technique to develop imitative speech in children with some rudimentary speech: establishing imitative control over vocal re- sponses already occurring in the child's repertoire. In this training sequence the first two steps above are retained and followed by 3. increasing the frequency of vocalizations, 4. introducing the experimenter's vocalizations as S the child's vocalizations, and 5. and 6. same as in first sequence of training steps. Ds for Baer and Sherman have done a large amount of work with motor imitation development in retarded persons. In an early study, (Sherman, 1965), an alternative procedure was developed to train motor imitative responses as a prerequisite to training vocal imitation in a subject pcssessing St was difficult actsr behavic' grcss motor te- SE'ted were c‘. ‘75:. blowing c SCJGC'S was pro . n’s: 26 possessing so few vocalizations that their differential reinforcement was difficult. Sherman then began to reinforce imitations of gross motor behaviors (raising an arm, etc.). Once motor imitation of novel gross motor behaviors occurred reliably, the imitative behaviors pre- sented were closer approximations to vocalizations: open mouth, cough- ing, blowing out a match, blowing (without a match), unvoiced sounds, and finally a voiced sound. Once this initial repertoire of imitative sounds was produced and consistently reinforced, the retarded subject began imitating other sounds; later sounds were chained into words and other new, whole words then imitated. At other times Baer, Peterson, and Sherman (l967) used imitative chains of motor responses ending with a vocalization to encourage subjects to make the sometimes difficult transition from motor imitation to vocal imitation. Baer, gt;§l,, (l967), Lovaas, Freitas, Nelson, and Whalen (l967), and Metz (l965) have trained children in motor imitation skills who have had no discernible imitation as a basis for the establishment of an elaborate imitative repertoire. Shaping procedures were used that involved complete physical prompts or "putting the child through" the desired action by physically guiding him and then rewarding the action thus prompted. Such complete prompts would gradually be faded out requiring the child to make successively closer approximations in his own imitation of the trainer's action. Baer, gt_al, used this pro- cedure as part of the training to establish generalized imitation. Generalized imitation is said to occur when a person imitates novel be- haviors which were never reinforced. These behaviors would continue to be imitated at a high strength as long as some of the other imitative responses in a training session were reinforced. Pt ghenon a c five responSe taUy retarce Social appro. r5;mwes. As 5 45"9uage ACOu‘f Wier. 1972 “tardates unA' leinitial cor TE4\ , "1' Hords p ,(Eh* ‘ "‘ 9t ,‘\a]'s I\) thr Sen 27 McReynolds (1970) found that food and social reinforcement given on a continuous schedule during the acquisition of vocal imita- tive responses (vowel sounds and consonant-vowel combinations) in men- tally retarded children increased the speed of initial acquisition. Social approval could then effectively maintain these imitative vocal responses. As an effort to sequence the vocal imitation phase in the Language Acquisition Program, the LAP, (Kent, Klein, Falk, and ‘ Guenther, l972) Guenther and Klein (l969) asked 33 institutionalized retardates under IO 30 and between the ages of ll and l7 to imitate a random list of mono and bi—syllabic nonsense syllable.words using all the initial consonants tested in the early speech research (Poole, 1934). Words were taken from the list of 200 most likely spoken nouns (Kent, gt_al,, l968). Each response was reinforced with food; two changes per word were given if there was no initial response. They concluded that, in the institutional population tested: l. the retardates were no more likely to respond imitatively to nonsense syllable stimuli than to word stimuli, nor were they more likely to correctly imitate the initial consonant in consonant-vowel nonsense syllables than in words. Therefore, they concluded, that in the teaching of vocal imitation to nonverbal retarded children, there appeared to be no rationale for the use of nonsense syl- lables. Also the words taught in the vocal imitative phase were those the child later was taught to produce functionally in the LAP. 2. Syllabification was more often correctly imitated in mono- syllabic words than in bi-syllabic words. Therefore, im- itation of mono-syllabic words was sequenced to be taught before imitation bi-syllabic or word combinations. 3. There appeared to be a positive correlation among the three rank orders of correctly imitated consonants in non- sense syllables, in mono-syllabic and in bi-syllabic words. And, upc ported ' conant s The faiconsonar tees (b,h,w,~ arfizblation g q‘:";~'¢ ~.p.al supp: sexence of de 1', :3 fears COFQEY Rigie 28 And, upon inspection, these rank orders corresponded to those re- ported in the literature on normal developmental sequence of con- conant sounds (Poole, 1934; Templin, 1957). Therefore words trained in the LAP were selected whose ini- tial consonants were among the first ten rank ordered consonant pho- nemes (b,h,w,m,p,n,k,j,g, and sh). Blanchard (1964) who studied the articulation growth patterns of 350 retarded residents at Pacific State Hospital supports Guenther and Klein's findings in that she found the sequence of development of consonant sounds in retardates between 8 to 15 years comparable to the sequence of development in normal children. Research on Control of Disruptive Behaviors During_Language Training A number of techniques to decrease disruptive behaviors oc- curring during language training of retarded children have been re- searched. Common procedures used are extinction (withholding positive reinforcers following inappropriate responses) and time out from posi- tive reinforcement (TO) (Lawrence, 1971; McReynolds, 1969; Risley and Wolf, 1968; Sloane, Johnson, and Harris, 1968). Disruptive behaviors (inattention, pushing stimulus objects away, tantrums, etc.) were usu- ally eliminated by these researchers when all possible reinforcers for these behaviors were removed contingent on their occurrence. TOs ranged from removal of the child to a T0 room for a short period of time to a mild TO: e.g., the trainer looks away from the child con- tingent on disruptive behavior. Risley and Wolf (1968) found that 105 could be used to de- crease disruptive behaviors once a child had spent some of the language training SEE notice the 3 Risley and k; it;‘.e~.ented c rim. with the :atiole with itate, etc.) ' Lah' ‘zed profouncj ‘Tn decreasing items, head hi‘ i” ”9" Opinion IE resident's tn all behavior .9.” procedures “may 0f 5+ tr. "'9 m1 dgive ”my wi W” h epuShed awa rfl“‘ "“‘lc‘like b1 ‘, amt. . “W: Vlng’ 9tt 29 training session sitting in a chair and had been reinforced enough to notice the absence of reinforcement during a T0. T0 as described by Risley and Wolf is essentially a systematic mild punishment procedure implemented contingent on disruptive behaviors. When used in conjunc- tion with the reinforcement of appropriate behaviors that are incom- patible with the disruptive behaviors (attending, using objects to im- itate, etc.) the disruptive behaviors usually are eliminated. Lawrence (1971), in the training of 3 older, institutional- ized profoundly retarded adults, did not find TO procedures effective in decreasing the disruptive behaviors they displayed (pushing away of items, head hitting, putting a foot on the table, table pounding, etc.). In her opinion TOs were too similar to the control procedures used in the resident's prevailing institutional environment. She found that they not only were ineffective but tended to have a negative influence on all behavior following the TO procedure. Therefore Lawrence changed her procedures and gave commands, presented new stimuli, or presented a variety of stimuli contingent on disruptive behaviors. For example, when the disruptive behavior of putting a foot on the table occurred she would give the command, "Put your foot down," which was reinforced contingently with the removal of the foot from the table. When items were pushed away by the resident Lawrence introduced a new item. Autistic-like behaviors of repetitive postural movements (block shaking, hand waving, etc.) were interrupted by touching the arm of the resident and then presenting a new task. It seems that the substitution of these procedures for TOs contingent on disruptive behaviors is behaviorally unsound. If consistently applied, the residents would learn to engage in disrupti ‘v‘ tnei, which ‘ tive behavio' contingent If to increase *3 to decrease tr freo coatings .'ES ently incr 3’3" exanple . 9 9w .ind aireafij (".5 r PS a. ‘ ErerLOlf‘l "agency in lLr‘n- gcse Corr Tite u 3": D. °°Er (is: “arinh ‘ 3 e’Deh‘ i alnlnn ‘1’. J Jul] -. zed rs. SQ ugShfi' 30 in disruptive behaviors when the task was too difficult or disliked by them, which in the long run would increase the frequency of the disrup- tive behavior. McReynolds (1969) applied TO from positive reinforcement, contingent TOs were also implemented to function as negative reinforcers to increase the emission of the correct response as well as punishers to decrease the incorrect response. During these instances a T0 was given contingent upon the continued emission of a formerly correct but presently incorrect response in imitation of the experimenter's model. (For example, TOs were given for the response "want pop," when the child already had demonstrated that the response ''I want pop," was in his repertoire, i.e., "I want pop," had occurred at a 30-40 percent frequency in training). TOs were terminated contingent on the emission of the correct response (i.e., "I want pop"). In general these TOs not only rapidly decreased the incorrect responses that they contingently punished but they also increased directly the frequency of the correct response. McReynolds replicated the effectiveness of T0 to decrease inappropriate responses and demonstrated the technique of using T0 to increase correct responses. The main disadvantage of T0 procedures is that they require time away from the ongoing training program. Sailor, Guess, Rutherford, and Baer (1968) suggest a novel approach to controlling temper tantrums during experimental verbal training of retardates that eliminates this lost training time. Vocal imitation training was given to an institu- tionalized severely retarded girl who displayed frequent tantrums throughout the day. The words to be imitated during training were graded by N we through ssonse. (Dr: t'e subject w. r: ‘Q‘ ' , . deiCult; 3162C0ntirge Cray-.4 \‘ our; K Cliff iate' +‘ '-iOns *1 . A; c ~FQL€ 35‘: J .F'Jr- d . ur“; a 4 ‘ n- ”“(ij 4. ' 'h'e dlfc ‘ i it??? e _ Paint 37.1: Il ’ OEr 1 ital, 31 graded by the experimenters according to difficulty: number of words, one through four word responses, and number of syllables in each re- sponse. (Prior to the introduction of the tantrum control procedures the subject was successfully imitating comparable words at all levels of difficulty.) When disruptive tantrums occurred they were ignored, but following the tantrum, two opposite conditions were alternately ap— plied contingent on the tantrums: Condition I and III: experimenter changed stimuli from high- difficulty to low-difficulty content for two trials immedi- ately following the tantrum. Condition 11 and IV: experimenter changed stimuli from low— difficulty to high-difficulty content for two trials immed- iately following the tantrum. Conditions II and IV both resulted in a decreased rate of tantrums thereby allowing more words to be introduced and imitated; however, these words, though new to the subject, were of less difficulty to her. Conditions I and III resulted in an increased rate of tantrums and far less vocalizations per session due to the longer amount of time spent in tantrums; however, the words imitated by the subject were of a greater difficulty level. The data seems to show that the subject's tantrums during conditions I and III served an operant function of re- ducing the difficulty of the stimulus units presented to her, and were therefore maintained at a high level of occurrence. During conditions II and IV, her tantrums seemed to serve only to increase that level of difficulty and therefore were greatly reduced. The experimenters argue that: . difficc rate ac stimuli. 7r: clirg disrupt Ever, it see" ‘ C =?za:. 011 of Var 3‘1“} LE", overloa Eflip . Vn S a ”SUV!“ :qr‘t * and M 13nd "”3311: J th~ . --~t 32 . the contingency between tantrums and consequent stimulus difficulty was the factor responsible for the control of tantrum rate achieved in this study, rather than prevailing conditions of stimulus difficulty or reinforcement density (p. 243). These findings suggest an alternate procedure to T0 for han- dling disruptive behaviors in retardates during language training; how- ever, it seems that these results need to be replicated and refined further. Research on Specific Aspects of Lagguage Training A large number of studies have been done to investigate spe- cific aspects of training retardates in early language skills from techniques of generalizing verbal behaviors to another environment (Jeffery, 1972) to receptive training of adjective inflections (Baer and Guess, 1971). These experimental studies provide operant techniques which are then incorporated into experimental and nonexperimental appli- cations of language training programs for mentally retarded persons. Jeffrey (1972) investigated techniques to program the gener- alization of verbal responses from training sessions to the classroom-- an often overlooked stage of language training programs. In training sessions a nonvocal, low trainable girl learned to imitate phonemes and to tact and mand pictures and objects at a high rate. To increase the probability that these behaviors would generalize to the classroom Jeffery 1. selected training words of relevance to and high frequency in the classroom, (A) 0; WI .6! 2. 33 implemented classroom practice through a self managed lan- guage training machine--subject was taught to put picture recording cards in a Language Master machine, label the picture recording her response, listen to her response and the prerecorded correct response, judge the correctness of her response, and sort cards accordingly into "good" and “not so good" piles, . trained peer trainers to use praise and to work daily with the subject on reviewing picture word cards, and . saw that the teacher was made aware of the subject's newly learned response which she then prompted and praised in the classroom. The results revealed increases in four areas: 1. an increase from 41 percent to 95 percent correctly imi- tated phonemes, which was still maintained at this level when measured one month after training had terminated; . an increase in the subject's verbal response in the class- room during treatment (tripled baseline rate) which was maintained through the two follow-up measurement periods (1 and 3 months after termination of training); . an increase in the teacher's and classroom students' audi- tory discriminative stimuli directed at the subject (i.e., any verbal sounds directed specifically at the subject or class in general); and Jet and maintainir out aiso 99”er creased the a; aniclass and vetalizations Risl edtialic spee reoa nd incre tanmq UES USE n! 34 4. an increase in the teacher's or subjects' verbal conse- quences directed at the subject's verbal or nonverbal re- sponses (any verbal sounds directed specifically at the subject or class which followed the subject's verbal or nonverbal responses within 5 seconds). Jeffery's program was successful in substantially increasing and maintaining the rate of verbal responses not only made in training but also generalized to the classroom; the generalization program in- creased the auditory stimuli directed toward the subject by the teacher and class and the verbal consequences they provided for the subject's verbalizations. Risley and Wolf (1967), in an often cited study, decreased echolalic speech in low functioning, nonverbal, institutionalized chil- dren and increased the use of functional speech. Behavior modification techniques used included 1. the elimination of disruptive behavior through extinction and TO procedures in conjunction with differential rein- forcement of appropriate responses that are incompatible with the inappropriate responses, . establishment of control over imitation, transition from imitation to naming, . expansion of the naming vocabulary, and 01-500“) . establishment of phrases and sentences. Echolalic children sporadically and under their own control imitate words and phrases. This imitative ability, to be useful in language tr: is done by l: inforcing ti Often the tr rears of inc involves inc' model. Both liabiy imitat children, not tive speech w reinforcement was silent bei torce self imi 35 language training, must be brought under the trainer's control. This is done by presenting a word repeatedly over a 5-second period and re- inforcing the child whenever he says the same word following the model. Often the trainer may use a word frequently spoken by the child as a means of increasing the probability of an imitation; another technique involves increasing the pitch, intonation level, and loudness of the model. Both these techniques are deleted as soon as the child is re- liably imitating the word. Another phenomena of training echolalic children, noted by Risley and Wolf, was the increase of other nonimita- tive speech with the establishment of imitation through differential reinforcement. This speech was decreased by waiting until the child was silent before introducing the imitative model so as not to rein- force self imitative and nonimitative speech with the next stimulus presentation. Once frequent imitation occurred (5 to 6 times per min- ute) extraneous behaviors of arm waving, turning in chair, etc. were extinguished and the presentation of reinforcement and imitative models was made contingent upon attending behavior. The transition from imitation to naming involves changing the trainer's stimulus control from a word model to be imitated to a stim- ulus object and a question ("What is this?"), in the presence of which, a particular verbal response is made. Basically this stimulus control is transferred by a process of fading in new stimuli and fading out verbal imitative prompts. This process is represented by the following steps: A. Vocal imitation: S? "Say 'train'." --------------------------- R1 "Train." Much of the r has used thes Risley and We iry rudimenta Bar fit a severely Speech consis neatly PGrSE :conected res iiSizine pim My One Que: each traim S 3'iige and ca 27555 36 B. Transition from imitation to naming: 53 "What is this, say, 'train'." ----------- R2 "Train." sg "What is this, say, 'trrr.'" ----------- R3 "Train." 52 "What is this, 'tr'." ------------------ R4 "Train." C. Naming: 5g "What is this?" ------------------------ R5 “Train." Much of the recent language acquisition research with retarded children has used these operant behavior modification techniques reported by Risley and Wolf (1967), with adaptations made to fit children without any rudimentary speech. Barton (1970) modified the bizarre and inappropriate speech of a severely retarded institutionalized boy. Though the child's speech consisted of a large clearly articulated vocabulary, he fre- quently perseverated responses or answered questions with random or un- connected responses, with advertising jingles, or with other inappro- priate responses. Behavior targets included the child's responses to magazine pictures and the experimenter's questions: 1. "Who, what, where is (that, doing )?" 2. "What, where (is, are) (he, she, it, they) (doing, having, playing, etc.) (here, there, on, in the )?" 9p. 300). Only one question was asked per picture while 50 pictures were presented each training session. Appropriate responses were reinforced with praise and candy while all inappropriate responses were given a T0: closed and covered the magazine and experimenter turned away from the child for 1C were ignorec. Tr or inappropr tie exclusio' successful tigl tions about p about things reinforCErs 5 in intennitte inforcgnEnt W training Dhas. the ”lack Of n 37 child for 10 seconds. The child's negative responses, “I don't know," were ignored. The results showed that "simply by reinforcing appropriate or inappropriate responses, their numbers can be increased, almost to the exclusion of the unreinforced class (p. 304). Attempts were not successful to get the response class of appropriate answers to ques- tions about pictures to generalize to general conversational questions about things and people not in pictures, Barton suggested that food reinforcers should have been faded out and social reinforcers put on an intermittent schedule prior to generalization testing where no re- V inforcement was given for appropriate responses. Also the five gener- alization tests, given throughout the training phase and following the training phase, did not have the variety of stimuli present during training but used the same 66 questions five times. This, along With the lack of reinforcement, may have made the generalization test so un- like training conditions that the child's newly learned appropriate re- sponses would not generalize. Research is needed to more fully develop Barton's techniques of generalization training. Stremel (l972) trained three trainable and severely retarded children with limited expressive noun vocabularies, in an operant lan- guage training program, to produce the basic grammatical relations, i.e., subject-verb-object responses. Three intermediate training goals were used based on Bloom's developmental data (1970). 1. action verbs and verb-object combinations a. receptive level: performing commands b. expressive level: answering questions ("What are you doing?") 38 c. expanded expressive level: giving verb-object rer I . sponses in order to receive things or events (“Go'bath- room," "Want drink," etc.). 2. Subject-verb combinations a. receptive level: pointing to pictures representing subject-verb combinations (”Point to 'boy run!’"). b. expanded expressive level: answering questions about pictures with subject-verb responses (0: "What doing?" R: "Boy run"). 3. Subject-verb-object combinations a. receptive level: pointing to pictures representing subjectsverb-object combinations ("Point to 'boy sit chair'" . . b. expanded expressive: answering questions about pic- tures with subject-verb-object combinations ("What boy doing?" R: "Boy sit chair?). The trainer always gave the child expanded feedback thereby providing advance exposure to the next response class to be trained. Such an example would be: trainer points to a picture and asks, “What doing?" Child responds: "Boy run." Trainer gives expanded feedback: "Boy run school." Tokens and pictures were used in a novel way to prompt expansion responses: V-O, S-V, and S-V-O. Stremel's expansion of the child's responses made use of tele- graphic speech rather than conversational or adult syntax speech. For example, adult syntax speech would state: "Ihg_boy i§_runnjgg_tg_ school," while the modified form of telegraphic speech would state: "Boy run school." Brown and Bellugi's research (1964) showed that the modified form of telegraphic speech characterized a child's first sen- tences. Miller and Yoder (1972) support Stremel's use of modified tele- graphic speech in their recommendations that adults working with lan- guage retarded children should: 1. re: whi~ 2. 9X2: 3. tali has: most mear Br (7972) state “he to pick the cgnCEDtua “:5“! does fig CORtEnt y l NflathnS. i 39 1. reduce their syntax to telegraphic speech whenever possible while talking to the child; 2. expand and model utterances made by the child; and 3. talk to the child as much as possible about the things that are happening at the moment--'at a moment when he (the child) is most likely to be attending to the cues that can teach that meaning' (Brown and Bellugi, 1964) (p. 302, 1972). Brown and Bellugi (1964) Miller and Yoder (1972) and Stremel (1972) state that this procedure prevents the need for the child to have to pick out from the adult syntax those forms or words which carry the conceptual meaning of the linguistic code, which the normal child capably does on his own. The language retarded child would then have the content words presented directly to him in phrases or two-word com- binations. Kent gt_al., (1972) and the language program used in the present study, the MLAP, also used expanded feedback to the child dur- ing receptive and expressive phases. However, in both programs the feedback is given in adult syntax rather than in the modified telegraph- ic form, since it was believed that adult syntax feedback would be given with more natural expression and rhythm. More research is needed to de- termine which type of expanded feedback is best suited to the severely and profoundly retarded child. Guess (1969) and Guess, Sailor, Rutherford, and Baer (1968) have done some interesting research in training severely retarded per- sons to understand and use plurals. Guess, gt_al,, (1968) demonstrated positive outcomes to the following questions posed by Baer, Guess and Sherman (1972): Can vocal tacting behavior--specifically, labeling-~be given a rule-bound grammatical organization, such that (the) child will use this rule to produce new instances of language which have not been directly trained, not previously modeled, yet exemplify the rule? such 0 behaVl; (p. 96 Aseverely '" :iter "ain'- otject pall‘S specific to ' inforcement C and after the child, the or theprevious was discovere rule were the cording to th The which demonst in; that it i fewel wGD a sin 40 rule? And the question continues: will the rule so produced have such organizing power that it will overturn or preempt any prior behavioral organization to which a label might have been subject? p. 96 . A severely retarded girl was taught generative productive plural usage. After training, she was able to correctly label new single objects and object pairs in the single or plural without further direct training specific to those objects. After the contingencies were reversed (re- inforcement of non-plural responses to object pairs, and vice versa), and after there was a corresponding reversal of the response by the child, the original correct usage of plurals was again recovered when the previous contingencies were re-implemented. During this time it was discovered that several plurals learned according to the reversed rule were then, without any direct training, corrected by the child ac- cording to the normal pluralization rule. These results suggest, as with Baer, gt_gl.'s (1967) research which demonstrated generalization of the "class" of imitative respond- ing that it is feasible to train nonverbal severely retarded persons to develop a single productive language class, the plural morpheme, by im- itation and differential reinforcement. According to Baer, gt_gl. (1972) in additional studies: Our data demonstrate simplistically generative grammatical accom- plishment in the case of adjective-noun phrases, noun suffixes, present tense/past tense usage, and the receptive transitional use of comparative and superlative forms of adjectives. Our subjects uniformly have been drawn from the ranks of institutionalized re- tardates who for many years previously have shown little or no linguistic skill or organization. Thus, Janet's study (Guess, gt_ 61., 1968) proves to be typical and prototypical. As a group, these studies then contribute to the context with which this argu- ment opened: a rule-bound organization of language can be achieved in language-deficient children by techniques no more complicated than ' which (P- 93 tarded subj ' tionally iru ductive acq. sition oppos is a prereo; that recepti ‘.nctionall; needs to be .ow EXDY‘ESS' fireversal i 41 than reinforcement and imitation applied to a sequence of items which are no more than ordinary exemplars of the desired rule p. 99 . In another study by Guess (1969) the performance of two re- tarded subjects indicated that their receptive comprehension was func- tionally independent of their expressive speech in the grammatical pro- ductive acquisition of the plural morpheme. This study suggests a po- sition opposite to the traditional (i.e., that receptive comprehension is a prerequisite to expressive speech) in its implicated conclusions that receptive language and expressive speech could be two separate and functionally independent classes of behavior. However, more study needs to be done with larger groups and using research designs that al- low expressive training prior to receptive training rather than using a reversal design. Language Training Programs for Lower FunctioninggRetardates Although a wide range of operant language training programs exist presently for retarded children only aspects of some recent ones will be reviewed (Bricker and Bricker, 1970; Gray, 1970; Gray and Ryan, 1973; Kent, 1972; McCarthy, 1969; McCubrey, 1971; Sailor, Guess, and Beer, 1973). An eighth program, the Language Acquisition Program, (Kent, Klein, Falk, and Guenther, 1972) upon which the present study's Modified Language Acquisition Program is heavily based, will be covered in more detail. Different researched applications of Gray's (1970) and Gray and Ryan's (l973) comprehensive and rather complex language program have been 5 children (3 deal' and ha"_ Gray and Rye siohais VerS the Percent ‘ Der ch“d pet the aCC“racy iessiona's' ' ;ort60 for DE aDDear when a me than Om Ba: based One of the process ‘ ing steps ‘50 grained step serial langUi th- ‘ .cse With C! ositi onss 6U Va .4 content 1 '42 have been successfully done with populations of young language delayed children (3 to 4 years delay), trainable and severely retarded children, deaf and hard of hearing students and bilingual children. In addition Gray and Ryan (1973) collected data on the effectiveness of paraprofes- sionals versus professionals trained in using the program, in terms of the percent of correct child responses and the total number of responses per child per session. The results indicated that there was no drop in the accuracy of the training when done by paraprofessionals but the pro- fessionals' responses per child were 28 percent higher than that re- ported for paraprofessionals. However, this difference tended to dis- appear when aides were working with the program as often as teachers-- more than once per week. Basically the program's delivery system was a behaviorally based one of "programmed conditioning" which incorporates branching as the process whereby the teacher used additional instructions and branch- ing steps to help a child who was having difficulty on a particular pro- grammed step. The content of the training is based on the sequence of normal language acquisition. Not only are content words taught (i.e., those with concrete referents) but also function words (articles, prep- ositions, auxillary verbs, etc.) which provide grammatical context for the content words. The child is trained at appropriate levels in the program based on the results of the Programmed Conditioning for Language Test (PCLT). This test operates on the premise that when asked to repeat a sentence, a child will use only those constructions which he himself has in his language structure. Thus, Ithel . say [bi test i. poses. Gray and Rye. the language ception of t if children a test when in lirited. Aft used as the S SSE-cific Will be taught. T SFESS ln trdl Mod Bad to prom; iron, the 11tide Sewing Other used in this reamy On a a. . ..n S PFOgra :K‘i Child Lining table J firi raise yo 3,513". 1778i 3"], SQrve 43 Thus, if you asked a child who did not have "is" or the article 'the' to repeat the sentence 'the ball is on the table' he would say 'ball on table.‘ Although this is not a formal, standardized test it has been found to be a valid and useful tool for our pur- poses. (Gray, 1970, p. 115.) Gray and Ryan report that the PCLT has accurately placed children in the language curriculum over the past eight years of use with the ex- ception of the echolalic and the bilingual child. Both of these types of children are so skilled in imitation that they may score high on the test when in reality their daily use of the language forms is quite limited. After the PCLT is given the first tested program failed is used as the starting point for training. Before training begins, a specific criterion test is given on the particular grammatical form to be taught. The criterion tests are re-administered to evaluate pro~ gress in training. Modeling of the correct response is a procedure frequently used to prompt the child in making the Correct response. Group train- ing is recommended as a means to avoid overtraining to allow benefit from the modeling hypothesis (Bandura, 1969): children learn from ob- serving other children. Tokens paired with social praise are commonly used in this program. Also a behavior management system which relies heavily on a token economy has been used with the program. Gray and Ryan's program seems to be geared to children with some imitative skills (child can imitate vocal prompts), attending (child can sit at training table), and receptive understanding skills (child understands "Are you ready to work?" "Sit at your desk," or "When you are ready to work, raise your hand," etc.). For these reasons, this particular pro- gram may serve as an excellent supplement to others (Bricker and Bricker, 1970; Kent: with CF‘is bet and CA and d institutions indll livec tal period w treatment, a til period ( sessions an< xhwfl sort 6 rd ce fl‘OiTl G :rocedu res fmctional . 44 1970; Kent, 1972; Kent, et_al., 1972; Sailor, Guess and Beer, 1973), which begin at the more elementary levels of training the prerequisite skills of attending and motor and vocal imitation. The next three studies were done in institutional settings. The first two dealt with the language training of mongoloid subjects (Buddenhagen, 1971; MacCubrey, 1971) while the third used attendants to train severely and profoundly retarded infants and young children (McCarthy, 93431., 1969). In MacCubrey's study, eighteen trainable mongoloid subjects with CAs between 4-6 and 7-10 and MAs of two years were matched for MA and CA and divided into three groups. All were drawn from one of two institutions and possessed some ability to produce speech. Groups I and II lived in a separate research facility for the 7-week experimen- tal period while Group III remained in their institution, without treatment, as institutional controls. During the seven-week experimen- tal period Group I received verbal operant conditioning in daily group sessions and Group II received daily activities of a traditional pre- school sort (for children of their MA) without conditioning and sepa- rate from Group I. Verbal operant conditioning sessions included the procedures of shaping and fading-in of nonverbal prompts to elicit functional speech used for object and picture naming, descriptions of action pictures in word combinations, and discrimination of colors and polar opposites. Pre- and post-experimental evaluations were made with two language measures, an IQ measure (Stanford-Binet) and a social rating scale. Although significant positive changes were noted for Group I on the Star interpreted statistical teacher and Med for re 8. nonverbal ing ‘Dr training 45 on the Stanford-Binet and the language measure, these results must be interpreted with caution due to methodological problems of design and statistical analyses used by MacCubrey (i.e., confounding of group, teacher and institution with treatment, insufficient matching substi- tuted for random assignment, etc.). Buddenhagen (1971) gave intensive individual training to four nonverbal institutionalized mongoloid children. The general strategies for training involved: 1. phonetic analysis or assessment of the S's entry behavior to determine which components of the S's extant repertoire could serve as a basis for a beginning and lead most rap- idly to viable speech; 2. providing social and material reinforcement for successful responses; and 3. providing a context which made the S's responses appro- priate in terms of meaning or content (p. 34). All four children exhibited a variety of behavior problems (fighting, stereotypic behaviors, masterbating, etc.); these behaviors were gen- erally treated by positively reinforcing appropriate behaviors which were incompatible with them, occasionally time-outs were used. CA ranged from 7 to 13 while all four subjects functioned in the severely retarded range. Since the entry behavior of each child was quite different and all training was done individually, Buddenhagen describes the proj- ect as four separate experiments in verbal conditioning. Generally though, Buddenhagen tried to establish in the subjects a verbal reper- toire that would have great utility in the cottages in which the sub- jects resided. Mands, such as "I wanna drink," "I want more," etc., were consid wards becau the request; verbal beha- whoMyW* Mlstimuli four reports hwever, the, experimental research with l. A chi propr fectei 2. Hongo' from s 3. Age cg from c . In orc skills ally c 0f the - Aversi duce t aDplau orcem - Social tdln t] - Use of Verbal ' IUtEFVQ should 46 were considered to have greater probability of being maintained on the wards because they provided pleasant consequences for the resident When the requests were granted and were independent of specific antecedent verbal behavior by the attendants (i.e., "fine," or "O.K.," are valu- able only when uttered in appropriate contexts of the antecedent ver- bal stimuli of "How are you?" or other similar questions). Each of the four reports incorporates the basic operant procedures already reviewed; however, they are valuable for their detailed accounts within an applied experimental setting. Buddenhagen's conclusions tend to reinforce prior research with nonverbal retardates: l. A child's history of nonstalking can readily be reversed if ap- propriate changes in the child's human environment are ef- fected . . . (p. 160); 2. Mongolism need not decrease the child's potential for profiting from speech training; 3. Age per se does not seem to be related to the amount of success from operant speech training programs; 4. In order for shaping of vocal responses to occur attending skills must be present in the subject. Visual attending gener- ally guarantees auditory attending and permits the observation of the visual aspects of vocalizing; 5. Aversive controls administered by the experimenter greatly re- duce the reinforcement value of experimenter-produced praise, applause, and similar manifestations of approval (social rein- forcement, p. 163); 6. Social reinforcement alone is generally insufficient to main- tain the trained vocal performances of nonverbal children; 7. Use of vocal imitation to establish vocal repertoires in non- verbal children is an effective and efficient technique; and it. Interverbal chaining or linking together of dissimilar syllables should be done prior to bringing the single syllable or repli— cated syllable response to a high strength. M: train and W tarded infar and fine ”10t principles 0‘ wards goal be sponse. Usin were instruct oi informally lined by Gese nept daily n: with Treatmer 47 McCarthy, gt_gl, (1968) used institutional attendants to train and informally assess progress of 55 severely and profoundly re- tarded infants in the developmental areas of language, social, gross and fine motor skills. Training procedures incorporated behavioral principles of positive reinforcement of successive approximations to- wards goal behaviors and prompting to initially encourage a desired re- sponse. Using the children they worked with.dai1y, ward attendants were instructed by psychologists and special educators in the methods of informally testing and assessing the four developmental areas out- lined by Gesell. Attendants were provided with training materials, kept daily notes on the progress of those they trained and were equipped with Treatment Card Decks which outlined l. the developmental task, 2. a test of the ability to perform that task, 3. the normative age level at which successful performance of that task could be expected, and 4. a suggested means of teaching that test task. In the language area the developmental tasks taught by the attendants ranged in normative age from two to 48 months. This area included tasks in a serial order such as: evidencing a response to another's voice, selected responsiveness to vocal output (e.g., one's own name), vocal imitation, following verbal directions, evidencing the meaningful use of words, naming objects, etc. Results at the end of one year showed significant gain (p<.05) in the treatment group over the group receiving regular institutional care. This gain was still significant in the remaining 30 subjects at the end of training in for two rea which, ever. year only), tioning inf; sane staff-r Br grain for the iFlDOFtance of “19. Recent stitutionaliz recePtion thr Hearingi‘ul he 1ricult to ad: the cth to 48 the end of three years. Generally the most consistent gains came with training in the language development area. These results are important for two reasons: paraprofessionals were able to provide a treatment which, even when applied directly on the ward setting (during the third year only), resulted in significant developmental gains in low func- tioning infants; in addition, this extra training was done with the same staff-resident ratio as was present during the control periods. Bricker and Bricker (1970) described an operant language pro- gram for the severely language-handicapped child which emphasized the importance of establishing hearing levels of children prior to train- ing. Recent surveys estimate that approximately 20 percent of the in- stitutionalized retarded population exhibit hearing losses for speech reception threshold of twenty decibels or more (Siegenthaler, 1970). Meaningful hearing tests given in the standard ways are generally dif- ficult to administer to severely retarded speech deficient children. Therefore, operant audiometry procedures are implemented by training the child to perform the skills sequentially listed below: 1. behavior control; 2. establishment of a button pressing response to sound/light stimuli; 3. establishment of behaviors needed to cooperatively wear head phones; 4. reestablishment of the button press response to sound/light stimuli with the sound coming from the headphones; 5. fading out the light stimuli (establishing tone control at 100 Hz. or greater if responding becomes unreliable); ihis operant tpa onia i in pushing A the Process bi-siep pro and trainir fill sentenc S and C0Wren deficient c ’Etby 0the Kent's (197; i- A cc Pre- thre demc 2. Rese or v ness 3' A be taki, leve' 4' A la! the ( OOOtr 49 6. generalization of tone control across frequencies (250 through 6000 Hz); 7. generalization of tone control across intensities (by changes of 5 to 10 st); and 8. standard pure-tone audiometric assessment. This operant training allows a standard audiometric test to be given to a child incapable of making a verbal response but capable of a but- ton pushing response, contingent on sound stimuli from earphones. A more recent language program by Bricker (1972), still in the process of completion, has 24 subprograms each with complex step- by-step procedures. The program begins with audiometric evaluation, and training of attending and imitative skills and extends to meaning- ful sentence production. Sailor, Guess, and Baer (1973) have developed an impressive and comprehensive language training program for retarded and/or speech deficient children which has four main targets of emphasis unmet as yet by other language programs, except perhaps Bricker's (1972) and Kent's (1972). These are. l. A comparison of various operant techniques of training a single pre-language or early language skill (i.e., vocal imitation has three main training sequences which have all been successfully demonstrated), 2. Research as to whether any of the children's preexisting vocal or verbal characteristics predict their differential responsive- ness to these techniques, 3. A behavioral language program which has detailed sequences for taking the nonverbal, nonimitative child to a thoroughly useful level of language competence, and 4. A language program which contains elements designed to enhance the child's transfer or generalization of language behaviors to nontraining situations (Sailor, 1972). lhe prograf (labeling). (handing fell by asking " (includes t control, an: four dinens: sons and thi tent in the color, size, Ke iii?) (1972) similar to S r‘ecelltiVe an Counting and introdUCed a acritEFiOn Th El‘ State p as: 50 The program is organized into four dimensions of language: reference (labeling), control (manding or requesting), self-extended control (manding further instruction in the case of specific unknown referents by asking "What's that?" or "What are you doing?"), and integration (includes the act of combining the language skills taught in reference, control, and self-extended control). Content taught in each of these four dimensions extends from persons and things and actions with per- sons and things, into extension of a basic verbal repertoire with con- tent in the verbal skills of understanding and expressing possession, color, size, and relation. Kent's latest revision of the Language Acquisition Program (LAP) (1972) is based on extensive research and appears to have targets similar to Sailor, gt_gl,'s (1972). The revised LAP has added advanced receptive and expressive levels of training which include aspects of counting and color and shape sorting and naming. In addition, Kent has introduced a noncriterion training procedure where trainees are given advanced exposure to new tasks in later parts of the program on a non- criterion basis. This has been found to facilitate later training to a criterion level of performance. The original LAP (Kent, gt_gl., 1972) was developed at Fort Custer State Home and Training School in Michigan. The LAP also in- cludes a preverbal and verbal section. The preverbal section empha- sized the acquisition of attending behavior, motor imitation, and vocal imitation, while the verbal section trains the receptive linguistic repertoire and a prelinguistic expressive repertoire. The program is further organized into phases and each phase into parts. Initial imentories prior to tr until he me peses. l' atobjects 51 inventories are given to the child by the experimenter or the attendant prior to training each part. The child is then trained on that part until he meets criterion on the final inventory. The attending phase is prerequisite to training in all other phases. It consists of teaching the child to remain seated, to look at objects on command, and to establish eye contact with the trainer. After completion of the attending phase, motor imitation training begins. During this phase the child learns to imitate spe- cific manipulations of objects and movements of body parts. The child may then proceed to the vocal imitation phase, or simultaneously receive this training plus basic receptive training. This is because no vocal responses are needed in the receptive phases. Vocal imitation tasks range from merely vocalizing when the trainer vocalizes to vowel, word, and phrase imitation. There is no final cri- terion for this phase since the trainer continues to expand and refine the vocal skills of the child through vocal imitation training while also giving training in the verbal section of the program. The verbal section consists of four phases: basic receptive, receptive expansion, basic expressive, and expressive expansion. There are matching parts in each phase which represent what are assumed to be hierarchies of linguistic or prelinguistic complexity. The child is trained in progressively more complex ways to respond to familiar ob- jects, to verbal stimuli which refer to objects not immediately pres- ent, to certain body parts, to room parts, and to action words. In the basic receptive phase, the child is trained to respond to a number of simple verbal commands such as: "Touch the shoe,“ "Show he the tab‘ appropriat [. by the chil tions which child would Md "louch t Du tative abili 450' reSponse The child 15 aCthns rECE tsight to re DOints to ar perfoms an In resPond VErl and I"Elatio respond app the t"a1lle’r .‘4 . Jame" has I“. ll Ti 52 me the table," "Show me your eye," and the child also learns to respond appropriately to action words such as "Jump," "Sit." During the receptive expansion phase, the key words learned by the child in the basic receptive phase are reused in verbal direc- tions which involve two key words rather than one. For example, the child would be trained to respond to: "Put the bgll_on the tgblg," and "Touch the baby's nose." During the basic expressive phase, for which some vocal imi- tative abilities are required, the child's response changes from a mo- tor response of touching, pointing, or doing, to a vocal word response. The child learns to name the same objects, room parts, body parts, and actions receptively learned in the basic receptive phase. The child is taught to respond to questions such as, "What is this?" as the trainer points to an object, or the question "What am I doing?" as the trainer performs an action. In the expressive expansion phase, the child is trained to respond verbally to more complex questions using the same key words and relationships as in previous phases. For example, he learns to respond appropriately to questions such as: "Whose hair is this?" as the trainer points to the child's or a doll's hair; and, after the trainer has placed a particular object in a particular place, the ques- tion "Where is the shoe?" is asked of the child. Throughout training, the techniques of shaping, physical and imitative prompting, and fading are used by the trainer to evoke from the child the criterion response levels needed for each part and phase of the program. Shaping procedures for new responses do not exceed 40 trials in a anew beha‘. the first r be shaped w as well as A; rined prior attending 5'. reinforcemer in the MLAP, token but on '19-'11 to rein at the end c 0r tOy play As is 1972) a dbie 0n 1911C 53 trials in any one session. When two correct consecutive responses on a new behavior occurs, another new item is trained and alternated with the first response shaped. This alternation of each new response to be shaped with learned responses allows some overlearning of responses as well as the intermittent programming of success items. Appropriate concrete reinforcers for each child are deter- mined prior to initial testing. Generally, once a child has learned attending skills, he receives token training and is placed on a token reinforcement system. The particular token system used in the LAP, as in the MLAP, is one which does not assign a particular value to each token but one in which tokens are freely given with social reinforce- ment to reinforce appropriate responses from the child and exchanged at the end of each training session for a single choice of food, drink or toy play (music box, talking pull toy, etc.). As with the other programs reviewed, both the LAP (Kent, gt .31., 1972) and the revised LAP (Kent, 1972) use behavioral principles to teach a small-stepped sequence of skills taken from the data avail- able on language development in the normal child. Usigg Peers as Helpers, Behavioral Engineers, and Tutors in Classrooms and Institutions Utilization of Normgl Peers as Change Agents for Special Education Students The "helper" therapy principle of using peers to assist other peers in improving academic skills or modifying deviant behaviors is an old practice which is still applied informally and in more scientific ways in public school settings (Gartner, Kohler, and Reissman, 1971; Reissman, . use of reg-.1 younger chi ‘Jlrick, anc as for stuc 1972; Davis Datterson, studies repc ing as tutor tion classrc W: the phys Dc 9511 using r, ior the Fete Extensive f 0f the 0the 1- Tea 335 1de 2. The tie Cle 3- Stu Str ' Tea 1P7; del 5. The bel 117:; . SLL aSe 54 Reissman, 1965). There are numerous recent examples of the successful use of regular class students as behavioral engineers and tutors for younger children with behavioral and/or academic problems (Surratt, Ulrick, and Hawkins, 1969; Willis, Crowder, and Morris, 1972) as well as for students within the same classroom (Conlon, Hall, and Henley, 1972; Davis, 1972; Harris, Sherman, Henderson, and Harris, 1972; Patterson, Shaw, and Ebner, 1969). In addition, a large number of studies report the effectiveness of regular classroom students assist- ing as tutors or behavioral engineers with students in special educa- tion classrooms for the mentally retarded, the emotionally disturbed, and the physically handicapped (Csapo, 1971; Doorlag, 1970). Doorlag (1970) systematically examined all programs in Michi- gan using regular class students as assistants to educational programs for the retarded, the disturbed and the physically handicapped. His extensive findings are joined below with the conclusions and findings of the other studies reviewed. 1. Teachers who have worked with regular students as classroom assistants were highly supportive of the student assistant idea. 2. The student assistants were utilized primarily in the instruc- tional portions of the classroom program rather than in the clerical and supervisory aspects. 3. Student assistants were judged to be capable of performing in- structional tasks in the classroom. 4. Teachers generally believed that the use of student assistants improved the educational program for handicapped, academically delayed, and disruptive students. 5. The students utilized as behavioral engineers or tutors were believed as well as shown to have made personal and/or academic improvements themselves. 6. Students used as behavioral engineers or behavioral change agents were reported not to have acquired through contagion the inappropriate behaviors of the students they modified. 55 7. Student assistants used as behavior change agents can produce behavioral changes in disruptive regular class students and in emotionally distorbed students in special education classes. 8. Student assistants used as tutors with academically delayed students were judged or shown to have a positive effect upon the academic performance and attitudes of students tutored. 9. The administrative problems encountered in conducting student assistant or student behavioral engineer programs were not re— garded as significant or extensive enough to discourage the development of problems to the teacher of the assistants them- selves. 10. The student assistant programs were judged to be effective in all special educational classrooms examined regardless of the handicap the students had, their academic level, or the aca- demic level of the student assistants used. 11. Service as a student assistant was reported to facilitate the recruitment of potential teachers into the special education field. Utilization of Retarded Persons as Helpers, Substitute "Mothers" Trainers, and Behav- ioral Engineers of Other Retardates in ~ Classrooms and Institutions Retarded persons have also been used in a variety of "helper" roles in institutional settings and in special education classrooms. The usual role in institutions has been as assistants to attendants in resident care and housekeeping tasks (Cleland and Swartz, 1970; Ayllon and Azrin, 1968). Less often the retarded have served other retar- dates in the instructional roles of ”substitute" mothers (Houglan, 1963; McKinney and Keele, 1963), behavioral engineers (Kazdin, 1971; Gelfand, Gelfand, and Dobson, 1967; Whalen and Henker, 1969, 1970), or as trainers of self care skills (Dilley, 1969), language skills (Jeffery, 1972; Whalen and Henker, 1969, 1970), and reading and writing skills (Brown, Fenrick, Klemme, 1971). A recent survey (Cleland, Swartz, Drew, and Talkington, 1972) of attendant attitudes towards working retarded residents (student helpers, S.- Siis worked cantly lar; attendants 56 helpers, SH) in an institution revealed some response trends. These SHs worked in resident care and housekeeping tasks only. A signifi- cantly larger proportion of the 46 attendant responses indicated that attendants on both the SH's home ward and their work ward. l. regarded SHs as less aware of the needs of the children than they were, 2. reported that SHs were less apt to discuss with them the nonambulatory residents they worked with, 3. reported that conversational topics the SHs discussed were: a. least often: "wanting to do a better job of caring," b. more often: "wanting a different job,‘l "difficulty in helping patients," or "messiness of patients," c. most often: ”when do I get paid?", and 4. indicated that SHs, if it were possible, would not be used by them as babysitters. Generally, the survey indicated high agreement on questions about SHs from attendants on home ward and work ward. But whenever a response dissonance did occur on individual SHs, the work ward attendant was the person most often sought out by the SH. These results suggest that a more responsive relationship, or one of differential reinforcement, exists between SHs and their work- dorm attendants than their home-dorm attendants. Conceivably this is a function of the SH feeling more important, and meaningfully participating, on the work-dorm than on the home-dorm setting p. 34 . An additional factor influencing this relationship was that the SHs were paid only by the attendants on the work ward and never by the home—ward attendants. The authors suggest from the results that al- though a comradeship may develop between the attendant and the SH, there is a For e and S? tarda. atten: move t: tween Th1 residents 1' housekeepiri level of he that have t HlWEQr h! ”tarded D dearer jO Vigion fad Ir “Omen WEr‘e bCys fOUr crEESing ‘ rOutjne C 1.nitiatln unldren C C: lic tors O +:e nt a‘id 57 there is always the interferring factor of job competition: For example, pessimistic attitudes may arise because of attendant and SH job similarity; hence, if attendant attitudes toward re- tardates are to be changed, removal of the SH5 from a quasi- attendant (and perhaps competitive) role into different jobs is a move that could excise a persistent barrier to communication be- tween professional and attendant staff (p. 34). This survey points out the difficulties of placing retarded residents into the unstructured helper positions of resident care and housekeeping duties, i.e., resident-attendant competition and a high level of negative job attitudes held by residents. Other work roles that have been reported in the literature do not seem to evoke compet- itive or negative feelings but rather more therapeutic benefits for the retarded person. Perhaps this is because each study or project had clearer job descriptions, prework training sessions, initial job super- vision fading into support--supervision, and reinforcement for success. In McKinney and Keele's study, educable and trainable adult women were given general instructions to "mother" two severely retarded boys four hours per day for one month. "Mothering" consisted of in- creasing physical attention given to the boys by assisting in their routine care, playing with them, teaching new skills and words, and initiating physical contact. After 200 hours of such treatment, the children demonstrated significant increases over a control group in factors of purposeful and verbal behavior and significant decreases in factors of social behavior and random activity. Another similar but nonexperimental project reported by Houglan (l963), instructed high-functioning retarded women and men to be pat tient aides in an institution for the retarded. Houglan's project gave specific e well as ir classroom - status bee were used Attendants Problems of but it seer: ties as wel acted to 5U In tarded fema fiEn and to SleS. Th lEterded re in Se” car alga; bEha outward 58p Ser other Fol e StUdy (Self. patlEnts, i' +' .lentg. The 58 specific assistance in caring for the young and bedridden residents as well as in bedmaking and dishwashing. The retarded aides received classroom instruction in their tasks and earned coffee breaks and some status because of their new roles. In addition, some resident aides were used to teach new groups of aides the skills of bedmaking, etc. Attendants were carefully selected to supervise and train the residents. Problems of job competitiveness with other attendants occurred initially but it seemed that the specific definition of patient training aide du- ties as well as the pretraining and follow-up supervision each received acted to subside complaints and promote the acceptance of the aides. In another nonexperimental project, Dilley (l969) taught re- tarded female residents to be "big sisters" to profoundly retarded wo- men and to provide instruction in toileting, dressing, and eating skills. The results indicated to the author that mildly and moderately retarded residents can successfully train lower functioning residents in self care skills. Dilley reported that the "big sisters" benefitted also: behavior problems decreased and appropriate social behaviors and outward aspects of self concept increased. Serving as behavior therapists, modifiers, or engineers is an- other role that has been filled by retardates. In a ward observation study (Gelfand, gt_al,, 1967) in a hospital for retarded and psychotic patients, it was found that the patients themselves, over nurses and attendants, were the best unprogrammed behavioral engineers towards pa- tients. That is, patients on their own ignored more often the inappro- priate psychotic behaviors of other patients, i.e., 79 percent of the time in comparison to 64 percent for attendants and 39 percent for nurses; and tient behav‘l otten (68 pe Kc adalts to se workshop set each client , verbal reper. in addition, ‘30“- Talkir tion. Anotne Elld ”ECk, was P'aiSed him 1 self was gm itcse he tal| Sineers lathe ‘ 1. Rein. 59 nurses; and, the patients also did well in reinforcing prosocial pa— tient behaviors (56 percent of the time) though nurses reinforced more often (68 percent) and attendants less often (44 percent). Kazdin (1971) went a step further and programmed retarded adults to serve as behavioral engineers in an institutional sheltered workshop setting. A token reinforcement system was utilized in which each client was assigned a particular color. Two clients with low verbal repertoires were given the color-coded tokens daily and in- structed to reinforce each other for talking to clients during the day. In addition, other clients gave them tokens for engaging in conversa- tion. Talking with others increased with this contingency in opera- tion. Another client, with self-controlled poor posture of the head and neck, was given tokens daily to reinforce his peers whenever they praised him for good posture. A fourth client who talked only to him- self was given tokens by other clients for talking and gave tokens to those he talked with. Kazdin found that using peers as behavioral en- gineers rather than only staff seemed to offer the following advan- tages: l. Reinforcement could be given more often; 2. When reinforcers were given only by staff, the staff came to serve as the only discriminative stimuli for the increase or decrease of the target behavior, while peer-administered con- tingencies tended to generalize the target behaviors throughout the workshop and across the workday; 3. Using clients as behavioral engineers appears to be rehabilita- tive and therapeutic to clients themselves in that they gain a greater awareness of desirable behaviors; 4. By giving clients a greater role in the social milieu of the institution the authors felt that some of the negative effects of institutionalization might be undone; and 5. The staff had more time to effectively supervise. In mined WC with W'Ve, correct r85: correCt res; and teacher following 59 1, Two 1 2. These othe 3, The l readl 4, Four sion5 5, Pracl with ing. lraining proc tion cot’malldS prompting of sheets, total ing totals or heeled by ti the procedure 60 In classrooms for the trainable mentally retarded, students have been effectively used as trainers of their peers. Jeffery (1972) trained two students in two 20-minute sessions to review picture cards with low-verbal students. They were taught to praise the student for correct responses while repeating with prompts the cards on which in- correct responses were made. In Brown, gt_al,'s (1971) study a teacher and teacher aide in a trainable classroom taught students using the following sequence of instruction to be reading and writing trainers. 1. Two students were taught to read two different lists of words. 2. These students were instructed in a specific procedure to train others in reading the words each had learned. 3. The two readers formed a team to train other classmates in reading the words they had learned. 4. Four students worked in pairs to conduct group word review ses— Sions: 5. Practice in writing words students could read was done in pairs with peers giving each other feedback on quality of handwrit- ing. Training procedures used by students included specific verbal presenta- tion commands, reinforcing feedback for correct responses, imitative prompting of incorrect responses, recording of all responses onto data sheets, totalling of correct responses at end of sessions, and enter- ing totals on progress charts. These training procedures were slowly modeled by the teacher aide. Then, as the trainers tried implementing the procedures, the aide's verbal guidance was slowly faded out, being completely eliminated by the end of the fifth, seventh, or sixteenth session depending on the type of training. Both group and individual training was done by the retarded trainers. Training results revealed student learning gains at every level of instruction as well as a no- ticeable increase in the level of motivation for learning in the class- room . As o'risticate:I ant and imi functioning 3 following 5- anean CA 0' lild. They v tending skil ing 0f Simpl WP“, anc a.'r3.'3l‘0Driate Strategy, tr dent's Spom Child lniti: The years and a been heard ter a 15 mi The the game Se “‘5 film the O tES. NEXt‘ trainlng~D' boring the Cedwes le. C ing the Pl; 61 A series of studies by Whalen and Henker (l969, T970) have so- phisticated this use of retarded trainers by teaching them to use oper- ant and imitative techniques as the method to train nonverbal, low- functioning children in attending, in motor and vocal imitation, and in following simple vocal directions. The five trainers, all females with a mean CA of T9 and mean IQ of 55, were trained over a 4 l/2 month per- iod. They were taught to decrease disruptive behaviors, to train at- tending skills, motor imitation, vocal imitation, receptive understand- ing of simple commands, to use techniques of prompting and fading of prompts, and to give tactual, verbal and tangible praise contingent on appropriate responding. Rather than use a standard task sequencing strategy, trainers were taught to attend to and incorporate the stu- dent's spontaneous behaviors. Therefore, tasks were changed when the child initiated other appropriate responses. The ten ambulatory trainees or students had a mean CA of 7 years and a mean IQ of 30. Final selection included only those who had been heard to make sounds and who were able to imitate one response af- ter a TS minute training session with the experimenter. The design used a matching procedure whereby two trainees of the same sex and almost equal CA and IQ were randomly paired with each of four trainers and one trainer was nonrandomly paired with two train- ees. Next, one trainee in each pair was randomly assigned to a training-play condition and the other to the play-training condition. During the training condition all the reinforcement and modeling pro- cedures learned by the trainer were used with the child. However, dur- ing the play condition, which acted as a control period, trainers were instructe: noncontin; 60 sessior sions of t' reversed b__ We trainers. Ceoonstrati WétClled and 0f the trai and the Big dOTle thT‘Oug Praise WEre the eXPErim a few trial alSo ObSErv Tra P one trainer Selation i a to: 62 instructed to play with, not teach their child and to give food on a noncontingent basis. During Phase I the training-play condition lasted 60 sessions. This was followed by Phase II which consisted of 60 ses- sions of the play-training condition where each child's treatment was reversed by his trainer. Whalen and Henker used various methods to teach the retarded trainers. Much verbal instruction, some peer modeling and occasional demonstrations were used by experimental staff. Often as a trainer from a former pilot study worked with the new trainer, the experimenter watched and provided a running explanation. Depending on the skills of the trainer the experimenter withdrew physically between the third and the eighteenth session to an observation room where viewing was done through a one-way mirror and needed instructions and trainer praise were given over an intercom. When new procedures were necessary, the experimenter reentered the training room, gave the instructions for a few trials, and returned to the observation room. Play sessions were also observed from this room and trainers were reinforced over the in- tercom for playing rather than teaching the child. Trainers were encouraged to observe other trainers at work as a means of additional instruction through modeling. Task order was loosely controlled by the stimulus materials provided for each session. One trainer, who could read, used a foot pedal-controlled slide projec- tor with slides of commands to be given to guide the order of task pre- sentation. In addition to staff praise, trainers were paid for their jobs and allowed to have coffee breaks with experimental staff. Du versed for play condi' hon. In A reduced so their other' Eac of the spec but childr Willy more in the play Slignificant] 24 Percent 1 Hhal l. ”her 2. All ding PFOn in ( Trai Crin bd E 4- Trgi dlSC may SEn 5. Thet 63 During the second phase of 60 sessions the conditions were re- versed for trainees with trainers teaching the child formerly in the play condition and playing with the child formerly in a teaching condi- tion. In this phase staff demonstration and feedback were drastically reduced so as to test trainers' ability to generalize their skills to their other child. Each child was measured on a Social Behavior Test consisting of the specific training items and generalization items. It was found that children who participated in the training group improved signifi- cantly more on the Social Behavior Test than did those who participated in the play control group. However, even the Phase I play condition significantly increased responsiveness as measured on the test for the five trainees in that condition. A 6-month retest revealed an average 24 percent loss in acquired repertoire at posttesting. Whalen and Henker drew a variety of conclusions from the study regarding retarded trainer effectiveness: 1. When not consistently supervised they are inefficient trainers; 2. All trainers had difficulties with the perceptual-motor coor- dination needed to manipulate training materials and provide prompts and reinforcement to the child, which often resulted in delayed reinforcement; 3. Trainers often verbalized too much, embedding the intended dis- criminative stimuli commands in the midst of their extra ver- bage; 4. Trainers tended to present difficult tasks too early and to discontinue other tasks before learning was stablized. This _ may be the reason that only 33 to 78 percent of all taSks pre- sented were learned by the children; and 5. The best method of trainer instruction was the brief but re- peated presession reviews. Beyond their learning of specific training procedures, the authors ob- served improvements in the trainers' general adaptive behaviors. ‘hat: 64 Summary Relevant to the PUrposes of the Study that: The research reviewed seems to conclude with little dispute . the difficulties of utilizing attendants in resident- training programs in institutions center around: a. low attendant-resident ratios, b. little or no provision of relevant specialized training for attendant staff,and c. a general absence of administrative reinforcement pro- vided to attendants for resident training behaviors; . if resident training programs are to have lasting effective- ness they must include attendants as instructors; . to equip attendants with the skills to instruct retarded residents, an inservice training program is needed which not only begins at the attendant's level of knowledge, but also involves active responding by the attendant, shaping procedures employed throughout training, opportunities to generalize learnings to actual training situations, and a follow-up program of supportive reinforcement and shaping procedures; . institutional settings for the retarded generally retard even further the cognitive and social growth of its resi- dents; delays in language development are an aspect of this retardation; 10. ll. 65 . language may be regarded as verbal behavior which is subject to the same principles as any other behavior; . the mentally retarded child is delayed in his speech develop- ment but appears to follow the same sequence of development as the normal child; . imitative training is an important aspect of operant lan- guage training in retardates; . commonly used, effective procedures to control disruptive behaviors of retardates occurring during language training are extinction and time out from positive reinforcement; . the majority of recent language training research with the mentally retarded incorporates the sequence of normal lan- guage acquisition as its content within a framework of be- havioral principles; consistently supervised retardates can be effectively uti- lized with other retardates as substitute "mothers" behav- ioral engineers, and reading, writing and language acquisi- tion trainers in classrooms and institutional settings; and retardates in trainer roles appear to make gains in general adaptive behaviors and social interaction skills. orstl lard: Deer: seek: and i can t (level le‘lnf CHAPTER II METHODOLOGY Purpose This study is applied research, the purpose of which is to dem- onstrate and test the effectiveness of a strategy for utilizing re- tarded residents of a residential institution to train their retarded peers under the guidance of the attendant staff. Specifically, it seeks to determine whether significant gains in language functioning and in the ability to respond to a standard measure of intelligence can be produced by retarded trainers who are trained to administer a developmentally ordered language program incorporating principles of reinforcement and modeling. D_es_i.os mcwxaou oo :o mm mm "cow“ -muwsw eczom szo> venom xcm maexes x3 uczom m.emcweeu mcwxqou use?» -muwew _muo> mmoeo .upw eweco m on pcwom co pm; a use xUPpm _euwuem> a co mew; m use apnep mcwuczoa guage: .e .m N .— "muumnno saw: mpcmsm>oe mcwxaou N ow ow Apumpcou wxm Fewepimgav cm>wm we ucaesou comm «Lemma emceegu um mewaOA .u mgzumwm can uceEEou co mpumnno pm mcwxooJ d Ammmcoam -mg Lopes mpnwueaeoucw eo :owu -ecwewpmv mcwcwmeu cue: «Lowemp -:_ peep mucmsm>oe mo u_e mcwuuma .n 8.5.5 .e .wcwppao Eegmoee cowpwm_:co< mmeamceb cwwmwnoz .m mezmwe o 88 e E 3 # e e 2 8 A 61 lms E 8 m S o... J - xmx .Fpen .neou xmx .P_mn .xamn .mogm xmx .Fpmn .anu .anu .xaen .mocm .cooam .Lmu .znmn .mosm .cooam .cooam .Lmu .ae; .um; “pumnno .Lmo .um; "pumwno .1? "Ampnwmw> uocv mcwmmws e mergez .m umpemucou e mewsmz .m pomwno cm mcwucwm .m _Pen we» moczon .Fpen ms“ Ppoe Ppen 6:» .aem .uwm .coems Fpen mzp muczoa .uum mucaon .Fan wen .853n "xpw>_uue .ppen we» P—oe .umm .mogm use ppmn .awx Fpoe .umm .cueee :e seemewa op .ywm .soees .asaw use pen Am_n_m_> “ocv .pwm .aszw "weep? emcweep mcwp_mp .u "wwwuv>wuom mewsmz .u muumwno 03p acrncwm .u i>wuue newsgomgma .u .uuw .mmo: .upm .me m.p=wu as .mes m.zamn izum .mmo: m.%gen .wmoc no» umucwoa sumo» m.pcmu:pm Am.xaen Lo gummy .Lwe; we puma xuon .Lwe; .mmo: .mxu m.p:mc=pmicowmmmmmoev .mmoc .mxu "magma mmocz weeppmp .u "mpgma Avon mewsez .u "magma xeoa op mc_gcwoq .u xuon op mcwucwoa .o .uum .m_aep .opm co mogm .megu co xon .xon :_ same .eoopm xon .e_m;u .coon Fpen ”m? pumnno .mecu .Loo—e .mpnep co ppen Hagen Eco; .mpnep "mugma ca mews: mew—Pep .a ”magma Eco; mewsmz .a co “cameo cm mcwuepa .a Soc; 0» mc_p:_oe .a xnen .Fpen .emu aux .F—mn .xwx .oozm .cooqm Aux .Fpen .Lmu .xamn .upm .cooam use Leo .emu .mo;m .cooam .neou .um; "gum“ .mocm .cooam .nsou .oogm use 3206 ”macaw .xan .anu .um; .no em so; mcpxm< .e .ue; ”mpumnao mewsez .m -no oz» emcweep mcw>Pm .e "mpumnno mcweuaoh .m mews: ecu mmogz ppmp even; _mne4 page mega: eo pmneb page mega: op .xm< op .Pmne4 ow o» muse: m_mcwm mcwma meowpecpneoo mcwuceumemuc: mpmcwm mcwucepmgoccs Amvueo: mcwm: "amen; ”mange w>wmmmeaxm uwmmm .o "mmega cowmcmaxm m>_pnmumm .m "mmmce m>wunmumm uwmem .e cowmceaxm m>mmmwgaxm .n 89 in the diagram approximate the training sequence. For example, the at- tending phase is prerequisite to training in any other phase of the pro- gram. Placement of a student into MLAP training sequence and that child's movement from part to part is determined both by his initial and final test results and by the chart in Appendix C which describes the prerequisite language skills for entry into each part of the pro- gram. The initial test for any part of the program is the same as the final test for that same part. Both tests have the same tasks which generally include two trials on each task presented in a random order. The two different labels, initial and final, are meant only to describe the times of testing a student. Initial testing is done before train- ing on a part so as to determine an appropriate teaching order of tasks within that part: tasks on which a student made approximate responses are taught prior to incorrect response tasks, while those the student made no response to are usually the hardest for the student to learn and are taught last. Final testing is done before passing a student on a part and ending that particular training. While training a student, two methods are used to adjust the instructional order of training tasks to fit student progress. One of these is intermediate testing which is different from initial and final testing only because a portion of a part in the MLAP is tested rather than the entire part. When a student seems to be responding correctly on the same tasks in a given part then the student would be tested on only those items. Each is presented twice in a random order. If the child is able to make correct responses on 90 percent of the items then 90 training on those items would be discontinued, except for intermittent review, and would concentrate on the remaining items in that part. The student would be given a final test when all items in a part seem to have been learned. In addition to intermediate testing, training observations are useful in updating the training order to fit the student's progress. These observations are merely a written record of the student's perfor- mance during a training session. The particular task, the student's response, and whether or not a prompt was given by the trainer are the areas of information recorded by the observer. The attendant examines the training observation to determine if any tasks need intermediate testing--that is, if the student makes a consistent, correct response to a task without the trainer's prompting. An example of a training observation is included as Appendix D. The testing procedures on all tests, initial, intermediate, and final, are the same. The command for each item is given when the child appears to be attending, and is not repeated; no prompting is given to the student, all responses are reinforced and no reinforce- ment is given following those tasks on which the child makes no re- sponse. This latter procedure prevents differential feedback to the student during testing. Criterion score for passing initial and final test is generally the same as for intermediate testing: 90 percent correct. In order that training results be statistically analyzed and for the efficient employment of the language program by attendants and retarded trainers, a number of modifications in the LAP were necessary. 91 Broadly outlined, these changes occurred in the motor imitation phase, the training suggestions in the program manual, the description and use of the initial, intermediate, and final tests, the addition of the ob- ject, "spoon," to the last four phases of the program, and the specific teaching procedure involving pre-organization of training trials by the arrangement of picture task cards for the retarded trainer. This latter modification was necessary for two reasons. First, during a pilot study, trainers without task cards tended to rely on the experimenter for changes in task presentation; that is, they would present one task then stop and ask what to do next, or present many tasks one time each, or change the task itself. Therefore, train- ing cards were introduced to lessen the dependence of the trainer, to prevent inappropriate task changes and to program the needed repetition of tasks so that new skills could be shaped in the student. The com- mand, materials and expected response for each training task were rep- resented by a picture. Examples of three training cards are located in Appendix E. These pictures were prearranged for the trainer by the su- pervising attendant or experimenter in a picture-holding circular file. The instructional task order was determined by the student's initial and final test results as well as his training progress evidenced in intermediate tests and training observations. The trainer was taught to look at the first card, present the command and materials associated with the picture, and praise the expected response or an approximation of it. If necessary, he prompted the correct response in the student. Then the trainer turned over the picture card to reveal the next task for presentation. 92 The former LAP training procedure included the use of shap- ing procedures on a single new task for up to 40 consecutive trials in a session. When a new task was correctly responded to, two times con- secutively without prompts, then it would be alternated with another new task. Also when a student made an incorrect response on a new item, the alternation procedure was immediately eliminated and train- ing dropped back to an easier level. These more sophisticated train- ing procedures, although resulting in the most efficient training, could not be incorporated with the retarded trainers used in this study. Therefore, the use of pre-programmed training cards solved one problem, but presented another--it excluded the immediate adjustment in task or- der for a student in relation to that student's responses, unless ver— bally directed by an observant trainer or supervisor. Phase Structure The basic organization employed in both groups during the eight week study is outlined in Figure 6. Group E was trained and su- pervised by the experimenter while Group C was run jointly by two ex- perimental assistants. Both assistants were well qualified for their job; each was a teacher, a graduate student in special education (men— tal retardation), familiar with Howell State Hospital, and had some ex- perience in behavior modification, and in testing and training the se- verely and profoundly retarded. Phase I lasted the first two weeks (Mondays through Fridays); Phase II extended for the following three weeks (Mondays through Fri- days); and Phase III, though originally intended to extend for 5 weeks, 93 PHASE I; 2 geeks Phase Ia. Initial training Phase Ib. Initial train- Phase Ic. Initial of retarded trainers in ing of Group E atten- training of GroupC both groups. dants: familiarization attendants: obser- with MLAP testing and vation of trainers. training. » attendant I Iexperimenter I Iexperimenter I - A observes I experimental trainer I I attendant staff : \' 7 i student J: I stu'dent Iward resident Iward resident” student III studentj ll _ 1 | student A"‘ student and meetings with ex- *— ee_ *7 perimental staff student H student IIexp. staff attendant PHA§E II; 3 weeks PHASE III: 3 WEEKS Training attendants to Training attendants to supervise supervise one trainer two trainers at a time with minimal at a time. assistance from experimental staff. I exp. staff I Iexp. staff 9 ' ‘ i . - a 12 a I attendant I I attendant I trainer I_trainer I trainer , .|_ _ a, _ Istudent I|Istudent.I student I student student]: student I Figure 6. Training Organization and Procedure for Groups E and C During the 8-week Period of Training. 94 was decreased to an intensive three week period (seven days per week) due to the pending placement of 5 of the 12 trainers in the community. General Controls Common to Both Groups Outside of the experimenter, the assistants, the speech ther- apist, and the participants, no other institutional staff were informed of the group membership of any of the participants. Attendants were asked to refrain from revealing specific information about the study until after its completion. The basic differences between the training received by each group were generally outlined to the attendants. Since past studies (McKinney and Keele, 1962, etc.) regarded the inter- action employed in the present study's Group C as a treatment of "mothering" or play, this group was never labeled or regarded as a con- trol group during the study. In fact the attendants in both groups were challenged and repeatedly encouraged to employ their retarded trainers and their own ideas as much as possible so their students could achieve the identified language training goals made available in both groups. In addition, attendants were informed that at the study's end, the specific language training procedures used in Group E (called Group I in the study) would be made available to those in Group C (called Group II) through special training workshops. The following procedural treatment variables for Groups E and C were controlled, i.e., made equal for both groups. 1. All-over phase organization and procedures (except for the attendant portion of Phase I), 0101-500 11. 12. 13. 14. 95 . Amount of time spent with attendants, trainers, and stu- dents by experimental staff, i.e., by experimenter in ex- perimental group and by assistants in control group, . Training rooms, Training equipment, Student edible and toy reinforcers, . Trainer payment systems and portable stores for spending earnings, . Length of training sessions, . Average number of training sessions per student, . Basic organization of each training session, . Participant titles and amount of praise and exposure in institutional newsletter write-ups for attendants, train- ers and students, Enforcement of the policy that the majority of direct in- teractions with the students be provided by trainers, Amount of videotaping for experimenter purposes, Explanation to attendants and trainers of the language training goals (based on pretests) for each student in Phase I; and updating his or her initial language train- ing goals, and Testing and observation procedures. The main differences in procedure for the two groups lay in the content taught trainers and supervising attendants and ultimately taught the students. 96 Phase Ia: Initial Instruction Procedures Common to Trainers in Both Groups During Phase Ia each of the twelve trainers received indivi- dual attention one hour each day for 9 days by the experimenter (Group E) or the assistants (Group C). In both groups a trainer worked 20 minutes with each of his students. The remainder of the training hour was spent instructing the trainer to perform tasks secondary to work- ing with his students. These included equipment pick-up and put-away, etc. Also when the student needed to be disciplined, the trainer was directed to give the discipline with guidance as needed from the ex- perimental staff. This was done to encourage independence of the trainer in his ability to control the student from the beginning of their contact together. The careful prior matching of the student to the trainer in areas of speech, aggressiveness, and physical limita- tions avoided many problems later between the two. Trainers were en- couraged to physically accompany their students to and from training sessions, whether this meant holding their hands, pushing their wheel- chairs or having the student push the trainer's chair. And, to further encourage independence, the trainer was taught to assist in those tasks secondary to training the student: getting and putting away training equipment and student reinforcers, setting timer for 20 minutes at ses- sion's beginning, and, to the extent possible, meeting experimental staff in an assigned place at the correct time prior to each daily training session as well as returning to his own ward, and receiving his payment for work at the end of each training session. 97 Phase Ia: Initial Instguctional Proceduresufer Grqup E Trainers< In Group E the Burke-Rowland inservice training method was employed to instruct trainers. Therefore as the trainer observed, the experimenter demonstrated a language instructional procedure on the student. This first demonstration proceeded slowly and was always geared to the trainer's comprehension level. Continuously checking to see if the trainer was watching, the experimenter gave the demonstra- tion in the following step-by-step order. 1. Looked at the picture task card, verbalized the task and set up any materials needed for that task, 2. Presented the appropriate command for that task to the student when he was attending, 3. Waited for a response, prompting only if needed, and 4. Praised all correct and partially correct responses with words, smiles, a pat, and either a token of food (depend- ing on the student) immediately after the response. Next, the trainer was asked to imitate all the steps as dem- onstrated. As much assistance as needed was provided and any correct or partially correct attempts by the trainer were praised. Depending on the trainer's ability to reproduce the teaching procedure, one of the following steps was taken by the experimenter next: 1. If the trainer made no attempt to copy the demonstration with the student or gave a totally incorrect imitation, then same demonstration was repeated more slowly and in 98 smaller steps; e.g., look at training card, verbalize task, ready materials and then let the trainer imitate procedure thus far. 2. If the trainer made a partially correct attempt to copy the demonstrated training procedure with the student, then the same entire procedure was repeated, expecting and helping the trainer to give a better imitation of the training than in his first attempt. 3. If the trainer correctly copied most or all steps of the demonstrated training task, then the task card was turned and the next different task was demonstrated showing all the 4 steps (1 through 4 above) and allowing the trainers to imitate again. 4. If the trainer correctly copied the entire demonstration, then the next different task card was located and explained and perhaps demonstrated in isolation or with the student. With these trainers it was not necessary to repeat all steps of the demonstration, i.e., presentation to student, prompting and reinforcement could be omitted in the dem- onstration. However, the trainer was observed as he tried this second task with the student so as to check his sec- ond performance to see if further explanation was needed. It was very important to provide much vocal assurance and praise to the trainer as he worked with the student in the initial training sessions. If the student appeared to be responding positively to the trainer, comments to emphasize this were given (e.g., “Look how 99 he is working and smiling for you; I think he likes his new teacher!“). From the very beginning of their instruction, Group E trainers were taught to follow the definite structure of reinforcement and modeling techniques and the MLAP sequence. Phase Ia: Initial Instguctional Procedures for Group C Trainers During Phase Ia, Group C trainers were not intensively in- structed in specific techniques of working with their students. In- stead, they were encouraged to play and verbally interact with their students. Experimental assistants followed these Phase Ia guidelines with Group C trainers. Make the trainer understand what kinds of things he should and could do with his students by l. telling him that he is a teacher to the students assigned to him and that . . . 2. he should try hard to teach each student the language training goal (however this is mentioned only during first session), . he should play with the student and get him to play, he should teach him new things, he should talk to him, he should use all the equipment provided, . he should keep him in the room for the whole 20 minutes, . he should give him physical contact (touching, pats, hugs), komVOSU'l-P-w . he should give the student food or toys whenever the trainer wants them during the sessions, 100 10. he should not hurt the student, and 11. he should be encouraged to ignore the student and remove toys when the student begins disruptive behaviors (throw- ing, hitting, etc.). Although tokens were provided, no token reinforcement system was implemented in Group C, nor were picture task training cards used by trainers. In addition, as mentioned earlier, Group C trainers were given the benefit of knowing their students' language training goals. During the first training session the specific training goals for each of his students as determined by the MLAP pretests were explained to the trainer by the assistant in clear and standardized ways. However, following the demonstration, the trainer's performance was not shaped. Instead, all his interactions with the student which approached the guidelines for Group C trainers were reinforced by the assistant. A large percentage of Group C time was spent in perfecting the trainers' abilities to carry out the duties secondary to training and in encouraging and, at times when necessary to prevent student dis- ruptive behaviors, suggesting and even showing play activities to be done with the students. Appendix F outlines the general procedures taken by the assistants with Group C trainers during this phase. Phase Ib and c: Initial Training, Procedures Common to Supervising Attendants in Both Grougs Phase Ib, illustrated in Figure 6, involves the initial train- ing of Group E supervising attendants while Phase Ic concerns the ini- tial training of Group C attendants. Both Phase Ib and Ic occurred lOl simultaneously with Phase Ia during the first two weeks. This is the only phase of attendant training where the organization differed be- tween Group E and C. However, time spent with experimental staff was equal for both groups: one hour each day on duty, Mondays through Fridays--averaging to 4.5 hours per attendant. Phase Ib: Initial Training of Group E Supervising_Attendants The attendants in Group E spent the entire Phase learning the assessment and training procedures and the organization of the MLAP. First one or two residents were selected from the attendant's ward to be jointly assessed and then trained by the experimenter and the atten- dant, next all four students assigned to that attendant training unit were assessed and their training programs determined and practiced. Generally one or two ward residents plus four students provided a wide variety of training levels and gave the attendant practice in most phases and parts of the MLAP. Group E attendants were trained to meet an informally deter- mined level of beginning skill in two areas: general abilities in language assessment and training procedures and use of operant and mod— eling methods. Both areas of skill were probed intermittently by the experimenter during the end of the first phase when testing and train- ing the remaining two or three students. At these times, the experi- menter withdrew all demonstrations and withheld vocal prompting so as to informally test the attendant for her comprehension and skill in the following general areas: 1. training versus testing procedures; 102 . phase organization and sequence of the MLAP; . terminology of the MLAP; . administration and scoring of the initial tests; determination of training levels; use of training procedures in the MLAP; VOW-#00“) . reinforcement procedures (token, praise, and direct food methods); 8. shaping techniques (reinforcement of successive approxi- mations, fading of prompts, etc.); 9. use of physical and imitative prompts; 10. familiarization with picture task training cards and MLAP manual; and 11. control of disruptive student behavior. The remaining training time was structured to concentrate on any of the attendant's skills that appeared weaker than others. Phase Ic: Initial Training of Group C Supervising Attendants The three Group C attendants spent the two weeks of Phase I involved in various activities with the assistants for an average of 5 1/2 hours. The first hour-long meeting revolved around the details of the training sessions: reviewing Group C procedure rules (see Ap- pendix G), and discussion of daily log sheets in attendant notebooks, assigned trainers and students, rooms for equipment storage and train- ing, and the attendant's daily work schedule. Attendants were shown and encouraged to use daily log sheets during Phase II and III to 103 record the activities each trainer engaged students in and to discuss any observable progress. It was emphasized that these onld aid ex- perimental staff in describing the day-to-day activities of Group C. Second, the attendants made observations of four 20-minute training sessions supervised by the experimental assistants, observing each trainer-student combination assigned to them. During these obser- vations attendants were discouraged from interacting with the students or trainers but instead were asked to observe. At this time, the at- tendants were given the written training goals for each student. These goals based on the pre-MLAPFT results, were the same as was demonstrated to the trainers during the first training session. The goals listed immediate areas for training as well as areas for later work. Examples are listed in Appendix H. Conversation with the attendants regarding the student goals was limited to rephrasals of the written statements rather than elaborations upon training procedures or demonstrations by the assistants. Third, another meeting, a half hour in length, was held by assistants with the attendants. During this meeting the discussion centered around possible activities they would encourage trainers to carry out with students, potential problems that might occur between trainers and students and solutions to those problems. In addition, it was emphasized that trainers should perform the majority of direct interaction with the students (including discipline); exceptions to this rule might include brief demonstrations by attendants as a means of suggesting new training ideas to trainers. Also attendants were reminded that trainers needed praise and continual support for their new role as teachers. 104 Finally another series of observations was scheduled during which the attendant was permitted to interact with trainers and stu- dents. Phase II: Procedures Common to Both Groups During Phase II the attendant in both groups learned to fill the position of instructor and supervisor of the retarded trainer taken during Phase I by the experimenter in Group E and the assistants in Group C. Phase II extended over the weekdays of three weeks. Because of the attendant's ward duties, the time allowed for their participa- tion in the study was restricted. During Phase I and II each attendant was allowed one hour every day they were on duty, Mondays through Fri- days.* Rather than train every day, trainers worked a single 20-minute session with each of their students every other day during this phase. The second trainer in each training unit was alternated with the first, giving each trainer an average of 7.5 hours of work during this phase. Attendants averaged 10 hours of work with the time equally divided be- tween their two trainers. *In Phase III, two hours were allowed each attendant everyday she was on duty, including weekends. These hours the attendants par- ticipated were always scheduled during the least busy times of the ward where she worked. In addition, the trainers, many of whom had other jobs (5 fed residents, 1 washed cups, and l folded laundry) had to have their schedules coordinated with their supervising attendants in such a way as to allow coverage by experimental staff, equal time invested in training sessions by all participants, and availability of training rooms. The unpredictable occurrences of illness and medical appoint- ments by any of the participants and isolated incidences of behavior problems in trainers and students, added to the many difficulties of scheduling training sessions. 105 General goals were again formulated and distributed to atten- dants in both groups to guide their efforts during the three weeks of this phase. These goals are included as Appendix I. Both experimental and control attendants continued the work towards teaching trainers in— dependence in those tasks secondary to instruction of students: getting students and returning them to their wards, setting timers, etc. Some skills of instructing trainers seemed to be acquired more slowly by at- tendants than others. Among these were the ability to provide enough vocal praise for the trainer's appropriate instructional interactions with their students, shaping the trainer's skills to directly handling the misbehaviors of his students, and keeping daily logs or records of training interactions (Group C attendants kept diary accounts in daily logs while Group E attendants recorded their impressions and the MLAP parts trained by each trainer). The first two behaviors seemed to run opposite to "survival behavior" often developed by overworked atten- dants in many institutions while the latter difficulty appeared to be due-to lack of time in Group E. In Group C, it appeared due to a lack of reinforcement from experimental staff for keeping the logs. However, experimental staff attempted to openly encourage these behaviors as well as model them repeatedly during Phase II. Often, unanticipated problems arose with trainers. In these cases many supervising attendants were able to contribute as much to creative solutions as did experimental staff. One example included two trainers, one from each group, who appeared unaffected by praise from attendants or experimental staff, i.e., their performance remained at a low level of quality with many errors in carrying out Group E trainer 106 procedures or little initiative in Group C interaction with the student. One attendant began withholding trainer pay when the trainer's perfor- mance was poor; next, with the experimenter's suggestion, she began paying the trainer, one penny at a time, during the training session when he showed some interaction with his student. This shaping proce- dure was also tried with some success on the Group E trainer. Phase II: TrainingiGroup E Aggendants to SuperViSe One Trainer at a Time - Group E attendants were expected to achieve in other areas as well as those mentioned for control attendants. As outlined in Ap- pendix J, increasing their skill levels from Phase I in the use of MLAP and in the application of reinforcement and modeling principles. In addition to these performance goals listed for Phase Ib, the atten- dants were expected to be able to adjust picture task cards for train- ers, to maintain a high level of correct instructional technique in trainers, and to be able to teach new language training procedures to trainers using the Burke-Rowland method. Generally their abilities in- cluded language assessment, implementation of training programs through a retarded trainer, and reassessment. The flow chart in Figure 7 out- lines this process of supervising the trainer. Steps 1 through 3 were done for each student before the study. Therefore Group E attendants were involved during Phase II in learning to carry out steps 4 through 10. Although Group C attendants could probably have supervised two trainers rather easily during this phase the attendants in Group E 107 Student begins language train- ing program here. i. l determined. IReinforcer for student is J f 2. l a Initial tests are given to the student and scored; parts to be trained are determined. 3.. I If a student passes at least the initial test for parts l-A and 1-8 in the attending phase, he may go to the next step. 1 3b. \L If a student fails initial test in the attending phase for either part l-A (Sitting) or 1-8 (getting rid of move- ments that interfere with training) or both, the attendant trains the student directly until the final tests for these are passed. Figure 7. 4. l Training cards for each part to be trained are arranged by the attendant in a teach- ing order to fit the stu- dent's initial test results. .. l With the student and the ma- terials present, the trainer is shown and taught by the attendant how to train the student on each part of the program being used with that student Flow Chart of the MLAP for Attendant Supervised Retarded Trainer- Student Pairs. 108 10a. 10b. 1036 tOI If the student gets the appro- If the student misses StepeZ.. priate number correct for that enough tasks on that “:ri part in order to pass, new part to fail the final parts may be added to his test, training on that training as outlined in the part is continued as part-by-part program descrip- needed. tion. 9 l When the student has learned all the tasks in a part of the program, the final test for that part is given and scored by the attendant. ,. I As the training continues, training cards for each part being trained‘are adjusted as needed (some added, re- moved, or rearranged) by the atten- dant. This occurs when the student learns or begins to learn new tasks or fails tasks once known. 7. T As training continues, the student's progress is periodically noted by the attendant by: a. observing his performance and keeping track of successes on each part being trained. b. giving intermediate tests on the tasks being trained to the student after training sessions. .. T For the rest of the program, the trainer trains the student under attendant super- vision; as the trainer works, he is ‘-9 praised (and later paid) for good work and is shown and taught to correct mis- takes as they occur and to try new teach- ing methods with the student as needed. /\ 10b. back to Step 6 109 needed the additional three weeks to practice their new skills with one trainer at a time before progressing into Phase III. ng§e_II: Training_Groug C Attendants to SugerviiOne TrainerTat a lime Group C attendants spent all of Phase II implementing the general goals for that time period (Appendix I). Their work, there- fore, centered around teaching the trainers to interact and play with their students, to train towards the language goals, and to become more proficient in the tasks secondary to student instruction. Phase III: Procedures Common to Both Groups Apart from intermittent assistance to experimental attendants on language training, Phase III procedure was essentially the same in both groups and included a large decrease in the direct instruction and constant support from experimental staff. It included, however, the provision of intermittent reinforcement to attendants and to trainers (although trainers received most of their support from the attendants supervising them). The sequence of the first two phases was employed as the means to prepare retarded residents and paraprofessional staff for the positions they held in Phase III. By Phase III Group E atten- dants were equipped with enough skills in behavior control and language training to independently operate the retarded trainer programs in lan4 guage acquisition training. Group C attendants had developed enough of the skills of praising trainers, instructing them in play activities when needed, and efficiently operating the training units so that they 110 were also able to independently operate their retarded trainer programs in play or "mothering" of students. This attendant independence did not include the withdrawal of experimental staff reinforcement which, as previously discussed, is of major importance. Phase III involved an increased intensity of training from two sources: number of persons supervised and frequency of supervision. That is, in this phase both trainers worked simultaneously in the same training room and each student received two training sessions per day. Also, due to a pending community placement of five of the twelve train- ers, training was intensified and operated on a Monday through Sunday basis but in accordance with the attendant's 5 day schedule. Whenever an attendant was on duty, including weekends, she held training ses- sions. All training sessions except two make-up sessions were held solely by attendant staff during Phase III. This procedure was not used in Phase I and II when training sessions were not held on weekends but were held by experimental staff on all Mondays through Fridays in- culding those when the attendant of a particular training unit was off duty. Experimental staff visited most training sessions for at least a few minutes during this phase. Their main duties were to pro- vide reinforcement for appropriate supervisory behaviors in the atten- dants and to handle the daily problems which often arose and delayed or prevented scheduled training sessions from being held. These prob- lems included the location of missing trainers or students (taken on walks, to dentist, etc.), refilling student reinforcer supplies, and making adjustment in training room locations. Although the institution 111 was extremely supportive and cooperative with the demands of experimen- tal staff, the lack of space and the tendency for different professional and ward staff to set different priorities for the same resident often resulted in conflict. For example, one ward wanted to confine a trainer to the ward for the day because of her misbehavior on the ward, apart from her performance as a trainer. This discipline then interfered with her availability as a trainer. Another case included a student sent for cataract surgery one third of the way through the study. The doctor did not know of the student's status in the study nor did the experimenter know of the long awaited plan for surgery. Also it was necessary in a few cases to request that a student be removed from other training pro- grams to avoid contamination of the experimental treatment. Because of a desire to train on the wards and a lack of space, all training ses— sions were held either in resident dining rooms or bedrooms. Therefore training schedules were adjusted to begin only after the housekeeping staff finished after-meal mopping. When residents whose beds were in training rooms were sick, other beds had to be used or training ses- sions were moved to other bedrooms. Provision of solutions to these problems by experimental staff was often necessary in order that train- ing sessions be held with some degree of certainty. Besides resolving the daily problems mentioned, experimental staff offered praise and encouragement to the attendants and assistance only if needed. Student gains were proudly shared by attendants and trainers with experimental staff, as well as the less obvious gains of trainers made in training and social skills. However, for the students and trainers whose improvements were slow and minimal, more encouragement 112 was needed to stimulate the continuation of effort invested by atten- dants. Videotaping of all trainer and students was done near the end of each phase. The taping served two purposes. First, it allowed many to see training sessions without the constant disruption of visi- tors. These tapes were shown to ward supervisors and programming staff as well as to a number of graduate classes. And this in turn provided an indirect source of reinforcement to all participants. The attendants and trainers were informed of these showings and appeared honored. Sec- ond, the taping itself seemed to result in a bit of status for the par- ticipants since videotaping was not commonly done in the institution. Another source of indirect reinforcement was directed towards partici- pants in the study. The participants, though not identified by the group, were discussed in three articles in the monthly institutional newspaper which is read by all staff. These articles described the study prior to its implementation, during its second phase and follow- ing an analysis of the main results. Description of Measures and Testing Procedures Overview of Testing In Figure 8, the testing procedures are outlined as they took place during the study. The Modified Language Acquisition Final Tests (MLAPFT) were used on a pretest, posttest basis to measure student gains in early language skills. The Cattell Infant Intelligence Scale was also given to students on a pretest, posttest basis to estimate men- tal age gains during the study. The Behavioral Observation Record (BOR) 113 was used to observe the instructional interactions between trainer and student in two different settings: during actual training sessions with retarded trainers and students, and during experimenter structured training sessions with supervising attendants and again with retarded trainers instructing an individual student not assigned to that atten- dant or trainer during the study. The BOR was used three times in the first setting near the end of each experimental phase. In the second setting the BOR was used on a pretest, posttest basis to record the in- structional interactions between a supervising attendant and a chosen student and again between a retarded trainer and a chosen student. TestingiDuring,Study Pretesting, Phase I Phase II Phase III Posttesting_ (2 weeks) (3 weeks) (3 weeks) 1. Students 1. Students a. MLAPFT a. MLAPFT b. Cattell b. Cattell BOR: Trainers observed during . 2. Trainers actual training sessions with 2. Trainers a. BOR* their students a. BOR* 3. Attendants 3. Attendants a. BOR* a. BOR* *Experimenter-structured training sessions were observed rather than the regularly occurring tra1n1ng 5e551ons. Figure 8. Testing Procedures Before, During and After the Experimental and Control Treatment. Modified Language Acquisition Program Final Tests (MLAPFT) description.--The MLAPFT tests the preverbal skills of attending, motor imitation and vocal imitation as well as the early verbal skills of 114 receptive and expressive language. The complete test form is presented in Appendix K. This test was adapted from the final inventory tests used in Kent, ejuels's (l972) LAP. As stated previously, the LAP was developed for institutionalized severely and profoundly retarded per- sons who were either nonverbal or in early stages of emergent speech. The final inventories were used in the LAP to determine when criterion levels of performance on parts of the program had been reached. How- ever, in this final inventory form the test scores were not addable to yield a total test score nor had the test's reliability for retarded populations been determined. (In conversations with Kent she reported the test-retest correlations were obtained on young normal populations to compare the sequencing of training items for normals with retardates. Test-retest correlations on the LAP with normals ranged from .8 to .9.) For this study a procedure, described below, was developed whereby the scores for the 29 tests could be added to produce a total test score. In addition the test-retest reliability was determined. The general procedures for administering and scoring the MLAPFT are given in Ap- pendix L. The total score for any subject on the MLAPFT is a converted total of his correct responses on each of the final tests for each of the 29 parts in the program. Since the final test for each part con- sists of unequal numbers of test items and therefore different vari- ances, the final test scores were converted to the same metric by cal- culating the standard deviation which was pooled across groups and/or testing times on each of the 29 final tests and then by dividing each subject's scores on a single final test by that test's standard 115 deviation. A subject's 29 final tests were all similarly divided by the various standard deviations and the quotients were added to yield a total MLAPFT score. Reliability.--A procedure was developed to answer two ques- tions relating to these test scores. First, do two examiners observ- 3 ing the same test performance agree on the scoring? A check of exam- iner agreement was regarded as necessary since assigning a score to a subject required close observation by the examiner of the child's stream of behavior. Second, what is the conventional test-retest reliability of the total.scores used in the hypothesis testing? To answer these questions a group of 20 retardates, similar in characteristics to those subjects in the study, was selected from the study institution. Table 6 summarizes the characteristics of this sample, and provides data for a comparison of this group with Groups E and C. It will be noted that this sample closely matches the study groups. However, the MLAPFT first test means for pairs of subjects ap- pear to differ greatly: 38.21 (reliability group) and 66.12 (Groups E and C). This apparent discrepancy is a function of the differences in the pooled standard deviation divisors used for the reliability group and for Groups E and C. The standard deviations for the reliability group were higher than those for Groups E and C, thereby resulting in larger divisors and lower total scores in the reliability group. Though very similar to Groups E and C in IQ and MLAPFT scores, a few students in the reliability group functioned below the lowest stu- dents in Groups E and C but generally more reliability group students 116 functioned above the highest students in Groups E and C. And it was this wide variability of the reliability group which caused its large standard directions on the MLAPFT. Table 6. Group Ranges and Means for the Reliability Group and Com- parison with Combined Group E and C Means. Reliability Reliability Group E and C Group Group Combined Ranges Means Means CA 11.68-40.27 yrs. 19.48 yrs. 17.23 yrs. IQ 5-38 18 11 Years Institutionalized 3.08-28.27 yrs. 12.30 yrs. 12.06 yrs. MLAPFT lst test O.27-80.03 38.21 66.12 MLAPFT retest O.66-78.53 39.88 ----- To answer the first question, the reliability group subjects were tested in a random order by the examiner who later did the test- ing for Groups E and C. During testing three of the 20 subjects were on a continual praise and intermittent food reinforcement schedule while the remaining 17 were on praise and token reinforcement. A sec- ond examiner, the writer, observed each testing session and indepen- dently scored the tests. This was done during the first testing per- iod only and was not repeated in the retest. The sets of scores were then correlated to provide indices of rater agreement. The total score, 117 as well as the scores for each of the 29 tests of the MLAPFT were cor- related. These correlations are presented in Table 7. Table 7. Inter-Examiner Correlations for MLAPFT Scores for the Reliability Group (n=20). Test r la. ...... lb. ...... 1c. 1.0 1d. 0.952 2a. 0.994 2b. 0.993 3a. 0.989 3b. 0.990 3c. 0.991 3d. 0.977 4a. 0.996 4b. 1.00 4c. 0.990 4d. 0.843 4e. 1.00 5a. 0.998 5b. 1.00 5c. 0.987 5d. 0.997 6a. 0.997 6b. 0.984 6c. 0.997 6d. 0.989 6e. 0.899 7a. 1.00 7b. 0.734 7c. 0.798 7d. 0.954 7e. 0.991 Total scores 0.9961 ------ Scores were perfect for all subjects. 118 It is apparent that high scorer agreement was achieved on these tests. On test 7b., "Telling Where an Object Is," the subject was required either to label the room part on or in which an object was placed or to point to the object and give a demonstrative expression ("There," "Here," etc.). It appears that judgment of the correctness of the vo- cal response may have been more difficult for the demonstrative and the room part words than for any other group of words. (chair, table, floor, box, there, and here.) The study examiner then readministered the MLAPFT to the same subjects approximately a week following the first testing. The first test and the retest total scores of the same subjects were correlated to determine the test-retest correlation coefficient, i.e., the stabile ity of the test's scores when given to the same subject twice in a short period of time. The test-retest correlation coefficient was 0.9965, representing very high test-score stability. This correlation coefficient was later used in the statistical analysis of the MLAPFT results. Testing procedures.--All 24 students involved in the study were tested in their own training room on the ward or as close to this room as possible. Each student was individually tested by an author of the original Language Acquisition Program (Kent, e§_gl,, 1972) who was intimately familiar with LAP and MLAP testing and training procedures as well as with low-functioning retarded populations. (In prior con- versations with Kent and Klein it was reported that unless the examiner was somewhat familiar with severely and profoundly retarded populations, 119 the students seemed less responsive and spurious testing results were often obtained despite the use of reinforcement during testing.) Two weeks prior to testing, each student had 5 to 15 minutes of token training with the examiner and the experimenter. During this time it was informally determined whether a token reinforcement system could be used during testing and later training sessions. The parti- cular type of token economy described here is one in which each token does not have a unit value but rather has general association with pri- mary reinforcers; any amount of tokens can be exchanged for a rein- forcer during token trade-in times. Briefly, the procedure consisted first of defining one or more responses in the student that occurred often or was easily evoked. This response then would either be com- manded by the examiner (and prompted if needed) or the examiner would wait a few seconds for the response to be emitted by the child. These responses included eye contact, hand folding, motor imitative tasks, etc. As soon as the response occurred the child was handed a poker chip token and praised, with an immediate exchange of the token for a choice of edibles (candy, salty foods, drink, soft food, etc.). When a child had no observable desire for edibles, toys were used as rein- forcers: e.g., a few seconds of a music box, play with a toy car, or one pull of a pull—string talking toy. This process was repeated a riumber of times depending upon the child, and then gradually the stu- (dent was required to hold onto two, then three and four tokens before van exchange was granted. Therefore, a larger number of responses was required for each primary reinforcement, but social praise and tokens r‘emained on a continuous schedule. Next, a token can was introduced 120 so tokens could be given out more efficiently and retained by the stu- dent between exchanges. Most students demonstrated an association of the tokens with the primary reinforcers. That is, it was observed that these students eventually tried to hand the examiner tokens prior to an exchange time as a means of obtaining primary reinforcement; that they cooperated with the introduction of the token can and the gradually increasing delay between token exchanges; that the frequency of the reinforced response increased; and that attending responses in- creased. The students exhibiting these behaviors at the end of one or two, 5 to 10 minute token training sessions were tested (and later trained) using token reinforcement: i.e., all correct, partially cor~ rect, and incorrect test responses were reinforced with praise and a token while token exchanges were given occasionally. The three to four students not displaying behaviors to indicate understanding of a token reinforcement system were tested (and later trained) using praise and intermittent edible or toy reinforcement. The token system was more efficient allowing more trials to be presented since there were less delays for edible reinforcement and continual token reinforcement did not result in satiation on primary reinforcers. In order to avoid the occurrence of an instructional shaping process or low response levels during testing, all correct, partially correct and incorrect responses of the student were reinforced. How- ever, no instances of a "no response" or disruptive behaviors were re- inforced. This contingency resulted in an increase in the number of responses made by the child during testing and a decrease in the dis- rUDtive behaviors and no response trials. Also, since the child did 121 not receive selective feedback on his test performance, his allover performance was greatly improved by testing alone. The objects used during testing (labeling objects, imitation toys, etc.) were not identical to those used during training. Testing usually was done in one sitting unless it was apparent that the child's inattention interferred with testing. In these few instances testing was ended quickly and repeated in its entirety later that day or the following day. Testing rarely extended beyond an hour and, on the av- erage, lasted less than 30 minutes. The group to which each student had been assigned was not known by the examiner and the testing order was random. All students appeared to relate well with the examiner and it was felt that an Optimal sample of each student's early lan- guage behaviors had been obtained in all testing interactions. Cattell Infant Intelligence Scale Description.--The Cattell Scale tests from the second month of mental development through the fourth year. At the third year, the scale links with items on the Stanford Binet Intelligence Test. There are a total of 113 subtests in the entire scale. Motor items are the main emphasis from the second through the seventh month. Attending to sounds is tested at the second, fourth and fifth months. Verbal con- cepts are tested at the eighth month (combining two sounds). Although motor imitation tasks predominate until the 23rd to the 24th month, the receptive and expressive verbal factors of understanding commands and using labels are intermittently tested through 30 months. 122 Generally the Cattell has greater item interest for the young child than other infant tests (London, 1961). This is an important fac- tor in securing optimum performance from young normal children as well as institutionalized profoundly retarded children. Despite its frequent use, the Cattell has not been widely investigated with normal or re- tarded children (London, 1961). However, it has been shown to have high predictive validity for young mentally retarded children (Erickson, 1968), high test-retest correlation for retarded children, and the sen- sitivity to discriminate between groups of mongoloid and organically retarded children (Kralovich, 1959). Reliability,--Using a large sample of mildly retarded chil- dren, London (1961) correlated their mental age scores on the Cattell with their retest scores on the Cattell, or another infant scale, or on the Binet, if their MA was beyond 3.5 years. Since testing data was gleaned from medical records, there was no control for the interval be- tween the first test and the retesting. The results showed that a sam- ple of mentally retarded children will demonstrate greater stability of IQ than a sample of average children. London's findings further sug- gested that the more retarded a child is the more stable would be his IQ. The correlation coefficient obtained for the lowest scoring third (of the mildly retarded sample tested was .78. The sample of retardates in this study was drawn from the profoundly retarded population func- 'tioning more than three standard deviations below London's sample. Considering this, it is reasonable to assume, on the basis on London's data and without further testing of profoundly retarded subjects, that lflie test-retest correlation on the Cattell for profoundly retarded 123 subjects is approximately .80. Therefore, the stability of the IQ in the profoundly retarded subject is assumed to be slightly more stable than it is in the mildly retarded subject. This correlation was later used in the statistical analysis of the results. Testing procedures.--The Cattell was given to the 24 students in the study on a pretest, posttest basis by an examiner thoroughly fa- miliar with the educational assessment of lower-functioning retarded populations. In addition, the examiner had had some contact with many of the students. All students were tested in a random order; the ex- aminer was unaware of their group membership. All testing was done in a small building within walking distance from the institution. Praise was used by the examiner to reinforce all student responses so as to keep the response level high and to avoid instructional feedback to the student. Tokens and food were used intermittently by the examiner to reinforce students. Testing was generally begun at a point estimated by the ex- aminer as the student's basal level. Then, as necessary, testing was extended downward to establish a basal and upward to establish a ceil- ing. Ten of the 24 students were tested on some early items in the Stanford-Binet Intelligence Scale since a ceiling had not been reached by the second half of the third year at which point the Cattell extends irito the Stanford-Binet. The testing period ranged from 15 minutes in leength to 45 minutes, with an average testing period of 20 minutes. A few students were tested in two shorter sessions due to inattention de- weloping after the first 20 to 30 minutes of testing. However, most 124 students were given a brief rest from testing by the examiner after 10 to 15 minutes of testing. Test breaks included walks down the hall, food reinforcers, etc. It was the examiner's opinion that the test results, both pre- and posttests, reflected an accurate estimation of the student's mental ages. As evidence, the examiner noted his familiarity with the students as well as their general high level of cooperation which did not appear to change much from pre- to posttesting. Behavioral Observation Record (BOR) Description.--The BOR was developed to record in behavioral terms the ongoing instructional interactions between a trainer and a student. A similar observation technique was reported by Lawrence (1971) to record language training sessions between an experimenter and retardates. The BOR is a coding technique which allows an observer to record the stimuli presented by a trainer to a student, the prompts given to the student, the student's response, and the consequences pro- vided by the trainer after the student's response. An observer, skilled in using this coding technique, can record most instructional interac- tions occurring within a training dyad. The BOR observation key and a sample recording sheet are found in Appendices M and N respectively. The basic observation unit in the BOR is called a trial and at: the least consists of a stimulus presentation. However, most trials also contained a student response and a consequence provided by the trainer. Five main event categories could occur in each trial in the fOTlowing sequential order: Attention of the child; Stimulus 125 Presentation by the trainer; Prompting the child; Response of the child; and Consequence occurring immediately following the child's response. If attention-evoking stimuli were provided by the trainer these were merely classified by the amount of effort exerted (e.g., trainer calls child's name, turns child's head) rather than further recording the attention-evoking procedure as a trial within itself. Stimulus presen- tation was that category in which the trainer's commands and actions were categorized (e.g., "Show me the car," "What is this?"). Prompts used by the trainer included all cues and physical and imitative assis- tance provided by the trainer following a stimulus presentation which acted to further evoke a correct or a partially correct response from the child. The student's responses were classified as correct, approx- imate, incorrect or as a no response. Consequential events were fur- ther subdivided into classes of Feedback, Reinforcement, and Other Con- sequences. Feedback was limited to language training feedback to re- inforce learning of specific words (e.g., trainer says, "That's the BALL," after child correctly touches the asked for object). Reinforce- ment was categorized as to its type: praise, tactual, food, etc. Other consequences included negative comments (e.g., "No,“ "Bad boy") and temper tantrum intervention procedures or those actions taken by the trainer specifically to control disruptive behaviors (e.g., time out). In addition to these five main events the interventions by supervising attendants or experimental staff occurring during the in- structional interaction between trainer and student could be coded. These generally took the form of praising or assisting the trainer. 126 Reliability.--During the pilot development of the BOR, it was determined that 3-minute observations were long enough to obtain an ac- curate sample of the instructional behaviors of a trainer during a par- ticular training session whether it be a retarded trainer or an atten- dant trainer. Five minute observations did not give much more addi- tional information while 8 or 10 minutes of observation yielded only redundant information. Therefore all BOR observations were timed with a stopwatch to be three minutes in length. A simple percentage of agreement was calculated to determine how well observers could agree on using the BOR. Whenever two observ- ers recorded the same response this was recorded as an agreement. The total number of responses consisted of the agreement responses plus the total of the nonagreement responses given by each observer. That is, if observer A coded a consequence to include tactual and vocal rein- forcement while another observer coded the same consequence to include only the vocal reinforcement, one disagreement would be counted out of a total of two observations. This would yield 50 percent agreement. This is a conservative estimate of agreement since the total count did not include the times when both observers were in agreement that a be- havior did ne§_occur and no code was recorded, e.g., both judged, in the example cited, that feedback had not been given by the trainer and therefore coded nothing in the feedback column of that trial. The final percentage of agreement between observers is based on the average of four pairs of observations made simultaneously but recorded independently by the experimenter and the study's single ob- server. The observations, each three minutes in length, were done on 127 the interactions between four different retarded trainers and students in two observation settings: experimenter-structured using students new to the trainer and actual training sessions with regularly assigned students. Three of the observations were made in the former setting while one was made in the latter setting. Two of the retarded trainers observed were trained to use Group E procedures while two remaining used Group C procedures. Percentage of agreement combining all ob- server responses ranged from 74.0 percent to 94.5 percent across the four pairs of observations. The mean percentage of agreement, 86.4 percent, indicates that the BOR can be used with a high level of con- sistency by different observers to record the same instructional inter- action between a retarded trainer and a student. Separate rater agree- ment percentages were not calculated for the categories of information later drawn from observations, such as number of student responses during a single observation. Observationgprocedures.--As indicated in Figure 8, the BOR was used to observe instructional interactions in two different set- tings: during actual training sessions with retarded trainers and stu- dents and during training sessions especially arranged by the experi- lnenter so as to get samples of attendant and retarded trainer instruc- tional styles before and after treatments. The observer who used the BOR during the training sessions participated in pilot work preceding the study during her last two terms of her senior year. As a psychol- OQy'major, she was thoroughly familiar with behavioral terminology and Quickly gained competence in using the BOR. Since the trainers, stu- dents, and attendants had become used to her presence during pilot work 128 and since she entered training rooms only before training sessions had begun, her presence during observation periods was not regarded as a threat nor as a disruption to training sessions. The application of the BOR in actual training sessions was repeated for each trainer during the last few days of each of the three experimental phases. These samples, taken at the end of each phase, were meant to reveal the progressive differences that occurred over the study in the instructional techniques used by trainers in the two groups. During these times the observer entered a training room at the beginning of a 20-minute session and, in a predetermined random order, observed the trainer as he or she worked with their student. All ob- servation periods were three minutes in length and the observer had 10 to 20 minutes (depending on whether there were one or two trainers in the room) in which to take the three minute continuous sample. The first few minutes of a training session were never used as part of this observation and generally neither were the last few minutes. This pre- caution was to avoid recording unrepresentative training periods dur- ing which a trainer was either preparing to begin or a student and/or trainer might begin to show some fatigue. The observer waited until it appeared that the trainer and student were settled and into the begin- ning of an instructional interaction. Once an observation was begun it was continued for three minutes. Therefore, when a student was asked by a trainer to exchange his tokens for a reinforcer, when time outs were given for student misbehavior, or when an attendant gave the train- er an instructional demonstration, training between the trainer and the student stopped but the observation continued. These latter events 129 occurred primarily in Group E. Because of these interruptions some ob- servations taken during the Group E training sessions may appear far less instructional than others where no interruptions occurred. A strict arrangement of selecting a randomly determined three minute observation period did not yield good samples of instructional interactions and were very difficult to schedule. For these reasons the observer generally had a 10 minute period of time within which to take a three minute observation. Group membership of the trainers, though not directly re— vealed to the observer, became rather obvious because of her familiar~ ity with the MLAP and the obviousness of the differences in training procedures used in the two groups. This difficulty could have been controlled by employing at least two observers either blind to the dif- ferences or viewing only one treatment. However, this alternative was not available. The observer in the experimenter-arranged training settings was the experimenter. These observations were done on a pretest, post- test basis. Each of the retarded trainers and attendants in both groups instructed specially selected students. These students were either nonparticipants in the study with similar levels of functioning or were participant students in training units other than the unit to 1which the attendant or retarded trainer belonged. Standard instruc- tions were used with the person in the instructor role to inform him of the task, the length of the observation and the reason. These stan- dardized instructions are found in Appendix 0. Basically each person Observed in the instructor role was asked in a randomly determined order 130 to train a student on either motor imitation tasks or on receptive la- beling tasks. The training levels of the students selected had already been identified by their results on the MLAPFT so that they did not readily perform the tasks presented to them. The persons doing the training were given a choice of three imitative or labeling tasks on which to train the student. They could work with a supply of rein- forcers for that child to be used at their discretion. Those latter observations, as well as those taken of retarded trainers during training sessions, were expected to reveal what train- ing techniques would be used by the instructor and whether the tech» niques incorporated modeling and reinforcement principles. It was not considered important that the student learn a response during an obser- vation but rather that the student be exposed to consistent, instruc— tional techniques which embodied the reinforcement and modeling prin- ciples conducive to learning. After the BOR observations were taken during training ses- sions, eleven different categories of information were drawn from the data collected. When observations were taken of attendants and train- ers during experimenter-structured situations only nine of these eleven categories of information were drawn from the data collected. The two categories omitted during these observations were the number of rein- forcements and the number of instructional demonstrations and comments given to the trainer by supervisory attendants or experimental staff. This was because these did not occur during structured situations. .All the categories were calculated in one of three forms: frequency icounts of an event observed, percentage of times an event was observed‘ 'f‘n‘fi‘ff‘d rim “ 131 to occur, or judgment ratings on an event. Appendix N, which gives a sample observation and a frequency, percentage or judgment on each of the eleven categories, should be consulted to illustrate the following explanations. Five categories of information were collected from the data in a frequency count form. 1. Number of complete trials gresented to the student. A complete trial was defined as consisting of at least a stimulus presen- tation, a correct or approximate response by the student and a rein- forcing consequence. It was assumed that the more frequent were the complete trials during a training session the more likely student learn- ing would occur. Therefore, whenever all three events, stimulus pre- sentation, correct or approximate response, and reinforcing consequence were observed to occur in that sequence and during a single trial, one complete trial was counted. Intervening events such as attention- evoking prompts, response prompts, or response feedback could also occur during a complete trial, by definition, but were not necessary' for a trial to be regarded as complete. 2. Number of times a student was attending prior to a stim- ulus presentation. It was assumed that the more often a student was attending the more likely it was that the student would be able to re- spond correctly, approximately on his own, or that he would be able to benefit from being prompted by the trainer. Therefore, whenever the student's attention was observed to be present, a check was recorded in the attention column. To estimate this category these checks were totalled. In addition, whenever the student's attention was 132 successfully evoked by the trainer just prior to stimulus presentation, these times were also included in the total attention count. 3. Number of student responses following a stimulus presen- Eggipp, It was assumed that the more times a student responded, the more likely he would be to make the correct response, thereby facilitat- ing learning. Even incorrect responses, when not reinforced, provide learning feedback to the student. Therefore this count included all student responses, correct, approximate or incorrect, made following a stimulus presentation. Observations of no responses were not included in this count. However, responses prompted by the trainer were counted. 4. Number of times the supervising attendant or experimental staff praised or otherwise reinforced a trainer. It was assumed that these reinforcements were appropriate and acted to increase the quality of instruction given to the student by the trainer. An asterisk was used to record the times a trainer was reinforced by supervisory per- sons. These asterisks were totalled to estimate this category. 5. Number of times the supervisory attendant or experimental staff provided instructional demonstrations, comments or commands to the trainer. It was assumed that trainers would need varying levels of intermittent instruction throughout the entire study in order that the quality of instruction remain high. Only their frequency and not the comprehensiveness of instructions given to trainers was judged by tflie observer. All demonstrations on the student by supervisory per- $ons and verbal comments or commands were counted together in this cat- Eflaory. 133 Three categories of information were collected from the ob- servation data in a percentage form. 1. Number of correct and approximate student responses rein- forced by the trainer divided by the total number of correctand approx- imate responses. Since the frequency of responses that are consistently and immediately followed by reinforcing consequences increases over I time, this percentage was important in the evaluation of trainer effec- u..- ev- -".-- -. tiveness. It was not the purpose of the observer to judge the actual reinforcing value of the consequences provided by the trainer. There~ L fore praise, food, toys, and tokens, when given by the trainer, were recorded as assumed reinforcers. (In all cases trainers and attendants were provided with objects and/or food determined to be reinforcing to the'student.) Only rarely were there no positive responses occurring in an entire training session. Since a trainer can easily create ope portunities for positive reinforcement to be given (e.g., reinforcing attending responses, prompting correct responses) these rare incidences were scored as 0 percent rather than a "no opportunity." Other cases of 0 percent scores occurred when some correct or approximate responses occurred but none were reinforced. 2. Number of tangible reinforcers given the student simultan- eous with praise divided by_total number of times tangible reinforcers were given. Since for many students praise was not reinforcing, it was regarded as a desirable procedure to always associate praise with tangible reinforcers as a method of establishing praise as a general- ized reinforcer. Therefore, each time was counted that tokens, toys, or food was given to a student; this total was then divided into the 134 number of times praise was given along with the tangible reinforcer(s). An exception in scoring occurred with trainers not using tangible re- inforcers during an observation. This happened quite often in Group C although as mentioned earlier all students were provided with tangi- ble reinforcers. When this occurred a "no opportunity" score was given the trainer. However, when all tangible reinforcers were given without ever being paired with the trainer's praise, this resulted in a 0 per- cent score. 3. Number of trials when a student made an approximate; an incorrect, or a no response on theipreceding trial and the same stimu- lu§_presentation was repeated along with student prompting or the same stimulus presentation was mereLy repeated without prompts on the next trial but this resulted in an improved level of response (approximate or correep) divided by the total number of times these opportunities togprompt the student occurred. It was assumed that the more often a student made a response the more quickly learning would occur: prompt- ing to evoke correct or approximate responses was taught Group E train- ers as a strategy to get an unresponsive student to perform, to be re- inforced and to eventually associate a particular stimulus presentation with a particular response. Many times opportunities for prompting occur when a stimulus presentation is given and a response delay occurs: i.e., the student does not appear to know how to respond. At these times appropriate prompts could immediately be given without repeating the stimulus presentation. In these cases the student's response in the preceding trial necessarily had had the same stimulus presentation. However, from the observation data these cases could not reliably be 135 discerned to be appropriate or necessary prompting, but only to be prompting. Since the observer did not make instant judgments as to the appropriateness of the prompt, the information taken from the observa- tion data on prompting had to be limited to those incidents when re- sponses were affected or not affected by the prompt or repetition of a stimulus presentation. In scoring this category when the trainer observed a need for prompting, and did so, and the resulting response was not improved, this incident was still regarded in the score as an appropriate prompt. However, the next trial following was also regarded as an opportunity for prompting the student and perhaps with stronger prompts. The three remaining categories of information drawn from the observation data consisted more of judgments made about the particular observation rather than actual frequency or percentage counts. These three categories, therefore, should be interpreted with more caution than the preceding categories which are less judgmental in character. 1. Consistency and dpprqpriate repetition of stimulus presen- tations. Especially with nonverbal or low-verbal students, it was as- sumed to be of particular importance that stimulus presentation be con- sistently given by trainers and that the acquisition of new responses be facilitated by repetition of the same stimulus presentation. Since all stimulus presentations could not be recorded word for word, the ac- tual changes in the verbal form of a stimulus presentation were not ob- Searvable from the data. But since the observer noted the key word ac- ‘tivity or object used in the stimulus presentation, it was possible to tefll how often and in what order stimuli were presented. When the same 136 stimulus presentation was repeated twice consecutively, a single point was awarded; when the same stimulus presentation was repeated three times consecutively, two points were given; and when the same stimulus presentation was repeated four times with at least one prompt during this series of presentations, three points were given. In the latter case if prompts were not given over a series of four repetitions, only two points were assigned and it was regarded as a stimulus presentation repeated three times. To further avoid assignment of points to inap- propriate repetition of stimulus presentations which occurred more of- ten in Group C (e.g., student successfully commanded without prompts to put tokens in a can for 10 minutes straight), points were not given to a series of the same stimulus presentation repetitions longer than four with a prompt or three without a prompt. In addition to the above rules, a single point was awarded when eight consecutive trials con- sisted of stimulus presentations in the same language task category (e.g., labeling common objects, touching body parts, imitative tasks). All points were totalled to yield a score for this category. The higher the score, as in all categories, the better the trainer was as- .sumed to be in this category. Because of the limitation of the obser- ‘vation data collected it is recognized that efforts to rate this cate- igory are less reliable than other categories. 2. Appropriate fading of prompts. When prompts or cues are IJSEd as a strategy to evoke a correct or an approximate response, it is important that they be gradually faded out so as to bring the re- Sponse under stimulus control. To judge the trainer's skill in applying 137 appropriate fading procedures a rough judging technique was employed. One of four ratings could be assigned depending upon a trainer's obser- vation data. a. No opportunity (N.O:). This rating was given when there was no observed opportunity to fade prompts. This occurred when no prompts were necessary to evoke correct responses, as in review train- ing, and when so few trials occurred during an observation that fading was not possible. The trainer was not penalized for failure to prompt when there did not appear to be an opportunity to do so. b. Minus (-I. This rating was given when the trainer prob- ably could have attempted to fade out prompts but did not do so. For example, a trainer with this rating might continuously give a full prompt (P3: putting the student through the entire correct response) over a large portion of the observation without ever testing to see if a withdrawal or a decrease of the prompt could be made. c. Plus (+I. This rating was given when prompts were inap- propriately increased rather than kept at the same level unnecessarily (as in b.). For example, a trainer with this rating might use a medium level of a prompt (P2: helping the student through the initial portion ()f the response to evoke a correct response) and then in the next trial Izse a full prompt to evoke the same correct response (P3: helping the sstudent through the entire response). Also, a trainer could unneces- ssarily begin prompting a response formerly evoked without prompts in a Preceding trial. d. 945;. This rating was given when there was some evidence 0f successful fading or of attempts to fade by the trainer. In the 138 first case a trainer could decrease the level of prompts or cues given the student or could fade them out altogether while the student would continue to respond at an approximate or correct level. In the second case a trainer could attempt to fade by "testing" the student's re- sponse strength and decreasing or omitting a prompt or cue but rein- stating it when the student stopped responding or began responding in- I correctly. I 3. Apprgpriateness of task to the student. If the language 1 training is initiated atoa level just above the child's independent i level of performance there is a higher probability that learning will occur than if training is begun below or too far above that independent level of performance. Since all trainers and attendants in both groups were informed of the language training levels appropriate for each stu- dent, this rating was deemed fair. One of three ratings was roughly assigned to an observation depending upon the student's observed re- sponses. a. Minus - . Training tasks were regarded as too easy when all the observed responses were correct and no prompting or cues were given by the trainer. b. Plus (+2. Training tasks were regarded as too difficult if the student's responses were all incorrect or were nonoccurring and prompts ue§e_given by the trainer at least once during the observation (if no prompts in this case were given a + rating would not be given to the trainer but rather an Appropriate rating since one could not be sure that the tasks were too difficult; this trainer's score in the cat- egory of percentage of appropriate prompting would reflect his failure to prompt). 139 c. Appropriate (.I- When the tasks presented to the student appeared appropriate to his level of performance an "A" rating was given; the student's responses probably included a combination of cor- rect, approximate, incorrect and no responses and may have included some prompted responses. W As previously mentioned, the unit of analysis in this study for both groups was the mean score on the dependent variables of the two students trained by the same trainer. This was so because neither student's score in a pair of students was independent of the other's score since they were both trained by the same trainer, and therefore, the comparison of the two groups may be actually regarded as a compar- ison of trained trainers (Group E) versus untrained trainers (Group C). Each trainer was rated by the mean growth score of his pair of students on both measures. Therefore the two main hypotheses tested by this study are: 1. For the Modified Language Acquisition Program Final Tests, the overall group mean of the mean score of each pair of students trained by a single trainer in Group E will be higher than the group Inean of mean scores for pairs of students in Group C. 2. For the Cattell Infant Intelligence Scale, the overall group mean of the mean scores of each pair of students trained by a single trainer in Group E will be higher than the group mean of mean scores for pairs of students in Group C. 140 Treatment of the Data Relating to the ijorpflypotheses To test the two main hypotheses comparing the effectiveness of the Group E trained trainers with Group C untrained trainers, the statistical analysis was based on the mean scores on both dependent variables of the pairs of students trained by the same trainer for each I group. A one-way ANOVA was done on the index of response for each of the two dependent variables: the Cattell Infant Intelligence Scale scores and the MLAPFT scores. Since the unit of analysis was the pair of subjects trained by each trainer, the scores used in the analysis were the mean scores of the two students on both variables. The index of response analysis requires that Z scores, ob- tained in the following way, be analyzed by a one-way analysis of vari- ance rather than merely analyzing pure gain scores: Z = Y - KX where: Y = the posttest of the dependent variable, .X = pretest on the dependent variable, and K = the known test-retest correlation coefficient for the dependent variable. Use of this particular analysis, in addition to the random assignment of attendants and trainer-child pairs to groups, improved the precision of the analysis of results by subtracting the retest error from the pretest, posttest difference prior to analysis, i.e., an effort to de- crease the probability of committing a type II error (acceptance of the null hypothesis when it is false) while holding the probability of a type I error (rejection of the null when it is true) constant. 141 Where probability values were determined, the alpha level was set at .05. No attempt was made to statistically analyze the related ques- tions involving the observational data. Because of the complexities and problems of interpretation inherent in post hoc statistical analy- ses, these questions, concerning the nature of the treatment, were in- formally answered by comparing the mean category scores of both groups of attendants and trainers observed during different phases of the study. CHAPTER III RESULTS Hypothesis It was hypothesized that after eight weeks of training the students in Group E would demonstrate more gains in prelanguage and early language skills than would students in Group C. These skills were measured by the MLAPFT before and after the training period. As previously explained, the pretest scores (X scores) for both groups were first averaged across pairs of students trained by the same trainer. This same procedure was repeated with the posttest scores (Y scores). Next, the error inherent in the measure itself, estimated.by the teSt- retest correlation coefficient, or K was removed from the gain made by the students by first multiplying the pretest score by the test-retest correlation coefficient. The test-retest correlation coefficient for the MLAPFT, as previously mentioned, was .9965. Finally, this product was subtracted from the Y scores to yield a Z score for each pair of students. iThe formula below summarizes this transformation: 2 = Y - KX Next, a one-way analysis of variance wae done on these Z scores to de- termine in a rather precise way which group's mean post score was greater. The mean Z score on the MLAPFT for Group E was 18.33 with a standard deviation of 13.49 while the mean Z score for Group C was 0.65 142 143 with a standard deviation of 3.09. Raw pre- and posttest scores for the total test and its seven phases for all pairs of students are re- ported in Appendix P. The results of the one-way analysis of variance of the index of response are reported in Table 8. (The decision to employ an analysis of variance of index of response was made prior to the determination of the near perfect test- retest correlation. Despite this fact this analysis was used rather than an analysis of variance of gain scores.) Table 8. Analysis of Variance of Index of Response on the MLAPFT Group Means of Mean Scores for Pairs of Students. T ‘_——‘ 4 Sources of Variation ' df MS F p less than Between groups 1 937.32 9.78 .01 Within groups 10 95.83 Total 11 These results are interpreted to mean that the gain in language skills in Group E was significantly greater than the gain in Group C at the .01 level of significance. Therefore the null hypothesis of no differ- ence between the groups in MLAPFT Z scores can be rejected. These results indicate that retarded trainers using a develop- mentally ordered language training program which incorporates princi- ples of modeling and reinforcement and being supervised by attendants skilled in applying the language program are significantly more effec-' tive at training language skills in profoundly retarded students than 144 are similar trainers untrained in these techniques and using unstruc- tured play methods of training with untrained attendant supervision. An examination of the raw gain scores (posttest score minus pretest score) of pairs of students in Table 9'reveals some interesting information on the effectiveness of specific trainers. In Group C it appears that the lack of structured training procedures resulted in little gain by students and no differentiation of trainer effectiveness. Table 9. Raw Gain Scores on the MLAPFT for Pairs of Students in Groups E and C. Group E Trainer Raw Group C Trainer Raw Gain Score Gain Score E-l +4.76 C-l +2.11 E-2 '+23.39 C-2 +2.57 E-3 -2.45 C-3 -3.41 E-4 +26.42 C-4 -3.7O E-5 +26.93 C-5 +1.67 E-6 +29.49 C-6 +3.30 However, in Group E, it is apparent that four of the trainers were re- sponsible for the significant mean gain of the entire group. Also, the students of two of the six trainers made little or no gain. These two trainers had the lowest IQ scores of all 12 trainers (l9 and 21) and, in addition, their students' 10 scores and attending skills were in the lower range of the student participants. One of these two trainers con- tinually forgot common aspects of the routine training procedure such as turning the training card, turning only one training card at a time, giving tokens along with praise, etc. It is difficult to know whether 145 the lack of gain in language skills is due to the less attentive and intelligent students, the less skillful and intelligent trainers, or an interaction between the two. In addition, these mean gain scores could mask a trainer who was effective with one student but ineffective with another student. However, an examination of the raw MLAPFT scores for single students did not reveal this occurrence in either group. Efforts were made to control the variable of training time. This variable affected all participants in the study. Students re- ceived instructional time from their trainers, trainers received in- structional time from the experimental staff and their supervising at- tendants while attendants received instructional time from experimen- tal staff. The time spent by students with trainers in language training sessions was not significantly different across all phases nor during any single phase and therefore this variable cannot account for the difference in gains made by the two groups. The F-tests on the mean number of training hours given to students, averaged first across pairs of students trained by a single trainer, then averaged across the groups, are reported in Table 10. The amount of time trainers spent with experimental staff during Phase I and with experimental staff and their supervising atten- dants during Phases II and III also was not significantly different be- tween the two groups as reported in Table 11. The amount of time attendants in both groups spent with ex- perimental staff, perhaps most important during Phase I and II, was 146 Table'HL. Group Means for Number of Training Hours During Phases for Groups E and C Students. Variables Group E Group C F* (n=6) (n=6) Phase I 2.56 2.83 .87 Phase II 2.45 2.75 .70 Phase III _ijEL_ ._Z;31_ 2.33 Total 13.22 12.48 .76 *No F-values were significant at the .05 level. Tablell. Group Means for Number of Hours with Experimental Staff and Supervising Attendants for Group E and C Trainers During Phases of the Study. Variables Group E Group C F* (n=6) (n=6) Phase I 8.33 7.50 .39 Phase II 7.67 7.33 .31 Phase III 21:90 ggpeg 3L2§_ Total 43.00 37.67 3.69 *No F-values were significant at the .05 level. also not significantly different between the two groups and cannot ac- count, therefore, for the differences in student gains. Table 12 re- ports these results; the total figure is a total taken across the two instructional phases, Phase I and 11, while Phase III, although listed, 147 did not include much contact time between experimental staff and atten- dant staff and merely represents the mean amount of time each attendant spent independently supervising their training units during Phase III. Table 12. Group Means for Number of Hours of Contact with Experimental Staff for Attendants in Groups E and C During the Phases of the Study. Variables Group E Group C F* (n=3) (n=3) Phase I 4.67 4.00** .40 Phase II 10.67 9.00 1.14 Phase III 26.33 24.67 .28 Total I and II 15.57 13.33 1.69 *No F-values were significant at the .05 level. **For two of the three attendants in Group C these hours of training were given during Phase II's time period although they were spent in Phase I activities: this was due to the replacement of two Group C attendants during the second and fourth weeks of the study and the need to make up training hours lost. In the interpretation of these results, the preceding vari- ables, alone or in combinations, cannot be regarded as those responsi- ble for significant differences between the groups, since they were controlled across the two groups. Therefore, the significant differ- ence in language skills demonstrated between the two groups on the post- test is interpreted as being due to the reinforcement and modeling meth- ods and the developmentally ordered, language training program in which Group E attendants and trainers were instructed. 148 It also was hypothesized that after eight weeks of training the students in Group E would demonstrate more gain in mental age than would students in Group C. Mental age was measured by the Cattell In— fant Intelligence Scale. The same procedure used to obtain 2 scores for the language measure just reported was repeated with the pretest, posttest mental age estimates. The test-retest correlation coefficient for the Cattell used in this transformation was .80. The results of the one-way analysis of variance are reported in Table 13. Table 13. Analysis of Variance of Index of Response on the Cattell Mental Age Group Means of Mean Scores for Pairs of Students. Sources of Variation df MS F p less than Grand Mean 1 247.88 131.18 .0001 Between Groups 1 1.87 0.99 .34 Within Groups 10 1.89 Total 12 The between group results in the second line of Table 11 are inter- preted to mean that the gain in mental age in Group E was not signifi- cantly greater than the gain in Group C at the .05 level of signifi- cance. Therefore, the null hypotheses of no difference between the two groups in the mental age Z scores cannot be rejected. The failure to find a significant difference between groups as reported in Table 13 can be partially explained in a number of ways. First, the length of treatment may not have been enough to affect men- tal age gains, although language skill gains were made. In a prior 149 chapter, the high stability of the MA in lower functioning retarded persons was discussed (London, 1961). In this study, the trend noted in the MA gain scores for both groups, 1.1 months in Group E and 0.8 months in Group C, is in the predicted direction, making plausible the hypothesis that the difference between the groups would increase as training time increased. A second explanation may be given for the results reported in Table 11 which concerns the skills tested by the Cattell. However, this explanation is somewhat incongruous with the preceding discussion regarding length of treatment. It was noted earlier that a smaller number of items in the Cattell test prelanguage and early language skills than test fine and gross motor skills, social skills, and cog- nitive skills other than language. For example, an examination of the test manual (Cattell, 1960) reveals that during the second to the sev- enth month, the main skills tested are motor skills. An informal task analysis of language related skills in the Cattell reveals that some attending skills are tested during the second to the seventh month and some motor and vocal imitation skills are tested from the eight to the twentieth month; but, the early receptive and expressive language skills are not tested until later months, primarily from the sixteenth through the thirtieth month. An examination of the students' mean pretest mental age re- veals that most students in both groups reached ceilings on the Cattell prior to the sixteenth month. Therefore, the average mental age for students in both groups was not as heavily dependent upon the language 150 skills trained in Group E as they were upon skills such as motor, so- cial, and cognitive skills other than language. It is of interest to examine gains in mental age apart from the above comparison of differences between the two groups. On the basis of previous research it would be predicted that retardates at this profound level would make no MA gains. Fisher and Zeaman, (1970, p. 161) referring to MA growth, state, "The Profounds flatten off be- fore 15 years." However, as reported in Table 14, gains were preva- lent in both groups when all but four of the 24 subjects were close to or over 15 years of age. Also three out of these four younger students made gains below the mean gain of 0.9 months which cannot, therefore, account for the growth in both groups. Eight of the subjects made zero gain or gains less than 0.5 months, the result that would be antici- pated. But many made gains over this short period that represent ap- preciable growth. When the grand mean MA gain for both groups is compared to the expected gain of O in a one-way analysis of variance of the index of response, the significance of this gain is illustrated. As shown in the first line of Table 13, this difference is significant at the .0001 level of probability. These results are interpreted to mean that the combined group gain in MA is significantly greater than the 0 gain which would be expected for similar students not spending time with re- tarded trainers or not being exposed to other instructional treatments. Despite the preparation of their trainers, students beyond the MA growth ceiling are still apt to make significant MA gains by merely spending time with retarded trainers. As mentioned earlier, this 151 Table 14. Chronological Ages, Pre- and Post-Mental Ages and Mental Age Gains for Students in Groups E and C. Student Pre CA (yrs.) Pre MA (mo.) Post MA (mo.) Gain (mo.) Group E l 16.08 18.2 18.8 0.6 2 18.75 16.4 17.2 0.8 3 20.08 13.6 16.8 3.2 13.00* 23.6 27.2 3.6 5 15.58 5.8 6.8 1.0 6 6.83* 7.2 7.6 0.4 7 19.83 22.4 22.4 0.0 17.75 26.4 27.8 1.4 9 19.42 19.2 19.2 0.0 10 21.33 24.2 26.0 2.2 11 19.33 29.8 29.8 0.0 12 25.33 25.6 25.6 0.0 Group Means 17.78 19.37 20.43 l.l** Group C 1 15.25 12.8 15.4 2.6 2 15.67 14.2 15.2 1.0 152 Table 14. Continued. Student Pre CA (yrs.) Pre MA (mo.) Post MA (mo.) Gain (mo.) 3 12.08* 7.4 8.0 0.6 4 14.67 21.0 21.2 0.2 5 19.08 15.2 16.0 0.8 6 19.17 18.2 18.2 0.0 7 19.75 23.4 24.4 1.0 8 20.08 27.8 27.8 0.0 9 14.50 20.4 21.6 1.2 10 24.50 23.2 22.2 1.0 11 17.75 18.6 19.4 0.8 12 7.33* 7.8 8.4 0.6 Group Means 16.65 17.50 18.15 0.8** *Students with CAs under the MA growth ceiling. **The transformed Z score gain difference between groups as reported in Table 13 is not significant at the .05 level. 153 gain could be due in part or total to an improvement in the students' test-taking behaviors--e.g., sitting, working with objects at a table, and attending to people and objects. Informal Questions To determine the quality of the training given to students by Group E trainers and indirectly by attendants, observations were gathered with the BOR. These data were evaluated to more carefully de- fine the actual procedures used by trainers and attendants in Group E as compared to Group C in order to provide objective evidence of the treatment differences. All the three-minute observations were taken in one of two different settings. 1. Regular training sessions with assigned students (Phase 1, Phase II, and Phase III observations). 2. Experimenter-structured training sessions with newly as- signed students (pre- and post-observations). Only retarded trainers were observed in the first setting, while both retarded trainers and attendants were observed training new students in the experimenter structured training sessions. The results of observations taken in regular training ses- sions and the related questions are reported and discussed first. Some of these observations were missing because of the illnesses or absences of trainers or students. Table 15 reports the number of observations taken in each group during each phase of the study. Since each trainer was observed with both students, a total of twelve observations was POS- sible for each group in each phase. Mean group scores were computed on 154 the number of observations taken in each group for each phase with the- number varying from 8 to 12. Table 15. Number of BOR Observations Taken of Trainers During Regular Training Sessions. Phase I Phase II Phase III Group E 8 12 12 Group C 8 10 11 Next, the results of the observations taken of trainers and of attendants in experimenter-structured training sessions and the re- lated questions are presented and discussed. No observations for at- tendants or for trainers were missing from the data collected in the experimenter-structured setting. In all the graphs Group E trainers and attendants are represented by solid lines while those in Group C are represented by broken lines. It was not the intent of the writer to statistically analyze these observational results, but instead to treat this data as explora- tory in nature and to seek out general trends rather than to test spe- cific hypotheses. Questions and BOR Results for Trainers in Regular Training_Sessions 1. Would Group E trainers have a greater average frequency or percentage of the following categories of training behaviors than trainers in Group C? 155 a. Number of complete trials presented. Figure 9 presents the results of the observations taken to answer this question. c—J Nwhmmuoosoo Number of complete trials presented —I . --------- —h ----- C ‘__. I II III Phases Figure 9. Mean Number of Complete Trials Presented by Trainers During Three-Minute Observations Taken During Regular Training Sessions at the End of Each Phase of the Study. Complete trials consist of a stimulus presentation given by the trainer, a correct or approximate response by the student, and a reinforcing con— sequence. Group E trainers consistently presented more complete trials than did trainers in Group C. Therefore, the students in Group E had more opportunities for learning to occur, i.e., correct and approximate responses given by the student after a stimulus presentation were posi- tively reinforced by the trainer. The most frequently missing event in 156 the Group C trials were the reinforcing consequences provided by the trainer. Although they presented stimuli to their students in the form of questions and commands with or without materials, Group C trainers more often failed to reinforce their student's approximate and correct responses. Therefore, although many trials may have been presented in Group C, they were often incomplete because the reinforcing consequences were not provided. Another factor contributed to the recording of in- complete trials: namely, the omission of stimulus presentations in the form of commands or questions. For example, in a few observations the student repeatedly dropped poker chips into a can while the trainer in; termittently talked to the student but did not attempt to direct the student's responses. The drop in mean number of complete trials presented in both groups over the three phases is a trend noticeable in most of the cate- gories of training behavior observed. This phenomenon will be discussed in the final chapter. b. Number of trials when the student was attending prior to a stimulus presentation. Figure 10 presents the results of the observations taken to answer this question. Group E students were observed to have a higher mean number of attending responses made just prior to the trainer's stimulus pre- sentation than did Group C trainers. Attending included the times a student attended on his own as well as the times he was prompted by the trainer to attend preceding the presentation of stimulus materials, question or command. The failure to reinforce attending and other 157 responses less often in Group C probably resulted in the observed lower frequency of attending responses. Also, the lowered attention level would most likely result in a lower rate of learning and therefore cor- rect responses would be made less often. 10 U1 co 3 9 pm a: ‘U: 8 :E 1333 can 7 K - E : er \ 30 6 \ "" \ £5. \ no.0 5 \ EC: \ 44:13:34 \ earn up“ 5.11:3 mcm .omm EHO) :us. 2 zone. 1 I II III Phases Figure 10. Mean Number of Trials when the Student was Attending Prior to a Stimulus Presentation During Three-Minute Observations Taken During Regular Training Sessions at the End of Each Phase of the Study. The decreasing trend was again observed in both groups across the first two phases. However, the frequency of attending response re- mained stable from the second to the third phase in both groups. c. Number of student responses. 158 Figure 11 presents the results of the observations taken to answer this question. 11 10 in (I) 2 9 O Q. 8 8 S. +e E .92; 7o\\ - - 3 12’ 6 \\ \ “as x w \, .Q 4 ’.—o e \x ,3” z 3 .,-’ 2 1 I II III .Phases Figure 11. Mean Number of Responses Made by the Student During Three-Minute Observations Taken During Regular Training Sessions at the End of Each Phase of the Study. Responses counted included incorrect as well as correct and approximate responses. Students in Group E made more responses than did Group C students. Although this graph does not reflect the intent or correctness of the student's responses, mere frequency still reflects a group difference. A large decrease in responses exists in both groups from Phase I to Phase II. 159 The differences between Figure 9, number of complete trials, and Figure 11, number of student responses, for both groups illustrates some interesting points. Since Figure 9 includes only correct and ap- proximate responses and Figure 11 includes incorrect responses as well, it seems that Group E students made few incorrect responses during the observations. The close similarity in the curves for Group E in these two figures may be explained by the fact that trainers were not only taught to train at levels just above the student's independent level of responding but also because these trainers quickly prompted the stu- dent when he failed to respond on his own. Both these factors resulted in more approximate and correct responses, and the potential for more complete trials. However in Group C, the curves in Figure 11 and Fig- ure 9 are not parallel. This may be due to the lack of reinforcing consequences or stimulus presentations needed for complete trials or perhaps it is due to the lower attending levels and higher level of in- correct responses. d. Percentage of correct and approximate student responses reinforced by the trainer. Figure 12 presents the results of the observations taken to answer this question. The data in Figure 12 suggests that Group C trainers fail to reinforce students for correct and approximate responses while Group E trainers display a consistently high tendency to reinforce students. Despite the general reinforcing quality of the daily training sessions for students in both groups, the additional differential reinforcement given by Group E trainers was necessary for specific learning to occur. 160 8 100 rum % L E 5.: - - §'”' 90 ‘7' 82’ Etta 80 tom 8 Bo 70 cosi— +4c 6'5 60 ms. to. o m 50 Um : “5g 40 832’ a o /\ 8449.3 [I \\C cccu / \\ mmczo // \ 83".; ,/ ‘\ was. ' \o mind-3 10 Phase I Phase II Phase III Figure 12. Mean Percentage of Correct and Approximate Student Responses Reinforced by the Trainer During Three- Minute Observations Taken During Regular Training Sessions at the End of Each Phase of the Study. d. Percentage of tangible reinforcers given simultaneously with praise. Figure 13 presents the results of the observations taken to answer this question. Tangible reinforcers consisted of food, toys, or tokens, while praise generally took the forms of "Good boy (girl)," or "Good," Ioften coupled with smiles and/or pats. Tangible reinforcers had to be given in order for a percentage score to be calculated for a trainer. Some trainers did not give any tangible reinforcers during an observa- tion. This is illustrated in Figure 13 by Group C during Phases II and III where no tangible reinforcers were given, thereby resulting in an 161 100 90 .777777777777777“-_2 E 80 7O 60 50 40 30 20 Percentage of tangible reinforcers given simultaneously with praise I II III Phases Figure 13. Mean Percentage of Tangible Reinforcers Given by the Trainer Simultaneously with Praise During Three-Minute Observations Taken During Regular Training Sessions at the End of Each Phase of the Study. absence of percentage scores for these times. Table H5 presents the number of trainers giving tangible reinforcers upon which the percent- ages in Figure 13 are based. For example, in Figure 13 the mean per- centage score of 99 percent for Group E during Phase I was based on the performance of seven of the eight trainers observed. One trainer out of eight did not have an opportunity to give tangible reinforcers, while the seven that were observed to give tangible reinforcers to students praised them simultaneously on the average of 99 percent of the time. In this category, Group E trainers generally gave tangible re- inforcers simultaneously with praise. Group C trainers were rarely 162 Table 16. Number of Trainers Giving Tangible Reinforcers During Observations of Training Sessions. Phase I Phase II Phase III Group E 7 of 8 10 of 12 9 of 12 Group C 3 of 8 O of 10 O of 11 observed to give tangible reinforcers and when they did, during Phase I, no praise was coupled with the tangible reinforcers. Since many students were not observed to be reinforced by praise alone, food or toy reinforcers had been identified and were made available for all students. These were rarely used by Group C trainers. Token reinforce- ment procedures, taught to Group E trainers, were never used by Group C trainers. Therefore, the opportunity to teach students to associate praise with tangible reinforcers was not used by Group C trainers. The information from Figures 12 and 13 together leads one to conclude that the majority of "reinforcers" given by Group C trainers, especially in Phase II and III, were praise; and the reinforcing power of this praise, because it was not coupled with tangible reinforcers, can be questioned. However, Group E students by being exposed to tan— gible reinforcement and praise simultaneously were learning to regard the trainer's praise as reinforcing. e. Percentage of appropriate prompts. Figure 14 presents the results of the observations taken to answer this question. 163 m 100 .p § 9o & a, 80 13’ 'C 70 D. 2 60 - E - 2' q_ so 0 g, 40 ’lo_____c _____ . 3 ,I’ g 30 .,,” t.“ 33 20 10 I ' II III Phases Figure 14. Mean Percentage of Appropriate Prompts Given by the Trainer During Three-Minute Observa- tions Taken During Training Sessions at the End of Each Phase of the Study. Appropriate prompts included those times when a student made an approximate, an incorrect, or a no response on the preceding trial and the same stimulus presentation was repeated along with student prompting; or the same stimulus presentation was merely repeated with- out prompts on the next trial, but this resulted in an improved level of response (approximate or correct). Prompts had to be given by a trainer in order for a percentage score to be calculated. However, some observations had no incidents of prompting--either the student was responding well and needed no assistance or the trainer failed to prompt every time it was needed. Table 17 presents the number of trainers giving prompts, appropriate or inappropriate, during the 164 observations and upon which the percentages in Figure 14 are based. Therefore in Figure 14 the mean percentage score of 93 percent for Group E during Phase I was based on the prompting performance of seven trainers out of eight observed. One trainer did not have an opportun- ity to prompt while the other seven observed, prompted appropriately on the average of 93 percent of the time. Table 17. Number of Trainers Prompting Students During Observations of Training Sessions. Phase I Phase II Phase III Group E 7 of 8 6 of 12 7 of 12 Group C 5 of 8 5 of 10 7 of 11 Based on this smaller number of observations Group E trainers still tended to prompt appropriately more often than did Group C train- ers. The rules for appropriate prompting perhaps were more difficult for retarded trainers to implement than were some of the other catego- ries of instructional technique taught to Group E trainers. Group E demonstrated a rather large drop during Phase II in their observed abil- ity to appropriately prompt while Group C trainers improved somewhat. 2. Would Group E trainers have higher average judgment ratings than Group C trainers in the following areas? a. Consistency and appropriate repetition of stimulus presen- tations. 165 Figure 15 presents the results of the observations from which judgments were made to answer this question. 04-, 44:6 IUQJ .I—m 0L0) 50.5- 5 “CO. L . ’58:; War-4 #0 seem :xo 7 4 LU: 2 :0 HQ)!- CH-IJ C GU)!- E.PH] -_-’—---.- ------- . mend) ....... UCD. 3°C) 7 Phases Figure 15. Mean Judgment Ratings for the Consistency and Appropriate Repetition of Stimulus Presenta-' tions by Trainers During Three-Minute Observa- tions Taken During Regular Training Sessions at the End of EaCh.Phase of the Study. Points were given to an observation depending upon how often and in what order stimuli were presented to the student. Group E trainers were programmed to present tasks in a particular order with certain repetitions depending upon how the picture task training cards were arranged by the experimenter or the supervising attendant. There- fore they tended to present tasks more consistently and with more ap- propriate repetitions than did Group C trainers who did not have train- ing cards as guides. b. Appropriate fading of prompts. Table 18 presents the results of the observations from which the judgments were made to answer this question. 166 Table 18. Mean Percentage of Appropriate Fading of Prompts, Over- prompting, Unnecessary Prompting, or No Opportunity to Prompt Done by Trainers During Training Session Observa- tions. 3 Phase I Phase II Phase III Group E Group C Group E Group C Group E Group C Judgment: OK good fading of prompts 25 O 0 0 17 9 Judgment: + overprompting 12 38 17 O 17 9 Judgment: - Unnecessary Prompting 25 25 O 0 8 18 Judgment: N.0. No opportunity to fade prompts 38 38 83 100 58 63 On the average Group E trainers tended to fade prompts more often than Group C trainers. The percentages also seem to indicate that there were slightly more times Group C trainers did not have an opportunity to fade prompts. This could in turn be due to an inappro- priate level of task difficulty; when tasks are too difficult there would be no opportunity to fade prompts and when tasks are too easy there would be no opportunity to prompt. No trend is apparent in the two inappropriate fading techniques used by trainers. The skills of prompting and fading of prompts are difficult even for the nonretarded trainer and, as is illustrated by these results, also proved difficult for the supervised trained retarded trainer. 167 c. Appropriateness of task to student. Table 19 presents the results of the observations from which the judgments were made to answer this question. Table 19. Mean Percentage of Trainers Providing Appropriate, Difficult, or Easy Tasks for Students as Observed During Training Ses- sions. Phase I Phase II Phase III Group E Group C Group E Group C Group E Group C Judgment: Appropriate 88 38 75 50 75 27 Judgment: too difficult O O 0 20 O 18 Judgment: too easy 12 62 25 30 25 55 Although differences are not as large as with some other cat- egories, it is still apparent that Group E trainers performed better than Group C trainers--they presented tasks at appropriate levels of difficulty more often than Group C trainers. That is, their students tended to make many correct and approximate responses but still needed intermittent prompting when errors were made. Group E trainers also made errors in the level of task difficulty less often than Group C trainers. It is interesting to note that Group E trainers were never judged to make the tasks too difficult during the observations taken, although at times they did make the tasks too easy. The judgment that the tasks were too easy refers to a performance in which a student made 168 no errors or very few, and in which he gave all correct responses need- ing no prompting (i.e., prompts given following an error on the same task). This would appear to be a lesser training error than making tasks too difficult. In fact, review training or overlearning of newly acquired tasks is used as an effective strategy to increase the re- tardates' retention of learned responses (Denny, 1966; Kent, 1972). The retarded appear to benefit more from repeated presentations (trials) during original learning (overlearning) than do normals, and like normals, this repetition gives the biggest return when it is fairly well spaced within a session and across sessions. Such a procedure permits many different stimulus elements to partici- pate in the control of the approximate response (meaning). With the repeated use of many different stimulus contexts, the effects of stimulus generalization are greatly increased, and transfer be- comes possible (Denny, 1966, p. 8). It is of interest to contrast the mean BOR scores of indivi- dual Group E trainers in the nine categories of training behavior ob- served during regular training sessions with the mean untransformed MLAPFT gain made by each trainers' students. This information is pre- sented in Table 20. The skill of Group E trainers to produce gains in the language scores of their students is reflected in varying degrees by the first three BOR categories listed in Table 20, 1. number of complete trials presented 2. number of trials when the student was attending prior to a stimulus presentation 3. number of student responses In Table 20 trainers 1 and 3, whose students made little or no language gain, consistently demonstrated lower mean frequencies of these train- ing behaviors in an average three-minute observation taken during reg- ular training sessions. 169 ms.m~+ mo.u~+ Ne.m~+ me.~i mm.m~+ om.e + . smswmsh «saw us» xn umspssh masmuaum u ssosw eo mewss sue museum swag hsswm msmusoeswms anwmswp mo waspsmusma .m ROOF soap Nmm gum gmm New swswmsp xn umusoeswms mmmsosmms musswxosase use uumssou mo mmepsmusws .e m.~_ m.m m.m m.m m.u m.m mmmsosmms usmuaum eo,smnEus .m n.—F m.pp ~.m o.m w.o w.m sowueu ismmmss mzpzswum e on Loves oswuswuum mm: usmuaum msu smsz mpswsu eo smasas .N o.__ m.~ 5.x m.s o.m m.m umuswmmss mpmwsp mumpasou we smsE:s .P o m e m N P smswesp mmwsommuuo msoum mom .mswms «smusum m ssoso sum museum spew Pss<42 uwssoemsssusu use smswosp zoom on umsmwmm< musmuzpm o3» use mummss mmssh asp mmosu< uomssm>< use msowmmmm mswswus» sopzmmm ms_s=o smxm» msmspmsp m asoso sou museum mom .om mpnmb 170 To a lesser degree, the following BOR score categories listed in Table 20 reflected somewhat less consistently the effectiveness of a Group E trainer to obtain language gains in his students; 5. percentage of tangible reinforcers given simultaneously with praise 6. percentage of appropriate prompts 8. appropriate fading of prompts Categories 4, 7, and 9, listed in Table 20. did not show this differentiation between the effective and the ineffective trainers in Group E. 3. 0n the average did Group E attendants and experimental staff reinforce trainers more often than Group C attendants and experi- mental staff? Figure l6 presents the results of the observations from which judgments were made to answer this question. Number of times trainer was reinforced by attendant or experi- mental staff Figure 16. NCO->01 1 . -------- - ‘- ~~ ‘~ ~. I ‘II III Phases Mean Number of Reinforcements Given to Trainers by Attendants and Experimental Staff in Groups E and C During Three-Minute Observations Taken at the End of Each Phase of the Study. 171 These reinforcements generally consisted of praise given in a nonspe- cific form, e.g., "Good work, Sally," or in a specific form, e.g., "You got Alex to say 'ball'! Great!" The trend again favors Group E atten- dants and experimental staff in that they gave slightly more praise to their trainers than did Group C. It should be remembered that during Phase I these results are reported for experimental staff since attendants did not interact with trainers then; during Phase II the results reflect experimental staff and attendants while during Phase III, the results reflect only the at- tendant's performance in praising trainers. As discussed earlier, at- tendants tended to praise trainers less often than experimental staff and had to be encouraged to increase their efforts. The decline in Phase III could reflect this. however, during Phase III attendants were busy supervising two trainer-student pairs rather than only one. In addition, attendants were encouraged and able to lessen their praise and instruction to trainers since they were more capable and confident at this time. 4. On the average, did Group E attendants and experimental staff interrupt training with instructional comments directed towards the trainer more often than Group C attendants and experimental staff? Figure 17 presents the results of the observations from which judgments were made to answer this question. Instructional comments to trainers included demonstrations on the student, verbal explanations and verbal commands to modify the trainer's techniques of working with a student. Again the experimental staff and attendants in Group E tended to give slightly more instructional 172 Number of times attendants and experimental staff gave instructional com- ments to trainers Phases Figure 17. Mean Number of.Instructional Comments and Demon- strations Given to Trainers by Attendants and Experimental Staff in Groups E and C During Three-Minute Observations Taken at the End of Each Phase of the Study. demonstrations and comments to their trainers than did Group C. In both groups more instruction was given during the first phase. As with question three, the Phase I results reflect the performance of the ex- perimental staff rather than the attendants. These results could mean that experimental staff were more apt to give instructions to trainers than were attendants and/or that more instruction was needed in the first phase than in other phases. The number of instructional comments did decline as was expected since trainers in both groups became more independent in their skills over the eight weeks of the study. Questions and BOR Results for Trainers in Experimenter-Structured Training Sessions l. Would Group E trainers have a greater average post-study frequency or percentage of the six categories of training behavior ob- served than trainers in Group C? 173 Table 21 presents the results of the observations taken of the frequency and percentage categories during experimenter-structured training sessions and repeats the results from regular training ses- sions. Group E trainers did consistently better on all post-study fre- quency and percentage scores for the six training behaviors observed. The numbers of trainers upon which percentage scores e and f are based are reported in the footnotes of Table 21. 2. Would Group E trainers have higher average post-study judgment ratings than Group C trainers in the three categories of train- ing behavior observed? Table 22 presents the results of the observations taken of the judgment categories during experimenter-structured training sessions and repeats the results from regular training sessions. Again Group E trainers did consistently better on all post-study judgment scores for the three training behaviors observed. 3. In the skill categories listed in questions 1 and 2, did the average frequency, percentage, and rating scores of the trainers during the post-study observations in Groups E and C decrease from their scores during the training observations? Rather than decrease, the majority of scores in the post-study observation for both groups increased over their Phase III training ses- sion observations. All post-study scores marked by a single asterisk in Tables 2l and 22 reflect this increase while those in Table 22 marked by a double asterisk are judgment scores whose improvement over Phase III scores are reflected by a decline, e.g., inappropriate fading of prompts. 174 0": m": megeum HHH amuse can» empemgm mmgeum xeeumsumeee "xeeumuumee o eeecu mu: "xeepm-mce o eeeew ”meepmuumee m eeegw mu: "xeepm-eee m eeeeo epeemce> menu :e ee>cemee eeecm see c we “so mcmcweeu me geesecm ”aeepmupmee u eeecw _uc ”xeeumueee u eeeee ”aneumuumee m eeecw on: "xeepm-ece m eeeew .epeewce> menu.ce em>eemee eeeem Lee c we use meeewecu we geese: C: _.N M" _ «MN MW WM mm mm w mmueeeee eueweeeeeee ee mmeueeeeme .w «cop .o.z .o.z o em 0 Fem_ece sew: x~meeeceppeswm acop mm mm mm .o.z m ce>wm mgmueemcwme mFewmceu $0 wmeucmueee .m mm mm Fm mp _e o emecemcwee memceemec emu um we mm mm m mpeewxeeeee ece ueeeeee we emepeeeeme .e «o e m n o_ u m_ n N PF 0 w mmmceemmm uceeeem we geese: .e «w e e N m_ e cewpeucemece wepeswum e» Levee «mp N N _F e u meeeceeea we: eeaeeem can; m_~wee eo consaz .e mm m M m_ m w eeucemece mpewcw mumpeseu we emesez .e lumen emcee omega 5 ewes; .necn. mmeucmucme ece xueeeemcuv A.mweeum .mcevmmem mewcweeh Lepemwm cw nee mcevmmmm me_c_eep emeeueecum Iceucmsweeexm cw memeweeh we mcewue>eemeo mom we mupemem mmeuceeeee use xeeeeeeee new: .FN m_eeh 175 eewcwpuwe an mmceom HHH emcee em>e ee>ecesv museum xeepmuemeesk meeeUm HHH omega can» eeueecm meceUm xeeum-umeek ea¢. mm On mm o u **o mm mm up up u m kko o o o o u o o o o o + u «oo_ mm om mm oo_ o «cop mu mu mm mm < m pceeeum e» xmeu we mmecmue_eeeeee< .e «kmm mo cop mm mm 0 «exp mm mm mm mm .o.z m *«o mp 0 mm up u e«o m o mm mm - u sue a 0 mm o u «to up up NP 0 + m esp m o o o u eme N. e mm o ee e measoeg co mcweee weareaoeaa< .a «m p F P m o meewaeucmmmee mepeswum me «e m m e N u cewuwueeme wueweeeeeee ese auceumwmeeo .e xeaem HHH He H xe=em aaoce numee mmese emcee emcee -eee A.meeeem peeEmeeev .m:e_mmem mcwcwech Lepemem cw ece mcevmmem merevee» emeeuoecum-eeuceewcoexm cw mcmcwech we meewpe>cmmeo mom we muyemem «emsmeee new: .NN epeeh 176 Trainers in both groups were able to successfully generalize the training behaviors from their assigned students in regular train- ing sessions to a new student in a training session structured by the experimenter. Reasons for this successful generalization may be the changed working situation where the trainer was told specifically what activities to do with the new student, worked only a few minutes, and was positively encouraged not to hurry but to spend all his time work- ing with the student. Before these observations, efforts were taken to make the trainer feel relaxed, special, and valuable as a teacher. All of the trainers had a friendly working relationship with the ex- perimenter who did these observations. Each was individually taken to a quiet room on their ward and individually observed. Each generally received a large amount of attention and vocal reinforcement before and after the observation. Only if a trainer demonstrated fear did the ob- server offer encouraging words during the observation time; and, in most observations, nothing was said to the trainer during the three min- utes so as not to influence his performance. In summary, during the post-study observation trainers were very'aware yet comfortable that they were being observed while this same awareness was not generally present when trainees were observed during the regular training sessions. During those latter observations, train- ers were not told their teaching performance was being observed and were not necessarily aware the observer was even present. It seems that the trainer's performance improved when, in a well-defined and positive sit- uation, he was aware of being observed. 177 4. Did these average frequency, percentage and rating scores of the trainers in Group E decrease less than those of trainers in Group C? As discussed above, the average post-study scores of the trainers increased for both groups, but all the scores of the Group E trainers increased more than did those of the Group C trainers. 5. Were the training skills observed to be approximately equal in both groups of trainers during the pre-study observations? As measured by the pre-stUdy Observations shown in Table 22, the training skills of both groups of trainers were approximately the same. However, it should be noted that Group C trainers performed bet- ter than Group E trainers in five of the nine categories observed: num- ber of trials where student was attending; number of student responses; percentage of tangible reinforcers given simultaneously with praise; and appropriateness of task to student. Group E trainers performed bet- ter than Group C trainers in only one category: percentage of appro- priate prompts. However,it is not known whether these differences are meaningful. In any event, it appears that Group E trainers did not have a pre-study advantage over Group C in the training behaviors ob— served. Questions and BOR Resultsfgr Attendants in Experimenter-Structured Training Sessions 1. Would Group E attendants have a greater average post-study frequency or percentage of the six categories of training behavior ob- served than attendants in Group C? 178 Table 23 presents the results of the observations taken of the frequency and percentage categories during experimenter-structured training sessions. Except one score which was the same for both groups, all post-study scores were better for attendants in Group E than Group C attendants. The number of attendants observed that gave tangible re— inforcers (e) and prompts (f) upon which these group scores are based are indicated in the footnotes of Table 23. 2. Would Group E attendants have higher average post-study judgment ratings than Group C trainers in the three categories of train- ing behavior observed? All post-study judgment scores in Table 23 were the same or better for Group E attendants than for Group C attendants. Throughout the observations, Group E attendants improved in their skills of di- rectly training a student although the majority of their instruction and supervised practice during Phase I and II was at the level of teach- ing retarded trainers rather than students. It is important to note that there was less difference of training skills between groups of attendants than between groups of trainers. A comparison of attendants and trainers within groups also reveals some interesting findings. Group C attendants performed better in all nine post-study categories of training behavior than did Group C trainers; while Group E attendants performed about the same or some- what better than Group E trainers in all but one of the nine categories observed. This exception in superiority of scores occurred when Group E attendants faded prompts appropriately 67 percent of the time while Group E trainers did so 83 percent of the time. 179 Table 23. Mean Frequency, Percentage and Judgment Results of BOR Observations of Attendants in Experimenter-Structured Training Sessions. Pre-study Post-study .Group E Group C Group E Group C Frequency and Percentage Scores a. Number of complete trials b. Number of trials where student was attending prior to stimulus presentations c. Number of student responses d. Percentage responses reinforced e. Percentage tangible reinforcers paired with praise1 f. Percentage of appropriate prompts? 1 0 12 9 11 12 17 14 9 11 17 16 15% 41% 94% 76% 25% N.0. 100% 100% 40% 33% 89% 56% Judgment Scores a. Consistency and appropriate repetition b. Appropriate fading of prompts c. Appropriateness of task to student 7 7 8 7 0K 33% 33% 67% 33% + 0% 0% 0% 0% - 67% 67% 0% 33% A 100% 67% 100% 100% + 0% 0% 0% 0% - 0% 33% 0% 0% 1 this variable: Group E preestudy: n=2 Group C preestudy: n=O number of attendants out of 3 in each group observed on Group E post-study: n=3 Group C post-study: n=l 2All attendants in both groups were observed on this vari- able (n=3). 180 A final comparison can be made in the post-study scores of Group C attendants and Group E trainers. These attendants' and train- ers' scores were the same or only slightly different in all categories except the two concerning prompting. Group E trainers prompted appro- priately 75 percent of the time in the post-study observations while Group C attendants did so only 56 percent of the time. Appropriate . fading of prompts was done by Group E trainers 83 percent of the time M1 but only 33 percent of the time by Group C attendants. In this same category, the trainers in Group E were not observed to give any unnec- essary prompts while Group C attendants did so 33 percent of the time. From these results one can conclude that if retarded trainers are ex- posed to intensive instruction in reinforcement and modeling procedures and MLAP training sequence, as was provided by the Group E treatment, their training skills are comparable to attendants exposed to Group C treatment; and, in the skill areas of prompting and fading of prompts, these trainers perform better than the attendants. 8. Were the training skills observed to be approximately equal in both groups of attendants during the pre-study observations? As shown in Table 23, both groups of attendants, when ob- served before the study, were similar in their performance in all but two of nine categories of training skills. Group E attendants were better in training at the appropriate level of difficulty for the stu- dent, while Group C attendants reinforced the student's correct and approximate responses 4l percent of the time as compared to Group E's reinforcing 15 percent of the time. 181 Informal Accounts of Training Sessions_by Attendants and Experimental Staff Informal results described in daily notes taken by experimen- tal staff and supervising attendants reveal additional information on the day-to-day content and progress of training in both groups. Daily logs were used mainly by Group C attendants and experimental staff to describe, in more detail than Group E accounts, the activities between trainers and students, progress, complaints and whenever training ses- sions were cancelled. Group E experimental staff and attendants used their training session schedules to record the MLAP parts trained each day and some brief descriptions of significant occurrences. During Phase I all these written accounts were made by the experimental staff. During Phase II accounts were recorded both by experimental staff and by attendants while during Phase III only attendants recorded descrip- tive accounts of the training sessions. Examples Taken from Group_E Accounts Each example begins with the phase in which it was recorded, followed by the person recording, and the actual comments; students and trainers are identified in parentheses. Phase I, experimenter: Joanie (trainer) remembered to say "Do this," today when training imitation tasks with Lyle (student). Phase I, experimenter: Went well today; Ralph (trainer) needs only one demonstration with David (student) and then he can duplicate it per- fectly! Phase I, experimenter: Bob (trainer) always begins a session by greet- ing his student: "Hi. How are you today Arthur?” And his student always smiles back. 182 Phase I, experimental staff: Matt (student) is imitating all the tasks with objects. Often without apparent reason he grabs onto Roger's (trainer) arm; Roger gently releases his grip and continues. Will try a time out with Matt tomorrow whenever he picks at his face or stares at his hand--both of which interfere with training. Phase II, attendant: David (student) started using the two-syllable "ta-ta," instead of the usual "ta" for table today! Phase III, attendant: Bob (trainer) used time outs with Connie (stu- dent) once at the table and once in a chair in the corner. She was not very cooperative. Phase III, attendant: Kerry (student) would look at Linda (trainer) for part 3a (gross vocal imitation) and open his mouth, watching her carefully! That's progress and an approximation! Phase III, attendant: Stevie (student) touched his eye on Linda's com- mand (trainer) without any prompt--the first time! Phase III, attendant: Joe (student) began whispering labels again to- day; Roger (trainer) gave him a time out. Phase III, attendant: Matt (student) knows "march!" Did it three times in a row without help. Examples from Group C Accounts Phase I, experimental staff: Tom (trainer) really needs a lot of en- couragement still and tends to have a negative attitude toward work sometimes. Mark's (student) attention and eye contact are better than yesterday. Phase I, experimental staff: Tom (trainer) immediately got out the cowboy hat for Mark (student) remembering how much he loves it. Then he combed Mark's hair--something he loves. Tom is doing amazingly bet- ter today. Hopefully his self-confidence is increasing. Phase 1, experimental staff: The second day and Sally (trainer) is al- ready going to the ward and getting Danny (student) by herself, setting timer, getting objects out, and giving Danny juice on her own. Phase I, experimental staff: When we picked up Joe (student) today the attendant looked happy to get rid of him. He was engaging in varied stereotypic behaviors. Once he was sitting down with Sally (trainer) and had things to do, this behavior terminated. Phase 1, experimental staff: Kathy (student) ran around room scream- ing, threw reinforcers, and toys across the room. 183 Phase 1, experimental staff: Implemented behavior modification pro- gram today with Kathy (student). Ruth Ann (trainer) cooperated. The three of us played ball on the floor together until Kathy threw ball and began screaming; we ignored her and continued to play ball and left her out. When Kathy was quiet we allowed her to play with us. Phase II, experimental staff: Ruth Ann (trainer) has better control over Kathy (student) than I do--as long as she is not yelling. When Ruth Ann yells, Kathy gets upset and either runs out of the room or throws things. Kathy seems to enjoy walking hand-in-hand with Ruth Ann. Kathy seems to throw things so Ruth Ann will get after her to pick them up. Phase II, experimental staff: Cindy (trainer) now calls out “Betty Jean" before working on any specific tasks and waits for her attention. She also uses very intermittent reinforcement--says, "Good." Phase II, attendant: Cindy (trainer) waited for suggestions. Opened toy box herself and took out ball. Played with ball and doll herself ignoring her student. Following my suggestion and demonstration she showed Betty Jean (student) how to put chips in can, then handed Betty the chips. Betty and Cindy both laughed frequently and speeded up their movements. Phase II, attendant: Cindy (trainer) showed Steve (student) how to play chimes. Gave him a cracker. Gave him bell to ring. Showed him the car, bell, and comb. He made "oh" sounds and said "car" and made sounds for bell and comb. Phase II, attendant and experimental staff: Ruth Ann (trainer) was playing with Gino (student) on the mat--tickling him and putting his socks on. Ruth Ann complained about her health often in previous ses- sions. Did not while playing with Gino. She attended to Gino instead of relating his news or gripes to attendant. Phase 11, experimental staff: Belinda (trainer) held up objects and asked Mary (student) "What is this?" Belinda's praise was unenthusi- astic. Belinda asked her to name body parts, but Mary always repeated "eye . ll Phase III, attendant: All students and trainers did very good today. Mitch (student) will get his own chair. Joe and Danny (students) were good at stringing different objects with Sally (trainer) to help. Phase III, attendant: Tom worked better when I told him he would not get paid unless he tried harder with his students. It worked very well. I think maybe they are getting a little bored with the same things. I will get some different toys from C-building next week. Phase III, attendant: Videotaping today. Mitch and Danny (students) very good. Tom (trainer) is still working very good when he thinks he will not get paid unless he does! 184 Phase III, attendant: Gino (student) took hammer from Ruth Ann (trainer) and pounded peg board which she held for him. He held hammer and moved it without direct prompting! Then repeated several more times without prompting! The value of these written accounts lies mainly in their il- lustration of the day-to-day problems and successes occurring within the individual training units. Also, these written accounts tend to support the BOR findings. Summar Based on the results of the formal and informal analyses and restricted to the population from which the sample for this study was drawn, certain generalizations may be stated. 1. Retarded trainers using a developmentally ordered language training program which incorporates principles of modeling and rein- forcement and supervised by attendants skilled in applying the language program are a. more effective at training prelanguage and early language skills in profoundly retarded students; but b. are not more effective at increasing the mental ages of profoundly retarded students than are similar trainers un- trained in these techniques and using unstructured play methods of training with untrained attendant supervision. 2. Regardless of being trained in operant and developmental language training techniques or being supervised by trained or untrained attendants, students spending time each day with supervised retarded 185 trainers are likely to make significant mental age gains over students not exposed to retarded trainers. 3. In comparison to trainers using structured play methods of training with untrained attendant supervision, retarded trainers, using operant and developmental language training techniques and super- vised by attendants trained in the same skills, were observed to con- sistently and more frequently a. present complete trials to students, b. have students attending to stimulus presentations, c. evoke student responses, d. demonstrate appropriate reinforcement of students' re- sponses, e. couple tangible reinforcers with praise, f. demonstrate appropriate prompting of students, 9. provide consistent stimulus presentations with the appro- priate repetition, h. fade prompts appropriately, i. teach tasks appropriate to the students, and j. successfully generalize their training skills to new stu- dents in experimenter-structured training situations. 4. In comparison to untrained attendants, attendants skilled in the application of an operant, developmentally ordered language training program were observed to consistently and more frequently a. provide reinforcement and instruction to the retarded trainers they supervised, and 186 b. demonstrate a high quality of training skills when asked to instruct a student in a particular early language task (these training skills are described in 3a through 3i). CHAPTER IV SUMMARY AND CONCLUSIONS Summary of the Study Within most institutions for the retarded are barriers which prevent adequate intellectual stimulation of its residents. These bar- riers include the low ratio of attendants to residents, the low ratio of education and speech professionals to residents, the inadequate preparation and motivation of attendants to train residents, and the administrative priorities which in practice generally do not result in adequate educational programming for residents. As a consequence of these factors, institutions for the retarded have been shown to retard even further the social, cognitive, and language skills of their resi- dents. The purpose of this study was to demonstrate and compare the effectiveness of a strategy for utilizing the higher functioning resi- dents of an institution to train the lower functioning residents under the guidance of the attendant staff. Specifically, this study sought to determine whether significant gains in language functioning and in the ability to respond to a standard measure of intelligence could be produced by retarded trainers who were trained to administer a develop- mentally ordered language program which incorporated principles of re- inforcement and modeling (Group E). In order to determine the 187 188 effectiveness of the instruction given to these retarded trainers and their supervising attendants in Group E, a control group (Group C) was employed where trainers were encouraged to provide their students with reinforcement on a noncontingent basis, verbal and tactual stimulation, training based on their own or their attendant's ideas, and play inter- action. Identical in both groups, the organization consisted of basic training units. Each of these units was composed of one supervising attendant who directed two retarded trainers. Each trainer, in turn, was assigned to two retarded students, although all training was done with one student at a time. Each group included three of these train- ing units or a total of three attendants, six trainers and twelve stu- dents. Trainers and students were paired together first and then ran- domly divided into the two groups. Attendants were randomly divided into two groups and then paired with two trainer-student trfiads. Regarding the characteristics of the students averaged across both groups, the mean chronological age was 17 years; mental age was one year and 10 months; IQ was ll and there was an average of 12 years of institutionalization. Retarded trainers had a mean CA of 28 years; MA of 7 years 4 months; IQ of 43 and l4 years of institutionalization. With the exception of one characteristic of the trainers, neither the two groups of students nor of trainers were significantly different on these characteristics. The exception occurred when Group C trainers were found to be institutionalized significantly fewer years than were Group E trainers; a factor which could have acted to favor Group C. 189 Attendants participating in the study had a mean age of 29 years and worked a mean of 2 l/2 years in institutional settings. Neither group was significantly different from the other on these two characteristics. The experiment consisted of three phases. Phase I, lasting two weeks, was an initial training period which varied for Groups E and C. Phase II, lasting three weeks, involved the training of atten- dants in the supervision of one triad at a time, while Phase III, last- ing three weeks, involved each attendant supervising her two triads with maintenance levels of supervision and instruction from experimen- tal staff. During Phase I and II Group E trainers and attendants were trained to teach toward specific language goals by applying condition- ing principles as embedded in the Modified Language Acquisition Pro- gram (MLAP). During this same time, Group C trainers were also given language training goals for their students, but without training in the use of this program or behavioral principles. Insofar as possible, all other training variables were made equivalent for the two groups. Thus, the independent variable was the application of a developmentally or- dered language training program which incorporated the principles of re- inforcement and modeling. In both groups the daily training sessions between trainer and student lasted 20 minutes. Each trainer worked with one of his as- signed students for a single session and then with the second student for another single session. Throughout the entire B-week study each student received a mean of l3 hours of training. Except for brief 190 demonstrations given to the trainers by attendants and experimental staff and for some extreme cases of behavior control, all the direct interaction with the students during these training sessions was done by the trainers. The main dependent variables were measures of language abil- ity (The Modified Language Acquisition Program Final Tests, MLAPFT) and intelligence (Cattell Infant Intelligence Scale). The design analyzed the pretest, posttest scores on both dependent variables of all pairs of students trained by a single trainer by a one-way analysis of vari- ance of the index of response. The scores of the two students in a triad were combined in the analysis since they were considered not to be independent but to be, in part, a function of the skill of the trainer in the triad. A less formal dependent variable, the Behavioral Observa- tional Record, BOR, was utilized to gather data in both groups regard- ing the instructional interactions between trainers and students in regular training sessions and between trainers and students and atten- dants and students in experimenter-structured training sessions which utilized students new to trainers and attendants. It was the main hypothesis of this study that the intensive training of Group E at the attendant and trainer levels in the use of reinforcement and modeling techniques and the Modified Language Acqui- sition Program would yield greater growth in language skills by the Group E students than by the Group C students who were paired with un- trained retarded trainers and supervising attendants. This hypothesis was supported since the mean gain in language skills in Group E stu- dents was significantly greater than the gain in Group C. 191 An examination of the MLAPFT gain scores of pairs of students in both groups revealed great differentiation between trainers in Group E and little or no differentiation between trainers in Group C. That is, the pairs of students for four trainers in Group E made all the gain that contributed to the significant difference between the two groups. However, the gain scores for students of the two remaining Group E trainers were minimal or negative as were all the gain scores for pairs of students in Group C. Also, the two Group E trainers whose students made little or no language gains were observed to be less skilled and both had 105 of 20, the lowest of all trainers in the study. It was further hypothesized that Group E students, being dif- ferentially reinforced for attending skills and early language skills commonly tested by infant intelligence tests, would demonstrate greater mean scores on the Cattell Infant Intelligence Scale than would the stu- dents in Group C. The results did not support the hypothesis. Two somewhat contradictory explanations were offered for these results. First, since the trend in MA gains slightly favored students in Group E and due to the high stability of MA in profound retardates, it was hypothesized that an extension of the training may have resulted in significant MA gains in Group E students as compared to Group C stu- dents. Second, the Cattell Test items to which the students were ex- posed were not largely concerned with early language skills. Therefore, the MAS measured by the Cattell were not largely affected by Group E's gains in language skills. 192 Although this hypothesis was not supported by a significant mental age gain difference in favor of Group E students, the results of the analysis did reveal that both groups of students made signifi- cant MA gains as compared to the expected zero gains for the population of older, profoundly retarded students from which they were drawn. This finding was interpreted to mean that regardless of being trained by operant and developmental language training techniques or being su- pervised by trained or untrained attendants, students spending time each day with supervised retarded trainers are likely to make signifi- cant mental age gains over students not exposed to retarded trainers. The series of questions posed to investigate the actual in- teractions between an instructor and a student was answered informally by an examination of the data gathered with the BOR. Some of this data were collected from three-minute observations made of each student and his trainer during regular training sessions at the end of each phase of the study. The remainder of the observations, also each three- minutes in length, was taken before and after the study in training sessions structured by the experimenter. These latter observations in- cluded not only trainers and students but also attendants and students from both groups. All students trainedduring these observations were new to the attendant and the trainer. The results of these observations of attendants and of train- ers in both groups are summarized below. A. In comparison to trainers using unstructured play methods of training with untrained attendant supervision, retarded trainers us- ing operant and developmental language training techniques and supervised 193 by attendants trained in the same skills were observed to consistently and more frequently 1. 4300“) 8. present complete trials to students, have students attending to stimulus presentations, evoke student responses, . demonstrate appropriate reinforcement of student's re- sponses, . couple tangible reinforcers with praise, . demonstrate apprOpriate prompting of students, . provide consistent stimulus presentations and with appro- priate repetition, . fade prompts appropriately, . teach tasks appropriate to the students, and . successfully generalize their training skills to new stu- dents in experimenter structured training situations. In comparison to untrained attendants, attendants skilled in the application of an operant, developmentally ordered language training program were observed to consistently and more frequently 1. 2. provide reinforcement and instruction to the retarded trainers they supervised, and demonstrate a high quality of training skills when asked to instruct a student in a particular early language task (these training skills are described in A.l through A.9). 194 Discussion The wide array of conclusions drawn from this study can be logically grouped into four categories: those'concerning trainers, attendants, students, and the MLAP. Discussions relating to each of these categories will be followed by recommendations for related re- search or further implementation of the retarded-trainer program. As strongly supported by the language gains and the BOR ob- servations, retarded trainers can be taught to successfully instruct under supervision early language skills according to the principles of reinforcement and modeling. This provides strong support to Whalen and Henker's similar findings (l97l). This is so even though, on the average, trainers in this study as compared to Whalen and Henker's trainers were 8 years older, l3 IQ points lower, and had been institu- tionalized 8 years longer and the students in this study were l0 years older, l9 IQ points lower and had been institutionalized l0 years longer. Recommendations. As a successful means of supplementing staff and stimulating lower functioning retardates, the higher functioning retarded adults should be trained and em- ployed to work in training units under the supervision of the attendant staff. Training areas could be expanded into the self-care skills of eating, dressing, toileting and groom- ing. In such expansions, the trainer would learn to instruct new content, but the reinforcement and modeling methods would remain the same. As support for McKinney and Keele's results (1963), gains in the mental ages of profoundly retarded students occurred after spending a mean of 13 hours with trainers regardless of whether the supervised trainers were taught specific instructional technidues. 195. 1 Recommendation. To aid in the prevention of further retarda- tion due to a lack of stimulation and continuous contact with a single person, institutional wards could be arranged such that students and trainers lived together with formal daily supervised interactions occurring between pairs. Also self- care training, scheduling of sessions, and unplanned informal interactions of trainers and students would be facilitated by this arrangement. As was apparent from the BOR observations for all trainers and from a breakdown in language gains for Group E students, some train- ers were more skilled than others in their interactions with students and more successful in their training. In Group E those more success- ful in applying the reinforcement and modeling procedures were also those whose students made larger language gains; and the trainers who were less skilled, according to their BOR observations, were those whose students did not make language gains. In Group C the BOR obser- vations revealed that some trainers, even without special instruction, were better than others. However, there was not enough differentiation in Group C language gains for these to be correlated with observed trainer quality. Also, the trainers that scored poorly in some or most of the BOR categories observed before the study also scored poorly in later observations. In addition, trainers with physical handicaps, though carefully matched to less mobile students, had good and poor training results and BOR observations, thereby disqualifying this fac- tor as a means of trainer selection. Recommendation. BOR observations should be used to select the potentially successful trainers as well as to assess deficits in trainer skills that could be then more specifi- cally remediated. In support of Cleland, et al. (1972), Houglan (1963), and Whalen and Henker (1971), the retarded trainers in this study were more 196 successful in regular training sessions with supervision from experi- mental staff or attendants than they were without. Although this need for supervision occurred in varying amounts from trainer to trainer, supervision seemed to be necessary for two reasons. First, praise and attention from the supervisor tended to maintain or improve the train- er's skills and enthusiasm. And second, instructions and demonstra- tions from supervisors assisted trainers when they were unable to in- tuitively adjust training procedures to fit the student's progress or lack of progress. The need for supervision varied widely from trainer to trainer. Some trainers could function well on a very sparce schedule of supervi- sory reinforcement showing little concern for their monetary earnings, while others seemed to need almost continual praise plus tangible rein- forcers. The latter trainers rarely saved their money while the former trainers either did not spend their money or spent it only infrequently. In addition, some trainers learned to give appropriate time outs with- out supervisory direction and to shape responses and fade prompts. Other trainers, even after hours of instruction still depended greatly upon their supervisors for direction in some of the basic trainer tasks, e.g., taking toys out for the students, setting the timer, turning the training cards after each task presentation. VRelated to trainer supervision is the phenomenon observed in the BOR data of a consistent reduction in all the trainer's skills from Phase I to Phase II with little or no change from Phase II to Phase III. A number of explanations can be given. First, during Phase I the exper- imental staffs' time was devoted solely to one trainer at a time, but 197 during the second phase it was divided between the trainer and the at- tendant in that training unit. Along with this, the fact that the fre- quency of instructions given to the trainer was observed to decline from the first to the second phase may have resulted in lowered trainer skills although the frequency of praise remained at the same level. Perhaps more important and not reflected in the BOR results was the fact that the quality of instructions given to trainers also changed along with the frequency. Less demonstrations were given and more ver- bal instructions were used by experimental staff and attendants. There- fore, the trainer worked more on his own. Another factor affecting this decline in observed trainer skills could be the fact that trainers worked every other day during the second phase rather than daily as they did in Phase I. There is an additional possibility that trainer enthu- siasm could have lessened during Phase II although this was not apparent. Recommendation. Retarded trainers need some amount of super- vision so their performance can be maintained and the more difficult methods of adjusting training techniques to fit the student's progress can be incorporated. In addition, train- ers should work daily with days off given according to their supervising attendant's schedule. Also it seems that train- ers should be more slowly weaned from continuous schedules to intermittent schedules of reinforcement and instruction. Since no Group E trainers were without picture task training cards and no Group C trainers were given cards, the effect of these cards cannot be assessed apart from the additional instruction given to Group E trainers. Therefore, training cards may have been one of the important factors in the higher BOR trainer ratings and higher student gains in Group E. Some of the difficulties observed by Whalen and Henker (1971) in the training done by retardates may have been (I! .g____- %. -' “J, 198 eliminated or reduced by the use of training cards. Trainers who used these cards to guide their instruction had more task consistency and appropriate repetition, and presented more tasks, demonstrated less ex- cessive verbalizations, and presented tasks in an order determined by experimental staff or attendants rather than themselves. Recommendation. Picture task training cards should be used ‘.fi; to guide trainers instructing early language skills and could be expanded to guide their instruction of students in ‘ self care skills. ’ Although formal data are lacking, trainers, as in the research .J reviewed, were observed to develop pride in their work, to eagerly an- L ticipate work times, to gain confidence and patience in their interac- tions with assigned students and to develop assurance and self-reliance in carrying out trainer tasks and getting to work on their own. Along with these changes attendants appeared to allow trainers more privileges and responsibilities and to give respect for their ability as trainers. Institutional staff not directly involved in the project expressed sur- prise over the trainers' skills shown in the videotaped training ses- sions. | Recommendation. As a planned portion of programming to pre- vent the development of dependence upon others and of inade- quate self concepts, institutionalized retardates should be trained to handle increasing levels of responsibility and in— dependence. During Phase III when attendants independently operated the training units, the quality of training shown by the BOR ratings did not decrease from Phase II. Although attendants were encouraged by ex- perimental staff and given small amounts of intermittent instructions, 199 they essentially maintained the entire retarded trainer program. Atten- dants were never absent from their regular ward work for much more than two hours per day to participate in the study. These findings are im- portant in illustrating the untapped or inefficiently tapped resource institutions have in their attendant personnel. As pointed out by Harmutz (1973): If specific, visible, behavioral goals were set forth for the children's behavior and reinforcements of the staff were made con- tingent upon the child achieving these goals, no doubt this would effect a realignment of the staff time allocations (p. 557). In addition, it must be noted that attendants in this study volunteered to participate and that two attendants, both in Group C, decided during Phase I not to continue their participation. It appears that not all attendants would welcome a redefinition of their role to include educa- tional duties. Recommendation. Attendants with expressed interests should be given the inservice training, the time, the responsibility, and the reward necessary to participate in resident-training programs. Originally it was planned to include as students only those residents with basic attending skills of sitting without stereotypic movements and attending to objects. Attention training seemed to be a task too difficult for retarded trainers. However, although all stu- dents selected could sit without stereotypic movements, some students with poor object attending responses had to be included because of the limited population from which to select. Seven of the 24 students in the study had difficulty attending to objects plus no imitative skills. Of these students six held the lowest pretest mental age scores and all 200 but one made gains under the average. However, six of the seven did make small gains in attending and imitative skills. In summary, only small gains in language or in mental age were made by low attending students who also had no other early language skills. Recommendation. The basic attending behaviors of sitting in a chair, absence of stereotypic motor responses, and looking at objects on command should be present in a student prior to his inclusion in a retarded-trainer program. Withholding reinforcement and time out procedures were effec- E tive in the reduction of disruptive behaviors in students. In Group E :‘J these methods of control were used directly by the students while in L Group C intervention and assistance by the supervisor was often needed. Also a token reinforcement system where tokens have no unit value was used by five Group E trainers with 10 of the 12 students. (Two low- attending students trained by the same trainer were unable to be placed on a token system.) The token reinforcement system seemed to notice- ably increase the efficiency of training sessions. Recommendation. Programs for profoundly retarded persons should employ extinction and time out procedures to control disruptive behaviors and simple token reinforcement proce- dures to increase appropriate behaviors. The procedures of the MLAP were generally well grasped by the three attendants in Group E and easily adapted to retarded trainers. However, it seemed that some additional instruction time would have been profitable for attendants during Phase I. These areas for more instruction in the MLAP included understanding the differences between testing and training procedure, use of the three types of tests: ini- tial, intermediate, final, and the finer techniques of training retarded 201 trainers, and more practice in the direct training and testing of stu- dents. Some of the instruction time for Group E attendants could have been more efficiently done in a small group. It appeared that the MLAP manual served attendants as a source of information in the absence of experimental staff. Attendants in Group C were given MLAP training after the study so as to participate in a continuation of the retarded 1“; trainer project. Comments from both groups of attendants indicated that the program's features of assessment-based training to criterion performance levels, sequential levels of training and early language 53 ' 4- development were especially important and relevant as compared to other training programs they had been exposed to for retarded residents. Recommendation. Training programs for the retarded in which attendants are employed as teachers should be relevant, as perceived by the attendants, based on developmental evidence, consist of training levels that are defined by simply admin- istered tests of performance, and available to attendants in an understandable and organized manual. Implications for Research Additional research is necessary to explore the following questions and topics raised by this study. 1. What amount of a student's gains in language skills and in mental age takes place during the Phase III portion of a retarded trainer program, i.e., when the training units are operated by an at- tendant? 2. If the third phase of a retarded trainer program were ex- tended, would the decline in the trainer's instructional skills ob- served between Phase I and II in this study recur and continue its 202 decline during this time or would the trainer's skills remain stable during this extension of Phase III? 3. Besides the instructional skills, do retarded trainers make any measurable gains in the affective areas of self-confidence, independence, responsibility, and social awareness? 4. For those retarded trainers in the IQ range of 20 to 50, ‘~ . to what extent can IQ level be a factor in the prediction of their I)“ success, i.e., as reflected in gains made by their students or in their BOR ratings? J ‘ L 5. Which BOR ratings for a trainer are most related to the gains made by that trainer's students? 6. Which BOR ratings for an attendant are most related to the gains made by the trainers and students supervised by that atten- dant? 7. To what extent is the generalization of a student's lan- guage skills from training sessions to the ward facilitated by holding training sessions on the ward? What are other techniques of facili- tating this generalization? 8. Alternative assistive training devices besides the pic- ture task training cards need to be developed for the more skilled and the less skilled retarded trainers as well as for the trainer with little or no arm movement and/or speech. 9. The procedures of instructing supervising attendants and retarded trainers, as occurred during Phase I and II of this study, need to be standardized and defined with criterion levels of perfor- mance . 203 10. Since many of the higher functioning residents of insti- tutions are being placed back into the community, it is most important for future applications of retarded trainer programs that an explora- tion be made of the utilization of trainers in community programs for the severely and profoundly retarded as a supplement or an alternative. to the trainers' placement in sheltered workshops. BIBLIOGRAPHY BIBLIOGRAPHY Ashbaugh, L. L. "An Evaluation of an Attendant Training Program Based on Principles of Behavior Modification." Unpublished doc- toral dissertation, the Pennsylvania State University, 1971. Ayllon, T., and Azrin, N. The Token Economy: A Motivational System for Therapy and Rehabilitation. New York: Appleton-Century- Crofts, 1968. Azrin, N. H., and Foxx, R. M. "A Rapid Method of Training the Insti- tutionalized Retarded." Journal of Applied Behavior Analysis, Vol. 4 (1971). Baer, D. M., and Guess, D. "Receptive Training of Adjectival Inflec- tions in Mental Retardates." Journal of Applied Behavior Analysis, Vol. 4, (1971). Baer, D. M.; Guess, D.; and Sherman, J. A. "Adventures in Simplistic Grammar.“' Language of the Mentallpretarded. Edited by R. L. Schiefelbusch. Baltimore: University Park Press, 1972. Baer, D. M.; Peterson, R. F.; and Sherman, J. A. "The Development of Imitation by Reinforcing Behavioral Similarity to a Model." Journal of Experimental Analysis of Behavior, Vol. 10 (1967). Barton, Elizabeth Spindler. "Inappropriate Speech in a Severely Re- tarded Child: A Case Study in Language Conditioning and Gen- eralization." Journal oprpplied Behavior Analysis, Vol. 3 1970 . Bellugi, U. "The Acquisition of Negation in Children's Speech." Doc- toral dissertation, Harvard University, 1967. Bellugi, U. "Development of Language in the Normal Child." Language Intervention with the Retarded: Developing Strategies. Edited by J. E. McLean, D. E. Yoder, and R. L. Schiefelbusch. Baltimore: University Park Press, 1972. Bender, L. "Infants Reader in Institutions: Permanently Handicapped.“ Bulletin Child Welfare League America, Vol. 24, 1945. 204 205 Bensberg, G. J.; Barnett, C. D.; and Hurder, W. P. "Training of Atten- dant Personnel in Residential Facilities for the Mentally Retarded." Mental Retardation, June, 1964. Blanchard, Irene. "Speech Pattern and Etiology in Mental Retardation." American Journal of Mental Deficiency, Vol. 68 (1964). Blindert, H. Dieter. "Experiments in the Establishment of Verbal Be- haviors in Children Through the Training of Mothers in Be- havior Modifications." Unpublished doctoral dissertation, The University of Rochester, 1969. Bloom, L, Language Development; Form and Function in Emerging Grammars. Cambridge: MIT Press, 1970. Bloom, L. "Semantic Features in Language Development." Language of the Mentally Retarded. Edited by R. L. Schiefelbusch. Baltimore: University Park Press, , Blout, R. W. "Language and the More Severely Retarded." American Journal of Mental Deficiency, Vol. 73, No. 1 (July, 1968). Bricker, W. M. "A Systematic Approach to Language Training." Language of the Mentally Retarded. Edited by R. L. Schiefelbusch. Baltimore: University Park Press, 1972. Bricker, W. A., and Bricker, D. D. ”A Program of Language Training for the Severely Language Handicapped Child." ExCeptional Chil- dren, Vol. 37 (1970). Bricker, W. A.; Morgan, D.; and Grabowski, J. "Token Reinforcement of Attendants Who Work with Low-Functioning Children." Ap;_ stracts of Peabody Studies in Mental Retardation 1965-1968, Vol. 4, No. 2 (1968). Brodbeck, A. J., and Irwin, O. C. "The Speech Behavior of Infants Without Families." Child Development, Vol. 17 (1946). Brown, R., and Bellugi, U. ”Three Processes in the Child's Acquisition of Syntax." New Directions in the Study of Language. Edited by E. Lenneberg. Cambridge, Massachusetts: MIT Press, 1964. Brown, Lou; Fenrick, N.; and Klemme, H. "Trainable Level Retarded Stu- dents Teach Trainable Level Retarded Students." Unpublished manuscript, University of Wisconsin and Madison Public Schools, 1971. Browning, R. M., and Stover, D. 0. Behavior Modification in Child Igggtmggt. Chicago: Aldine-Atherton, 1971. Buddenhagen, R. G. Establishipg Vocal Verbalizations in Mute Mongoloid Children. Champaign, Illinois: Research Press Company, 1971. 206 Burke, 0., and Rowland, M. "An In-service Technique to Teach Ward Attendants How to Give Language Development Training to In- stitutionalized Retardates." Paper presented at the Annual Meeting, Michigan Speech and Hearing Association, Lansing, Michigan, October, 1971. Butterfield, E. C. "Basic Facts about Public Residential Facilities for the Mentally Retarded." Changing Patterns in Residential Services for the Mentally Retarded. Edited by R. Kugel and W. Wolfensberger. Washington: U.S. Government Printing Office, 1969. Cattell, Psyche. The Measurement of Intelligence of Infants and Young Children. New York: The Psychological Corporation, 1960. Centerwall, S. A., and Centerwall, W. R. "A Study of Children with Mongolism Reared in the Home Compared with Those Reared away from the Home." Pediatrics, Vol. 25 (1960). Cleland, C. C., and Swartz, J. D. Mental Retardation: Approaches to Institutional Change. New York: Grune and Stratton, 1969. Cleland, C. C.; Swartz, J. 0.; Drew, C. J.; and Talkington, L. W. "Charge Attendant Attitudes Toward Working Residents.“ Men- tal Retardation, Vol. 10, No. 4 (1972). Conlon, M. F.; Hall, C.; and Hanley, E. M. "The Effects of a Peer Correction Procedure on the Arithmetic Accuracy for Two Ele- mentary School Children." Behavior Analysis and Education-- 1972. Edited by G. Semb. The University of Kansas, 1972. Cortazzo, Arnold D.; Bradtke, L. E.; Kirkpatrick, W. J., Jr.; and Rosenblatt, K. P. "Innovations to Improve Care in an Insti- tution for the Mentally Retarded." Children, Vol. 18, No. 4 (July-August, 1971). Corte, H. E.; Wolf, M. M.; and Locke, B. J. "A Comparison of Proce- dures for Eliminating Self-Injurious Behavior of Retarded Adolescents.“ Journal of Applied Behavior Analysis, Vol. 4 (1971). Csapo, Margaret G. "Utilization of Normal Peers as Behavior Change Agents for Reducing the Inappropriate Behavior of Emotionally Disturbed Children in Regular Classroom Environments." Un- published doctoral dissertation, University of Kansas, 1971. Davis, M. "Effects of Having One Remedial Student Tutor Another Reme- dial Student." Behavior Analysis and Education--1972. Ed- ited by G. Semb. University of Kansas, 1972. 207 Denny, M. R. "A Theoretical Analysis and Its Application to Training the Mentally Retarded." International Reviewgpresearch in Mental Retardation, Vol. II. Edited by N. R. Ellis.‘ New YOrk: ‘Academic Press, 1966. Denver, Richard B. "The Use of Language by Mentally Retarded Children: A Review of the Literature." Unpublished manuscript, Center for Innovation in Teaching the Handicapped, Indiana Univer- sity, 1971. Dilly, G. M. "Retarded Women Teach Self-Help Skills." Hospital and Communitprsychiatry, Vol. 20 (1969). Distefano, M. K., Jr., and Pryer, M. W. "Basic Issues and Problems in Attendant Training." Mental Hygiene, Vol. 48 (1964). Doorlag, Donald Henry. "The Use of Regular Class Students as Teaching Assistants in Special Education Classes in Michigan." Un- published doctoral dissertation, Michigan State University, 1970. Dupan, R. M., and Roth, S. "The Psychologic Development of a Group of Children Brought Up in a Hospital Type Residential Nursery." Journal of Pediatrics, Vol. 47 (1955). Erickson, Marilyn T. "The Predictive Validity of the Cattell Infant Intelligence Scale for Young Mentally Retarded Children." American Journal of Mental Deficiency, Vol. 72, No. l. Fischer, L. "Hospitalism in Six Month Old Infants." American Journal of Orthopsychiatpy, Vol. 22 (1952). Fischer, L. "Psychological Appraisal of the Unattached Preschool Child." American Journal of Orthgpsychiatry, Vol. 23 (1953). Fischer, M. A., and Zeaman, D. "Growth and Decline of Retardate In- telligence." International Review of Research in Mental Re- tardation. Edited by N. R. Ellis. New York: Academic Press, 1970. Freud, A., and Burlingham, D. T. Infants Without Families. New York: International University Press, 1944. Gardner, J. M., and Watson, L. 5., Jr. "Behavior Modification of the Mentally Retarded, An Annotated Bibliography." Mental Re- tardation Abstracts, Vol. 6, No. 2, April-June, 1969. Gartner, A.; Kohler, M.; and Riessman, F. Children Teach Children. New York: Harper and Row, 1971. 208 Gelfand, D. M.; Gelfand, S.; and Dobson, W. R. "Unprogrammed Rein- forcement of Patient's Behavior in a Mental Hospital." Be;_ havior research and Therapy, Vol. 5 (1967). Gesell, A., and Amatruda, C. Developmental Diagnosis. (2nd Edition). New York: Harper and Row, 1947. Goldfarb, W. "Effects of Psychological Deprivation in Infancy and Subsequent Stimulation." American Journal of Psychiatry, Vol. 102 (1945). Gray, B. 8. "Language Problems: Language Acquisition Through Pro- grammed Conditioning." Behavior Modification: the Human Effort. Edited by R. H. Bradfield. San Rafael, California: Dimensions Publishing Company, 1970. Gray, 8., and Ryan, 8. A Language Prpgram for the Nonlanguage Child. Champaign, Illinois: Research Press Company, 1973. Groves, I. D., and Carroccio, D. F. "A Self-Feeding Program for the Severely and Profoundly Retarded." Mental Retardation, Vol. 9, No. 3 (1971). Guenther, H., and Klein, 0. "Imitation of Initial Consonants and Syl- labification in a Severely Retarded Population." Unpublished manuscript, Fort Custer State Home and Training School, Augusta, Michigan, 1969. Guess, D.; Sailor, W.; Rutherford, G.; and Baer, D. "An Experimental Analysis of Linguistic Development: The Productive Use of the Plural Morpheme." Journal of Applied Behavior Analysis, Vol. 1, No. 4 (1968). Guess, O. "A Functional Analysis of Receptive Language and Productive Speech: Acquisition of the Plural Morpheme." Journal of Applied Behavior Analysis, Vol. 2, No. 1 (Spring, 1969). Haggerty, A. D. "The Effects of Long-Term Hospitalization Upon the Language Development of Children." Journal of Genetic Psy- chology, Vol. 94 (1959). Harmutz, Morton G. "Observational Study of Ward Staff Behavior." Ex; ceptional Children, Vol. 39 (1973). Harris, V. W.; Sherman, J. A.; Henderson, D. G.; and Harris, M. S. "Effects of Peer Tutoring on the Spelling Performance of Ele- mentary Classroom Students." Behavior Analysis and Educa- tion--l972. Edited by G. Semb. University of Kansas, 1972. Hartung, J. R. "A Review of Procedures to Increase Verbal Imitation Skills and Functional Speech in Autistic Children.“ Journal of Speech and Hearing Disorders, Vol. 35, No. 3 (1970). 209 Haviland, R. T. "A Stimulus to Language Development: The Institutional Environment." Mental Retardation, Vol. 10, No. 2 (1972). Houglan, Marilyn. "The Mentally Retarded Contribute, Also." Nursing Outlook, Vol. II, No. 3 (1963). Ishiyama, T.; Grover, W. L.; and Paterson, J. H. "Milieu Changes Com- plement Education." Hospital and Community Psychiatry, V01. 17, No. 6 (1966). Jeffrey, D. 8. "Increase and Maintenance of Verbal Behavior of a Men- tally Retarded Child." Mental Retardation, Vol. 10, No. 2 1972). Kazdin, Alan E. "Toward a Client Administered Token Reinforcement Pro- gram." Education and Trainingyof Mentally_Retarded, Vol. 6, No. 2 (1971). Kazdin, A. E., and Bootzin, R. R. "The Token Economy: An Evaluative Review.“ Journal of Applied Behavior Analysis, Vol. 5 (1972). Keane, V. E. "The Incidence of Speech and Language Problems." Mental Retardation, Vol. 10, No. 2 (1972). Kent, L. R. "Language Acquisition Program." Unpublished manuscript, ‘ COldwater State Home and Training School, Coldwater, Michigan, 1972. Kent, L. R.; Klein, D.; Falk, A.; and Guenther, H. "List of 200 Most Likely to be Named Nouns in a Severely Retarded Population." Unpublished manuscript, Fort Custer State Home and Training School, Augusta, Michigan, 1968. Kent, L. R.; Klein, D.; Falk, A.; and Guenther, H. “A Language Acquisi- tion Program for the Retarded." Language Interventignwith the Retarded: Developing Strategies. Edited by J. E. McLean, D. E. Yoder, and R. L. Schiefelbusch. Baltimore: University Park Press, 1972. Kralovich, Anne Marie. "A Study of Performance Differences on the Cattell Infant Intelligence Scale Between Matched Groups of Organic and Mongoloid Subjects." Journal of Clinical Psychol- 991. Vol. 15 (1959). Lawrence, Joan Allen. "A Comparison of Operant Methodologies Relative to Language Development in the Institutionalized Mentally Re- tarded." Unpublished doctoral dissertation, Boston University School of Education, 1971. Lent, J. R. "Mimosa Cottage: Experiment in Hope." Readings in Clin- ical Psychology Today, Edited by B. A. Henker. California: CRM Books, 1967. . 210 London, S. K. "The Stability of the IQ of Mentally Retarded Pre-School Children." Doctoral dissertation, University of Michigan, Ann Arbor: University Microfilms, 1961. Lovaas, O. I.; Freitas, L.; Nelson, K.; and Whalen, C. "The Establish- ment of Imitation and Its Use for the Development of Complex Behavior in Schizophrenic Children." Behavior Research and Therapy, Vol. 5 (1967). Lowery, L. C. "Personality Distortion and Early Institutional Care." American Journal of Orthopsychiatry, Vol. 10 (1940). Matthews, J. "Speech Problems of the Mentally Retarded." Handbook of Speech Pathology. Edited by L. E. Travis. New York: Appleton-Century-Crofts, 1957. Matthews, J. "Communication Disorders in the Mentally Retarded." Handbook of Speech Pathology and Audiology. Edited by L. E. Travis. New York: Appleton-Century-Crofts, 1971. Mental Retardation Source Book of the Department of Health Education and Welfare. Washington, D.C.: Department of Health, Educa- tion, and Welfare Publication No. (OS) 73-81, 1972. Metz, J. R. "Conditioning Generalized Imitation in Autistic Children." Journal of Experimental Child Psychology, Vol. 2 (1965). McBride, H. J. "The Differential Effectiveness of Two Methods of Train- ing Institutional Attendants in the Technique of Behavior Modification." Unpublished doctoral dissertation, Michigan State University, 1972. McCarthy, J. J.; Stevens, H. A.; and Billingsley, J. F. Program De- velopment for Severelnyetarded, Institutionalized Children. Madison, Wisconsin: University of Wisconsin, 1969. MacCubrey, Jean. "The Effects of Verbal Operant Conditioning on Young Institutionalized Mongoloids." Doctoral dissertation, Boston College, Ann Arbor, Michigan: University Microfilms, 1970. McKinney, J. P., and Keele, 1. "Effects of Increased Mothering on the Behavior of Severely Retarded Boys." American Journal of Mental Deficiency, Vol. 67 (1963); McReynolds, Leija V. "Application of Timeout from Positive Reinforce- ment for Increasing the Efficiency of Speech Training." Journal of Applied Behavior Analysis, Vol. 2, No. 3 (1969). 211 McReynolds, Leija V. "Reinforcement Procedures for Establishing and Maintaining Echoic Speech by a Nonverbal Child." A Func- tional Analysis Approach to Speech and Language, Chapter 6, ASHA Monograph No. 14. Washington, D.C.: American Speech and Hearing Association, January, 1970. O'Brien, F., and Azrin, N. H. “Developing Proper Mealtime Behaviors of the Institutionalized Retarded." Journal of Applied Be- havior Analysis, Vol. 5 (1972). O'Brien, F.; Bugle, C.; and Azrin, N. H. "Training and Maintaining a Retarded Child's Proper Eating." Journal of Applied Be- havior Analysis, Vol. 5 (1972). Panyan, Marion; Boozer, H.; and Morris, Nancy. "Feedback to Attendants Parker, G. Patterson, Pizzat, F. Poole, I. Reissman, Rheingold, Risley, T. Risley, T. as a Reinforcer for Applying Operant Techniques." Journal of Applied Behavior Analysis, Vol. 3 (1970). O. "Attendant-Nurses for the Mentally Deficient: Some Evi- dence)" American Journal of Mental Deficiency, Vol. 55 1951 . Gerald R.; Shaw, David A.; and Ebner, Michael J. "Teachers, Peers, and Parents as Agents of Change in the Classroom." Modifying Deviant Social Behaviors in Various Classroom Set- ti gs. Edited by Arthur M. Benson. Monograph #1, Department of Special Education, College of Education, University of Oregon, February, 1969. Behavior Modification in Residential Treatment for Childrgn; Model of a Program. New York: Behavioral Publications, 1973. "Genetic Development of Articulation of Consonant Sounds in Speech." Elementany English Review, Vol. 2 (1934). E. "The 'helper' Therapy Principle." Social WOrk, Vol. 10 1965 . Harriet L., and Bayley, Nancy. "The Later Effects of an Ex- perimental Modification of Mothering." Child Development, V01. 30 (1959). R.; Hart, 8. M.; and Doke, L. "Operant Language Development: The Outline of a Therapeutic Technology." Language of the Mentally Retarded. Edited by R. L. Schiefelbusch. Baltimore: University Park Press, 1972. R., and Wolf, M. M. "Establishing Functional Speech in Echolatic Children." Behavior Research and Therapy, Vol. 5 1967 . 212 Sailor, W.; Guess, D.; and Baer, D. M. "Functional Language for Ver- bally Deficient Children: An Experimental Program." Mental Retardation, Vol. II, No. 3 (1973). Sailor, W.; Guess, D.; Rutherford, G.; and Baer, D. "Control of Tan- trum Behavior by Operant Techniques During Experimental Ver- bal Training." Journal of Applied Behavior Analysis, Vol. 1 1968 . Sheehan, J.; Martyn, M.; and Kilburn, K. "Speech Disorders in Retarda- tion." American Journal of Mental Deficiency, Vol. 73 (1968). Siegel, G. M. "Three Approaches to Speech Retardation." Language of the Mentally Retarded. Edited by R. L. Schiefelbusch. Baltimore: University Park Press, 1972. Siegenthaler, B. M. "Factors Relating to Speech-Hearing of Children of Below Normal Intelligence." University Park: Pennsylvania State University, 1970. Skeels, H. M.; Updefraff, R.; Wellman, B. L.; and Williams, H. M. "A Study of Environmental Stimulation: An Orphanage Preschool Project." University of Iowa Studies of Child Welfare, Vol. 15 (1938). Skinner, B. F. Science and Human Behavior. New York: Macmillan/Free Press, 1953. Skinner, B. F, Verbal Behavior. New York: Appleton-Century-Crofts, 1957. x Sloane, H. N.; Johnston, M. K.; and Harris, F. R. "Remedial Procedures for Teaching Verbal Behavior to Speech Deficient or Defective Young Children." Operant Procedures in Remedial Speech and Language Training. Edited by H. N. Sloane, Jr., and B. D. MacAulay. New York: Houghton Mifflin, 1968. Smolev, S. R. "Use of Operant Techniques for the Modification of Self- Injurious Behavior." American Journal of Mental Deficiengy, Vol. 76 (1971). Spradlin, J. E. "Assessment of Speech and Language of Retarded Chil- dren: The Parsons Language Sample." Language Studies of Mentally Retarded Children, The Journal of Speech and Hear- ing Disorders, 1963a (Monograph Supplement 10). Spradlin, J. E. "Language and Communication of Mental Defectives." Handbook of Mental Deficiency. Edited by N. R. Ellis. New York: McGraw-Hill, 1963b. 213 Standards for Residential Facilities for the Mentallngetarded. Chicago: Joint Commission on Accreditation of Hospitals, 1971. Stremel, K. "Language Training: A Program for Retarded Children." Mental Retardation, Vol. 10, No. 2 (1972). Surratt, P. R.; Ulrick, R. E.; and Hawkins, R. P. "An Elementary Stu- dent as a Behavioral Engineer." Journal of Applied Behavior Analysis, Vol. 2 (1969). Tarjan, George. "The Role of Residential Care--Past, Present, and Fu- ture." Mental Retardation, Vol. 9, No. 5 (1966). Templin, M. Certain Language Skills in Children. Minneapolis: Uni- versity of Minnesota Press, 1957. Thompson, T., and Grabowski, J. (eds.) Behavior Modification of the Mentally Retarded. New York: Oxford University Press, 1972. U.S. Department of Health, Education and Welfare, Office of the Secre- tary, Office of Mental Retardation Coordination, Mental Re- tardation Source Book of the Department of Health, Education and Welfare, Washington, D.C., 1972. Watson, Luke S. "Application of Operant Conditioning Techniques to In- stitutionalized Severely and Profoundly Retarded Children." Mental Retardation Abstracts, Vol. 4 (1967). Watson, L. 5., Jr.; Gardner, J. M.; and Sanders, C. "Shaping and Main- taining Behavior Modification Skills in Staff Members in a Mentally Retarded Institution: Columbus State Institute Be- havior Modification Program." Mental Retardation, Vol. 9, No. 3 (1971). Whalen, Carol K., and Henker, Barbara A. "Pyramid Therapy in a Hospital for the Retarded: Methods, Program Evaluation, and Long-Term Effects." American Journal of Mental Deficiency, Vol. 75 1970). Whalen, C. K., and Henker, B. A. "Creating Therapeutic Pyramids Using Mentally Retarded Patients." American Journal of Mental De- ficiency, Vol. 74 (1969). Willis, J.; Crowder, J.; and Morris, 8. "A Behavioral Approach to Re- medial Reading Using Students as Behavioral Engineers." .Be; havior Analysis and Education--l972. Edited by G. Semb. University of Kansas, 1972. Yarrow, L. J. "Maternal Deprivation: Toward an Empirical and Concep- tual Re-Evaluation." Psychological Bulletin, Vol. 58 (1961). 214 Yoder, D. E., and Miller, J. F. "What We May Know and What We Can 00: Input Toward a System." Language Intervention with the Re- tarded: Developing Strategies. Edited by J. E. McLean, D. E. Yoder, and R. L. Schiefelbusch. Baltimore: University Park Press, 1972. Zisk, P. R., and Bialer. "Speech and Language Problems in Mongolism: A Review." MnaLoLSoeesthdJeaninsulisoLdecs Vol. 32 (March, 1967). Li I -._ APPENDICES APPENDIX A LETTER TO PARENTS AND GUARDIANS OF STUDENTS LETTER TO PARENTS AND GUARDIANS OF STUDENTS June 14, 1972 Dear Your has been selected to participate with 23 other residents in a language training program this summer at Howell State Hospital. This program is directed towards improving and developing a variety of prelanguage and language skills ranging from attending to understanding and using a variety of words. We feel that all partici- pants will benefit from the training. This is an intensive training program and will be carried out either once or twice a day, Mondays through Fridays. Training will begin June 26th and continue through August 25th. It is imperative to have an evaluation of all participants before and after this training program in order to assess their progress and evaluate the effective- ness of the program. The 2 periods set aside for testing are from June 19th through June 25th, and from August 28th through September lst. It is necessary that be available at the institution during all these testing and training times in order to remain included as a participant in the program. However, all weekends and evenings will certainly be available for institutional and home visits. We realize that these summer months are often used for vacations; unfortunately, the present training program has been planned for this same time span. We greatly hope that will be able to participate during these times. If, however, this interferes with your summer plans for and you are unable to make adjustments, please notify the Social Services Department at Howell State Hospital as soon as possible on or before June 19th, so that another person may be included in the program in place of your . If you have any questions regarding the training program, please leave your name, telephone number, and when you may be reached with the Howell State Hospital Social Services Department so that I may contact you as soon as possible. Thank you. Sincerely yours, Martha S. Rowland M.S.U. Graduate Student Special Education 215 APPENDIX B LETTER TO PARENTS AND GUARDIANS 0F TRAINERS LETTER TO PARENTS AND GUARDIANS OF TRAINERS June 14, 1972 Dear Your has been selected to participate with 11 other residents at Howell State Hospital as trainers in a language training program this summer. will be assisting myself, other graduate assistants, and hospital attendants in the individual train- ing of lower functioning residents living at Howell State Hospital. Trainers will learn to train a variety of early language skills as well as how to become an effective worker with others. Each trainer will be paid a small amount for their work in the program which can be saved or spent on items chosen by the trainer. We feel that all par- ticipants will benefit from the training. Since the training program begins on June 26th and continues un- til August 25th, operating Mondays through Fridays, it is imperative that be available at the institution during these dates in order to participate as a trainer in the program. However, all weekends and evenings will certainly be available for institutional and home visits. We realize that these summer months are often used for vacations; unfortunately, the present program has been planned for this same time span. We greatly hope that will be able to participate during these times. If, however, this interferes with your summer plans for and you are unable to make adjustments, please notify the Social Services Department at Howell State Hospital as soon as possible on or before June 19th, so that another person may be in- cluded in the program in place of your . If you have any questions regarding the training program, please leave your name, telephone number, and when you may be reached with the Howell State Hospital Social Services Department so that I may con- tact you as soon as possible. Thank you. Sincerely yours, Martha 5. Rowland M.S.U. Graduate Student Special Education 216 APPENDIX C STUDENTS' PLACEMENT IN MLAP BASED UPON INITIAL 0R FINAL TEST RESULTS STUDENTS' PLACEMENT IN MLAP BASED UPON INITIAL 0R FINAL TEST RESULTS Direction of Training Prior to and Upon Completion of the Required Number of Correct Responses for Each Part Phase 1. Attending 2. Motor Imitation 3. Vocal Imitation Part *a. *b. sitting getting rid of motor be havior that interferes with traini . looking at objects on gesture and command . looking at trainer be- fore each command a. with object . with body parts . gross vocal imitation: copying the trainer's sound by making any sound . vowel imita tion n9 S 217 Program part(s) needed before training this part None la and 1b 1d 1d and generally 2a 2a, 2b 3a Program part(s) to be trained upon completion of this part la or lb (if no interfering be- havior exists) 1c 1d 2a 2b 3a, 4a, 4b, 4c, 4d 3b (if the child can imitate words, go to 3c) 3c 218 Program part(s) Program part(s) to be trained ' needed before upon completion Phase Part training this part of this part 3. (continued) c. word imita- 3b or some words 3d, 6a, 6b, 6c, tion imitated during 6d.(if 4a, 4b, 4c, initial testing 4d, respectively are completed) d. phrase imita- 3c no completion of tion training 4. Basic Re- a. touching ob- 2b 4e, 5a, 6a ceptive jects Phase: Un- derstanding b. pointing to 2b 5b (if 4a is com- Words that room parts pleted) 6b Label c. pointing to 2b 5c, 6c body parts d. performing 2b 6d activities e. finding ob- 4a 5d (if 4e is com- jects (not pleted) visible) 5. Receptive Ex- a. giving trainer 4a 5d (if 4e is com- pansion Phase: two objects pleted) Understanding Two Words that b. putting ob- 4a and 4b 7b (if 6a and 6b Label jects on room are completed) parts c. touching body 4c plus "baby" 7c (if 6b and parts: pos- from 4a I'baby" from 6a session-—stu- are completed) dent's and baby's d. finding two 4e and 5a 7e (if 6a is com- objects (not pleted) visible) 6. Basic Expres- a. naming objects 4a, 3c 6e, 7a sive Phase: Using Single b. naming room 4b, 3c 7b Words to Label parts c. naming body 4c, 3c 7c parts Phase 6. (continued) 7. Expressive Expansion Phase: Using Combinations of Words to Label, to Ask, to Tell Whose and Where 219 Part d. naming ac- tivities e. naming con- cealed ob- jects a. asking for O objects . telling where an object is . telling whose body part is pointed to . telling trainer to perform an activity . naming mis- sing objects Program part(s) needed before training thisnpart 4d, 3C 6a 6a 6b "baby" from 6a and 6c 6d 6a and 7a Program part(s) to be trained upon completion of thisypart 7d 7e (if 7a is r1 completed 7e (if 6e is completed none none none none APPENDIX D SAMPLE TRAINING OBSERVATION SAMPLE TRAINING OBSERVATION Student 31113; Date 7-18-72 Attendant Jill Trainer Mary Initial Test Key: + correct response Intermediate Test (+) partially correct - incorrect Final Test (1 2 3 4) 0 no response / TRAINING OBSERVATION Program Phase and Part 4a T CD" (D «H m «H m 3": 2' g a); 2' S .313 8 5% ‘82 E 3 c: QTask °' 3 "o Task °' :32 BALL + # .§EEN1N + HPt'E'. / + coma ML MCOMBL / + ._ tBAJJL + _ 429MB +41 I CAR ' CA3 / BALP + Iggy - I KEY */ + A SPOON "' HAT l -y HAT / + SPOON , + ( /'items correct without trainer prompting each time presented; these items are ready for intermediate testing.) 220 APPENDIX E PICTURE TRAINING CARD EXAMPLES PICTURE TRAINING CARD EXAMPLES 1. Used for parts 4c, 5c, 6c and 7c. 2. Used for parts 4b and b. 3. Used for parts 5a I H and 5e. 221 APPENDIX F GUIDELINES FOR EXPERIMENTAL ASSISTANTS' WORK WITH TRAINERS DURING PHASE I GUIDELINES FOR EXPERIMENTAL ASSISTANTS' WORK WITH TRAINERS DURING PHASE I General Group II Rules for Experimental Assistants: 1. Goals for trainers during Phase I (first two weeks) A. During the first session only, tell the trainers the specific 0. dd language training levels for each of their two students in clear simple statements. . Show the trainer, in the first session only, one simple demon- stration of the training procedure needed to train the level ex- plained in words above (A). Do not ask him to respond in imita- tion of your demonstration, but encourage him if he does no matter what his performance is (i.e., whether he attempts to copy your demonstration of appropriate training procedure or not . . Make the trainer understand what kinds of things he should and could do with his students by: 1. telling him he is a teacher to the students assigned to him and that . . .; 2. he should try hard to teach him the goal mentioned in A above (however, this should be mentioned only during the first ses- Sion ; 3. he should play with the student and get him to play; 4. he should teach him new things; 5. he should talk to him; 6. he should use all the equipment provided; 7. he should keep him in the room for the whole 20 minutes; 8. he should give him physical contact (touching, pats, hugs); 9. he should give the student the food or toy he especially likes whenever he wants during the training sessions; 0. he should not hurt the student; and 1. he should be encouraged to ignore the student and remove toys when the student begins disruptive behaviors (throwing, hit- ting, etc.). Instruct the trainer in how to pick up his student from the ap- propriate ward; accompany him initially giving help and instruc- tion as needed but aiming to make the trainer independent in, this skill by the end of the first phase. If the trainer is physically handicapped to the point where this is too time con- suming or if the student is quite a distance from the training room (playground, etc.) then it might be more convenient to make adjustments in this part of the trainer's job; for example, the trainer might bring the student part of the way to the training room and the assistant or attendant bring the student the first part of the way. 222 223 . Teach the trainer how to set the timers to 20 minutes at the be- ginning of each training session (a small piece of red tape at the "20" place might help initially). . Teach the trainers to end each session when the bell rings that the 20 minutes are over; at this time, the trainer should learn to put away equipment. . Teach the trainer to return the student to ward or play area after each session (see D). !! . Payment procedure: ~14; 1. . After the first few days (or sessions) trainers will be paid Initially the trainers will be paid at the end of each train- ing session so that they understand the payment for work rule; also trainers during the first one or two times will be able ‘ _ to buy items from the store box after each training session. L_J only once following both training sessions after the second m: student has been returned to the ward. . Trainers receive five cents per session; or 10 cents at the end of their second session. . It is best, if time allows, to count out the pennies for the trainer and try to get him to begin counting with or after you; then allow him to decide whether he wants to buy some- thing (showing him the store box) or whether he wants to put his money in his box and save it; if he decides to buy, help him count out the right amount. . Training calendar: 1. Teach the trainer how to locate the appropriate box that stands for that day and to place a check or a star (depending on whether the trainer enjoys stars over checks--some older trainers may think stars are "childish"). . As with payment, this should be done after each session only initially until the trainer has the idea of this type of re- ward and record for his work. . Encourage trainer independence by: 1. 2. 3. training the trainers to pick up and return both of their stu- dents to their respective wards; training the trainers to use timers; and by using screens to allow the trainer to work more by himself; at first these might not be used during the entire 20 minute session but before the first phase is finished, the trainer should be use to working behind a screen and getting only minimal comment and assistance from the assistants. 224 K. Trainers may work at the tables or they may work on the floor with their students; however, they should be encouraged to spend some time each session (probably at the beginning) working at the table while sitting in chairs. L. When the student is engaging in disruptive behaviors 50 percent or more of the training period, then a behavior modification pro- gram will be worked out and implemented using the trainer; how- ever, before any disruptive behavior reaches this point, the assistant or attendant should try to handle it through the trainer by suggesting various methods (ignoring child and removing toys-- "time-out"; saying a loud "No!", etc.); hopefully most trainers - will initiate their own forms of control and no intervention will be needed. If the trainer appears to be reaching the point of frustration, help should be provided in a way which allows the trainer to remain in his position of control with the student -. ‘ (let the trainer implement any suggestions you might have). Un- ' _ der no conditions should the trainers be allowed to physically [p3 control the students through punishment. M. Main job of the assistant during Phase I is to encourage all in- structional and play interactions of the trainer with his stu- dents and provide vocal support. N. Each trainer will meet daily in a one hour block with the assis- tants during Phase I, training each of his two students once (20 minutes a piece). APPENDIX G GENERAL TRAINER GOALS FOR ATTENDANTS IN CONTROL GROUP . .a."'2‘“"' rm 9" GENERAL TRAINER GOALS FOR ATTENDANTS IN CONTROL GROUP Goals for trainers during Phase I (first two weeks): Group I atten- dants. l. Trainers will learn what kinds of things they should do with their students, such as trying to play with the student and getting him to play, trying to teach each student his individual learning goal, trying to teach each student things he does not already know, talking to the student, using all the equipment with the student, changing activities when the student looks bored. keeping the student in the room for the entire 20 minutes, . giving the student hugs, pats, smiles and praise when he does something that the trainer thinks is good, . giving the student food during the session when he wants to, . not hurting the student in any way but letting him know that the trainer is boss and stopping the student when he hurts others or destroys things, and k. trying to ignore the student when he engages in small, minor, bad behaviors (removing his toys, and not giving him any food or attention, and possibly saying "No."). :m-thOU'W he ‘0 2. Trainers will learn how to pick up their students from where they live, how to return them and pick up the next students to be trained (this responsibility will vary depending on the trainer). 3. Trainers will learn to put out the equipment and put it away in the box; also they will learn to pick up after the student at the end of each session (or teach the student to assist). 4. Trainers will learn to set the timers at the beginning of each ses- sion; to quit when the bell rings and begin putting equipment away and then to return the student to his ward. 5. Trainers will learn about being paid for good teaching work and how to either save or spend their earnings. 6. Trainers will learn to use the training calendar to record training sessions. 225 ‘xb‘l- F EXAMPLES OF STUDENT LANGUAGE TRAINING GOALS PROVIDED FOR ALL ATTENDANTS DURING PHASE I Mitch presentlprelanguage traininglgoals: improve his ability to a. look at objects on command and gesture, b. look at the trainer before each command is given, and c. imitate movements with objects. following success on a, b, and c fptureprelanguage training_goals: develop his ability to d. imitate movements of body parts. Debby: present_prelanguage traininq goals: improve her ability to a. imitate movements of body parts. following success on a. futurevearly language training goals: improve her ability to . imitate vowel sounds of ah, ee, oh, oo, . show the trainer objects he calls for by name, . show the trainer room parts he calls for by name, . show the trainer bOdy parts he calls for by name, and . perform simple activities that the trainer commands. ‘fifDQOU' 226 FE V APPENDIX I GENERAL TRAINER GOALS FOR ATTENDANTS IN BOTH GROUPS GENERAL TRAINER GOALS FOR ATTENDANTS IN BOTH GROUPS Goals for Trainers During Phase II (July lO-July 28): Group I and II attendants 1. Teach the trainer to perform as independently as possible in each of the following areas: *a. getting training equipment, I! , *b. getting screen and setting up, ' *c. getting student(s) and informing attendant on ward, '1 d. setting timer, e. working directly with student on a number of different activities, f. stopping work when timer rings, . a *g. returning student(s) and informing attendant on ward, H g *h. putting away equipment, and w *i. meeting attendant on assigned ward in correct area at correct time and returning by self to own ward after training is over. *These areas will be the most difficult for some trainers and their independence will vary. , 2. Be sure to remember: a. to have trainer put check or star on training calendar following training, b. to help remind trainer of next work day, time and place, c. to pay trainer and help him learn about counting, saving and buy- ing if possible, d. to praise trainer as often as you can when he does well, and e. to let trainer do most of the direct work with the student and to also do most of the discipline with the student. 3. Try to teach trainer a number of different activities to do with each of his two students based on that student's ability and needs. Also get the trainer to be familiar with these different activities so when a change in activity is needed (and the trainer does not change on his own) the supervising attendant can simply name another activity which the trainer would be able to promptly engage in with the stu- dent. 4. Try to work out solutions to any misbehavior problems with the stu- dent or with the trainer; if the student is the misbehaver then try to get the trainer to do the appropriate discipline rather than con- trolling the student yourself directly. 227 228 5. Also whenever there is time, please try to write down the different things that happened between student and trainer and between trainer and attendant each day in the daily log notes. This will help to define how the students made the gains that they already are making because of the trainer's and your efforts and hard work. APPENDIX J PERFORMANCE GOALS FOR ATTENDANTS IN EXPERIMENTAL GROUP PERFORMANCE GOALS FOR ATTENDANTS IN EXPERIMENTAL GROUP Extra Goals for Group I during Phase 11 (July lO—July 28) 1. Learn to test students on any of the 27 training parts of the Mod- ified Language Acquisition Program to determine whether new parts need to be added to a student's training, old training procedures need to be dropped, or changes in parts being trained are needed. 2. Become familiar with the training procedures for all parts of the language program so trainers can be taught by you to train students in the correct way. 3. Learn to adjust the training cards for the trainers to fit the training needed for each student. 4. Be able to observe the student's performance as he is being trained so as to know when the student needs to be tested or when his train- ing program should be adjusted to fit his progress or lack of pro- gress. 5. Be able to interact with both trainers so that needed training changes can be made as soon as they are observed. Also become skilled at praising the trainers for all their improvements and good work to keep their quality of work with the students at a high level. 6. Be able to achieve all of the above with each trainer separately so that during the last month of the project (Phase III: July 31- August 25) you will feel able to supervise both trainers as they train simultaneously in one room. I will not provide much assis- tance during this phase but will be available to talk over the progress and changes you make with your trainers and students. 229 APPENDIX K THE MODIFIED LANGUAGE ACQUISITION PROGRAM FINAL TESTS TO ACCOMPANY THE MODIFIED LANGUAGE ACQUISITION PROGRAM THE MODIFIED LANGUAGE ACQUISITION PROGRAM FINAL TESTS T0 ACCOMPANY THE MODIFIED LANGUAGE ACQUISITION PROGRAM* Student Reinforcer: ' A. Check one: Date token no token Examiner B. Check one or more and specify: food: 1. solid Test Purpose: Check One 2. soft Initial Test 3' drink other: 1. toy Follow-up Final Test (l,2,3,4) 2. other GENERAL DIRECTIONS: The child's performance on each trial is recorded in the blank following the response as one of the following: + correct response (+) approximation - incorrect response NR no response With the exception of the first two parts in the Attending Phase (see the description for those parts in the testing manual) each correct response is worth one point and all other responses are not worth any points or fraction of points. SUMMARY OF SCORES: Number of possible points follow individual part, and number of points needed to pass are in par- enthesis. Total scores, enter them below, and circle training area(s). 1. Attending Phase A. Sitting - 30 (30) B. Elimination of ICMR - 30 (30) C. Looking at objects - 5 (5) D. Pre-Trial Eye Contact - 5 (5) *MLAP was adapted from: Kent, L., Klein, D., Falk, A., and Guenther, H., "A Language Acquisition Program for the Retarded." In McLean, J. E., Yoder, D. E., and Schiefelbusch, R. L., (eds.) Language Intervention with the Retarded: developing strategies, Baltimore, Maryland: Univer- sity Park Press, 1972; by Martha S. Rowland, Michigan State University, Department of Elementary and Special Education, 1972. The MLAPFT are meant to be accompanied by the testing and training pro- cedures described in: A Modifjed Language Aggujsitjon Program for use b A Cl'al S a d ; ‘l'el - -:r i :q i: 1 u-d t'l‘ - .1‘: P.’ , Martha 5. Rowland, 1972. 230 231 II. Motor Imitation Phrase A. With Objects - 8 (7) 8. With Body Parts - 8 (7) * — III. Vocal Imitation Phase ing score) A. Gross Vocal - 3 (3) B. Vowel - 12 {11) C. Word - 36 33) D. Phrase - 20 (no pass IV. Basic Receptive Phase (Understanding single words that label) A. Touching Objects - 16 (15) B. Pointing to Room Parts - 8 (7) C. Pointing to Body Parts - 8 (7) D. Performing Activities - 12 (11) E. Finding an Object - 16 (15) (Not Visible) V. Receptive Expansion Phase (Understanding combinations of words that label) A. Giving Trainer 2 Objects - lO (9) B. Placing Object on Room Part - lO (9) C. Pointing to Body Parts - 16 (15) (self and baby's) D. Finding 2 Objects - 10 (9) (Not Visible) VI. Basic Expressive Phase (Using single words to label) A. Naming Objects - 16 (15) B. Naming Room Parts - 8 (7) C. Naming Body Parts - 8 (7) D. Naming Activities - 12 (11) E. Naming a Concealed Object - 16 (15) VII. Expressive Expansion Phase (Using combinations of words to label) A. Asking for an Object - 4 (4) B. Telling Where Object Is - l6 . (15) C. Telling Whose Body Part is Pointed to - 16 (15) D. Telling Trainer to Perform an Activity - 6 (4) E. Naming a Missing Object - 8 (7) SCORING SECTION PERFORMANCE 1. Attending Phase I-A Sitting Still: Note whether child sits without prompts or receipts of reinforcers for 30 seconds. If less than 30 seconds, note number of seconds child sits: seconds 232 I-B Elimination of Incompatible Motor Responses (Getting rid of movements that interfere with training): ‘NOte whether child exhibits incompatible motor behavior within the 30 second sit- ting period; if so, note number of seconds child sits without ICMR: seconds Description of any ICMR: I-C Looking at Objects: Note whether child looks at correct ob- jects when trainer says, "Look at this," and points to the ob- ject. Total . key spoon gomp_ baby car key spoon comb papy. car key spoon comb baby gan . key spggn_ comb baby car 5. key_ spoon comb baby car hum-a HII I-D Pre-Trial Eye Contact: Trainer presents initial Inventory "C" again, delaying each trial slightly (may wait 5 seconds), giv- ing each child an opportunity to look at her before giving the command to "Look at this." Note whether child looks at trainer (without prompting) prior to each of the 5 "Look at this" trials. Total . spoon comb key baby 'gap . spoon ggmb_ key baby car spoon comb key_ baby car spoon comb key bapy_ car . spggn_ comb key baby car 0143de Hill II. Motor Imitation Phase: Trainer says, "Do this," and presents the following imitative models for the child to imitate. II-A Imitation with Objects Total Hammer a peg on a toy peg table Point to a chair Place a single ring on a supported vertical stick Point to a chair Put a hat on one's head Put a hat on one's head 0301-5 (”Nd 233 7. Place a single ring on a supported vertical stick 8. Hammer a peg on a toy peg table II-B Imitation with Body Parts Total Touch one's nose Stick tongue out of mouth Stick tongue out of mouth Touch stomach with both hands Touch one's nose Put both arms straight out horizontally at sides . Touch stomach with both hands . Put both arms straight out horizontally at sides cow mmth—a III. Vocal Imitation Phase III-A Grps§_yggal_lmfitation: Trainer says, "Do this, say 'ah'" presenting three times for the child to imitate. Total 1. "Do this, say 'ah'" 2. "Do this, say 'ah'" 3. "Do this, say 'ah'" III-B ygugl_1mi;g;ign; Trainer says, "Say 'ah'" presenting each vowel for the child to imitate. Total l. ah 7. ee 2. ee 8. oo 3. oh 9. oh 4. ee 10. ah 5. oo 11. oh 6. ah 12. oo III-C Word Imitation: Trainer says, "Say hat" presenting each word for the child to imitate. Total 1. hat l3. spoon 25. gone 2. gone 14. chair 26. table 3. floor 15. hair 27. baby 4. nose 16. box 28. teeth 5. comb l7. chair 29. key 6. car 18. key 30. ball 7. comb l9. floor 31. baby 8. table 20. ball 32. box 9. eye 21. shoe 33. eye 10. hat 22. nose 34. mine 11. spoon 23. hair 35. mine 12. shoe 24. car 36. teeth 2 (A) 4 III-D Phrase Imitation: Trainer says, "Say key and comb" present- ing each phrase for the child to imitate. Total 1. Key and comb 11. Baby's eye 2. Key on chair 12. Spoon baby 3. My hair 13. Baby floor 4. Baby floor 14. Key on chair 5. My hair 15. Baby's hair 6. Key and comb 16. My teeth 7. Hat and spoon l7. Baby's eye 8. Spoon baby 18. Baby's hair 9. Hat and spoon 19. My eye 10. My teeth 20. My eye IV. Basic Receptive Phase (Understanding single words that label) IV-A Touching Objects: Trainer places objects on table in front of child and within his reaCh and says, "Show me the comb." Total 1. comb 9. hat 2. car 10. spoon 3. comb 11. baby 4. key 12. key 5. ball 13. shoe 6. ball 14. shoe 7. car 15. baby 8. spoon l6. hat IV-B Pointing to Room Parts: Trainer places a box and chair in view of the child and says, "Show me the floor." Total 1. floor 5. floor 2. box 6. chair 3. table 7. table 4. chair 8. box IV-C Pointing to Body Parts: Trainer faces child "Show me (child's name) nose." Total n: :5 D. m n: ‘< m U l. nose 5. eye 2. eye 6. nose 3. teeth 7. hair 4. teeth 8. hair 2 (A) 5 IV-D Performing Activity: Trainer places a ball and food on the table in front of the child. Both trainer and child stand as trainer gives each command. Note response to: Total 1. jump 7. sit 2. sit 8. jump 3. roll the ball 9. eat 4. march 10. bounce the ball 5. roll the ball 11. bounce the ball 6. march 12. eat IV-E Finding Objects - Not Visible: Objects are placed behind the screen while the child watches. Trainer says, "Go get the ball." Total 1. ball 9. hat 2. spoon lO. shoe 3. key 11. comb 4. baby 12. car 5. spoon 13. comb 6. key 14. car 7. baby 15. hat 8. ball 16. shoe V. Receptive Expansion Phase (Understanding two words that label) V-A Giving Trainer Two Objects: Trainer places all objects on table and says, "Give me the ga§_and the baby." (Hold out both hands) Total 1. car and baby 6. shoe and comb 2. spoon and hat 7. baby and ball 3. key and car 8. baby and shoe 4. key and car 9. shoe and spoon 5. car and spoon 10. baby and comb V-B Placing objects on room parts: Trainer places all objects and the box on the table and says, "Put the spoon in the box." Total 1. spoon in box 6. baby on table 2. car in box 7. key on chair 3. spoon on floor 8. baby in box 4. shoe in box 9. comb on floor 5. car on table 10. car on chair 236 V-C Touching Body Parts (Self ang_Baby's): Trainer places a doll in front of child and says, ”Show me the baby's teeth.“ Total 1. baby's teeth 9. baby's teeth 2. child's teeth 10. child's eye 3. child's teeth 11. baby's hair 4. child's hair 12. child's nose 5. child's hair 13. child's nose 6. baby's eye 14. child's eye 7. baby's hair 15. baby's nose 8. baby's eye 16. baby's nose V-D Findingntwo objects (not visible): Objects are placed behind the screen while child watches. Trainer says, "Go get the comb and the shoe." Total comb and shoe key and hat hat and shoe hat and shoe car and spoon baby and car . baby and car shoe and baby comb and hat spoon and baby mth—a ooooxnos VI. Basic Expressive Phase (Using single words to label) VI-A Naming,0bjects: Trainer says, "What is this?" as each object is held up, one at a time. Total 1. key 9. hat 2. baby 10. spoon 3. car 11. key 4. baby 12. ball 5. car 13. hat 6. shoe l4. comb 7. comb 15. spoon 8. ball 16. shoe VI-B Naming Room Parts: Trainer notes the child's response while pointing to various room parts and says, "What is this?" Total 1. chair 5. floor 2. box 6. table 3. chair 7. table 4. box 8. floor VII. VI-C VI-D VI-E Expressive Expansion Phase (Using combinations of words 237 Naming Bodnyarts: Trainer notes child's response as she points to his body parts and to the child's body parts and says, "What is this?" Total 1. teeth 5. eye 2. nose 6. hair 3. nose 7. teeth 4. eye 8. hair Naming Activities: Trainer performs activity and then says, "What did I do?" Total 1. eat 7. roll the ball 2. sit 8. sit 3. roll the ball 9. march 4. bounce the ball 10. jump 5. jump 11. march 6. bounce the ball 12. eat Naming Concealed Objects: Trainer notes child's response while showing object to child, placing it in box, covering box, and saying, "What is in the box?" 1. ball 9. hat 2. baby 10. spoon 3. car 11. key 4. baby 12. ball 5. car 13. comb 6. shoe 14. key 7. comb 15. spoon 8. shoe l6. hat to label) VII-A VII-B Asking for Objects: Trainer notes child's response when he shows child box of eight objects, then tips it away or puts under the table, and says, "What do you want?" Total 1. "What do you want?" 3. "What do you want?" 2. "What do you want?" 4. "What do you want?" ______ Telling where an object is: Trainer puts the object on the room part as the child watches and then notes the ' response when asking, "Where is the baby?" Total 0 3' do —-I D. U, 5. spoon on table 6. spoon in box 7. car on floor 8. hat on floor baby on floor . spoon in box . comb on table . comb on chair thd ' a m-I' I L. I I... N 38 9. ball on chair 13. car on table 10. key on table 14. baby in box 11. key in box 15. shoe in box 12. ball on floor 16. hat on table VII-C Telling whose body part is pointed to: Trainer points to body part on either the doll or the child and notes whether the child correctly answers with either "baby's," "mine," etc. in response to the question asked, "Whose nose is this?" Total baby's nose 9. baby's teeth baby's teeth 10. child;s_eye baby's hair 11. baby's hair baby's nose 12. child's hair baby's eye 13. child's eye . ghjlg;s_teeth l4. child's teeth child's nose 15. Baby's eye cfiilflf§ nose 16. child's hair CDNO‘U‘I-th-d VII-D TellingnTrainer topperform an activity: Trainer notes child's response to the question, "What do you want me to do?" performing each activity requested by the child after the question. Total r30 What do you want to do? . What do you want me to do? . What do you want e to do? . What do you want me to do? What do you want me to do? . What do you want to do? (75(1'|«b(;s.>l\l--l B i‘ VII-E Naming Missing_Objects: As the child watches, trainer places 3 objects in a box and removes one out of his view and then notes whether child names object missing from the group of 3 that he has just seen. Trainer asks, "What is one?" Total l. spoon baby key 2. comb ball car 3. comb car baby 4. baby comb ball 5. key ball hat 6. car hat comb 7. hat spoon shoe 8 . key baby spoon Comments: APPENDIX L EXPERIMENTAL TESTING, RECORDING AND SCORING PROCEDURES FOR ADMINISTRATION OF THE MODIFIED LANGUAGE ACQUISITION PROGRAM FINAL TESTS (MLAPFT) EXPERIMENTAL TESTING, RECORDING AND SCORING PROCEDURES FOR ADMINISTRATION OF THE MODIFIED LANGUAGE ACQUISITION PROGRAM FINAL TESTS (MLAPFT) General Experimental Testing_Procedures A. Most of the final tests given for each of the 29 parts in the program include two trials on each item; since the test form ‘lists all_items to be tested, it therefore includes the appro- priate number of trials allowed for each part. All items in each part's test will be given in the randomly predetermined order which is listed on the test form. All materials (toys and objects) used in the pre- and posttest- ing will be different materials than those used during training. The examiner does not begin a trial unless the child is exhibit- ing appropriate attending behavior (sitting, no incompatible motor behaviors, looking at examiner or test materials); however, waiting need not extend beyond 10 seconds when that item is given anyway. The examiner waits until the child completes the response before reinforcing him. 1. All correct responses, approximations and incorrect responses are immediately reinforced with continuous enthusiastic praise (smile, words, and perhaps tactual reinforcement) and with continuous token reinforcement, or with intermittent food or toy reinforcement; this latter choice (token or non- token reinforcement) depends upon which reinforcement system the child being tested is operating on. This information will be determined prior to testing. If the child is on a token system, the accumulated tokens may be exchanged for a choice of edible reinforcers or toy-usage reinforcers (i.e., one pull on a pull-the-string talking toy). This token ex- change as well as toy or food reinforcement may be given in- termittently throughout testing by the examiner to keep at- tending and responding at a high rate. 2. No reinforcement, social or otherwise, will be given when the subject makes no response following a test trial. As is expected, no prompts of any kind will be used to evoke re- sponses from the child during testing. (Imitative or demonstra- tive models used during the vocal and motor imitation phases are not considered prompts.) 239 240 Test Ceilings 1. *3. **4 ***5. Attending Phase: All subjects will be tested first on all parts of this phase. If there are ng_correct responses for this en- tire phase, no further testing for that subject will be done. All other parts will be scored as zero. . All subjects with at least some correct responses in the atten- ding phase are tested in the following order on all the parts of the four phases listed below: (order exceptions*) Motor Imitation Phase (parts 2a and 2b) Vocal Imitation Phase (parts 3a, 3b, and 3c)* Basic Receptive Phase (parts 4a, 4b, 4c, and 4d)** Basic Expressive Phase (parts 6a, 6b, 6c, 6d, and 6e)*** Part 3d, Phrase imitation, is tested only if the child makes some correct responses in BC, word imitation. . Part 4e, Finding one object (not visible), is tested only if the child makes at least one correct response on part 4a, touching objects. If there are any correct responses in any of the parts of the Basic Receptive Phase, then the subject is tested on those parts in the Receptive Expansion Phase which are expansions of the Basic Receptive Phase (e.g., 5b is an expansion of 4a and 4b). Those tests in this phase are then given immediately following the Basic Receptive Phase Testing and just prior to giving the Basic Expressive Phase tests. If there are no correct responses on any of the parts in the Basic Receptive Phase, then no testing is done in the Receptive Expansion Phase and it is scored zero. Part 5e, Naming a concealed object, is tested only if the child makes at least one correct response on part 5a, naming objects. If there are any correct responses in any of the parts of the Basic Expressive Phase, then the subject is tested on those parts in the Expressive Expansion Phase which are expansions of the Basic Expressive Phase (e.g., 7d is an expansion of 6d). Those tests in this phase are then given immediately following the Basic Expressive Phase tests. If there are no correct responses on any of the parts in the Basic Expressive Phase, then no test- ing is done in the Expressive Expansion Phase and it is scored zero. APPENDIX M BEHAVIORAL OBSERVATION RECORD BEHAVIORAL OBSERVATION RECORD Observational Key: Trial: Each trial consists of a stimulus presentation at the very least; these are each numbered under the section labeled "trials." Generally, however, a trial includes a child's re- sponse and a consequence. Events within a trial: 1. Attention: the child is looking at the trainer or the training materials as the stimulus is presented. A-l In order to evoke the child's attention the trainer calls the child's name in an emphasized tone. A-2 In order to evoke the child's attention the trainer calls the child's name and says, "Look at me," while pointing towards himself. A-3 In order to evoke the child's attention the trainer repeats A-2 and physically helps the child turn towards the trainer to face him. - The child is not looking as stimulus is presented. / The child is already looking and no effort is made by the trainer (name may be called but should be regarded as part of the vocal stimulus presentation rather than as attention eliciter). 2. Stimulus Presentation: use of words and/or gestures which may or may not include a physical demonstration (as in imitation) and/or materials (objects, toys, etc.) to command a response by the child. VC Vocal command, i.e., "Touch the car." VGC Gestural and vocal command, i.e., "Touch the spoon," as the trainer takes hand to mouth to gesture “spoon." D A physical demonstration (model) of the response to be imi- tated by the child, i.e., motor and vocal imitative tasks. - An incorrect stimulus presentation (includes none). (Note: Under stimulus presentation the observer should also indicate the stimuli presented, i.e., ball-spoon-key [as in receptive labeling]). 241 3. Prompt: 242} different kinds and degrees of assistance given to the child by the trainer to get the child to make the correct response. IP a. Imitative physical or vocal prompt: used to get the child to imitate the correct response (motor response) i.e., in receptive labeling when the trainer gives a vocal command to touch an object and the child, after not responding, is prompted by the trainer who touched the correct object and says as he touches it, "Do this." (Vocal imitative prompts, "Say 'ball,'" are also in this category). Physical prompt: The trainer touches the child's arm or another part of his body only briefly as a means of getting the child to begin and complete the correct response. The trainer puts the child through the first half or three- quarters of the correct response but allows the child to complete the remainder of the response, i.e., trainer pushes the child's elbow toward the object asked for but does not push his hand to the point of contact with the object. The trainer puts the child through the entire correct re- sponse. Cueing: the trainer touches, points, taps or otherwise draws attention toward all the stimulus objects but does not cue the correct choice only; also there is no physical contact with the child during cueing as is during prompting physi- cally. The trainer gives a cue (tap, point to, touch, etc.) that in- dicates the correct choice to be made among the various stim— ulus objects; no physical contact with the child. Some form of prompt (cue, IP, etc.) should have been given but was not given by the trainer. The trainer gave some form of prompt which was unnecessary (too much, i.e., P rather than a P ); difficult to judge as it is based on the child's respoAses following the un- needed prompt. 4. Response: refers to the child's response to the trainer's presenta- tion of stimuli. .+ (+) correct response (whether done with or without prompts by the trainer an approximation of the correct response (e.g., child hits table with hammer rather than hitting pounding table in the motor imitation with objects part of the program) incorrect response 243 NR no response (i.e., the child does not respond on his own and/or no assistance is given so that he does not make a response) Consequence: refers to anything that happens to the child by the trainer immediately following the child's response or an NR (which was preceded by stimulus presentation) and occurs before the next stimulus presentation. a. Feedback: generally in the form of vocal feedback to the child by the trainer meant to emphasize the correct word or phrase connected with the correct response in that trial (in receptive labeling, "That's the BALL," is said following the child touch- ing the ball when asked to do so by the trainer). + correct vocal feedback (should have emphasis on key work) - incorrect vocal feedback or none when there should have been some b. Reinforcement: the positive reinforcement that trainer gives to the child. V Vocal praise, "Good boy," etc. 1 Tactual reinforcement; pat, hug, etc. G Gestural reinforcement; clapping, jumping up and down, etc. Toy Concrete toy usage reinforcement given to the child F Food reinforcement (solid or liquid) Tk Token reinforcement c. Other Consequences: NF Negative feedback; "No," "Bad boy," etc. TI1 Temper tantrum intervention: trainer says "NO!" loudly to the child upon first evidence of tantrum behavior. 2 Trainer says "NO!" more loudly and gives a short time out to the child (turns away from child, or turns chair of child around). TI3 Trainer does all the above plus holds child in chair or physically restrains child in chair for the time out. TI Intervention of supervisingnperson: attendant, assistant or experi- menter intervenes vocally and/or physically into the ongoing train- ing between the trainer and child. (Code appropriate to the inter- vention is made at the point of intervention on the observation form . (*) Supervisor praises the trainer *V Supervisor intervenes with vocal assistance to the trainer *D Supervisor intervenes with physical assistance or a demonstra- tion for the trainer APPENDIX N SAMPLE BEHAVIORAL OBSERVATION RECORD SAMPLE BEHAVIORAL OBSERVATION RECORD Trainer Date Group _y_i ___II Student Training Level 4a Attendant Phase X II III I Atten- Stimulus fir Feed-1 , Trial tion Presentation Prompt Response back (Consequence :Comment 1 / baby * VC + + V Tk 2 / shoe VC NR 3 *1 shoe _ *C VC NF 4 / shoe i * .VC‘ P, + V Tk 5 / shoe * VC P1 + + V Tk 6 / shoe VC + + V Tk 7 A - hat. ,3 1 VC _ _ C 8 A / hat ' 3 vc c? + + v Tk 9 / hat VC C1» + + V Tk 10 / car VC + + Tk 11 / car *C vc + + 12 / hat VC - NF 13 / hat * VC + + V 1k 14 / hat ' VC + + V Tk 244 245 Sample BOR observation and tabulation of 11 categories of events: 1. 2. 10. 11. Number of complete trials 9 Number of times student was attendingprior to stimulus presentation 13 Number of student responses 13 Number of times the supervising attendant or experimental staff reinforced the trainer 5 Number of times the supervisingnattendant or experimental staff provided instructional demonstrations or comments to the trainer 3 Percentage of correct and approximate responses reinforced by the trainer 90 percent (9/lO) Percentage of tangible reinforcers (tokens, toys, food,netc.) paired with praise 88 percent (8/9) Percentage of appropriate promptslprovided by the trainer 75 percent (3/4) Consistency of the stimulus presentation 9 Judgment regarding the trainer's abiliay to fadelprompts OK Judgment regarding the appropriateness of the trainingltask(s) to the student Appropriate APPENDIX 0 DIRECTIONS TO STRUCTURE TRAINING INTERACTIONS BETWEEN ATTENDANTS AND STUDENTS, RETARDED TRAINERS AND STUDENTS DIRECTIONS TO STRUCTURE TRAINING INTERACTIONS BETWEEN ATTENDANTS AND STUDENTS, RETARDED TRAINERS AND STUDENTS Structured use of BOR: Attendant instructing student A. Motor Imitation "Attendant's name I would like you to spend the next few minutes with student's name. I want you to try to get him (her) to imitate or copy you when you" (show cards). 1. put a block into a bowl 2. ring a bell 3. and when you hit the table. "Work on teaching any or all these three tasks in any order you want and repeat them as often as you want and use anything you can think of to try and train ____ to copy or imitate you when you put the block in a bowl, hit the table, or ring the bell. You don't need to hurry. I'll let you know when to stop so just spend all your time working with ____, OK? You can start." 8. Receptive labeling "I would like you to spend the next few minutes with student's name. I want to see you try to get ____ to show you an object you ask for. Here are some objects to use; ask nim_to show you or touch any one or more of these: spoon, baby, car. Work on teaching pim_to under- stand (not to say) these words; all or some of them. Train them in any order and repeat them as often as you want. Use any ideas you can think of to try and train _____to understand these words-~car, spoon, or baby--when you say them. You don't need to hurry. I'll 246 247 let you know when to stop so just spend all of your time working with ____. OK? You can start now." Structured use of BOR: retarded trainers instructing students A. Motor Imitation "This is name of student. I want you to try to get _____to do what you do. See if you can get him to ring the bell after you ring the bell; then try to get him to hit the table after you hit the table; and you can try this too--see if he will put the block in the bowl after you put the block in the bowl. You want to try to teach _____ to copy what you do. You just work real hard with ____ for a few minutes and I'll stand right here and just watch. Here's the bell to ring, and the block and bowl and the table to hit. OK? Let's go." B. Receptive Labeling “This is name of student. doesn't know what the word 'tooth- brush' means but he uses it every day. Could you teach him to get the toothbrush when you ask him? And here's two other things he does not know the words for: spoon and car. See if you can think of a way to teach _____ to know what the words 'toothbrush,‘ 'spoon,’ and 'car' mean. Try to get ______to show you which of these goes with the words 'toothbrush' and which goes with the word 'spoon' and which goes with the word 'car.' You just work real hard with for awhile and I'll stand here and watch what you do. OK? Let's go. It APPENDIX P PRETEST AND POSTTEST MLAPFT PHASE TOTALS AND TEST TOTALS FOR PAIRS OF STUDENTS IN GROUPS E AND C 248 em.mm o.e e.o o.o mm.o mN._ em.N NN.Pm umee mm.mm o.o o.o o.o o.o o.o mm.o N¢.Nm wee e ese m me.mm o.e o.e o.o ON.o o.o mm.~ Nw.mm umee Nm.mm o.o o.o 0.: o.o o.o mm.o eN.Nm wee N ece F u eeeew m_.o_p mm.m m_.m_ em.o~ mu.m pe.~ mp.o o_.Nm umee eo.om o.o mm.p mm.m mN.m Ne.m No.m mN.em wee NF use FF em.em em.o NN.m e—.m_ we.e_ me.m NN.m oF.Nm pmee me.NN o.o o.o mN.o Nm.m o.o mo.m eN.Nm wee OP use m FN.eo_ um.m em.op ew.o Pm.w Nm.N mm.m op.um umee mm.m~ o.o Fm.e mm.e me.m e¢.e No.m me.mm wee m use N em.me o.o o.o o.o o.o o.o 0N.o ee.we umee mN._m o.o o.o o.o o.o o.o o.o mN._m wee m use m No.mm em.o me.e ep.o_ om.w me.m NN.m m_.em umee me.me o.o ee.o o.o mo.m mm.m mo.m oo.mm wee e use m mm.oe o.o o.o o.o Nm._ o.o mm.e em.em umee NN.mm o.o o.o o.o om.e o.o pm.N om.Nm wee N eee P eeeem N e m e m N P .mcwee uceeepm _eue» emece mmece emece omega mmece emcee emeee m eeece .0 wen u eeece =_ meeaeaem ea meeaa Lee mPMeOH ewe» use mpeeoe amaze haawmmmcexm ”N emcee Aeoseo mpcee Eeew emceem we weueuv emcee e>wmmeeexm ewmem no emege Nemuem magma secw meeeum we wepeuv mmece :ewmeeexu e>wpememm ”m emcee Ameuee mucee Eecw emceem we weueuv mmege m>wpeeeem ewmem "e emcee Nun-em mucee secw emceem we weueuv wees; :ewaeuwsH _eee> ”m emege AeNueN mucee Eecw meeeem we wepeuv emcee :ewueuwsH Lopez ”N emese Nepuep acme seew meeeem we weueuv emcee mewecmuu< up emcee "eeemme mm.em o.o e.o PN.o mN.N ee._ mm.N NN.mm umee mN.em o.o o.o o.o Ne._ o.o em.w om.Nm wee NP new w_ mN.oN o.o ee.o me.e mN.e mo.m we.m Nm.mm umee pp.me o.o o.o Ne.m oo.e NF.N __.N op.mm wee ow use a mm.em om.o em.o we.m Fm.Nw mo.m we.m me.em umee moN.mop oo.ow em.N_ Nm.N ee.NP mm.e eN.e mN.om Lee m use N om.me o.o o.o o.o mo._ o.o N_.w mm.ee umee —N.Nm o.o o.o o.o Ne.o o.o mN.o em.om Lee e eee m «seem N e m e m N F .mewee aceeepm __euew emcee emege emeee emeee emege emcee mmeee .easeweeoe .eN apnaw