Behavioral Acquisition By Persons With Developmental Disabilities PDF

38 Chapter 1 Timberlake, W., & Farmer-Dougan, V. A. (1991). Reinforcement in applied settings: Figuring out ahead of time what will work. Psychological Bulletin, 110, 379-391. Tustin, R. D. (1994). Preference for reinforcers...

38 Chapter 1 Timberlake, W., & Farmer-Dougan, V. A. (1991). Reinforcement in applied settings: Figuring out ahead of time what will work. Psychological Bulletin, 110, 379-391. Tustin, R. D. (1994). Preference for reinforcers under varying schedule arrangements: A behavioral economic analysis. Journal of Applied Behavior Analysis, 27, 597- 606. Vollmer, T. R., & Iwata, B. A. (1991). Establishing operations and reinforcement effects. Journal of Applied Behavior Analysis, 24, 279-291 Watson, L. S., Orser, R., & Sanders, C. (1968). Reinforcement preferences of severely mentally retarded children in a generalized reinforcement context. American Journal of Mental Deficiency, 72, 748-756. Windsor, J., Piche, L. M., & Locke, P. A. (1994). Preference testing: A comparison of two presentation methods. Research in Developmental Disabilities, 15, 439-455. Witryol, S. L., & Fischer, W. F. (1960). Scaling children’s incentives by the method of paired comparisons. Psychological Reports, 7, 471-474. Behavioral Acquisition by Persons With Developmental Disabilities 39 Chapter 2 Behavioral Acquisition by Persons With Developmental Disabilities Anthony J. Cuvo and Paula K. Davis Southern Illinois University The application of behavioral principles to promote the acquisition of behavior by persons with mental retardation and other developmental disabilities dates back half a century. Some of the landmark studies capture the target behaviors, behavioral principles and procedures, and research methods that were of interest to the pioneers of applied behavior analysis. In one of the earliest studies, the “operant conditioning of a vegetative human organism” was investigated (Fuller, 1949, p. 587). The acquisition and extinction of arm raising by an 18-year-old male with profound mental retardation was studied. Other target behaviors for acquisition in those early studies included visual discrimination and differentiation (Barrett & Lindsley, 1962), academic behavior (Birnbrauer, Bijou, Wolf, & Kidder, 1965), vocalizations in a mute child (Kerr, Meyerson, & Michael, 1965), and verbal behavior (Salzinger, Feldman, Cowan, & Salzinger, 1965). During the past quarter century, the research literature demonstrating effective applications of behavioral principles to the instruction of persons with mental retardation has grown exponentially, such that strategies derived from applied behavior analysis represent the dominant approach in special education and adult habilitation. The recent issue of Behavior Analysis in Developmental Disabilities (Iwata et al., 1997) presents a selected collection of reprints from Journal of Applied Behavior Analysis of “classic” and “cutting edge” articles, according to its editors. Among the topics related to behavioral acquisition included in the volume are self-care and daily living skills, language acquisition and communication, leisure and recreation, academic performance, vocational skills, community preparation, as well as health, safety, and bio-behavioral applications. These topics represent the major domains of behavior acquisition for persons with mental retardation during the past several decades. Shifts in the philosophy of habilitation (e.g., normalization, inclusion, self- determination), bolstered by supportive legislation (e.g., Individuals with Disabili- ties Education Act, 1990) and litigation (Halderman v. Pennhurst, 1977) during the past 40 years, have resulted in the movement of persons with mental retardation from life-long institutionalization in public residential facilities to life in the community where they live, go to school, go to work, and take part in other activities 40 Chapter 2 that the community has to offer (e.g., eating in restaurants, taking public transpor- tation, engaging in recreational activities). Although the evolving philosophy and law enabled people with mental retardation and other developmental disabilities to find new lives in community alternatives to institutions, it has been behavior analysis that has provided the necessary scientific framework for a technology of instruction that helps people with disabilities function effectively in the community. Although the literature on behavior acquisition for persons with mental retardation is vast, there are common elements that can be illustrated in an instructional model used by Cuvo and his associates in a program of research that originated in 1973. The model was developed in the context of behavior analysis of community living skills (Cuvo & Davis, 1983); however, it has general relevance to acquisition by persons with mental retardation. That model will be summarized because it presents the major components that must be considered by behavior analysts and others interested in acquisition. The key components of the model relevant to acquisition of all target behaviors are those pertaining to stimulus control and its transfer. [For a more complete description of the model, see Cuvo and Davis (1983). Cuvo and Davis (1998) provides a detailed description of stimulus control procedures and a discussion of problems in the transfer of stimulus control.] A Model for the Acquisition of Functional Community Living Skills Preliminary Steps to Behavioral Program Development Assess the Learner’s Environment of Community Functioning An approach that guides assessment and training of community referenced skills is based on the demands of adult life in the community. An initial step is to conduct an “ecological inventory,” a systematic method of identifying and catalog- ing skills needed by people to function in a socially integrated fashion in their community (Brown et al., 1979). Strategies for assessing the demands for living in the community include performing direct observations in the community settings where learners will live and work, as well as conducting interviews with individuals who play a significant role in the lives of persons with disabilities. Assess the Learner’s Community Living Skills After the ecological inventory has been completed, learners should be assessed to determine which living skills are and are not in their behavioral repertoires. The assessment could be either a broad-based general assessment of many skills or a focused evaluation of certain high priority skills essential for providing key reinforc- ers. Select Instructional Goals and Objectives The results of the community-based assessment are likely to reveal that learners lack a number of skills necessary to function independently in the community. The ones that increase learners’ options and independence from service providers should have the highest priority as instructional goals (Brown & Snell, 1993). Other Behavioral Acquisition by Persons With Developmental Disabilities 41 variables to consider when selecting instructional goals and objectives include whether or not the skill would be: (a) used frequently, (b) used across different stimulus conditions, (c) maintained by natural environmental conditions, and (d) appropriate for the learner’s non-disabled peers. Consider Legal and Ethical Issues Prior to implementing a behavior analysis program, legal and ethical issues must be considered. In addition to adhering to the requirements of all pertinent statutes, litigation, and administrative policies, a behavior analyst must determine whether there are special risks involved in teaching community living skills (e.g., pedestrian and public transportation skill training involves safety considerations). Task Specification and Analysis When specifying tasks to be trained, behavior analysts must consider natural stimulus and response variations associated with community settings, as well as the requirements to live adaptively in those settings when defining instructional objectives. Will learners have a functional community living skill if they respond to a limited stimulus class or must they be able to respond to a broader stimulus class? Will a single response be appropriate or should a larger response class be taught? After the task has been specified, it is a common practice to develop a task analysis for behavioral chains. Task analyses have two primary uses (Moyer & Dardig, 1978). First, the written sequence of steps could serve as an assessment instrument (e.g., a checklist with each step scored to indicate correct or incorrect independent performance). A second use is to specify responses and their sequence to be trained. Task analyses, therefore, serve both an evaluation and training function. The steps of a task analysis should reflect the requirements of the pertinent instructional objective. The conditions, behavior, and criterion statements that comprise the instructional objective should be incorporated in the task analyses, as appropriate. The test of the adequacy of how well a task analysis has been written comes in its use with learners. Revising a task analysis tends to be idiosyncratic to individual learners and their specific errors. Task Sequencing After one or more tasks have been specified and analyzed, the next instructional design questions are how to teach responses within tasks, and how to sequence multiple tasks for training. Teaching a Single Response Chain Whole task instruction. A review of published research on community living skills training shows that the majority of experiments have employed the whole or total task sequence to teach the links or responses within chains. Learners are instructed through all steps of the entire task analysis each session in a forward sequence until the terminal behavior has been performed. Chaining. Instead of allowing a learner to move through the whole task without mastering one or more links in the chain each session, a chaining procedure would 42 Chapter 2 require training only one response to criterion at a time. In forward chaining, training is delivered only for the first step and continues until the learner reaches criterion on that step. Subsequently, the learner performs the first step and receives training on the second step until reaching criterion. Training continues in this cumulative fashion until all steps have been mastered. Backward chaining follows the same general strategy as forward chaining except that the order proceeds cumulatively from the last to the first response in the chain. Teaching Multiple Discriminations and Chains When teaching persons to respond to multiple examples from a single instruc- tional class (i.e., members of the same stimulus class), should the various examples be taught sequentially to criterion or should they be taught simultaneously? When teaching several tasks from different instructional classes (i.e., responding to different stimulus classes), should the tasks be taught concurrently or sequentially? Would learning be maintained if previously mastered responses were integrated with new responses to be acquired? Should learners receive massed, spaced, or distributed practice on responses to be learned? Although researchers have posed these task sequencing methodological questions using somewhat different terminology, these issues have a common conceptual framework and empirical database. The research on teaching multiple tasks and multiple training examples will be considered. Distribution of trials. The temporal distribution of training trials has been classified as massed, distributed, and spaced. Research appears to favor the distrib- uted trials approach to instruction in which training trials on one task are inter- spersed with training trials on one or more other tasks. Distributed practice should result in fewer training trials to attain an acquisition criterion, and less instructional time for tasks, especially those that are more difficult (Mulligan, Guess, Holvoet, & Brown, 1980). Interspersal training. Another effective task sequencing procedure has been termed “interspersal training,” which involves mixing training trials of items on which learners have performed at or above criterion with items on which they have performed below criterion (Neef, Iwata, & Page, 1977). Interspersal training should help maintain performance on responses for which learners have reached criterion. General case instruction. Multiple training and testing exemplars are taught concurrently rather than sequentially. Guidelines for selecting exemplars include: (a) define the instructional class, (b) define the range of relevant stimulus and response variation within that class, and (c) select examples from the instructional class for use in teaching and testing (Horner, Sprague, & Wilcox, 1982). Horner and McDonald (1982) demonstrated the superiority of general case instruction compared to single instance instruction for a vocational task. Stimulus equivalence methodology. Research on stimulus equivalence has demonstrated a methodology for using match-to-sample tasks to train a few conditional relations, and then test for the emergence of new untrained conditional relations (Sidman, 1971). New conditional relations that may emerge following training of a few conditional relations are also called equivalence relations if reflexive, Behavioral Acquisition by Persons With Developmental Disabilities 43 symmetric, and transitive relational properties are demonstrated. This approach has been fruitful in expanding the number of members of stimulus classes to which persons with mental retardation respond. Stimulus equivalence methodology, as the general case approach, has utility in this instructional model by broadening the stimulus classes to which learners will respond before emitting behavioral chains. The Training Environment Dimensions of the Training Environment The training environment is composed of several physical or interpersonal dimensions, each of which can vary. Physical setting. One dimension is the physical setting in which instruction takes place. At one end of the training setting continuum is the natural location where people typically perform community living skills (e.g., grocery store). At the other end of the continuum are artificial settings where discriminative stimuli and response topographies may be quite different from those in the natural setting (e.g., classrooms). Intermediate along this continuum are simulated (analog) settings that are structurally designed to approximate discriminative stimuli in natural settings (e.g., classroom area arranged like a store). Instructional materials. A second component of the training environment is the instructional material used in the physical setting. Analogous to the setting dimension, instructional materials can also vary along a continuum from natural to simulated. In natural settings, instructional materials typically would be the actual or natural stimuli. In contrast, artificial instructional materials are physically quite dissimilar to natural materials (e.g., worksheets). Modified natural materials fall between natural and artificial materials with respect to their representation of natural discriminative stimuli. For example, actual telephones may be used to teach dialing, but they may not be plugged into wall jacks during the initial stages of training. The trainer. A third dimension of the training environment is the instructor. Traditionally, human service professionals (e.g., public school teachers) or parapro- fessionals (e.g., public residential facility direct care workers) have been the trainers. Sometimes, however, learners’ peers, with or without disabilities, may have a major role to play in instruction. Another level on the trainer dimension is self-manage- ment, when antecedents and consequences are manipulated by individuals to influence their own behavior (Browder & Shapiro, 1985). Trainer-learner ratio. The fourth dimension of the training environment is the trainer-to-learner ratio, an interpersonal dimension. Individual learner instruction is characterized by one trainer providing antecedent cues and response consequences to one learner. Group instruction, which involves one trainer and multiple learners, can be not only effective, but also more efficient than individual instruction (Reid & Favell, 1984). Stimulus Control The process of instruction involves using procedures to bring the learner under the stimulus control of natural community discriminative stimuli and conse- 44 Chapter 2 quences. Stimulus control is accomplished initially by the trainer using various prompts to occasion the target responses, and then providing reinforcing stimuli after those responses. The instructional prompts and consequences that have been used to promote stimulus control of community living skills will be discussed. Instructional Prompts Verbal instruction. Verbal instructions have been implemented along a continuum from less-to-more explicit response specification. Providing the least response specificity are nonspecific verbal prompts such as asking, “What is next?” (Cuvo, Jacobi, & Sipko, 1981). A more controlling discriminative stimulus along the verbal prompting continuum is a succinct statement specifying the response to be performed (e.g., the next step of a task analysis). Although verbal instructions typically have been provided in person by the trainer, they also have been delivered by audiotape (Alberto, Sharpton, Briggs, & Stright, 1986). Visual cues. Another category of instructional prompt is the visual cue. One of the most frequently employed forms is the picture prompt (e.g., pictorial represen- tations of the steps of a task analysis). Physical gestures (e.g., pointing or manual motions by a trainer simulating the response the learner should make) also have been employed as visual cues. Modeling. In the vast majority of the community living skills training research, service providers (e.g., public residential facility employees, graduate student experi- menters, public school teachers) have functioned as live models. Also, peers with (Wacker & Berg, 1984) and without disabilities (Blew, Schwartz, & Luce, 1985) have served as models. Rather than using live modeling, an alternative has been to employ videotaped modeling. An alternative to either live or videotaped human models has been the use of dolls or figures as models to teach community living skills. These objects have been manipulated to demonstrate responses learners should make in a simulation related to community living (Page, Iwata, & Neef, 1976). Physical prompts. The most controlling type of prompting involves the trainer grasping the participants’ relevant body parts and manipulating them to occasion the appropriate response. Physical prompts can be on a continuum that ranges from full prompting to light touches that occasion responding. Permanent prompts. Permanent prompts may be especially useful when persons with limited behavioral repertoires are learning complex skills, or for individuals with physical or sensory disabilities. These prompts are used continu- ously without transfer of stimulus control to more natural discriminative stimuli. Examples of permanent prompts or prosthetic devices include a personal telephone directory of emergency telephone numbers and picture recipe cookbooks. Consequences In order for the initially neutral antecedent stimuli described above to become discriminative for responding, it is necessary that responses emitted be followed by differential reinforcement. A variety of consequences has been used to positively reinforce community living skill responses (e.g., verbal and visual performance feedback, social, edibles, money, tokens). Behavioral Acquisition by Persons With Developmental Disabilities 45 Transfer of Stimulus Control To the degree that stimuli and response topographies in both the training and natural environments are similar, generalization may readily occur (Striefel & Owens, 1980). This is an argument for training in natural environments. To the degree that stimuli in the natural environment differ from those in the training setting, there may be a failure to generalize, either in whole or in part. One of the major challenges facing behavior analysts, therefore, is to transfer stimulus control from the training environment to the environment of ultimate functioning. Fading, particularly prompt fading, has been used most frequently as a transfer of stimulus control procedure in the community living skills literature. Qualitatively different prompts have been sequenced to bring community living responses under stimulus control. The sequence frequently cited in the literature, termed “system of least prompts,” orders instructional prompts from lesser to greater artificial control (Snell & Brown, 1993). In addition to fading across qualitatively different types of prompts, some studies have reported fading within a prompt type (e.g., various levels of modeling). Prompt delay also has been used to transfer stimulus control in community living research. Transfer is accomplished by programming either a constant or an increasing delay between presentation of the natural discriminative stimulus first and then the trainer’s prompt (Snell & Gast, 1981; Striefel & Owens, 1980). Over trials, learners begin to respond to the natural discriminative stimuli and transfer of stimulus control takes place. Skill Maintenance After learners have acquired community living skills, behavior analysts should assess and program, as necessary, for the maintenance of that behavior in the environment of ultimate functioning. A common practice has been to hope that the behavior will maintain and provide additional instruction contingent on response problems. Alternatively, instructional procedures can be implemented during original training itself to help insure response maintenance (e.g., fading the instruc- tional program, transferring from contrived to natural reinforcers, interspersal training, overlearning, permanent prompts, self-management). Recent Research The behavioral model of skill acquisition described above has been employed in research conducted by the present authors for more than 20 years. In recent years, focus has shifted from determining whether we can teach functional skills to people with developmental disabilities to investigating variables that promote more effec- tive and efficient instruction of those skills. In this section, selected studies will be presented to illustrate efforts to make this behavioral model of skill acquisition more effective and efficient. Studies will focus on community living, academic, and self- determination skills. 46 Chapter 2 Community Living Skills Instruction Promoting Stimulus Control with Textual Prompts As stated in the instructional model, task analyses have been used as assessment instruments to evaluate learner performance, as well as to specify the behavioral chain for instruction. Most typically, trainers have task analyses on a clipboard and score responses as they are emitted by learners during assessment and training. The specific steps of the task analyses generally are not disclosed to learners. For learners with limited behavioral repertoires, this manner of assessment and instruction with task analyses may be appropriate. For learners with more advanced behavioral repertoires, including reading skills, the traditional procedure may not be very efficient. It might be more efficient to disclose all the steps of the task analysis to learners and present them a written listing of the steps to be performed. That was the hypothesis of a series of studies investigating whether stimulus control could be achieved by learners using written task analyses as self-administered textual prompts. The role of the trainer was initially confined to teaching learners how to use one set of task analyses as prompts, and then providing response contingent feedback. In addition, another research question inquired about the degree of specificity of the written steps (i.e., generic vs. specific) required for promoting effective stimulus control. Would people with mild disabilities respond appropriately with generic task analyses, which included statements about outcomes to achieve, or would learners require greater response description in the steps of task analyses? Furthermore, would we be able to promote effective stimulus control using the written task analyses as self-administered prompts without trainer feedback? Could the written prompts be used for self-management of instruction without having the trainer in the immediate vicinity to provide systematic performance feedback? These research questions were investigated with three groups of young adult rehabilitation clients with mild disabilities performing functional community living tasks (i.e., cleaning an oven and stove, and doing laundry; Cuvo, Davis, O’Reilly, Mooney, & Crowley, 1992). Participants’ IQs ranged from 69 to 84, and reading grade equivalents ranged from 3.5 to 8.0. The first group was used to evaluate whether written, generic task analyses and end-of-session feedback would be sufficient to promote stimulus control or whether it was necessary to employ more specific task analyses. This research was conducted in the context of a multiple- baseline design across two kitchen-cleaning tasks for individual participants. Results showed that performance with the generic task analyses and end-of-session feedback did not result in socially meaningful improvement over baseline. When the written, specific task analyses were administered with end-of-session feedback, learners reached the acquisition criterion quickly. These findings suggest that learning would be facilitated by using greater procedural specificity of written prompts, along with end-of-session feedback. With the above finding, the effect of the specific task analysis and feedback on acquisition was systematically replicated on a second group, with the addition of Behavioral Acquisition by Persons With Developmental Disabilities 47 generalization probes to untrained exemplars of kitchen appliances in novel settings. Results confirmed that self-administered specific textual prompts and end- of-session feedback were effective instructional procedures, and that participants demonstrated generalization by cleaning additional appliances without prompts and feedback. In an effort to extend the use of textual prompts for self-management, the third group was given specific textual prompts to use as antecedent cues. The trainer was not physically present in the setting and end-of-session feedback was not provided. Results showed that prompts without feedback resulted in partial acquisition, but it was necessary to add feedback to the textual prompts for participants to attain the acquisition criterion. The results support a basic principle of behavior analysis; prompts without consequences do not promote effective stimulus control. In a subsequent study, the utility of providing learners with mild disabilities written task analyses to be used as self-administered textual prompts was further investigated (McAdam & Cuvo, 1994). The main research question was whether participants who were trained to use a set of textual prompts for one task would respond appropriately without training when given novel task analyses for different tasks. To enhance stimulus control and the utility of the task analyses as self- administered prompts, the task analyses were enhanced compared to traditional ones. In this research, each step of the self-management task analyses consisted of three written sub-steps: (a) an antecedent step, (b) a behavioral step, and (c) a self- evaluation step. The purpose of the antecedent step was to orient participants to natural discriminative stimuli for the subsequent response step. An example of an antecedent step was “Find refrigerator side walls.” These antecedent sub-steps also prompted an overt response to help insure that participants attended to these textual stimuli (e.g., “When you have found refrigerator side walls, circle ‘yes’”). This served a self-monitoring function. Response sub-steps described the behavior to be occasioned by the preceding antecedent sub-step (e.g., “Wipe refrigerator side walls from top to bottom using a sponge and warm soapy water. Wipe refrigerator side walls until all dirt is gone”). Self-evaluation sub-steps prompted participants to evaluate and correct, if necessary, their performance on the preceding step (e.g., “Did you wipe refrigerator side walls from top to bottom using a sponge and warm soapy water? Did you wipe refrigerator side walls until all dirt was gone? If not, do so now”). Self-evaluation steps also contained the overt response instruction (e.g., “If you wiped refrigerator side walls from top to bottom using a sponge and warm soapy water until all dirt was gone, circle ‘yes’”). Initially, modeling, corrective verbal feedback, and contingent descriptive praise were employed to train participants to use a written task analysis to perform one home maintenance task (e.g., clean toilet). Subsequently, participants were tested on their use of different task analyses for novel tasks combined with end-of- session general feedback (e.g., “Good job Bill. You cleaned the toilet correctly”). No direct training was provided on these novel task analyses. Results indicated that the written task analyses served as self-administered written prompts and, along with 48 Chapter 2 general feedback, provided stimulus control for the second and third tasks. This research on textual prompts (Cuvo et al., 1992; McAdam & Cuvo, 1994) provides evidence that persons with mild disabilities may be able to use written task analyses with only general end-of-session feedback to perform functional community living tasks after they have been trained how to use the textual prompts. Behavioral Acquisition using a Personalized System of Instruction Another method that may be efficient for teaching learners with more advanced behavioral repertoires, especially if the learners can read and write, is a personalized system of instruction (PSI). With PSI, learners proceed at their own rate through the course material and progress to new material only after mastering the previous material (Keller, 1968). Instructors provide repeated testing, scoring, and tutoring. Zencius, Davis, and Cuvo (1990) investigated whether PSI would be effective in teaching adults with mild disabilities to write checks, make deposits, and reconcile bank statements. Two groups of rehabilitation clients between the ages of 18 and 23 served as participants. Six of the participants had IQs between 70 and 84. Both of the remaining participants were of average intelligence; one had cerebral palsy and one had a diagnosis of psychoneurosis. Each participant received a PSI workbook divided into three instructional units, one for each of three banking tasks to be trained. Each unit began with systematic written instructions and visual models for completing that unit’s task. To promote skill generalization, the models sampled the range of check, deposit, and reconcili- ation problems that the learners were likely to encounter when handling a checking account. The procedure was derived from general case instruction. In addition to the models, the workbook also contained problems requiring skills trained in that unit, which were interspersed with skills trained in previous units to promote response maintenance. Participants worked individually in a group with other students and progressed through the instructional units at their own pace. The instructor circulated around the room, providing feedback to participants on all problems completed since the previous observation. Praise was given for correct responses and errors were corrected in one of two ways. If the instructor observed the participant make an error, he used a system of least prompts (i.e., nonspecific verbal instruction, specific verbal instruction, a gesture plus specific verbal instruction, modeling plus specific verbal instruction) to transfer stimulus control to natural discriminative stimuli. Errors the instructor did not directly observe were corrected only with specific verbal instruc- tions. All participants demonstrated rapid acquisition of the monetary skills, demon- strating mastery of each unit in 3 to 7 trials per unit. All except two participants maintained those skills at an 8-to-10-week follow-up assessment. The two who did not maintain recouped the skills with brief remedial training. Additionally, all four who were assessed for generalization performed the skills in the community (e.g., bank, grocery store, post office). This study suggests that PSI can be effective for teaching persons with mild disabilities. It may be especially suitable for those skills Behavioral Acquisition by Persons With Developmental Disabilities 49 whose natural stimuli can be presented textually and whose natural responses are written. Cumulative versus Interspersal Task Sequencing In a follow-up to the PSI research described above, a study was conducted to determine the effectiveness of two different task-sequencing procedures (Cuvo, Davis, & Gluck, 1991). Task sequencing referred to the order in which training and test trials of the instructional package were presented. In this study, 20 young adults with mild disabilities (IQs ranged from 60 to 80) were taught seven money management tasks (e.g., depositing money into a savings account, withdrawing cash from a savings account, using money orders) using self-paced instructional books and training procedures similar to those described above. In this study, however, the manner in which the 56 workbook practice problems was presented was manipulated to determine whether sequencing of the problems affected learning. In the cumulative workbook, when a new task was introduced, participants initially completed four practice problems from that task, and then additional practice problems from all previously trained tasks. In contrast, the interspersal workbook presented practice problems that reviewed previously trained tasks integrated with, rather than after, practice problems from the new task. Results showed that both task-sequencing methods produced statistically significant im- provements in performance and that gains were maintained at one-week and one- month follow-up assessments. Both methods were highly and equally, effective, probably because they incorporated maintenance training with acquisition training. Teaching Complex Discriminations Persons with mental retardation and other disabilities have faced a history of discrimination. Their rights often have been defended by family members, friends, attorneys, and rehabilitation professionals (Herr, 1983). In one of the first studies to demonstrate that people with mild mental disabilities can learn self-advocacy skills, eight men and women with a variety of mild disabilities (i.e., mild mental retardation, psychiatric disabilities, cerebral palsy, learning disabilities) were taught to discrimi- nate whether or not their rights had been violated and, if so, to engage in the necessary steps to redress the rights violation (Sievert, Cuvo, & Davis, 1988). Training occurred in two stages. In the first stage, participants were taught to discriminate whether a rights violation had occurred. A list of 30 legal rights and their accompanying conditions was developed. The rights represented four general categories: personal (e.g., right to marry), community (e.g., right to minimum wage), human service (e.g., right to look at one’s own records), and consumer rights (e.g., right to buy safe products). For each of the 30 specific rights, scenarios were developed that described hypothetical interpersonal situations in which participants were denied a request. In some of the situations, the request was justifiably denied because a condition was not met (e.g., did not pay a fee when applying for a marriage license). In other situations, the request was not justifiably denied. 50 Chapter 2 Discrimination training consisted of the verbal presentation of a scenario in which a right was or was not violated. Participants were expected to determine whether a violation had occurred and to explain why or why not. Praise was provided for correct answers; verbal instructions and modeling were provided when errors occurred. Instruction began with the first general rights category. After participants received interspersal training on all rights within that category, training on the second rights category was provided followed by practice on all specific rights within that category. This was followed by interspersal training in which scenarios from the first and second rights category were presented in arbitrary order for practice and feedback. Training continued in this fashion until all four general rights categories and their specific rights had been trained. After legal rights discrimination training, participants were taught a behavioral chain to redress rights violations. Instructional procedures included written instructions, videotaped modeling, behavioral re- hearsal, verbal feedback, and praise. Participants in this study learned to discriminate and respond to up to 200 hypothetical situations in which their rights were and were not violated. In addition, they learned a general complaint response to possible rights violations. With only one exception, participants demonstrated generalization to four role-play situations in the community. Academic Tasks Effect of Response Practice Variables on Learning Spelling and Sight Vocabulary In a series of four experiments, Cuvo and colleagues (Cuvo, Ashley, Marso, Zhang, & Fry, 1995) investigated the effect of several response practice variables on the learning of spelling and sight vocabulary by young adult rehabilitation clients and adolescents in special education classes. The first experiment involved a component analysis of the “cover-write” method for teaching spelling to adults whose IQs ranged from 70 to 85 and whose spelling grade equivalent ranged from 2.8 to 3.7. The cover-write procedure involved 10 steps in which the learner looked at a word whose spelling was to be learned, said it, printed it twice, covered the model and printed it again, checked the spelling, and repeated this process of writing with the model both present and covered. There were two major research questions: (a) whether an equivalent amount of written practice would be just as effective with the model always visible (i.e., write method), and (b) whether oral practice would be as effective as written practice (i.e., oral method). Results showed rapid acquisition with no meaningful differences for all three conditions. This suggested that practice and reinforcement were the essential components of the cover write procedure, and that oral and written practice were equally effective. The covering component was not associated with enhanced learning over and above response practice. If practice and reinforcement were the key variables, how much practice is necessary? Experiment 2 compared three levels of written practice of words (i.e., 5, 10, 15 times) for students with mild or moderate mental retardation. Results showed Behavioral Acquisition by Persons With Developmental Disabilities 51 no difference in this response practice variable; more practice did not result in more rapid acquisition or better maintenance. Writing the words 15 times was not a more effective instructional technique than writing them 5 or 10 times. Experiments 3 and 4 further tested the effects of amount of practice and relevance of practice to the words to be learned. Students with behavior disorders were administered a spelling task (Experiment 3), and students with mild or moderate mental retardation were administered a sight vocabulary task (Experiment 4). Contingent on response errors for each task, students practiced the words to be learned either one or five times, or an irrelevant word not to be learned five times. Again, results showed that additional relevant practice (i.e., 5 times) did not facilitate learning and irrelevant practice was as effective as relevant practice. Taken as a whole, this series of experiments suggests a parsimonious procedure of limited response practice and positive reinforcement for the tasks and populations studied. Effect of Community versus Classroom Instruction on Learning Sight Words Instruction of persons with mental retardation traditionally has taken place in classroom settings, with the hope that these students would subsequently respond appropriately to natural discriminative stimuli and consequences outside the classroom. In recent years, proponents of community based instruction have argued that teaching should take place in natural community environments, in large part, to increase the probability that students would respond appropriately in those settings. If training occurs in natural environments, one does not have to hope for generalization from the classroom to the community. It is an empirical question, however, whether students would generalize from school to community settings. If generalization were to occur, then it may be more efficient to train in school rather than deal with transportation and student supervision issues that arise when teaching in the community. This issue was the focus of a study in which functional sight words were taught to students with mild and moderate mental retardation under three conditions (Cuvo & Klatt, 1992). Students received instruction either in the natural environ- ment where the words occurred (e.g., stores at a mall), in school with a videotaped presentation of the words as they appeared in the community, or in school with traditional flashcard instruction. Students not only had to read the words, but also state what they would do if they saw that word. The latter attempted to measure comprehension of the word. In each condition, a constant prompt-delay procedure was used to transfer stimulus control from the trainer’s prompt to the natural textual stimuli. Results showed rapid acquisition in all three training conditions and generali- zation from the flashcard and videotape conditions to the community sites. Although there may be social arguments in favor of community-based instruction, this research demonstrated that two forms of school-based instruction were as 52 Chapter 2 effective as community based instruction for teaching functional sight vocabulary to students with mild and moderate mental retardation. Effect of Stimulus Equivalence Instruction on Learning an Arithmetic Operation Research on stimulus equivalence during the past two decades has provided a means of training academic skills such as reading (Mackay, 1985; Sidman, 1971), spelling (Stromer, Mackay, & Stoddard, 1992), and pre-math (Green, 1993). Re- search on math skills was extended by Lynch and Cuvo (1995) using match-to- sample tasks to teach conditional relations between quantities represented as printed fraction ratios (e.g., 1/4), equivalent printed decimals (e.g., 0.25), and their pictorial counterparts (e.g., 1/4 or.25 presented as 25 shaded squares of a 100-square grid). Participants were fifth-and sixth-grade students who were experiencing difficulty with math. Post-test performance by all participants indicated the emergence of equivalence relations between fractions represented as ratios, decimals, and pictures. The conditional training procedures facilitated the emergence of 12 stimulus equivalence classes. All participants matched decimals to counterpart fraction ratios and fraction ratios to equivalent decimals. Limited generalization of fraction- decimal relations emerged. The results replicated previous stimulus equivalence research showing that teaching a few relations directly provided for the emergence of many relations without additional training. Self-determination The studies reviewed in the academic and community living skills areas were selected for inclusion in this chapter to illustrate a variety of the components in the behavioral model of skill acquisition. The studies reviewed now are included to provide an application of the model to a novel task, self-determination, a commu- nity living skill that has been the subject of much discussion but little experimental research. Self-determination has been defined as “the capacity to choose and to have those choices be the determinants of one’s own actions” (Deci & Ryan, 1985, p. 38), and it is a strong predictor of quality of life (Wehmeyer & Schwartz, 1997). People with mental retardation who are more self-determined experience a higher quality of life than those who are less self-determined. Most people with developmental disabilities have the ability to express their preferences and participate in daily decision making (Williams, 1991), but they may not be given opportunities to do so. A survey of the self-determination of people with mental retardation revealed that the opportunity to participate in decisions was related to the importance of the decisions (Wehmeyer & Metzler, 1995). For example, they were more likely to participate in decisions about what to wear than where to live. Despite the fact that choosing where to live is one of the major decisions that adults make, residential placement decisions are often made by persons who are legally responsible and reflect the caregivers’ preferences rather than those of the person who will live in the residence (Turnbull, Turnbull, Bronicki, Summers, & Roeder-Gordon, 1989). Behavioral Acquisition by Persons With Developmental Disabilities 53 Two recent studies have demonstrated that people with mental retardation can be active participants in the decision about where to live (Faw, Davis, & Peck, 1996; Foxx, Faw, Taylor, Davis, & Fulia, 1993). Institutionalized adults with mental retardation were assessed to determine their preferences with respect to characteris- tics of group homes in the community (e.g., having a private room, using the telephone anytime). After the preference testing was completed, they received small photo albums that had pictures of their strongest preferences. Using the photo albums as visual prompts, they were taught to go on group-home tours and ask questions about the availability of their preferences in the homes, to report that information to their social workers when they returned to the institution after the tours (Faw et al.; Foxx et al.), and to decide whether that home was a good place to live (Faw et al.). Training procedures in both studies were based on the instructional model outlined above. The procedures used in Faw et al. (1996) will be described to illustrate the research. A 31-step task analysis was developed for testing and training purposes. It included 10 steps pertaining to asking preference questions when touring group homes, 10 steps pertaining to asking for clarification if the answer to a preference question was vague, 10 steps for reporting the information obtained to the social worker at the end of the tour, and the final step requiring the participant to make a decision about the home that had just been toured. Prior to training, the institutionalized participants were assessed in community group homes and in vacant homes on their facility grounds that were similar to community homes. The assessments revealed that participants had skill deficits in determining whether their preferences were available in the homes and deciding whether they would like to live in the homes. Based on these results, training on the 31 steps of the task analysis was provided. On the first day of training, which took place in a conference room, participants were presented individualized preference booklets that had pictures of their 10 highest preferences. Under each picture was a corresponding question (e.g., “Could I use the phone anytime?”) and boxes labeled “yes,” “no,” and “maybe.” Participants were also given evaluation worksheets that showed a picture of a home and a sidewalk consisting of 20 squares that ended at the front door of the home. The 10 squares of the sidewalk closest to the home were shaded. Each “yes,” “maybe,” or “no” answer regarding a preference question resulted in 2, 1, or 0 squares of the sidewalk marked with an “X,” respectively. If any shaded portion of the sidewalk contained an “X,” participants were taught to make the decision that the home would be a good place for them to live because the home had at least half of their preferences. Next, the trainer modeled how to perform all 31 steps of the task analysis. Specifically, he demonstrated how to use the photo booklet to ask preference questions, to score responses, to request clarification for vague answers, and to use the evaluation worksheet to evaluate the home. After the trainer modeled the steps, training trials were conducted with participants individually using a total-task format. Each trial began by the trainer role-playing the completion of a group-home tour and asking participants whether 54 Chapter 2 they had any questions. The participant was then given the opportunity to ask questions and clarify ambiguous answers as the trainer continued to role play the group-home director who would answer questions. After the participants had asked all 10 questions, the trainer played the role of social worker and the participant completed the evaluation worksheet including the determination of whether the home would be a good place to live. During each training session, participants were given the opportunity to perform each step of the task analysis with no assistance. If a participant made an error or did not respond, a system of least prompts was used. A general prompt (“What’s the next step?”) was initially used. If that prompt did not evoke the response, a verbal description of the correct response was provided. If the participant still did not respond correctly, the trainer modeled the correct response and repeated the verbal direction. Descriptive feedback was provided immediately after each step regardless of the level of prompt necessary to cue performance. Additionally, general feedback and praise were provided at the end of each training trial. Criterion for completion of training was performance of 100% of all of the steps of the task analysis with no help across four consecutive trials over a period of no less than three days. After training criterion was met, participants were tested individually at the vacant homes on the facility grounds and at community group homes. The assessments were conducted identically to those in baseline, except that during post- test assessments participants were allowed to use their preference booklets and evaluation worksheets. Prior to training, participants did not independently com- plete the steps of the task analysis when taken on tours of group homes in the community or when participating in simulated tours. After training, all participants improved their performance substantially and maintained their skills at a 1-month follow-up assessment. Both self-determination studies revealed that people with mental retardation can identify preferences and can be active participants in decisions about where to live. In a similar fashion, preference information can be used to assist persons access a desired lifestyle in their current living arrangement (Faw et al., 1997). Using the preference identification procedures described above, three people with mental retardation living in a state facility selected their 10 highest preferences. Using role playing, modeling, instruction, and feedback, they were taught to direct their own interdisciplinary team meeting by beginning the meeting, reporting their prefer- ences using a picture booklet, asking if they could access their preferences, recording the answer in the booklet, and then ending the meeting. Before identifying their preferences and learning to conduct their own meetings, participants asked 0 to 4 of their preference questions. After training, the three participants asked all 10 questions regarding their preferences. Additional data from that study revealed that by changing the focus of the meeting from the traditional reporting of assessment information by professionals to discussion about the preferences of the person, the types of interactions occurring during the meetings changed considerably. Specifically, before treatment only 17% Behavioral Acquisition by Persons With Developmental Disabilities 55 of interactions were between team members and the individual. After the individu- als were taught to direct their own meetings and express their preferences, the team and individuals interacted 57% of the time. Areas For Future Study Self-determination Although self-determination is a strong predictor of a person’s quality of life (Wehmeyer & Schwartz, 1997), little empirical research has been conducted in the area. Future research could extend the studies described above in at least two ways. First, the Faw et al. (1996) study could be replicated when teaching people with developmental disabilities to make other similar decisions that have long term consequences (e.g., choosing between jobs as a stocker or fast-food worker) and those that are expected to be relatively stable over time. A second area of research would extend knowledge by teaching a more complex decision-making skill. In the Faw et al. (1996) study, persons with mental retardation learned to follow a simple model: express preferences, ask questions regarding those preferences, summarize the information obtained, and make a data-based decision. Participants were taught only how to decide if the home they toured was a good place for them to live based on the availability of their preferences. Future research could teach participants how to compare several homes with the intent of choosing one as the most suitable. That same method could then be evaluated to determine its effectiveness in teaching people how to make other decisions requiring comparisons between preferred items. Self-Advocacy Two of the studies described above illustrated that persons with developmental disabilities can be their own advocates. In Sievert et al. (1988), participants learned to discriminate between situations in which their rights were and were not violated and how to redress those violations if appropriate. In Faw et al. (1997), participants learned to lead the interdisciplinary team meeting in which training and service objectives were established for their habilitation plan. These studies demonstrated that persons with developmental disabilities can learn to advocate for their prefer- ences (Faw et al., 1997) and rights (Sievert et al.). Future research should focus on the interpersonal problem solving that is required when attempts at self-advocacy are unsuccessful. For example, in a habilitation-planning meeting, if a person with developmental disabilities makes a request that the interdisciplinary team does not honor, how should the person respond? In this example, persons with developmental disabilities need to be taught complex social skills that permit them to negotiate with team members until both sides are satisfied with the outcomes. Much of the research in the social skills area has focused on relatively simple skills that are unlikely to be sufficient to address the requirements of complex, ongoing interactions such as those described above or those required in other complex interpersonal situations (e.g., at work, on dates). Such complex social interactions often require social skills used in combination with 56 Chapter 2 problem solving skills in which one weighs the outcome of various social responses and engages in the chosen response. Research into the most effective and efficient manner of teaching complex interpersonal problem solving skills that generalize across situations within and across settings is needed. Teaching Complex Equivalence Relations Another manner that research on behavioral acquisition and generalization could be extended is to employ stimulus equivalence methodology to form more complex equivalence classes that promote the functioning of persons with mental retardation. Research during the past quarter-century has focused on equivalence classes primarily related to basic academic tasks (e.g., reading, spelling, pre-math). Few studies have been relevant to community living, such as applications of stimulus equivalence procedures to teach monetary skills to persons with mental retardation (Stoddard, Brown, Hurlbert, Manoli, & McIlvane, 1989) and name-face matching to adults with brain injuries (Cowley, Green, & Braunling-McMorrow, 1992). Stimulus equivalence methodology could be extended to promote the develop- ment of a variety of other complex stimulus classes related to the functioning of persons with mental retardation. It is highly adaptive for individuals to respond to stimulus classes with several members rather than single members of a potential class. For example, equivalence relations among the spoken word, equivalent written stimulus, and the actual physical stimulus would be most helpful for stimuli related to food, its purchase, and preparation. A person living in a group home may be verbally asked to buy bananas, read the word in a recipe, see a picture of it in a specially constructed picture recipe, and then locate and buy bananas at the grocery store. A related more complex equivalence class might entail the various foods that comprise nutritional food groups (e.g., grain, meat). The class includes various foods that are members of the same class, and the category label for that class. People with mental retardation may learn substitutable foods within a class (e.g., various fruits), and to discriminate members of one class from members of another class with stimulus equivalence methods. People with mental retardation also may have difficulty responding to subtle affective or emotional stimuli, such as facial expressions or equivalent verbal stimuli of happiness, sadness, anger, disgust, fear, and surprise. Consequently, these facial stimuli may remain neutral and individuals fail to come under their control. As a result, they may not respond in a contextually appropriate manner to others, such as housemates or staff members in a group home. In addition, individuals with mental retardation should be taught to emit verbal and nonverbal responses that cue these emotions so that others may respond appropriately to them. Affective or emotional stimuli, such as those cited above, could be taught and formed using stimulus equivalence methods. Researchers are encouraged to consider these more complex domains for the extension of stimulus equivalence methodology. Behavioral Acquisition by Persons With Developmental Disabilities 57 Reading Objectives 1. State the dominant approach for behavioral acquisition by persons with developmental disabilities. 2. Describe the essential components of the behavioral model of functional community living skills. 3. Describe how stimulus control is accomplished. 4. Describe and give examples of commonly used instructional prompts. 5. Explain the role of consequences in establishing stimulus control. 6. Describe procedures to transfer stimulus control from trainer-delivered prompts to naturally occurring discriminative stimuli. 7. Describe recent research that illustrates behavioral acquisition of community living skills. 8. Describe recent research that illustrates behavioral acquisition of academic skills. 9. Describe recent research that illustrates behavioral acquisition of self-determi- nation skills. 10. Describe three areas of behavioral acquisition in which future research could be conducted. References Alberto, P. A., Sharpton, W. R., Briggs, A., & Stright, M. H. (1986). Facilitating task acquisition through the use of a self-operated auditory prompting system. Journal of the Association for the Severely Handicapped, 11, 85-91. Barrett, B. H., & Lindsley, O. R. (1962). Deficits in acquisition of operant discrimi- nation and differentiation shown by institutionalized retarded children. Ameri- can Journal of Mental Deficiency, 67, 424-436. Birnbrauer, J. S., Bijou, S. W., Wolf, M. M., & Kidder, J. D. (1965). Programed instruction in the classroom. In L. P. Ullmann & L. Krasner (Eds.), Case studies in behavior modification (pp. 358-363). New York: Holt, Rinehart, and Winston, Inc. Blew, P. A., Schwartz, I. S., & Luce, S. C. (1985). Teaching functional community skills to autistic children using nonhandicapped peer tutors. Journal of Applied Behavior Analysis, 18, 337-342. Browder, D. M., & Shapiro, E. S. (1985). Applications of self-management to individuals with severe handicaps: A review. Journal of the Association for Persons with Severe Handicaps, 10, 200-208. Brown, F., & Snell, M. E. (1993). Meaningful assessment. In M. E. Snell (Ed.), Instruction of students with severe disabilities (4th ed.; pp. 61-98). New York: Merrill. Brown, L., Branston, M. B., Hamre-Nietupski, A., Pumpian, I., Certo, N., & Gruenewald, L. (1979). A strategy for developing chronological age-appropriate and functional curricular content for severely handicapped adolescents and young adults. Journal of Special Education, 13, 81-90. 58 Chapter 2 Cowley, B. J., Green, G., & Braunling-McMorrow, D. (1992). Using stimulus equivalence procedures to teach name-face matching to adults with brain injuries. Journal of Applied Behavior Analysis, 25, 461-475. Cuvo, A. J., Ashley, K. M., Marso, K. J., Zhang, B. L., & Fry, T. A. (1995). Effect of response practice variables on learning spelling and sight vocabulary by persons with intellectual disabilities. Journal of Applied Behavior Analysis, 28, 155-173. Cuvo, A. J., & Davis, P. K. (1998). Establishing and transferring stimulus control: Teaching people with developmental disabilities. In J. K. Luiselli & M. J. Cameron (Eds.), Antecedent control: Innovative approaches to behavioral support (pp. 347-369). Baltimore: Paul H. Brookes. Cuvo, A. J., & Davis, P. K. (1983). Behavior therapy and community living skills. In M. Hersen, R. M. Eisler, & P. M. Miller (Eds.), Progress in behavior modification (Vol. 14; pp. 125-172). New York: Academic Press. Cuvo, A. J., Davis, P. K., & Gluck, M. S. (1991). Cumulative and interspersal task sequencing in self-paced training for persons with mild handicaps. Mental Retardation, 6, 335-342. Cuvo, A. J., Davis, P. K., O’Reilly, M. F., Mooney, B. M., & Crowley, R. (1992). Promoting stimulus control with textual prompts and performance feedback for persons with mild disabilities. Journal of Applied Behavior Analysis, 25, 477-489. Cuvo, A. J., Jacobi, L., & Sipko, R. (1981). Teaching laundry skills to mentally retarded students. Education and Training of the Mentally Retarded, 16, 54-64. Cuvo, A. J., & Klatt, K. P. (1992). The effects of in vivo, videotape, and flashcard instruction of community referenced sight words on students with mental retardation. Journal of Applied Behavior Analysis, 25, 499-512. Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. New York: Plenum. Faw, G. D., Davis, P. K., McCord, B., Troutman, L., Holden, M., & Livesay, J. (1997, May). Teaching people with mental retardation to self-direct their habilitation planning meetings. Poster presented at the annual meeting of the Association for Behavior Analysis, Chicago, IL. Faw, G. D., Davis, P. K., & Peck, C. (1996). Increasing self-determination: Teaching people with mental retardation to evaluate residential options. Journal of Applied Behavior Analysis, 29, 173-188. Foxx, R. M., Faw, G. D., Taylor, S., Davis, P. K., & Fulia, R. (1993). “Would I be able to...?” Teaching clients to assess the availability of their community living preferences. American Journal of Mental Retardation, 98, 235-248. Fuller, P. R. (1949). Operant conditioning of a vegetative human organism. American Journal of Psychology, 62, 587-590. Green, G. (1993). Stimulus control technology for teaching number/quantity equivalences. Proceedings of the 1992 conference of the National Association for Autism (Australia) (pp. 51-63). Melbourne, Australia: Victoria Autistic Children’s and Adults’ Association, Inc. Halderman v. Pennhurst, 446 F. Supp. 1295 (E.D.P.A. 1977). Behavioral Acquisition by Persons With Developmental Disabilities 59 Herr, S. S. (1983). Rights and advocacy for retarded people. Lexington, MA: D.C. Heath and Co. Horner, R. H., & McDonald, R. S. (1982). Comparison of single instance and general case instruction in teaching a generalized vocational skill. Journal of the Associa- tion for Persons with Severe Handicaps, 8, 7-20. Horner, R. H., Sprague, J., & Wilcox, B. (1982). General case programming for community activities. In B. Wilcox & G. T. Bellamy (Eds.), Design of high school programs for severely handicapped students (pp. 61-98). Baltimore: Paul H. Brookes. Iwata, B. A., Bailey, J. S., Neef, N. A., Wacker, D. P., Repp, A. C., & Shook, G. L. (1997). Behavior analysis in developmental disabilities (2nd ed.). Lawrence, KS: Society for the Experimental Analysis of Behavior. Keller, F. S. (1968). “Good-bye, teacher.” Journal of Applied Behavior Analysis, 1, 79- 89. Kerr, N., Meyerson, L., & Michael, J. (1965). A procedure for shaping vocalizations in a mute child. In L. P. Ullmann & L. Krasner (Eds.), Case studies in behavior modification (pp. 366-370). New York: Holt, Rinehart, and Winston, Inc. Lynch, D. C., & Cuvo, A. J. (1995). Instruction of fraction-decimal relations via stimulus equivalence. Journal of Applied Behavior Analysis, 28, 115-126. Mackay, H. A. (1985). Stimulus equivalence in rudimentary reading and spelling. Analysis and Intervention in Developmental Disabilities, 5, 373-387. McAdam, D., & Cuvo, A. J. (1994). Textual prompts as an antecedent cue self- management strategy for persons with mild disabilities. Behavior Modification, 18, 47-65. Moyer, J. R., & Dardig, J. C. (1978). Practical task analyses for special educators. Teaching Exceptional Children, 11, 1-16. Mulligan, M., Guess, D., Holvoet, J., & Brown, F. (1980). The Individualized Curriculum Sequencing model (I): Implications for research on massed, distributed, or spaced trial training. Journal of the Association for the Severely Handicapped, 5, 325-336. Neef, N. A., Iwata, B. A., & Page, T. J. (1977). The effects of known item interspersal on acquisition and retention of spelling and sight reading words. Journal of Applied Behavior Analysis, 10, 738. Page, T. J., Iwata, B. A., & Neef, N. A. (1976). Teaching pedestrian skills to retarded persons: Generalization from the classroom to the natural environment. Journal of Applied Behavior Analysis, 9, 433-444. Reid, D. H., & Favell, J. E. (1984). Group instruction with persons who have severe disabilities: A critical review. Journal of the Association for Persons with Severe Handicaps, 9, 167-177. Salzinger, K., Feldman, R. S., Cowan, J. E., & Salzinger, S. (1965). Operant conditioning of verbal behavior of two young speech-deficient boys. In L. Krasner & L. P. Ullmann (Eds.), Research in behavior modification (pp. 82-105). New York: Holt, Rinehart, and Winston, Inc.