1 SOME DETERMINANTS OF VICARIOUS REINFORCEMENT EFFECTS By ERIN M. CAMP A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHIL OSOPHY UNIVERSITY OF FLORIDA 2009
2 2009 Erin M. Camp
3 ACKNOWLEDGMENTS Many thanks are due to those who helped me complete this dissertation project. I would like to thank my family and friends for their constant support and encoura gement. I thank my labmates, especially Gracie Beavers, and Jill Harper for their contributions throughout the course of this project. I thank the members of my committee, Drs. Scott Miller, Stephen Smith, and Timot hy Vollmer. And lastly, I would like to thank Dr. Brian Iwata for providing priceless guidance and instruction throughout my graduate education.
4 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 3 LIST OF TABLES ................................................................................................................................ 5 LIST OF FIGURES .............................................................................................................................. 6 ABSTRACT .......................................................................................................................................... 7 CHAPTER 1 INTRODUCTION ......................................................................................................................... 8 Social and Be havioral Perspectives ............................................................................................. 8 Antecedent Determinants of Vicarious Reinforcement Effects ............................................... 11 Characteristics of Consequences ................................................................................................ 14 Durability of Effects .................................................................................................................... 16 2 GENERAL METHOD ................................................................................................................ 18 Subjects and Setting .................................................................................................................... 18 Stimulus Selection ....................................................................................................................... 18 Response Measurement and Interobserver Agreement ............................................................. 19 3 EXPERIMENT 1: ANTECEDENT EFFECTS OF VICARIOUS REINFORCMENT ......... 23 Stimulus Control ......................................................................................................................... 23 Establishing Operations .............................................................................................................. 24 Results and Discussion ............................................................................................................... 25 4 EXPERIMENT 2: QUALITY OF REINFORCEMENT DELIVERED TO MODEL ........... 30 Participants, Design, and Procedure .......................................................................................... 30 Results and Discussion ............................................................................................................... 31 5 EXPERIMENT 3: MAINTENANCE AND GENERALIZATION OF VICARIOUS REINFORCEMENT EFFECTS ................................................................................................. 35 Participants, Design, and Procedure .......................................................................................... 35 Results and Discussion ............................................................................................................... 36 6 GENERAL DISCUSSION ......................................................................................................... 39 LIST OF REFERENCES ................................................................................................................... 45 BIOGRAPHICAL SKETCH ............................................................................................................. 51
5 LIST OF TABLES Table page 2 1 Subject characteristics. ........................................................................................................... 21 2 2 Results of HQ edible item preference assessments .............................................................. 22 2 3 Results of LQ edible item preference assessments .............................................................. 22
6 LIST OF FIGURES Figure page 3 1 Responses per minute of target responses across conditions during the assessment of stimulus control (Experiment 1). ........................................................................................... 28 3 2 Responses per minute of target responses across conditions during the establishing operations evaluation (Experiment 1) ................................................................................... 29 4 1 Responses per minute of target responses across conditions of the quality of reinforcement assessment (Experiment 2). ........................................................................... 34 5 1 Responses per minute of target responses across conditions in the assessment of maintenance and generalization (Experiment 3). ................................................................. 38
7 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SOME DETERMINANTS OF VICARIOUS REINFORCEMENT EFFECTS By Erin M. Camp August 2009 Chair: Brian Iwata Major: Psychology A lthough vicarious reinforcement effe cts (increases in ones behavior as a result of observing anothers behavior being reinforced) have been demonstrated under a variety of experimental arrangements, little research has examined the determinants of those effects from the standpoint of basic learning principles or the conditions under which they are more or less likely to be observed. This study examined some antecedent and consequent influences of vicarious reinforcement effects, as well as the maintenance and generalization of those effects. The results of Experiment 1 identified the influence of both stimulus control and establishing operations on vicariously reinforced responding. Experiment 2 examined the influence of differential consequences delivered to a model in a vicarious reinforcement arrangement. Experiment 3 showed the maintenance, over time, of a response acquired via vicarious reinforcement as well as the generalization of responses to new environments and the generalization of imitative behavior. The general results from all th ree experiments highlight a variety of determinants of vicarious reinforcement effects.
8 CHAPTER 1 INTRODUCTION Vicarious reinforcement refers to one type of learning that occurs in the absence of direct contingencies. In general, vicarious reinforcemen t describes a change in responding as a result of observing the effectiveness of a models response; more specifically, the reinforcement of a models response. Many fields of psychology have described the influence of vicarious reinforcement, although th e terminology associated with, and mechanisms of action attributed to vicarious reinforcement by these different fields vary widely. Especially significant differences can be seen between social psychology and a behavior analytic perspective. Social and Be havioral Perspectives The fields of social and cognitive psychology frequently examine a wide range of topics based on informal learning from other people including, but not limited to: prejudice and discrimination, gender roles, group dynamics and leaders hip, conflict and violence, and personal and familial relationships. These topics are certainly diverse, as are the theoretical interpretations and empirical evidence that explain them. However, the commonality across these topics has to do with behavior t hat is acquired and/or mediated by the influence of other people. Vicarious reinforcement is one way that other people may influence the behavior of an individual. Research on vicarious reinforcement in the area of social psychology is typically subsumed under the more general category of observational learning, or social learning theory (Bandura, 1977; Miller & Dollard, 1941), which attempts to combine behavioral and cognitive perspectives. Social learning theory is a multi -process conceptualization tha t accounts for learning based on the observation of other individuals. Four components (attention, retention, reproduction, and motivation) comprise this system and, according to social leaning theory, each
9 component is necessary for observational learning to occur. The four components include both cognitive processes and behavioral outputs. The social learning approach is most classically illustrated in a series of studies by Bandura on aggression (Bandura, 1965, 1973; Bandura, Ross, & Ross, 1961, 1963a, 1963b). For example, Bandura et al. (1963a) evaluated whether aggression was more likely to occur following the observation of a successful model, an unsuccessful model, or a neutral model. Results indicated that the group that viewed the successful model was mostly likely to engage in aggression, a vicarious reinforcement effect, whereas the group that viewed the unsuccessful model was least likely to engage in aggression, a vicarious punishment effect. Other studies by Bandura have examined various quali ties of the model (e.g., a live model vs. video model or cartoon, adult model vs. peer model) as well as different subject characteristics (e.g., gender, age). Social learning theorists also have examined the effects of vicarious reinforcement on performa nces other than aggression. For example, Marston (Kanfer & Marston, 1963; Marston, 1966) examined vicarious reinforcement in the context of verbal interactions between individuals, and Liebert and Fernandez (1970) studied selections of commodities that wer e influenced by the performance of a model. In these examples, some of the variables examined (e.g., reinforcement of the models behavior) were similar to those of interest to operant learning theorists. However, several procedural differences exist (e.g. group designs vs. single -subject designs, pre -post sampling vs. repeated measures, the use of inferential statistics, etc.), as well as differences in pre -experimental assumptions and post -experimental interpretations. For example, a social learning inte rpretation of vicarious reinforcement effects might emphasize the subjects ability to attend to the model based on characteristics of the model (e.g., what the model looks
10 like, whether the model is an authority figure, or the personality or affect of t he model) as well as the subjects level of arousal or perceptual set. Or, a social learning theorist might discuss the subjects retention of the information observed, which could be the result of processes such as cognitive mapping, symbolic coding, stor age and retrieval, or mental rehearsal. By contrast, a behavior analytic theorist emphasizes those features of the experimental context that are observable and directly tied to basic principles of learning. Behavior analysts have examined intensely the ef fects of direct contingencies on behavior. For example, well established empirical and theoretical analyses can be found on the influence of behavioral processes like reinforcement and punishment (Cooper, Heron, & Heward, 2007; Pierce & Cheney, 2003). Most often, behavior analytic research has focused on the subjects direct experience with environmental contingencies. Indirect effects, such as vicarious reinforcement, have been studied less extensively. A few studies have examined vicarious reinforcemen t from a behavioral perspective (Broden, Bruce, Mitchell, Carter, & Hall, 1970; Greer, Dorow, Williams, McCorkle, & Asnses, 1991; Kazdin, 1973, 1977, 1979; Kazdin, Silverman, & Sittler, 1975; Ollendick, Dailey & Shapiro, 1983; Van Houten, Nau, MacKenzie -Ke ating, Sameoto, & Colavecchia, 1982). Some early examples of vicarious reinforcement research in applied behavior analysis focused on the classroom behavior of young students. Broden et al. and Kazdin (1973) both examined changes in the attending behavior of one student when a teacher delivered consequences for the attending behavior of a classmate. Similarly, Christy (1975) evaluated the effects of reinforcing the in-seat behavior of one student on the behavior of other nonreinforced students. Other type s of social behavior also have been studied with respect to vicarious reinforcement. For example, Ihrig and Wolchik (1988) used vicarious reinforcement to teach social responses to questions, and Van
11 Houten et al. (1982) used vicarious punishment to decrea se the disruptive behavior of pairs of children. Video modeling is a procedure that utilizes filmed models to enhance skill acquisition and may be based, in part, on vicarious reinforcement. That is, video -modeling examples that depict the consequences of a response can be considered vicarious reinforcement (e.g., Charlop, Shreibman, & Tryon, 1983; CharlopChristy, Le, & Freeman, 2001; Conyers, et al., 2004; Ihrig & Wolchik, 1988; LeBlanc et al., 2003; OConnor, 1969). For example, OConnor (1969) used video taped interactions of peers whose social behavior resulted in praise, acceptance to the group activity, access to toys, etc. to increase childrens interactions with peers. The influence of vicarious reinforcement in video modeling arrangements is unkno wn, especially when the subject receives direct reinforcement for imitating the performance demonstrated by the model. Antecedent Determinants of Vicarious Reinforcement Effects Given a number of demonstrations of vicarious reinforcement, a fundamental qu estion still remains about why responding occurs under conditions in which the subject never receives reinforcement. The social learning approach typically attributes behavior change to cognitive processes. For example, observation of the outcome of a mode ls behavior leads to the development of some expectation (Rotter, 1954) about the outcome of the subjects behavior. Expectation as an explanatory concept describes a persons belief or prediction that an event will occur in the future. For example, most people expect to receive a beverage from a vending machine after having made several previous purchases from vending machines; that is, through direct contact with consequences. Through observational learning, a person could develop the same expectation af ter observing another person purchase a beverage. Expectancy-value theory (Fishbein & Azjen, 1975) might suggest that the observation of a model could provide novel
12 information that affects ones belief about the observed action. Thus, expectancy is a func tion of beliefs (information about an action) and values (the benefit attributed to the belief). In either of the above examples, expectation is a construct that requires identification of its operational features as well as its determinants. Operationa lly, expectation is inferred when a subject has been exposed to a particular event and then behaves as if that event were to happen. In the case of vicarious reinforcement, the subject is exposed indirectly to reinforced responding (exhibited by the model ) and then behaves as if the subjects behavior were to be reinforced even though it is not. In other words, some feature of a models reinforced responding generates unreinforced responding on the part of the subject and, as such, implicates the influence of antecedent variables. This operationalization is consistent with a basic experimental preparation used to examine a common antecedent influence the process of stimulus control. It is a wellknown fact that a history of reinforced responding in the presence of certain stimuli produces a situation in which those stimuli acquire discriminative properties, and that the subsequent presentation of such discriminative stimuli occasions responding (Dinsmoor, 1995a, 1995b; Eckerman, 1969; Halle & Holt, 19 91; Maglieri, DeLeon, Rodiriguez Catter, & Sevin, 2000; Morse, & Skinner, 1958; Terrace, 1966; For example, Tiger and Hanley (2004) trained young students to request attention from a teacher only in the presence of a particular stimulus (a lei) by providin g reinforcement for responses made in the presence of the stimulus and by withholding reinforcement for responses made in the absence of the stimulus. After this history of reinforcement and extinction, requests for attention were much more likely to occur in the presence of the stimulus associated with reinforcement.
13 In the case of vicarious reinforcement, the subject may likely have an established history of receiving reinforcement when others are also reinforced. Given this history, the presence of a models reinforced responding could act as a discriminative stimulus. Thus, the observation of preferred consequences presented to a model in a vicarious reinforcement arrangement might occasion similar responses by the subject. That is, the availability o f reinforcement for a models behavior may be predictive of the availability of reinforcement for the subject, if the subject has a history of receiving reinforcers when reinforcement is provided to someone else. Another antecedent determinant of learned behavior is the extent to which a stimulus actually serves as a reinforcer, which, for a given stimulus, often is a function of degree of exposure. Deprivation and satiation, two ends of the continuum describing amount of exposure to a reinforcer, are exa mples of the more general influence of establishing operations (Michael, 1982, 1993; Skinner, 1953). Establishing operations (EOs) are antecedent events that alter the value of a reinforcer and increase the occurrence of behavior that historically produced that reinforcer. Deprivation (lack of exposure) is a classic example of an EO because it increases the value of a reinforcer and occasions behavior that has been maintained by it. By contrast, satiation (excessive exposure) decreases the value of a reinf orcer and its relevant behavior. Vollmer and Iwata (1991) examined the effects of EOs for several different types of reinforcers (food, music, and social interaction) on the performance of individuals with developmental disabilities and observed reliable i ncreases and decreases in behavior under conditions of deprivation and satiation, respectively. Responding that occurs under vicarious reinforcement contingencies may be influenced at least in part by the presence or absence of an EO. That is, because rei nforcement is not available
14 to the subject in vicarious reinforcement conditions, the subject is necessarily deprived to some extent, which may occasion responding. In summary, at least two antecedent influences stimulus control and establishing operations that are common determinants of learned behavior may contribute to the occurrence of vicariously reinforced responding. The purpose of Experiment 1 was to evaluate the separate effects of these variables in the context of vicarious reinforcement arrangem ents. Stimulus control was examined by correlating unique stimuli with the availability or unavailability of reinforcement for responding under direct reinforcement contingencies, and then testing for vicarious reinforcement in the presence of those stimul i. The role of establishing operations was examined by comparing vicarious reinforcement conditions presented during periods of deprivation vs. satiation. Characteristics of Consequences Although vicarious reinforcement is likely to be influenced in part by antecedent events, the role of consequences should not be overlooked. Even though the subject does not receive reinforcement under vicarious reinforcement conditions, the subject observes consequences delivered to a model. If vicarious reinforcement does have a discriminative stimulus function, then the effect may not be the result of just any consequence delivered to the model but instead is likely to be the result of a particular characteristic of the consequence. Previous research has demonstrated that different effects may be observed when different vicarious contingencies (i.e., reinforcement vs. punishment) are arranged (e.g., Bandura, Ross, & Ross, 1963a), but this effect may extend to characteristics of consequences within the same behavioral conti ngency. A large body of literature, both basic and applied, has identified key determinants of direct reinforcement effects (Fisher & Mazur, 1997; Herrnstein, 1961; Holland & Davison, 1971; Neef, Shade, & Miller, 1994, Neuringer, 1967; Todorov, 1973). Spe cifically, the quality, magnitude,
15 duration, immediacy, and/or schedule of reinforcement, as well as the effort of the response, can alter response strength. For example, Neef et al. examined different values of reinforcer rate, quality, and delay, as well as response effort, on completion of basic math problems and found that subjects allocated their responses according to the parameters of reinforcement that were available (e.g., allocating more responses towards reinforcers that were higher quality, or m ore immediate). Although there have been many demonstrations of vicarious reinforcement effects, the parameters of reinforcement are typically static and not manipulated within the experiment. Some vicarious reinforcement studies have manipulated reinforce r dimensions in combination with other manipulations (e.g., magnitude, Bruning et al, 1965; quality, Kelly, 1966; and schedule, Rosekrans & Hartup, 1972; Rosenbaum & Bruning, 1966), but the isolated effects of specific reinforcer dimensions are unclear. Given that differential responding has been observed when parametric manipulations are made in the context of direct reinforcement, it seems likely that responding also would be influenced under changing conditions of vicarious reinforcement. Reinforcer q uality has been identified as a highly influential determinant of reinforcement effects in applied studies when examined alone (DeLeon & Iwata, 1996; Fisher et al., 1992; Pace, Ivancic, Edwards, Iwata, & Page, 1985; Roane, Vollmer, Ringdahl, & Marcus, 1998) and when compared to other dimensions of reinforcement such as magnitude, schedule, and immediacy (Fisher, Thompson, Piazza, Crosland, & Gotjen, 1997; Mace, Neef, Shade, & Mauro, 1994; Neef, Mace, & Shade, 1992; Neef, Mace, Shea, & Shade, 1992; Neef, Sha de, & Miller, 1994). Experiment 2 examined the effects of reinforcer quality (as determined by preference ranking) on responding in the context of vicarious reinforcement by evaluating the effects of presenting different qualities of reinforcers to a model
16 Durability of Effects Once a vicarious reinforcement effect has been observed, it is unclear whether the effect will maintain over time or will extend, or generalize, to other circumstances. When direct consequences are delivered to a subject, maintenance and generalization of an acquired response often are observed or can be carefully programmed to ensure that they will be observed (Stokes & Baer, 1977). Given that the subject never directly contacts reinforcement for responding in vicarious reinforcemen t arrangements, subsequent maintenance or generalization may or may not be observed. A few studies that have utilized vicarious reinforcement procedures have presented some data on maintenance or generalization ( Charlop, Shreibman, & Tryon, 1983; Ihrig & W olchik 1988; Ollendick et al, 1983) but the results of these studies were mixed. For example, Charlop et al. compared the effects of trial and-error learning versus peer modeling on the acquisition of receptive labeling and reported similar response acquis ition across conditions but better maintenance and generalization for the modeling condition. Ihrig and Wolchik (1988) reported maintenance and generalization data following a study designed to evaluate differences between adult and peer modeling on a lang uage task. However, in some conditions of this study, both the model and the subject received reinforcers for responding, so the maintenance and generalization reported may have been a result of direct reinforcement contingencies as well as vicarious reinf orcement. By contrast, Ollendick et al. (1983) observed initial acquisition of puzzle completion via vicarious reinforcement alone but found that performance did not maintain. Only the subsequent addition of direct reinforcement was effective at recapturin g high rates of puzzle completion. The purpose of Experiment 3 was to assess maintenance and generalization of responses acquired through vicarious reinforcement. Maintenance was assessed by first establishing a response via vicarious reinforcement, and t hen continuing to conduct sessions in which the
17 subject never received direct reinforcement, to determine how quickly the response would extinguish. Two types of generalization were also assessed. The first test of generalization involved assessing the occ urrence of the response initially acquired under vicarious reinforcement conditions by conducting a test conditions in a new setting without a model present. This was a test for generalization of the specific response. The second generalization test involv ed substituting a new model who engaged in a new task in the absence of reinforcement contingencies to assess generalization of imitative behavior.
18 CHAPTER 2 GENERAL METHOD Subjects and Setting Thirteen individuals diagnosed with developmental disabili ties participated in Experiments 1, 2, and/or 3. Table 2 1 lists specific information about the subjects ages and diagnoses. Each subjects teacher or supervisor was interviewed to confirm that the subject was observed to attend to the behavior of others as well as to imitate the performance of gross motor tasks. Sessions were conducted in therapy rooms at either a day program for adults with developmental disabilities or a special education school. All sessions were 5 min in length and were conducted 3 to 6 times per day, 4 to 5 days per week, based on subject and experimenter availability. Trained confederates served as the models in all three experiments to ensure that model performance would be consistent across sessions and that they would respond even when the consequence involved delivery of non-preferred items. In each session, the model and subject were seated together on the same side of a table across from a therapist. Two identical sets of task materials (one for the model and one for the subje ct) were available on the table. Stimulus Selection Prior to the beginning of each experiment, several assessments were conducted to identify specific types of stimuli to be included in each condition. For all experiments, it was possible that subjects mi ght engage in the target response simply because no other activity was available during the sessions. As an activity control, a moderately preferred leisure activity was available throughout each session. Leisure activities were identified via multiple sti mulus without replacement (MSWO) preference assessments (DeLeon & Iwata, 1996). An array of leisure items was presented to the subject, who was instructed to select one item from the array. The subject was allowed to interact with the selected item for 30 s, after which that item was
19 removed from the array, and the subject was again instructed to pick an item from the remaining array. This continued until each item in the array was selected. Following all selections, preference for each item was determined by dividing the number of times each item was selected by the number of times that item was available in the array. A moderately preferred item was defined as an item that was ranked between the 4th and 6th selection of a 9 -item array. For all three exper iments, high quality (HQ) edible items were identified by conducting a paired stimulus preference assessment (Fisher et al., 1992). Pairs of edible items were presented to the subject, who was instructed to select one of the two items. The subject was allo wed to consume the selected item while the non -selected item was removed. Each edible item was paired with every other item until all possible pairs of items had been presented. After items had been presented, preference rankings were based on the percenta ge of trials on which each item was selected. HQ edible items were defined as those selected on at least 80% of trials. Preference assessments were conducted until 2 3 HQ edible items were identified; these are listed for each subject in Table 2 2. Experi ment 2 also required low quality (LQ) edible items, which were identified by conducting single stimulus preference assessments (Pace et al., 1985). Edible items were presented to the subject singly at least 3 times. Subjects had the opportunity to consume or not consume each item. Edible items were determined to be low quality if they were never approached or consumed by the subject (selected on 0% of presentations). The LQ edible items identified for each subject are presented in Table 2 3. Response Measur ement and Interobserver Agreement During the preference assessments, data were collected on subjects selections of items. Trained observers recorded each subjects response during each stimulus presentation. A second independent observer recorded subject selections on at least 30% (range, 33.3% to 100%) of the
20 trials presented. Interobserver agreement was calculated by comparing the records of the two observers on a trial bytrial basis. An agreement was scored for each trial on which both observers repor ted the same item selected. The number of trials with agreement was divided by the total number of trials (agreements plus disagreements) and multiplied by 100%. The mean agreement across subjects was 97.1% (range, 87.5% to 100%) for leisure activity asses sments, 97.9% (range, 91.6% to 100%) for HQ edible item assessments, and 100% for LQ edible item assessments. For Experiments 1, 2, and 3, target responses consisted of vocational tasks that were chosen for subjects on an individual basis. The subjects te achers or supervisors were interviewed to determine that the subjects had no known history of working with the specific task materials yet had the physical ability (motility and coordination) to complete the tasks. Target responses included: folding paper (Ben, Jenna, Jeff, Justin, Evan, Max, Jimmy), placing index cards in envelopes (Cal, Brad, Dave, Ted, Evan, Lisa, Jimmy), placing erasers on pencils (Evan, Cal, Ben, Jenna), or placing index cards in a basket (Jimmy, Ellen, Ben). Trained observers collected data on the frequency of subject responses and model responses as well as on the frequency of reinforcer delivery to the subject and the model when applicable. A second independent observer collected the same data on an average of 49.2% of the sessions (range, 34.8% to 78.6%) for each subject. Interobserver agreement was calculated by dividing each 5 -min session into 10 -s intervals and comparing the records of two observers on an interval byinterval basis. The smaller number of recorded responses for e ach interval was divided by the larger number of recorded responses. The proportion calculated for each interval was summed across all intervals in the session and was divided by the total number of intervals in the session to obtain a percentage of agreem ent between the two observers. The mean
21 percentage agreement across subjects was 97.2% (range, 91.1% to 100%) for subject responses, 93.0% (range, 88.5% to 99.5%) for model responses, 95.4% (range, 90.0% to 97.8%) for reinforcers delivered to the subject a nd 92.7% (range, 86.3% to 96.8%) for reinforcers delivered to the model. Table 2 1. Subject characteristics. Subject Age Diagnosis Experiment Max 40 Moderate mental retardation 1 Justin 19 Partial agenesis of the corpus callosum, Tourettes syndrome, ADHD, speech/language impaired 1 Jeff 21 Educable mental handicap, speech/language impaired 1 Ted 47 Mental retardation 1 Lisa 24 Autism 1 Ellen 14 Downs syndrome, educable mental handicap 1 Jenna 4 Developmental delay, speech/language impaired 2 D ave 14 Autism 2 Brad 10 Dandy Walker syndrome, other health impaired, ADHD, language impaired 2 Cal 19 Trainable mental handicap, speech/language impaired 2 Evan 5 Developmental delay, speech/language impaired 2, 3 Ben 5 Autism, speech/language impair ed 2, 3 Jimmy 49 Congenital cerebellar agenesis, moderate mental retardation 2, 3
22 Table 2 2. Results of HQ edible item preference assessments Subject HQ edible items Max Pretzels, Hot Tamale candies Justin Reeses Pieces, M&Ms Jeff Reeses Pieces, M&Ms Ted M&Ms, pretzels, wasabi peas Lisa M&Ms, Smarties candies Ellen Cheetos, ranch rice cakes Jenna Goldfish crackers, Reeses Pieces Dave M&Ms, Reeses Pieces, pretzels Brad Reeses Pieces, M&Ms Cal Smarties candies, Twizzlers Evan Study 2: R eeses Pieces, M&Ms Study 3: Pretzels, Smarties candies Ben Study 2: Fruit snacks, Smarties candies Study 3: Fritos, applesauce Jimmy Study 2: Hot Tamale candies, almonds Study 3: M&Ms, Reeses pieces, Cheetos Table 2 3. Results of LQ edible i tem preference assessments Subject LQ edible items Jimmy Green olives, mushrooms Ben Cauliflower, zucchini Jenna Broccoli, green olives Dave Feta cheese, green olives, cottage cheese Brad Bell pepper, feta cheese Cal Cauliflower, green olives Evan B roccoli, raisins
23 CHAPTER 3 EXPERIMENT 1: ANTECEDENT EFFECTS OF VICA RIOUS REINFORCMENT Six subjects participated in Experiment 1. The purpose of Experiment 1 was to determine the extent to which vicarious reinforcement effects are influenced by speci fic antecedent influences. The effects of stimulus control were assessed by conducting discrimination training and then testing for vicarious reinforcement effects. Establishing operations were assessed by presenting vicarious reinforcement conditions duri ng periods of deprivation and satiation. Stimulus Control The test for stimulus control involved training a subject with a history of reinforcement for responding in the presence of one stimulus but not in the presence of another, and then testing whethe r responding would continue to occur without direct reinforcement but in the presence of the stimulus that was correlated with reinforcement. Sessions were conducted in two different rooms, by two different therapists, and each room was paired with unique stimuli: colored response materials, tablecloths, therapist shirts, and plates on which edible items were placed. One set of stimuli was designated as the discriminative stimuli (SD) and the other, the stimuli delta (S). Specifically, SD sessions were con ducted by therapist A (wearing a green shirt) in room A, which contained a green table, green task materials, and green plates. S sessions were conducted by therapist B (wearing a red shirt) in room B, which contained a red table, red task materials, and red plates. The same model was present in both types of sessions. During sessions in which the model engaged in the target response, the rate of responding was similar across sessions, approximately 6 responses per min (i.e., about 1 response every 10 s). Sessions were alternated between the two rooms.
24 Baseline (BL): Task materials were present during baseline sessions in both rooms (SD and S). The model did not engage in the target response, and no programmed consequences were delivered to the model or t o the subject. Discrimination training: During discrimination training sessions, the model engaged in the target response. In the SD condition, both subject and model responses resulted in HQ edible items and brief praise delivered by the therapist on a v ariable ratio (VR) 3 schedule. In the S condition, neither subject nor model responses resulted in edible items or praise. That is, responses were reinforced in the presence of SD but not in the presence of the S Discrimination test: During discrimina tion test sessions, the model again engaged in the target response. In the SD condition, model responses resulted in HQ edible items and brief praise delivered on a VR3 schedule; however, subject responses resulted in no consequences (vicarious reinforceme nt only). In the S condition, neither subject nor model responses resulted in edible items. Establishing Operations The effects of establishing operations were examined by conducting vicarious reinforcement under two conditions. In each vicarious reinfo rcement session, a model engaged the target response. A therapist delivered contingent HQ edible items and brief praise to the model on a VR 3 schedule. No consequences were delivered for subject responses. In each contingent reinforcement session, both model and subject responses resulted in HQ edible items and brief praise on a VR 3 schedule. The effects of deprivation and satiation were assessed using and ABAB reversal design.
25 Contingent reinforcement (C -Sr): During this condition, the model engaged in t he target response, and a therapist provided HQ edible items and brief praise. In addition, responses made by the subject also resulted in the delivery of HQ items and praise. Vicarious reinforcement, Deprivation (V -Sr DEP): During this condition, the mod el engaged in the target response and a therapist provided HQ edible items and brief praise to the model. Subject responses resulted in no programmed consequences. These sessions were conducted after at least 60 min had elapsed since the subject had consum ed a meal or snack. Vicarious reinforcement, Satiation (V -Sr SAT): Prior to each satiation session, the subject was allowed 10 min of free access to the HQ edible items Satiation sessions were identical to the deprivation sessions; however, HQ edible item s were available to the subject continuously (not contingent on any response) for the duration of the session. Results and Discussion Figure 3 1 shows the rate of responding for the 3 subjects (Max, Justin, and Jeff) who participated in the stimulus contro l assessment. None of the subjects engaged in any responses during baseline in either the SD or the S condition. When discrimination training began, high rates of responding were observed for Max, Justin, and Jeff in the SD condition. In the S condition, Max and Justin responded initially, but their rates of responding quickly decreased to zero, whereas Jeff engaged in zero responses. After stable and differential rates of responding during discrimination training were observed for all 3 subjects, the dis crimination test was conducted. Max and Justins data show sustained responding during the SD condition of the discrimination test, thereby demonstrating a stimulus control effect. That is, the green stimuli, although no longer correlated with reinforceme nt, occasioned responding due to a history of correlation. Jeff responded initially in the SD test condition, but his responding quickly
26 extinguished. This pattern was replicated in subsequent discrimination training and test conditions. Jeffs data indica ted that stimulus control may have been observed, but that extinction in the test condition seemed to have a greater influence on his response rate. Figure 3 2 shows the rate of responding for the 3 subjects (Ted, Lisa, and Ellen) who participated in the establishing operations assessment. All 3 subjects engaged in similar patterns of responding. In the baseline condition, none of the subjects engaged in the target response. Ted engaged in high and increasing rates of responding in the first contingent rei nforcement conditions. Lisa and Ellens rates were higher than Teds in contingent reinforcement, although the data were slightly more variable. In the first vicarious reinforcement condition, the deprivation condition, Teds initial rate of responding was much lower than it was in contingent reinforcement, and his responding was extinguished by the fourth session. Both Lisa and Ellen showed sharp decreases in responding in the first vicarious reinforcement condition. Lisa stopped responding in the second s ession. Ellens continued to respond throughout the condition, although her rate of responding was much lower than in the previous condition. A second contingent reinforcement condition demonstrated high, stable rates of responding for all three subjects. In the satiation vicarious reinforcement condition, Ted, Lisa, and Ellen engaged in zero responses across the entire condition. The effects of deprivation and satiation on vicarious reinforcement were replicated for all 3 subjects, and a nearly identical p attern of responding was replicated for each subject when each condition was presented a second time. Ted and Lisa responded in only the first session of the deprivation condition, and ceased responding for the following sessions. Ellen engaged in the targ et response in the first 2 deprivation sessions, although her rate of responding was much
27 lower than the rate observed in the first presentation of the deprivation condition. Again, zero responses were observed for all 3 subjects in the second satiation condition.
28 Figure 3 1. Responses per minute of target responses across conditions during the assessment of stimulus control (Experiment 1).
29 Figure 3 2. Responses per minute of target responses across conditions during the establishing operations eva luation (Experiment 1)
30 CHAPTER 4 EXPERIMENT 2: QUALIT Y OF REINFORCEMENT D ELIVERED TO MODEL Participants, Design, and Procedure Seven subjects participated in Experiment 2, which involved presenting differential consequences to the model during vicarious reinforcement. HQ and LQ edible items were identified for each subject as described previously. A baseline condition was conducted in which the model did not engage in the target response, and no consequences were provided to the model or the subject. Fol lowing baseline, low quality vicarious reinforcement was assessed first, followed by an assessment of high quality vicarious reinforcement. During vicarious reinforcement conditions, the model engaged in the target response, and edible items identified as either HQ or LQ for the subject were delivered to the model along with brief praise on a fixed ratio (FR) 1 schedule. Prior to the first vicarious reinforcement session of each day, a therapist provided the subject with the edible items that would be delivered to the model such that the subject could sample each item. If little or no responding was observed across either LQ or HQ vicarious reinforcement conditions, the subject was exposed to a contingent reinforcement phase, during which HQ edible items an d brief praise were delivered to both the subject and the model on an FR 1 schedule. Prior to the first contingent reinforcement session of each day, a therapist prompted the subject to engage in the response and delivered the edible items to experience the contingency in effect. A subsequent HQ vicarious reinforcement condition was examined following the contingent reinforcement condition for any subject who experienced the contingent reinforcement condition. The effects of the quality of reinforcement for the model were evaluated in a multiple baseline across subjects design. Baseline (BL): During this condition, the model did not engage in the target response. No consequences were delivered for any responses made by the model or the subject.
31 Vicarious reinforcement (V -Sr) LQ: In this condition, the model engaged in the target response. A therapist delivered the LQ edible items (identified as low -quality for the subject) and brief praise for the models responses on an FR 1 schedule. No consequences were delivered for any responses made by the subject. Vicarious reinforcement (V -Sr) HQ: This condition was identical to the previous condition, except that HQ edible items (those highly preferred by the subject) were delivered to the model instead of LQ item s. Contingent reinforcement (C -Sr): During this condition, the model continued to engage the target response, and a therapist provided HQ edible items and brief praise. In addition, any responses made by the subject also resulted in the HQ items and prai se. Results and Discussion Figure 4 1 shows the results for Jimmy, Ben, Jenna, Dave, Brad, Cal, and Evan. During baseline, 6 out of 7 subjects engaged in zero rates of responding. One subject, Evan, engaged in a low rate of responding in one baseline sess ion, and zero rates of responding for the remaining baseline sessions. In the low quality vicarious reinforcement condition, 4 subjects (Jimmy, Ben, Dave, and Cal) engaged in zero responses, and 2 subjects (Brad and Evan) initially engaged in a low rate of responding that subsequently decreased to zero. The remaining subject (Jenna) initially engaged low rates of responding, but her responding gradually increased as the condition continued. In the first exposure to the high -quality vicarious reinforcement condition, Jimmy and Dave again engaged in zero target responses. Ben, Brad, and Evan responded initially but stopped responding after several sessions. Cal began to engage in target responses during this condition, and Jennas responding increased somewha t from the previous (LQ) condition and occurred at more stable rates.
32 Five subjects whose responding did not increase during the high-quality vicarious reinforcement condition (Jimmy, Ben, Dave, Brad, and Evan) were exposed to the contingent reinforcement condition. When reinforcers were delivered to the subject contingent on responding, all five subjects engaged in high, stable rates of responding. Following a history of contingent reinforcement and a subsequent return to baseline, the highquality vicari ous reinforcement condition was re introduced. Jimmys responding increased during his second exposure to high-quality vicarious reinforcement. By contrast, Ben, Dave, Brandon, and Evan responded in a manner similar to that observed during their first exposure to high-quality vicarious reinforcement: These subjects engaged in some responses initially but eventually stopped responding. The results of Experiment 2 are somewhat difficult to interpret due to variability in responding across subjects. Howe ver, the observed between -subject variability did not seem to be correlated with the subjects ages, diagnoses, or other characteristics. Five of the 7 subjects (Ben, Jenna, Brad, Cal, and Evan) engaged in some responding during their initial exposure to e ither the LQ or HQ vicarious reinforcement condition. Of these, Ben and Cal showed higher rates of responding under the HQ condition (although Bens responding eventually ceased). Jenna showed an increase in responding during the LQ condition, which maintained under the HQ condition. Finally, Brad and Evan initially engaged in some responding but eventually stopped in both conditions. Thus, the role of reinforcer quality in determining a vicarious reinforcement effects seems tenuous based on the present data. Although not the initial focus of this study, a history of direct reinforcement seemed to be a much more influential variable. Five subjects were exposed to this condition because they either never responded under LQ or HQ vicarious reinforcement ( Jimmy and Dave) or their responding
33 ceased under HQ vicarious reinforcement (Ben, Brad, and Evan). All five showed immediate increases in behavior under direct (contingent) reinforcement, although only one subjects (Jimmys) maintained when HQ vicarious reinforcement was reinstated. These data raise the question of whether a vicarious reinforcement effect will maintain for any given period of time.
34 Figure 4 1. Responses per minute of target responses across conditions of the quality of reinforcement assessment (Experiment 2).
35 CHAPTER 5 EXPERIMENT 3: MAINTENANCE AND GENERALIZA TION OF VICARIOUS REINFORCEMENT EFFECT S Participants, Design, and Procedure Three subjects, Evan, Ben, and Jimmy, who showed various patterns of responding in the vicarious rei nforcement condition of Experiment 2, participated in Experiment 3. The purpose of Experiment 3 was to assess maintenance and generalization of vicarious reinforcement effects. The vicarious reinforcement conditions were similar to those of the previous tw o experiments. Maintenance of vicarious reinforcement effects was assessed by repeatedly presenting vicarious reinforcement conditions until either extinction or long term persistence of responding was observed. Two types of generalization were assessed. T he first generalization test was conducted to determine whether a response acquired under vicarious reinforcement conditions would occur under novel conditions and in the absence of a model (a test for generalization of the response under new stimulus cond itions). The second generalization test was conducted to determine whether a novel response would be acquired in a novel condition with a model but in the absence of vicarious reinforcement (a test for generalization of imitative behavior). Baseline (BL): During the baseline condition, task materials were available to the model and the subject. The model did not engage in the target response, and no programmed consequences were delivered to the model or the subject. Baseline probe 1 (alone): During this probe, task materials were available to the subject. No model was present, and no programmed consequences were delivered for subject responses. Baseline probe 2 (imitation): In this baseline probe, a set of task materials for a second response was available t o the model and the subject. The model did not engage in the target response, and no programmed consequences were delivered to the model or the subject.
36 Vicarious reinforcement (V -Sr): During vicarious reinforcement conditions, the model engaged in the tar get response. A therapist presented HQ edible items and brief praise to the model on an FR 1 schedule. No programmed consequences were arranged for subject responses. Generalization probe 1 (alone): In this condition, the subject was seated in front of tas k materials for the target response in the vicarious reinforcement condition. No model was present, and no programmed consequences were delivered for subject responses. Generalization probe 2 (imitation): During this probe, the subject and model had acces s to materials for the second target response. The model engaged in the target response. No programmed consequences were delivered for model or subject responses. Results and Discussion Figure 5 1 shows rates of responding for Evan, Ben, and Jimmy for Expe riment 3. During the baseline condition, none of the subjects engaged in any responses in the baseline sessions or the baseline probes. Evan responded at low rates in the first few vicarious reinforcement sessions. However, maintenance was not observed ove r time in this condition. In fact, Evan usually responded once time at the beginning of each block of 3 vicarious reinforcement sessions and then did not engage in any additional responses for the duration of that session and the subsequent sessions. In addition, zero or near zero rates of responding were observed in both types of probe sessions, demonstrating the absence of generalization. Bens responding in the vicarious reinforcement condition was slightly more variable, but overall a low rate was obse rved and responding quickly extinguished. Ben engaged in one response at the beginning of the second generalizationimitation probe but then ceased responding for the rest of that session. Zero responses were observed in the generalizationalone probes in which no model was present.
37 Jimmy did not engage in the target response during the first two vicarious reinforcement sessions. Following these first 2 sessions, a moderate rate of responding was observed for 6 subsequent vicarious reinforcement sessions. However, ultimately, Jimmys responding extinguished. In addition, responding was initially observed in the generalizationalone probes but this performance also quickly extinguished. Zero responses were observed in the generalization imitation probes. The results of Experiment 3 showed that maintenance of a response acquired through vicarious reinforcement was quick to extinguish. As in Experiments 1 and 2, it is likely the case that initial responding was a result of stimulus control exerted by some prope rty of the experimental arrangement. However, these data again indicate that the extinction component of vicarious reinforcement conditions is highly influential on the responding of most subjects. That is, sustained responding was not observed because env ironmental conditions were not arranged to support continued responding. Generalization to novel conditions was not observed. Given that zero responses were observed, little can be said about the effect of vicarious reinforcement per se, but the data do su ggest that the absence of reinforcement for responding again had a suppressive effect.
38 Figure 5 1. Responses per minute of target responses across conditions in the assessment of maintenance and generalization (Experiment 3).
39 CHAPTER 6 GENERAL DISCUSS ION Vicarious reinforcement seems to capitalize on the influence of a number of variables, some of which were investigated in this study. The general purpose of this study was to examine some potential determinants of vicarious reinforcement effects, spec ifically, the antecedent functions of vicarious reinforcement, the influence of varied consequences provided to models, and the durability and generalization of vicarious reinforcement effects. The combination of these three experiments attempted to identi fy a range of behavioral processes involved in the occurrence of vicarious reinforcement. The results of Experiment 1 demonstrated the antecedent influences on responding during vicarious reinforcement. Results of Experiment 2 demonstrated that the quality of reinforcement delivered to a model during vicarious reinforcement was not influential for 4 of 7 subjects but it may have been influential in some way for the remaining 3 subjects. Finally, the results of Experiment 3 indicated that maintenance and generalization of vicarious reinforcement effects are not likely to be observed. Two types of antecedent influences were specifically evaluated in Experiment 1. Max, Jeff, and Justins data showed that stimulus control produced vicarious reinforcement effec ts, at least initially. Max and Justins data showed strong stimulus control effects, whereas Jeffs data highlighted the more potent control by direct contingencies. That is, stimulus control is likely to be responsible for initial vicarious responding, but the direct consequences for responding (extinction) seemed to be a more potent controlling variable. Although stimulus control was explicitly examined in Experiment 1, the results of Experiments 2 and 3 lend support to the conclusion that discriminative stimuli may initially occasion responding that is subsequently extinguished in the absence of direct reinforcement contingencies. For example, in Experiment 2, Evan engaged in a few responses at the onset of each vicarious reinforcement condition (the LQ
40 condition, and the first and second HQ conditions). A similar pattern also was observed in at least one condition for Ted, Lisa, and Ellen in Experiment 1; for Ben, Dave, and Brad in Experiment 2; and again for Evan, Ben, and Jimmy in Experiment 3. Teds, Lisas, and Ellens data from Experiment 1 revealed the antecedent influence of establishing operations on vicarious reinforcement effects. Their data indicated that at least initial vicarious reinforcement effects would be observed during conditions of de privation. By contrast, responding under vicarious reinforcement conditions was never observed during periods of satiation. In other words, some motivation to respond must be present if any vicarious reinforcement effect will be observed, even if that eff ect is temporary. Again, although tested explicitly in Experiment 1, other data also demonstrate the importance of motivational variables. In each subsequent experiment, any vicarious reinforcement condition that was presented was similar to the deprivatio n condition of Experiment 1. Thus, any effects of vicarious reinforcement that were observed were likely influenced, at least in part, by the presence of establishing operations. The results of Experiment 2 were somewhat surprising given the large amount of data indicating that reinforcer quality is an important determinant of learned behavior. That is, few subjects demonstrated a sensitivity to the quality of reinforcement provided to the model. However, the results were not completely unexpected in light of the absence of direct reinforcement contingencies. Even the conditions of HQ reinforcement delivered to the model had limited effects when the subjects did not receive those HQ items. This is demonstrated by the absence of any responding for some subjec ts (Jimmy, Dave) and the rapid extinction of responding observed for the remaining subjects (Ben, Brad, and Evan). In other words, it is not surprising that the consequences delivered to a model had little effect when no consequences
41 were delivered to a su bject. In fact, several subjects showed a great deal of sensitivity to the absence of reinforcement for responding. These data suggest that the subjects quickly contacted the extinction contingencies arranged for responding. When responding was not observe d, it seemed logical to create a history of contingent reinforcement for the response to determine whether responding would later occur under repeated conditions of vicarious reinforcement. However, this effect was observed for only one subject (Jimmy). Th ese data add evidence to the notion that some amount of intermittent, contingent reinforcement may be necessary to observe persistent responding during vicarious reinforcement conditions. The results of Experiment 3 reiterate that the effects of vicarious reinforcement are often transitory at best. Sustained maintenance was not observed for any of the subjects. This conclusion is also evidenced by Jeffs, Teds, Lisas, and Ellens data in Experiment 1 and Bens, Daves, Brads, and Evans data from Experi ment 2. In addition, the results of Experiment 3 demonstrate that the effects of vicarious reinforcement are not likely to generalize to other contexts, at least in the absence of arranged contingencies of reinforcement for generalized responding. One va riable that appeared to be highly influential for producing vicarious reinforcement effects seemed to be a previous history of direct reinforcement. This potentially influential variable was not specifically evaluated in this study, but a combination of da ta sets obtained seems to indicate a required history of reinforcement. A specific learning history was required for the stimulus control assessment in Experiment 1. As such, those data do not depict what responding may have been observed prior to any hist ory of reinforcement for responding. Experiment 2 purposefully examined the effects of LQ and HQ vicarious consequences in the absence of any programmed learning history, and some subjects failed to engage in any
42 responses or engaged in very low rates of r esponding. As a remedial strategy, a contingent reinforcement condition was implemented to determine whether subsequent vicarious reinforcement conditions would be more or less effective. However, only one subject (Jimmy) showed a different pattern of responding in a vicarious reinforcement condition subsequent to a contingent reinforcement condition. It is important to note that interpretation of the results from these studies may be limited to application for subjects with developmental disabilities. A nu mber of studies have reported positive results of vicarious reinforcement with nonclinical populations, and it may be that nonclinical subjects already have sufficiently extensive and diverse histories with both contingent reinforcement and vicarious reinf orcement, such that they are more likely to show vicarious reinforcement effects. If so, additional research on vicarious reinforcement for subjects with developmental disabilities may be especially helpful given the more limited historical influences with in this population. For example, early research in the area of language acquisition was determined to be best suited to the area of developmental disabilities given the more limited historical influences of this population. In experiments on language acqui sition in nonclinical populations, subjects tended to quickly acquire the target responses (Cuvo & Riva, 1980), and the effects of treatment may have been masked by other variables such as ongoing training in other contexts or maturation. Although the results of this study identified some determinants of vicarious reinforcement effects, more research is needed to examine a range of related behavioral processes involved in vicarious reinforcement. For example, in combination, the results of all 3 experiments indicate that vicarious reinforcement alone may not be sufficient to establish or maintain behavior. It is probably the case that previous demonstrations of vicarious reinforcement effects have involved
43 responses that continue to produce some schedule of intermittent reinforcement. For example, Charlop et al (1983) reported consistent maintenance of a receptive labeling task acquired via vicarious reinforcement procedures; however, it may be that the nature of their task made it likely that similar respons es would produce direct social reinforcement for the subject (i.e., teachers are likely to respond with praise if a subject subsequently labels correctly). By contrast, Ollendick et al. (1983) used a relatively arbitrary task (puzzle completion) that may h ave been similar to the arbitrary tasks selected for the current study, and they observed a lack of maintenance. It seems less likely that completion of these arbitrary tasks will be met with contingencies of reinforcement outside the experimental arrangem ent. Future research might attempt to further evaluate the history or availability of lean schedules of contingent reinforcement to enhance the effectiveness of vicarious reinforcement. In typical application, the isolated use of a pure vicarious reinforc ement arrangements does not seem to have much utility. That is, if a teacher or therapist is able to deliver reinforcers to a model, it seems that he or she should be equally able to deliver reinforcers to the subject. However, in combination with direct r einforcement contingencies, vicarious reinforcement procedures could certainly be used to enhance behavioral programs. It may be the case that vicarious reinforcement is most useful as a tool to address the problem of first occurrence. That is, if an individual does not demonstrate a particular skill independently, vicarious reinforcement as an antecedent event may be sufficient to occasion an initial response. However, it is probably necessary for an individuals initial responses to contact direct reinfor cement contingencies in order to ensure that additional responses will occur. Finally, if vicarious reinforcement is used a strategy for producing initial responses, it is important to note that the intended observers must probably have at least some gener alized imitative repertoire (Baer, Peterson, & Sherman, 1967;
44 Garcia, Baer, & Firestone, 1971). In other words, without a history of receiving reinforcers for imitating the behavior of others, the observer may be unlikely to reproduce the behavior of a mod el, regardless of the consequences provided to that model. Observational learning and imitation are certainly frequently reported behavioral occurrences. However, it is becoming increasingly clear that additional research is needed to provide a technological account of how these processes operate. In fact, although vicarious reinforcement is a widely accepted behavioral phenomenon, to date, no thorough operational account of vicarious reinforcement effects has been reported. It might be the case that the effects of direct reinforcement are too easily, and mistakenly, translated to account for the occurrence of vicarious reinforcement.
45 LIST OF REFERENCES Baer, D. M., Peterson, R. F., & Sherman, J. A. (1967). The development of imitation by reinforcing be havioral similarity to a model. Journal of the Experimental Analysis of Behavior, 10, 405416. Bandura, A. (1965). Influence of models reinforcement contingencies on the acquisition of imitative responses. Journal of Personality and Social Psychology, 1, 589 595. Bandura, A. (1966). Vicarious classical conditioning as a function of arousal level. Journal of personality and social psychology. 70, Bandura, A. (1971). Vicarious and self reinforcement processes. In R. Glaser (Ed.), The nature of reinfor cement. New York: Academic Press. Bandura, A. (1973). Aggression: A Social Learning Analysis. Englewood Cliffs, NJ: Prentice Hall. Bandura, A. (1977). Social Learning Theory Englewood Cliffs, NJ: Prentice Hall. Bandura, A., Ross, D., & Ross, S. (1961). Transmission of aggression through imitation of aggressive models. Journal of Abnormal and Social Psychology, 63, 575582. Bandura, A., Ross, D., & Ross, S. (1963). Imitation of film -mediated aggressive models. Journal of Abnormal and Social Psychology, 67, 274281. Bandura, A., Ross, D., & Ross, S. (1963). Vicarious reinforcement and imitative learning. Journal of Abnormal and Social Psychology, 67, 601607. Broden, M., Bruce, C., Mitchel, M. A., Carter, V., & Hall, R. V. (1970). Effects of teacher attention on attending behavior of two boys at adjacent desks. Journal of Applied Behavior Analysis, 3 199203 Bruning, J. L. (1965). Direct and vicarious effects of a shift in magnitude of reward on performance. Journal of Personality and Social Psychology 26, 278282 Carr, E. G., & Durand, V. M. (1985). Reducing behavior problems through functional communication training. Journal of Applied Behavior Analysis, 18, 111126. Charlop, M. H., Shreibman, L., & Tryon, A. S. (1983). Learning through observation : The effects of peer modeling on acquisition and generalization in autistic children. Journal of Abnormal Child Psychology, 11, 355366 Charlop -Christy, M.H., Le, L., & Freeman, K. (2001). A comparison of video modeling with invivo modeling for teaching children with autism. Journal of Autism and Developmental Disorders, 20 537552.
46 Conyers, C., Miltenberger, R.G., Peterson, B., Gubin, A., Jurgens, M., Selders, A., Dickinson, J. & Barenz, R. (2004). An evaluation of in vivo desensitization and video m odeling to increase compliance with dental procedures in persons with mental retardation. Journal of Applied Behavior Analysis, 37, 233238. Cooper, J. O., Heron, T.E., & Heward, W. L. (2007). Applied Behavior Analysis 2nd ed. Upper Saddle River, N.J.: Pearson Prentice Hall. Cuvo, A. J., & Riva, M. T. (1980). Generalization and transfer between comprehension and production: A comparison of retarded and nonretarded persons. Journal of Applied Behavior Analysis, 13, 315331. Dinsmoor, J. (1995). Stimulus c ontrol: I. Behavior Analyst, 5186. Dinsmoor, J. (1995). Stimulus control: II. Behavior Analyst, 253269. DeLeon, I. G., & Iwata, B. A. (1996). Evaluation of a multiple -stimulus presentation format for assessing reinforcer preferences. Journal of Appli ed Behavior Analysis, 29, 519533. Eckerman, C. O. (1969). Probability of reinforcement and the development of stimulus control. Journal of the Experimental Analysis of Behavior, 12, 551559. Fisher, W. W., & Mazur, J. E. (1997). Basic and applied resea rch on choice responding. Journal of Applied Behavior Analysis, 30, 387410. Fisher, W. W., Piazza, C. C., Bowman, L. G., Hagopian, L. P., Owens, J. C., & Slevin, I. (1992). A comparison of two approaches for identifying reinforcers for persons with sever e and profound disabilities. Journal of Applied Behavior Analysis, 25, 491498. Fisher, W. W., Thompson, R. H., Piazza, C. C., Crosland, K. & Gotjen, D. (1997). On the relative reinforcing effects of choice and differential consequences. Journal of Applie d Behavior Analysis, 30, 423438. Friman, P. C. (1990). Concurrent habits: what would Linus do with his blanket if his thumbsucking were treated? American Journal of Diseases of Children, 144, 1316 1318. Garcia, E., Baer, D. M., & Firestone, I. (1971). The development of generalized imitation within topographically determined boundaries. Journal of Applied Behavior Analysis, 4, 101112. Greer, R. D., Dorow, L., Williams, G., McCorkle, N., & Asnes, R. (1991). Peer -mediated procedures to induce swallowing and food acceptance in young children. Journal of Applied Behavior Analysis, 24, 783790.
47 Hagopian, L. P., Fisher, W. W., Sullivan, M. T., Acquisto, J., & LeBlanc, L. A. (1998). Effectiveness of functional communication training with and without extincti on and punishment: A summary of 21 inpatient cases. Journal of Applied Behavior Analysis, 31 211235. Halle, J. W. & Holt, B. (1991). Assessing stimulus control in natural settings: An analysis of stimuli that acquire control during training. Journal of A pplied Behavior Analysis, 24, 579589. Hamilton, M. L. (1970). Vicarious reinforcement effects on extinction. Journal of Experimental Child Psychology, 9, 108114. Herrnstein, R. J. (1961). Relative and absolute strength of response as a function of freq uency of reinforcement. Journal of the Experimental Analysis of Behavior, 4, 267272. Holland, V. & Davison, M.C. (1971). Preference for qualitatively different reinforcers. Journal of the Experimental Analysis of Behavior, 16 375380. Ihrig, K. & Wolch ik, S.A. (1988). Peer versus adult models and autistic childrens learning: acquisition, generalization, and maintenance. Journal of Autism and Developmental Disorders, 67 79 Kanfer, F. H. & Marston, A. R. (1963). Human reinforcement: vicarious and direc t. Journal of Experimental Psychology, 65, 292296. Kazdin, A. E. (1973). The effect of vicarious reinforcement on attentive behavior in the classroom. Journal of Applied Behavior Analysis, 6, 7178 Kazdin, A.E. (1977). Vicarious reinforcement and direct ion of behavior change in the classroom. Behavior Therapy, 8, 57 63 Kazdin, A.E. (1979). Vicarious reinforcement and punishment in operant programs for children. Child Behavior Therapy, 1, 13 36. Kazdin, A.E., Silverman, N.A., & Sittler, J.L. (1975). The use of prompts to enhance vicarious effects of non -verbal approval. Journal of Applied Behavior Analysis, 8, 279286. Kelly, R. 1966. Comparison of the effects of positive and negative vicarious reinforcement in an operant learning task. Journal of Educational Psychology 307310. LeBlanc, L.A., Coates, A.M., Daneshvar, S., Charlop Christy, M.H., Morris, C. & Lancaster, B.M. (2003). Using video modeling and reinforcement to teach perspective -taking skills to children with autism. Journal of Applied Behav ior Analysis, 36 253257.
48 Lennox, D. B., Miltenberger, R. G., Spengler, P., & Erfanian, N. (1988). Decelerative treatment practices with persons who have mental retardation: A review of five years of the literature. American Journal on Mental Retardation, 92 492501. Liebert, R. M., & Fernandez, L. E. (1970). Imitation as a function of vicarious and direct reward. Developmental Psychology, 230232 Liebert, R. M., & Fernandez, L. E. (1970). Effects of vicarious consequences on imitative performance. Ch ild Development 847852 Mace, F. C., Neef, N. A., Shade, D., & Mauro, B. C. (1996). Effects of problem difficulty and reinforcer quality on time allocated to concurrent arithmetic problems. Journal of Applied Behavior Analysis, 29 1124. Maglieri, K. A., DeLeon, I. G., Rodriguez Catter, V., & Sevin, B. M. (2000). Treatment of covert food stealing in an individual with Prader -Willi syndrome. Journal of Applied Behavior Analysis, 33, 615618. Marston, A. R. (1966). Determinants of the effects of vicarious reinforcement. Journal of Experimental Psychology, 71, 550558 Mazaleski, J. L., Iwata, B. A., Vollmer, T. R., Zarcone, J. R., & Smith, R. G. (1993). Differential reinforcement with and without extinction in the treatment of self injury maintained by atte ntion. Journal of Applied Behavior Analysis, 26, Michael, J. (1982). Distinguishing between the discriminative and motivational functions of stimuli. Journal of the Experimental Analysis of Behavior, 37, 149155. Michael, J. (1993). Establishing operati ons. The Behavior Analyst, 16, 191206. Miller, N. & Dollar, J. (1941). Social Learning and Imitation. Yale University Press. Morse, W. H., & Skinner, B. F. (1958). Some factors involved in the stimulus control of operant behavior. Journal of the Experim ental Analysis of Behavior, 1 103107. Neef, N. A., Mace, F. C. & Shade, D. (1993). Impulsivity in students with serious emotional disturbance: The interactive effects of reinforcer rate, delay, and quality. Journal of Applied Behavior Analysis, 26, 37 5 2. Neef, N. A., Mace, F. C., Shea, M. C., & Shade, D. (1992). Effects of reinforcer rate and reinforcer quality on time allocation: extensions of matching theory to educational settings. Journal of Applied Behavior Analysis, 25, 691699 Neef, N. A., Shad e, D., & Miller, M. S. (1994). Assessing influential dimensions of reinforcers on choice in students with serious emotional disturbance. Journal of Applied Behavior Analysis, 27 575583.
49 Neuringer, A.J. (1967). Effects of reinforcement magnitude on choic e and rate of responding. Journal of the Experimental Analysis of Behavior, 10, 417424. OConnor 1969. Modification of social withdrawal through symbolic modeling. Journal of Applied Behavior Analysis, 2 15 22. Ollendick, T. H., Dailey, D., & Shapiro, E. S. (1983). Vicarious reinforcement: Expected and unexpected effects. Journal of Applied Behavior Analysis, 16, 48549 Pace, G. M., Ivancic, M. T., Edwards, G. L., Iwata, B. A., & Page, T. J. (1985). Assessment of stimulus preference and reinforcer valu e with profoundly retarded individuals. Journal of Applied Behavior Analysis, 18, 249255. Phillips, R. E. (1968). Vicarious reinforcement and imitation in a verbal learning situation. Journal of Experimental Psychology, 76, 669 670. Pierce, W. D. & Cheney, C. D. (2003). Behavior Analysis and Learning, 3rd ed. Mahwah, NJ: Lawrence Erlbaum Associates. Roane, H. S., Vollmer, T. R., Ringdahl, J. E., & Marcus, B. A. (1998). Evaluation of a brief stimulus preference assessment. Journal of Applied Behavior A nalysis, 31, 605620. Rosekrans, M. A., & Hartup, W. W. (1967). Imitative influences of consistent and inconsistent response consequences to a model on aggressive behavior in children. Journal of Personality and Social Psychology, 429434. Rosenbaum, M. E., & Bruning, J. L. (1966) Direct and vicarious effects of variations in percentage of reinforcement on performance. Child Development, 959966. Rosenbaum, M.E., & Hewitt, O. J. (1966). The effect of electric shock on learning by performers and observe rs. Psychonomic Science, 81 82. Rotter, J.B. (1954). Social learning theory and clinical psychology. New York: Prentice Hall. Smith, R.G., & Iwata, B.A. (1997). Antecedent influences on behavior disorders. Journal of Applied Behavior Analysis, 30, 343375. Skinner, B.F. (1953). Science and human behavior New York: Macmillan. Stokes, T. F., & Baer, D. M. (1977). An implicit technology of generalization. Journal of Applied Behavior Analysis, 10, 349367. Terrace, H. S. (1966). Stimulus control. In W. K Honig (Ed.), Operant Behavior: Areas of Research and Application. New York: Appleton -Century Crofts, 271344.
50 Todorov, J.C. (1973). Interaction of frequency and magnitude of reinforcement on concurrent performances. Journal of the Experimental Analysis of Behavior, 19, 451458. Van Houten, R., Nau, P.A., MacKenzie -Keating, S.E., Sameoto, D. & Colaveccia, B. (1982). An analysis of some variables influencing the effectiveness of reprimands. Journal of Applied Behavior Analysis, 15, 65 83. Vollmer, T. R., Iwata, B. A., Zarcone, J. R., Smith, R. G., & Mazeleski, J. L. (1993). The role of attention in the treatment of attention -maintained self injurious behavior: Noncontingent reinforcement and differential reinforcement of other behavior. Journal of Applied Behavior Analysis, 26, 9 21 Vollmer, T. R., Roane, H. S., Ringdahl, J. E., & Marcus, B. A. (1999). Evaluating treatment challenges with differential reinforcement of alternative behavior. Journal of Applied Behavior Analysis, 32, 9 23 Weisberg, P., & Cl ements, P. (1977). Effects of direct, intermittent, and vicarious reinforcement procedures on the development and maintenance of instruction-following behaviors in a group of young children. Journal of Applied Behavior Analysis, 10, 314.
51 BIOGRAPHICAL SKE TCH Erin Camp initially studied behavior analysis at the University of North Carolina Wilmington (UNCW) under the supervision of Dr. Carol Pilgrim. She received a B.A. in psychology from UNCW in 2000 and subsequently entered the behavior analysis gradua te program at the (UF) under the supervision of Dr. Brian Iwata. While at UF, Erin participated in a variety of research and clinical activities. She was involved in a number of different research projects that covered a wide range of topics including the assessment and treatment of problem behavior and skill acquisition. While at UF, Erin also provided clinical behavior services to individuals who lived at a residential treatment facility for developmental disabilities (e.g., autism, mental retardation, and Prader -Willi syndrome). In addition, she engaged in several teaching activities including instructing an introductory behavior analysis course and serving as a teaching assistant for a behavior analysis l aboratory course. Erin completed her M.S. at UF in 2007. Following completion of the Ph.D. degree, Erin hopes to begin a research/clinical career specializing in behavioral applications in the field of developmental disabilities.