|
|
|||
 |
 |
 |
|
|
|||
|
|
|
|
|
|
Abstracts | Publications | Grants April 2004: Bejarano & Williams - KABA May 2000: Pietras & Hackenberg - SQAB May 2000: Vaidya & Hackenberg - ABA May 2000: O'Donnell & Saunders, K. - ABA May 2000: Dymond & Rehfeldt - ABA May 2000: Mahabir, Peláez, Cardenas, Calvani, Carasquillo, Carvalho, & Johnson - ABA May 2000: Peláez & Gewirtz - ABA Jul 1999: Peláez, Gewirtz, Sanchez, & Mahabir - International Conference for Infant Studies Aug 1999: Williams & Schroeder - American Psychological Association May 1999: Napolitano, Jack, Yoo, Weithof, Peyton, Williams, Zarcone, & Shores - ABA May 1999: Yoo, Jack, Napolitano, Peyton, Weithof, Williams, & Shores - ABA May 1999: Williams, Weithof, Yoo, Peyton, Napolitano, Jack, & Shores - ABA May 1999: Saunders, Johnston, O'Donnell, Williams, & Spradlin - ABA Mar 1999: Williams, Saunders, & Johnston - Gatlinburg Mar 1999: Dube & McIlvane - Gatlinburg Mar 1999: Pilgrim - Gatlinburg May 1998: Benedict, Johnson, & Keogh - ABA
Rafael Bejarano and Dean C. Williams, University of Kansas, Parsons Research Center Developmentally disabled individuals touched icons presented on a computer screen in exchange for money delivered on two signaled schedules that alternated (i.e., on a two-component multiple schedule). In the rich component, 10 screen touches produced a quarter, and in the lean component, a larger number of screen touches (e.g., 100) produced a point that could be exchanged for one penny after the session. Four types of transitions between components were arranged: lean-to-rich, lean-to-lean, rich-to-rich, and rich-to-lean. Experimental conditions differed from one another in terms of whether or not a timeout was interpolated between the components. With no intercomponent timeouts, the longest postreinforcement pauses occurred during the rich-to-lean transitions. This effect disappeared, however, when intercomponent timeouts were arranged. These findings suggest that off-task behaviors may occur during transitions from a more preferable activity to a less preferable one, but that it may be possible to eliminate such behaviors by interpolating a short break between activities. Poster presented at the 2nd convention of the Kansas Association for Behavior Analysis, Lawrence, KS, April, 2004. Stimulus generalization of punishment has been demonstrated with animals and college students in the laboratory and with individuals with developmental disabilities in natural settings. Lacking, however, are controlled experimental analyses of punishment generalization with individuals with MR in the laboratory, analyses that are necessary to inform behavioral intervention technology. The first step in assessing generalization is establishing control over punished responding with the stimulus correlated with punishment conditions. There are several procedural options for achieving this, all of which must prevent the punishing stimulus itself from also becoming the discriminative stimulus. The present paper describes an attempt to apply laboratory procedures used with animals and typically-developed humans to study generalization of punishment effects with individuals with mental retardation. Paper presented at the 26th Annual Convention of the Association for Behavior Analysis, Washington, D.C., May, 2000 The energy-budget rule, a model of risky choice developed by behavioral ecologists, predicts that risk sensitivity will vary as a function of rate of energy gain, daily energy requirements, and energy reserves. Under positive energy-budget conditions, in which selections of the certain option will fulfill daily energy requirements and selections of the variable option will do so only probabilistically, the model states that choice should be risk averse. Under negative energy-budget conditions, in which selections of the certain option will not fulfill daily energy requirements, choice should be risk prone. The present experiment assessed these predictions with adult humans under laboratory conditions. Positive and negative energy-budgets were arranged by manipulating the number of points that had to be accumulated within a block of trials (a criterion analogous to an energy requirement) for points to be exchanged for money. Three participants were given repeated choices between fixed and variable point amounts. Sessions consisted of twelve blocks of five trials. The first six blocks were forced-choice trials, in which only the fixed or variable option was presented on all five trials of a block; the second six blocks were free-choice trials, in which both options were presented concurrently. Exclusive preference for the fixed option could meet the criterion under positive, but not negative, energy-budget conditions. Consistent with the predictions of the model, choice was risk-averse under positive energy-budget conditions and risk-prone under negative energy-budget conditions. Poster presented at the 23rd Annual Conference of the Society for the Quantitative Analysis of Behavior, Washington, D. C., May 2000. Research on the extension of stimulus function via equivalence classes typically consists of four phases: a) prerequisite conditional relations are established, b) tests for the emergence of derived conditional relations are carried out, c) a particular function is established for a subset of the stimuli, and d) trials are presented to determine whether the trained functions have extended to other stimuli. These phases typically are presented sequentially. The results from previous experiments show extension of stimulus function prior to the development of equivalence-consistent conditional relations. The interpretations of the data, however, emphasize the mediating role of equivalence relations in the extension of stimulus function. In the current study, three subjects were exposed to trials from each of the four phases described above together beginning with the first session. This procedure allowed us to measure the development of equivalence relations and the extension of stimulus functions simultaneously. Results showed that, for different subjects, the trained discriminative functions extended to the stimuli nominally related via equivalence before, during or after the development of equivalence-consistent conditional relations. These data suggest that the relation between stimulus equivalence and extension of stimulus functions is more complex than previously believed. Paper presented at the 26th Annual Convention of the Association for Behavior Analysis, Washington, D.C., May, 2000 Accurate performance on conditional discrimination may not reflect development of sample-S+ relations (i.e., A1-B1 and A2-B2 on an AB task). For example, selections of B1 in the presence of A1 might be under sample-S- control (i.e., reject B2 in the presence of A1). Predominant or exclusive control by sample-S- relations may be a source of low accuracy on tests for symmetry, transitivity, and equivalence. Ensuring sample-S+ control may be one way of promoting emergent stimulus control. The purpose of the present study was to determine whether poor symmetry performance was related to a lack of sample-S+ control, and to determine whether training sample-S+ control directly would improve symmetry performance. Two subjects with severe mental retardation and a history of failure on equivalence tests participated. One subject showed high accuracy with both sample-S+ control and symmetry. A second subject showed low accuracy with both sample-S+ control and symmetry. When this subject was given direct training on sample-S+ and sample-S- relations, there was a slight increase in symmetry accuracy. The present research points to the importance of analyzing the sources of stimulus control underlying conditional discriminations and emergent stimulus relations. Paper presented at the 26th Annual Convention of the Association for Behavior Analysis, Washington, D.C., May, 2000 The transformation of stimulus functions is said to occur when the functions of one stimulus are altered or transformed in accordance with the derived relation between it and another stimulus, without additional training. This derived effect has been demonstrated in accordance with mutually and combinatorially entailed relations with a number of behavioral functions, experimental preparations, and subject populations. The present paper reviews the existing research on the transformation of stimulus functions and outlines a number of important methodological and conceptual issues that warrant further attention. We conclude by advocating the adoption of the generic terminology of Relational Frame Theory to describe the derived transformation of functions specifically and relational responding more generally. Paper presented at the 26th Annual Convention of the Association for Behavior Analysis, Washington, D.C., May, 2000 Research for a Behavior Analysis of 'Lying' 'Lying' has received little attention by behavior analysts. The limited existing research has, for the large part, focused on the unidirectional relationship between reinforcement and reports of 'lying' in animals and humans. Recent developments in derived stimulus relations research suggest that 'lying' may be better considered as a form of generalized operant behavior with bidirectional properties typical of verbal stimuli. This contemporary definition increases the scope of a behavior analysis of 'lying' by including research on rule-governance, derived relational control, and the effects of social conditions on, and reports of, operant behavior. The present paper will discuss the implications of findings from these research areas for a behavior-analytic account of 'lying'. Paper presented at the 26th Annual Convention of the Association for Behavior Analysis, Washington, D.C., May, 2000 We report results from several experiments on infant emotional and cognitive phenomena approached from an operant-learning paradigm. Specifically, we studied attachment, social referencing, language acquisition, equivalence relations, fear of the dark, fear of strangers, infant jealousy and infant depression. Analyses of these phenomena as operant-learning processes constitute remarkable departures from the traditional conception of process in the developmental and clinical literatures, where such phenomena are conceptualized as biologically based and maturational. The developmental literature emphasizes that there are universal emotions present at birth that emerge as the individual matures (e.g., Izard, 1992). Although results from our experiments do not rule out such unlearned or hereditary explanations, they do indicate that much of the variance in infant behavior could be accounted for by its relation with proximal antecedent and consequent stimuli provided by parents. When we studied infant attachment, fear of the dark, fear of strangers, jealousy, and depression phenomena under laboratory conditions, and with maternal behavior under experimenter control, these phenomena were shown to result from an operant-learning process involving contingencies of reinforcement provided by parents. These results are at odds with traditional explanations found in the literature on infant learning. Invited symposium at the 26th annual convention of the Association for Behavior Analysis, May, 2000 Experiments by Peláez et al. (1995, 1996) demonstrated that contingent maternal vocal imitation produces the highest rates of infant vocalization compared to noncontingent vocal stimulation and to contingent maternal speech. The present experiment expands on these findings by using an intervention that combines maternal speech as an antecedent stimulus and maternal vocal imitation as a reinforcer. Nine 4- to 10-month-old infants participated in daily experimental sessions for approximately one week. Each session lasted approximately 30 minutes and consisted of 10 to 20 one-minute trials. A single subject A1B1A2B2 reversal design was implemented across four phases: Baseline (A1), speech/imitation treatment (B1), reversal (A2), and speech/imitation treatment (B2). Seven of 9 subjects displayed higher rates of vocalization in the speech/imitation treatment phases compared to the baseline and reversal phases. Poster presented at the 26th annual convention of the Association for Behavior Analysis, May, 2000 Synchronized Reinforcement Procedure In Social Interactions Paper presented at ABA 2000 The synchronized reinforcement procedure,
developed by Pelaez et al. (1996, 1997), is a contingency-based
operant procedure that specifically allows the experimenter to
compare the reinforcing efficacy of different types of adult
stimulation in maintaining infant social behavior. It permits
a systematic comparison of the effects of different compound
stimuli (e.g., auditory, visual, tactile) provided by adults
during face-to-face interactions. In the first experiment, infant
preference for social stimulation that included touch during
a face-to-face situation with an adult was investigated in ten
1.5- to 3.5-month-old infants (M = 2.6 months). Two treatment
conditions were compared in an alternated, counterbalanced order
with each infant. A social stimulus compound that included touching
the infants functioned as a more effective reinforcer for infant
eye-contact behavior than a stimulus compound that did not include
touch. In a second experiment with twelve 2- to 4.5-month-old
infants, contingent rhythmic stroking was a more effective reinforcer
than tickling and poking when presented contingent on eye contact
with an adult female. Compared to tickling and poking, stroking
increased eye contact with the experimenter, and subjects smiled
and vocalized more and frowned and cried less. A systmatic replication of the Devany,
Hayes and Nelson (1986) and the Augustson and Dougher (1992)
studies was conducted with infants to explore the relation between
stimulus equivalence and language development. Nine typically-developing
infants (21 to 25 months) participated in 5 to 15 experimental
sessions, each session on a different day. Expressive language
skills were assessed (REEL Scale 2, 1991) before exposure to
a visual-visual conditional discrimination with animal-like figures.
In a single-subject design, infants were taught four conditional
discriminations: if A, then B; if A, then C; if D, then E; and
if D, then F. The order of presentation and the left-right position
of correct response were counterbalanced across training and
testing trials. Once an infant learned these mixed relations
under various reinforcers, tests for emergent stimulus relations
were given. Equivalence was established when a child matched
B and C, and E and F. Every subject attained criterion on the
four independent conditional discriminations and on the mixed
training. Five subjects who attained transitivity (at 80% or
above) performed below chance level on at least one of the four
symmetry tests. Eight of 10 subjects performed between 80 and
100% correct on the transitivity tests. We found a significant
negative correlation between the total number of trials to criterion
during the conditional discrimination training and the combined
receptive and expressive language quotient. Those infants with
higher language-skill scores required fewer trials to complete
the conditional-discrimination training. The results suggest
that language skills play a role in stimulus equivalence. In persons with mental retardation, relatively little is known about drug effects on behaviors other than the undesirable behaviors targeted for reduction. This information is important for clinical decision making and for understanding mechanisms of drug action in reducing aberrant behaviors. Research in this area is hampered by difficulties developing observation systems that are sensitive to drug effects in the MR population and that produce data that are comparable across a range of functional abilities. We use computerized laboratory tasks involving arbitrary stimuli, responses, and relations, to assess basic learning, memory, and motivational processes during controlled and clinical drug trials. The tasks are specifically designed so that difficulty can be adjusted for individual participant's intellectual level. This allows drug assessment on functionally equivalent baselines across subjects. The laboratory procedures also have the benefit of generating data comparable to the much larger drug research literature with animals. Paper presented at the 107th Annual Convention
of the American Psychological Association, Boston MA, August,
1999 Neuroleptics' Effects on Aberrant Behavior Robert T. Peyton, Dean C. Williams, Jennifer R. Zarcone, and Richard E. Shores This is the first of three posters that present preliminary data from a double-blind controlled clinical trial of risperidone with persons with mental retardation. This project is unique because it assesses several behavioral domains simultaneously in individual subjects. Here we present the effects on the undesired behaviors targeted for reduction. Two methods were used, an indirect rating of problem-severity using the Aberrant Behavior Checklist (ABC), and direct observation using the SCOPE observation system. Three school-age persons with severe to moderate mental retardation participated. ABC forms were completed weekly by primary caregivers who were instructed to evaluate behavior over the previous week. Mean ABC scores were computed for each subscale (irritability, lethargy, stereotypy, hyperactivity, and excessive speech). Direct observations of the frequency and duration of multiple categories of undesirable and desirable behaviors were recorded continuously during a structured activity. At least five, 30-minute observations were taken (one per day) in each phase. ABC problem-severity ratings generally decreased for all subjects under drug conditions. For all three subjects, inconsistencies across the direct and indirect measures were apparent. For example, Participant BA's aberrant behaviors increased during drug phases. These differences between blind-direct observations and informant ratings may be due to the influence of extraneous factors on raters' subjective judgments. Direct observation data are relatively immune to such factors. Presented at the 25th annual convention of the Association for Behavior Analysis, May, 1999 Measuring the Behavioral Effects of Psychotropic Drugs II: Atypical Neuroleptics' Effects on Social Interactions A. Weithof, Dean C. Williams, and Richard E. Shores Poster 1 illustrated systems for measuring the effects of psychotropic drugs (e.g., risperidone) on targeted behaviors in three individuals with mental retardation participating in a controlled clinical trial. When reductions in targeted behavior are shown, it is important to assess behavioral selectivity through concomitant measurement of appropriate behavior classes such as social behaviors. Data are from the three participants in Poster 1 during the 30-minute observation sessions using the SCOPE observation system. This program allows an in-depth analysis of social interactions for individual participants. These interactions are quantified as the percentage of observation time spent in interactions initiated by the participants and others (i.e., caregivers), as well as whether the interactions were positive, neutral, or negative. The conditional probabilities of interactions and the responses to those initiations, as well as compliance to staff demands, were also calculated. For MR and KN, overall interaction time did not change across conditions but these interactions became more positive, varied, and complex under drug than placebo. Participant BA's self-initiated interactions decreased during both drug conditions. This decrease may be related to reductions in vocalizations and self-talk that initiated the majority of his interactions. No noticeable changes in compliance to staff mands were noted for the three participants.
Presented at the 25th annual convention of the Association for Behavior Analysis, May, 1999 Measuring the Behavioral Effects of Psychotropic Drugs III: Atypical Neuroleptics' Effects on Learning and Memory Susan L. Jack, and Richard E. Shores Cognitive behavioral side effects of pharmacotherapy are seldom studied in persons with mental retardation. We describe a repeated acquisition of simple discrimination task (RA) and a delayed matching-to-sample task (DMTS), and demonstrate their sensitivity to risperidone in individual participants. Participants were those described in Posters 1 and 2. Stimuli were presented on a touch-sensitive computer monitor. In the RA task, pairs of abstract shapes were presented on the screen. New stimuli were presented each session, and the participant learned which shape was correct for that session. Difficulty was manipulated by adjusting the number of discriminations presented. On the DMTS task, retention intervals of 0, 2, 4, 8, and 16 s, were randomly programmed. Sessions were either a trial-unique procedure or a conditional procedure. KN showed dose-related increases in correct response rates on the RA task. MR showed no change in rates but decreased accuracy under drug. BA showed no change in RA rate and accuracy but dose-dependent decreases in DMTS accuracy under risperidone. In summary, BA showed equivocal effects on undesirable behaviors and negative effects on social behavior and memory performance. KN showed decreased undesirable behaviors and positive effects on social behaviors and learning. MR showed decreased undesirable behaviors, positive effects on social behaviors, but negative effects on learning. Presented at the 25th annual convention of the Association for Behavior Analysis, May, 1999 Interpreting Stimulus Generalization of Reinforcement and Punishment There are two major interpretations of stimulus-generalization data: the descriptive stimulus-control interpretation and the quantal interpretation. The descriptive approach maintains that there is a continuous relation between stimulus and response dimensions. According to this approach, on tests of stimulus generalization, training stimuli occasion trained response rates, and intermediate stimuli occasion intermediate response rates. The quantal approach, however, holds that stimulus-response units established in training occur in different proportions on generalization tests, and that intermediate levels of responding that occur in the presence of intermediate stimuli are an artifact of averaging across multiple controlling stimulus-response relations (i.e., varying proportions of trained responses). According to a review by Bickel and Etzel (1985), the quantal interpretation is a more accurate characterization of stimulus generalization based on reinforcement. The present analysis considers stimulus generalization of punishment from both interpretations by comparing traditional gradients of response rate with alternative measures of responding (e.g., interresponse times). Stimulus-generalization data are from a recent study with humans. The present analysis suggests that human punishment generalization data are consistent with both the descriptive and the quantal approaches. Association for Behavior Analysis, Chicago, IL, May, 1999 Recombinative Generalization within Consonant-Vowel-Consonant Syllables We show within-syllable recombinative generalization in a word selection task and in a reading task with nonreading adults with mild mental retardation. In each train-test cycle, participants were taught to select each of four printed words upon hearing the words (e.g., sat, mat, sop, sug). Then, tests determined whether the participant selected novel recombinations of the trained words (e.g., mop, mug) and whether the participant read aloud the words. There are two participants. One participant has completed seven train-test cycles. Beginning with the second cycle, she selected novel recombinations with near perfect accuracy. Reading tests, however, were not positive. Additional testing with showed that, (1) when presented a single word, she could select one of two auditory choices, and (2) when presented a single word, she could indicate whether a single auditory choice was correct or incorrect. In addition, gains in word reading were shown after the subject performed a printed-word-to-photo matching task. The second participant completed four train-test cycles. Beginning with the second train-test cycle, she also showed high accuracy in the word selection task, but little gain in word reading. Association for Behavior Analysis, Chicago, IL, May, 1999 Interference Effects on Discrimination Performance In conditional discrimination such as a typical matching-to-sample procedure, the contingencies for emitting a given discriminated response change (i.e., reverse) frequently--signaled by the conditional stimulus. To acquire or perform conditional discriminations, behavior must be flexible and sensitive to these rapidly changing contingencies. Behavior shows inflexibility when previously acquired stimulus control persists in the face of contingency change. We have been studying this proactive interference effect in delayed identity matching to sample. We do this by comparing performance across sessions in which (a) a limited number of stimuli are repeated within a session, and thus comparison stimuli change functions across trials, and (b) each trial presents new stimuli (a "trial-unique procedure"). Across a number of procedural variants, and in participants with mental retardation ranging from mild to severe, greater accuracy is reliably obtained under the trial-unique procedure than under procedures using small stimulus sets. Analyses of error patterns show the contribution of inter-trial interference when small stimulus sets are presented. Specifically, behavior becomes less flexible in that the stimulus that was correct on the previous trial tends to be selected on the current trial. This interference effect or inflexibility increases with the length of the retention interval and severity of retardation. Presented as part of a symposium entitled "Behavioral Flexibility and Changing Contingencies," Gatlinburg Conference on Research and Theory in Mental Retardation and Developmental Disabilities, Charleston, SC, March, 1999. This presentation reported research examining sensitivity to differences in reinforcement contingencies in individuals with mental retardation. A concurrent operants procedure was implemented in the context of a computer game. The game provided participants with two concurrently available response options, selections of two different types of animated figures displayed in the left and right halves of a touchscreen monitor. Over a series of conditions, the relative rates or magnitudes of reinforcement were changed in ratios of 5:1, 3:1, 1:1, 1:3, and 1:5. Sensitivity was quantified by application of the generalized matching equation, in which the slope of a regression line fitted to the data indicates sensitivity. Results with three nonclinical young adults showed high sensitivity to differences in reinforcer rates and rapid adjustments in behavioral allocation when rates changed. Results with six individuals with mental retardation showed pronounced individual differences in sensitivity, greater variability, slower accommodation to changes, and idiosyncratic differences in sensitivity to rate and magnitude variation. The results have implications for assessing potential reinforcers in applied settings. A comprehensive reinforcer assessment may need to examine not only response to reinforcer quality, but also to delivery parameters. The research also has implications for the design of programs to teach more effective choice-making to individuals with mental retardation. Presented as part of a symposium entitled "Behavioral Flexibility and Changing Contingencies," Gatlinburg Conference on Research and Theory in Mental Retardation and Developmental Disabilities, Charleston, SC, March, 1999. Developmental Level and Equivalence Class Flexibility in Children Traditionally, concept or category organization has been viewed as the outcome of two conflicting tendencies - one for new experiences to modify categories, and the other to resist changes that require effort and result in loss of continuity. The present study investigated the stability and flexibility of experimental equivalence classes in the face of class challenges in typically developing children across a range of ages. Three-choice arbitrary match-to-sample tasks were presented by a computer. Children first mastered two conditional discriminations (A1B1, A2B2, A3B3, A1C1, A2C2, A3C3) and then demonstrated the emergence of three three-member equivalence classes (A1B1C1, A2B2C2, A3B3C3) on unreinforced tests for reflexivity, symmetry, and transitivity/equivalence. When all test performances were stable, reinforcement contingencies were modified for the AC conditional discrimination (A1C2, A2C3, A3C1). After mastery of the altered baseline, the equivalence tests were presented again: Would performances remain consistent with the original classes or indicate reorganized classes (A1B1C2, A2B2C3, A3B3C1)? In general, younger children (2.5-8 yrs) showed erratic choice patterns on test trials following the class challenge, while children age 8 and older showed systematic patterns that were consistent with either the modified or the original contingencies. These results indicate that the organization of arbitrary classes may be easier to disrupt in young children, and they call for further study of variables that may influence class stability. Presented as part of a symposium entitled "Behavioral Flexibility and Changing Contingencies," Gatlinburg Conference on Research and Theory in Mental Retardation and Developmental Disabilities, Charleston, SC, March, 1999. The current study evaluated the effect of a self-management treatment package (Phase 1) and the effect of individual components of the treatment package (Phase 2) on the stereotypic behavior of a 13-year-old male with pervasive developmental disorder. In Phase 1, a multiple-baseline design was implemented across academic, lunch, and break settings. Additionally, an ABAB design was implemented in the academic setting. The self-management treatment package consisted of two concurrent elements: self-recording (marking with a pencil on a prepared data sheet with blocks corresponding to 15-second intervals signaled by a timer), and differential reinforcement of other behavior (DRO) by the delivery of a preferred stimulus if no stereotypy occurred during a prescribed amount of time. In baseline, stereotypy occurred in 44.1-58.8% of intervals in each setting. With the treatment package, stereotypy was decreased to near zero (0-0.9%). In Phase 2, a brief component analysis was conducted in one setting to determine which element of the treatment package was responsible for the behavior change. Both DRO and self-recording conditions alone maintained near-zero occurrences of stereotypy. Association for Behavior Analysis, Orlando, FL, May, 1998. |
