can it be done by tonight 11 pm
Iwantalldegrees
1555353898_1.pdf
EDUCATION AND TREATMENT OF CHILDREN Vol. 37, No. 3, 2014
Pages 407–430
An Evaluation of a Treatment Package Consisting of Discrimination Training and
Differential Reinforcement with Response Cost and a Social Story on Vocal Stereotypy for a
Preschooler with Autism in a Preschool Classroom
Amanda P. Laprime and Gretchen A. Dittrich Simmons College
Abstract
The purpose of the study was to evaluate the use of a treatment package com- prised of a social story, discrimination training, and differential reinforcement with response cost on the vocal stereotypy of one preschooler diagnosed with an autism spectrum disorder. The study took place in a preschool classroom of a public school and was implemented by the classroom teacher and staff. A changing criterion design was employed to evaluate experimental control. The results of this study demonstrated a clear decrease in vocal stereotypy as compared to baseline.
Keywords: Vocal Stereotypy, Response Cost, Differential Reinforcement, Changing-Criterion Design
A defining characteristic of autism spectrum disorders (ASDs) involves repetitive vocalizations or body movements, otherwise de- fined as stereotyped behavior (American Psychiatric Association, 2000). Rapp and Vollmer (2005) conducted a review of the literature and concluded that the majority of stereotypic behaviors are main- tained by automatic reinforcement. Skinner defined verbal behav- iors as “behavior reinforced through the mediation of other persons” (1957, p.2). Though repetitive vocal responses may topographically appear to be verbal, if the consequences maintaining such responses are not mediated by a listener, they then could instead be defined as vocal stereotypy maintained by automatic reinforcement. Vaughn and Michael (1982) defined automatic reinforcement as a contingency not mediated by a listener. Therefore, vocal stereotypy may be maintained
This study was conducted by the first author in partial fulfillment of the requirements for the Ph.D. degree in behavior analysis at Simmons College. We thank the staff at the LABBB Collaborative for their dedication and hard work with their students and this study.
Address correspondence to Amanda Laprime, Department of Behavior Analysis, Sim- mons College, 300 Fenway, Boston, MA 02115; e-mail: [email protected]
408 LAPRIME and DITTRICH
by the sensory byproduct it produces (Ahearn, Clark, MacDonald, & Chung, 2007). Because the reinforcing consequence cannot always be separated from the response, stereotypic behaviors may be difficult to extinguish. Vocal stereotypy is especially problematic in that the audi- tory product of the response is often challenging to block or place on sensory extinction. In addition, stereotypy may often persist when the individual is alone or void of social contingencies (Falcomata, Roane, Hovanetz, Kettering, & Keeney, 2004; Iwata, Dorsey, Silfer, Bauman, & Richman, 1982,1994; Rincover, 1978).
Based on the difficulty of separating the auditory byproduct of vocal stereotypy and the response itself (i.e., the movement of the vo- cal musculature), behavior change of vocal stereotypy achieved by sensory extinction may be attributed to punishment contingencies and not extinction (Ahearn et al., 2007). Ahearn and colleagues evaluated the efficacy of a punishment contingency on reducing vocal stereoty- py with response interruption and redirection. The participants were given vocal demands contingent on any instance of vocal stereotypy, which effectively reduced responding for all four participants (Ahern et al., 2007). Rapp and colleagues (Rapp, Patel, Ghezzi, O’Flaherty, & Titterington, 2009) extended the literature on mild punishment proce- dures by assessing whether vocal stereotypy could be brought under stimulus control of arbitrary stimuli. Their results demonstrated that vocal stereotypy was suppressed in the presence of a red card after the implementation of the response cost contingency when the card was present. The authors concluded that the negative punishment procedure was effective at establishing the red card as a discrimina- tive stimulus related to punishment (SDP) by being paired with both positive and negative punishment (Rapp et al., 2009).
Researchers have also examined the effects of negative punish- ment on vocal stereotypy; specifically, response cost was found to be effective at reducing vocal stereotypy in individuals with ASDs. Fal- comata and colleagues (2004) compared noncontingent reinforcement (NCR) and NCR with response cost to decrease vocalizations main- tained by automatic reinforcement. Initially, levels of vocalizations decreased moderately during the NCR conditions, but with the addi- tion of a response cost in which a Walkman© was removed contingent on vocal stereotypy, responding decreased to near zero levels for the participant (Falcomata et al., 2004).
Previous studies have also employed response cost procedures with token boards (Kostinas, Scandley, & Luiselli, 2001; Truchlicka, McLaughlin, & Swain, 1998). During this procedure, tokens, which function as conditioned reinforcers, are removed contingent upon the occurrence of a target response. At the end of a predetermined
409VOCAL STEREOTYPY
amount of time, remaining tokens can be exchanged for a backup re- inforcer (Cooper, Heron, & Heward, 2007). For example, Kostinas and colleagues (2001) employed a response cost procedure with a token system in which the participant was given 15 tokens at the start of each day. Contingent upon the occurrence of a perseverative verbal- ization, a token was removed. By the end of the day, if at least one token remained on the board, a preferred snack was delivered. Re- sults demonstrated a decrease in perseverative vocalizations for the one participant (Kostinaset al., 2001). In a similar study, Truchlicka and colleagues (1998) decreased errors during spelling exams and increased accuracy in spelling performance in three participants fol- lowing implementation of a token-based response cost system. Fur- thermore, the authors concluded that the token economy with the response cost component was en effective intervention for use in edu- cational settings, such as a middle school.
Social stories have also been used in conjunction with response cost and token board systems (Kuttler, Myles, & Carlson, 1998). So- cial stories describe a situation and the contingencies related to de- sired responses for that situation. The contingencies are described in a manner by which they are easily understood by the person for whom the social story was written (Grate & Garand, 1993). Social sto- ries were demonstrated to increase the effects of contingencies related to token economies (Kuttler et al., 1998). Additionally, social stories are often recommended not as a behavior change agent, but instead to help facilitate a better social understanding of the contingencies related to improvements in behavior and functioning in social situ- ations (Kokina & Kern, 2010). The American Psychiatric Association (2000) has suggested that social stories may address the need for predictability for children with autism in social situations and that visually cued instruction that may be more discernable than verbal instructions (Quill, 1997). This may be additionally important in situ- ations such as public school settings, where students may not have constant support and are required to share their teachers. In a meta- analysis conducted by Kokina and Kern (2010) on the use of social stories, the authors found that 21% of the participants in studies that employed social stories included preschoolers, while 60% of the par- ticipants were elementary aged. The majority of the studies surveyed utilized social stories which focused on a reduction of inappropriate behaviors and an increase in social skills (Kokina & Kern, 2010). The research on social stories have described their effects as potentially ef- fective, at best (e.g., Ali & Frederickson, 2006; Rust & Smith, 2006), yet they are often viewed positively by teachers and parents as a feasible and effective intervention (Smith, 2001). However, meta-analytic data
410 LAPRIME and DITTRICH
suggest that they social stories are often utilized in conjunction with behavior change procedures leading to difficult with determining the isolated effects of social stories on behavior alone (Kokina & Kern, 2010). While response cost systems with token boards and social sto- ries have been demonstrated to be effective at decreasing behaviors, such as precursors to tantrum behavior, perseverative vocalizations, and calling out in the classroom (Kostinas et al., 2001; Kuttler et al., 1998), to date, the efficacy of a treatment package utilizing these meth- ods has not been evaluated with vocal stereotypy in a public school classroom.
When evaluating methods to decrease behavior, reinforcement- based procedures, either alone or in conjunction with punishment procedures, may be considered. For example, differential reinforce- ment interventions have been empirically demonstrated to mitigate the negative side effects of punishment or extinction procedures (Catania, 1997). However, the combination of punishment procedures with differential reinforcement has been demonstrated to be more ef- fective at reducing problematic behavior than differential reinforce- ment alone (Buckley & Newchok, 2005; Kostinas et al., 2001). Differ- ential reinforcement of low rates of behavior (DRL), first described by Fester and Skinner (1957), effectively reduced disruptive behaviors in individuals with disabilities, including but not limited to requests for teacher attention (Austin & Bevan, 2011), stereotypy (Singh, Dawson, & Manning, 1982), and physical and verbal aggression (Alderman & Knight, 1997). A DRL procedure involves the delivery of positive reinforcement contingent uponthe frequency of a target behavior oc- curring less than or equal to a predetermined limit during a specified period of time (Deitz, 1977). A DRL procedure is often used when be- haviors occur too often to be susceptible to other reinforcement con- tingencies (e.g., differential reinforcement of other behavior; Martin & Pear, 2010).
Kostinas and colleagues (2001) compared the effects of a DRL procedure to a DRL with a response cost procedure on perseverative verbal behavior in an adult diagnosed with obsessive-compulsive dis- order. Results demonstrated that the DRL with response cost proce- dure produced greater reductions in perseverative speech than the DRL procedure alone (Kostinas et al., 2001). The authors concluded that DRL alone did not effectively reduce perseverative verbaliza- tions, and that the absence of inappropriate responses may be the re- sult of the participant avoiding negative consequences (i.e., the loss of a token), rather than accessing positive reinforcement (Kostinas et al., 2001). Though the authors did not discuss the function of the problem behavior, in the case of behaviors maintained by automatic
411VOCAL STEREOTYPY
reinforcement, it may be difficult to either substitute a functionally equivalent reinforcer or find a reinforcer that will compete with the byproduct of stereotypy when using differential reinforcement alone (Falcomata et al., 2004).
The purpose of the current study was to extend the literature on establishing SDP control of vocal stereotypy using arbitrary stimuli and a DRL with a token-based response cost procedure. A social story was used to describe the contingencies in place during the experimental sessions. All experimental sessions were conducted within a public school environment, with public school staff implementing the pro- cedures.
Method
Participant and Setting
David was a 4-year-old male who had been diagnosed with an autism spectrum disorder. He had been diagnosed at the age of 2-years-old. One year prior to the study, the Early Intervention De- velopmental Profile (Rogers et al., 1981) had been administrated to David. His social emotional skills were scored at the age equivalent of 33 months, self-care 25 months, cognition was scored at 30 months, perceptual fine motor was 24 months, gross motor was 32 months, receptive language was 24–31 months, and expressive language was scored at 26 months. The most recent speech and language testing was conducted during the study. Results from the Clinical Evaluation of Language Fundamentals® preschool second edition (Semel, Wiig, & Secord, 2008) suggested that David had a receptive language index of 61 (which was 2 standard deviations below the mean), and the results of the Peabody Picture Vocabulary Test fourth edition (Lloyd & Dunn, 2007) yielded a score of 74 (which was also 2 standard deviations from the mean).
He had been nominated for the study by his classroom teacher and mother due to the high rate of vocal stereotypy at both school and home. David attended an integrated preschool classroom in a pub- lic elementary school. Four to six other students with developmental disabilities attended his classroom five days per week. During two of the five days the classroom was integrated with neurotypical pre- schoolers. The classroom included one special education teacher and three classroom assistants. The special education teacher had taught in a preschool classroom for children with autism for 11 years and had a master’s of science in intensive special needs at the time of the study. She had taken post-masters classes in Applied Behavior Anal- ysis (ABA) after completing her degree. The other three classroom
412 LAPRIME and DITTRICH
assistants had a range of education from a high school diploma to an associates and bachelor’s degree. They had been working for 6 to 14 years with children with developmental disabilities. None of the classroom assistants had any classes or specific training in ABA other than workshops intermittently provided by their employer.
All sessions were conducted in the preschool classroom through- out the school day. The classroom included three individual student work areas divided by partitions, a circle area comprised of a rug and individual seats, and three tables, each identified by their shape and color. Sessions were conducted by the classroom assistant or teacher who was working with David for that period of time, and the class- room assistant or teacher also worked with one or two other students while running the experimental sessions. The study was conducted from September through February of the academic school year, ex- cluding programmed school breaks and weekends.
Response Measurement and Data Collection
Vocal stereotypy was defined as any instance in which David engaged in non-contextual vocal word or sound output for three con- secutive seconds. An example included, David verbally repeating the plot of a television show during lunch while looking at his food. A non-example included if David responded to a question, attempted to ask a question, or engaged in contextual or non-contextual vocal output when making eye contact with an adult.
Partial interval recording (Johnston & Pennypacker, 1980) data were collected during two to four experimental sessions per day. Per- centage of intervals in which vocal stereotypy occurred was calcu- lated by dividing the number of intervals in which vocal stereotypy occurred by the total number of intervals and multiplying by 100 per- cent. Intervals were between 1 to 2 min in length for a total of 5 to 10 min depending on the experimental condition. The length of the intervals was determined at the onset of the experiment after the re- sults of the descriptive assessment demonstrated a high frequency of vocal stereotypy with little to no breaks in between episodes. While average frequency was not formally recorded, a 1 min session was determined to be the initial interval to evaluate if such a measure was representative of the overall occurrence of vocal stereotypy and was determined to be feasible for the staff in terms of data collection. The total session time was determined based on two criteria, first, sessions would start at small enough intervals so David would access reinforcement frequently, second, the total session duration was long enough to be effective while being manageable for implementation by classroom staff.
413VOCAL STEREOTYPY
Experimental Design
A changing conditions design (Kazdin, 1982) with an embed- ded changing criterion design (Weis & Hall, 1971) was employed to evaluate the effects of discrimination training and differential rein- forcement of low rates (DRL) of behavior with response cost on vocal stereotypy. Criterion was initially set at 80% based on baseline lev- els of responding. The criterion was increased by 20% systematically when David engaged at or below criterion specific levels of respond- ing for two consecutive sessions. In order to demonstrate a functional relationship the criterion was increased to a previously attained level and then systematically decreased again.
Descriptive Assessment
Antecedent, behavior, consequence (Bijou & Baer, 1963) data were recorded by the classroom behavioral consultant prior to the beginning of the study during targeted weekly observations. Each ob- servation occurred for approximately 30 min. The classroom staff col- lected daily data on (a) percent of intervals in which vocal stereotypy occurred for up to five 5 min sessions, (b) the activity (Figure 1), and (c) the day of the week (Figure 2). An analysis of these variables was
Figure 1. Average percentage of 1 min intervals during a 5 min session where vocal stereotypy occurred across classroom activities during descriptive as- sessment.
0
10
20
30
40
50
60
70
80
90
100
Group PT Snack Choice Story Music Circle Lunch Cooking 1:1 Work
OT Group
Art
Activity
Pe rc
en ta
g e
of I
n te
rv al
s in
w h ic
h V
oc al
S te
re ot
yp y
O cc
u re
d
414 LAPRIME and DITTRICH
conducted in order to develop a hypothesis regarding environmental events that potentially maintained vocal stereotypy. The results of the assessment suggested that vocal stereotypy was maintained by posi- tive automatic reinforcement. This conclusion was made for the fol- lowing reasons: (a) teachers reported that no consequence appeared to result in an increase or decrease in the occurrence of the behavior, and that when left alone, David was likely to engage in continuous vocal stereotypy; (b) responding was variable with no clear trend throughout the assessment (see Figure 3), and no clear patterns in re- sponding among days, activities, or time; and (c) responding occurred at lower levels during activities in which there were other auditory stimuli present (see Figure 1).
At the onset of the study, the behavioral consultant met with the guardian and the teacher to determine goals for the intervention. Goals identified were as follows: 1) the intervention would begin with the criterion set so that David could contact reinforcement (i.e., 5 min sessions comprised of 1 min intervals), and 2) the session length would be systematically increased to 10 min total duration(comprised of five 2-min intervals). The terminal duration was set to approximate the duration of work sessions in kindergarten, to which David was transitioning the following year.
Figure 2. Average percentage of 1 min intervals across 5 min sessions where vocal stereotypy occurred across days of the week.
0
10
20
30
40
50
60
70
80
90
100
Monday Tuesday Wednesday Thursday Friday
Pe rc
en ta
g e
of I
n te
rv al
s in
w h ic
h V
oc al
S te
re ot
yp y
O cc
u re
d
Day of the Week
415VOCAL STEREOTYPY
Staff Training
The classroom teacher and assistants were trained to implement all experimental conditions by the classroom behavioral consultant who was also the primary experimenter of the study. During the first phase of training, the consultant reviewed the written behavior plan of implementation procedures with the teacher and assistants. In the second phase, the behavioral consultant modeled the implementation of the experimental conditions for the staff in vivo. In the third phase, each staff was individually observed implementing the experimental conditions and then given immediate feedback. The classroom teach- er and staff were required to go through the training sequence prior to running the experimental procedures.
Figure 3. Percentage of intervals in which vocal stereotypy occurred during baseline (BL) and treatment (DRL/RC) across 1 min intervals during 5 min ses- sions, 1.5 min intervals during 7 min sessions, and 2 min intervals during 10 minute sessions. Arrows indicate criterion (C=criterion) levels on the graph. C1 represents 80% of intervals, C2 represents 60% of intervals, C3 represents 40% of intervals, C4 represents 20% of intervals, C5 represents 40% of inter- vals, C6 represents 20% of intervals, C7 represents 0% of intervals, and C8 represents 20% of intervals.
0
10
20
30
40
50
60
70
80
90
100
Baseline
C 6
DRL/RC
C 3
C 4
C 5
C 7
5 min 7.5 min 10 min
C 8
Pe rc
en ta
g e
of I
n te
rv al
s
Sessions
C1
C2
1 11 21 31 41 51 61 71 91 101 111 121 131 141 151 161 17181
416 LAPRIME and DITTRICH
Procedures
General Procedures. Across all experimental conditions, class- room teachers and staff were instructed to implement the treatment procedures two to four times per day. The frequency was determined based on teacher reports of how often they could implement the pro- cedures, based on staffing and the number of students in the class- room (which changed at times).
David’s school day was broken up into activities in which there were no specified programmed consequences for vocal stereotypy and activities in which programmed consequences were imple- mented following the occurrence of vocal stereotypy. The teacher and classroom assistants met with the behavioral consultant to de- termine these activities. Playground, choice time, and free play were isolated as times when David was allowed to engage in vocal stereo- typy without the occurrence of a programmed consequence. Circle, small group activities, lunch, snack, and individual work time were determined to be times when vocal stereotypy competed with other desired responses, and it was therefore targeted for decrease utilizing the treatment package.
It was determined that changes in criteria would be communi- cated to the classroom staff by the behavioral consultant after an anal- ysis of the current data. The behavioral consultant was scheduled to be in the classroom one to two times per week. For this reason, David may have met criteria and then remained in that phase of the study until the next time the behavioral consultant was in the classroom for consultation. Therefore, the number of days in a particular phase of the study was variable.
Baseline. The baseline condition occurred throughout the entire school day. During baseline, staff redirected David to have a quiet voice (via verbal instruction paired with a finger put to the lip) con- tingent on any instance of vocal stereotypy. The social story, token board, and bracelets were not utilized during baseline.
Discrimination training. The no programmed consequence condi- tion was labeled the green condition, and a green circle was placed on the classroom schedule for those activities in which there would be no programmed consequence for vocal stereotypy. The DRL/RC condition was identified as the red condition, and the activities that resulted in the programmed consequence for vocal stereotypy were labeled with a red circle on the classroom schedule. The red and green colors were to serve as conditional stimuli associated with differen- tial consequences. In addition, the labels on the schedules served as prompts for the classroom staff regarding which experimental con- dition to run. It was determined that during some designated red
417VOCAL STEREOTYPY
activities it was not feasible for the red condition to be implemented due to insufficient staffing; therefore, during these times the green condition was implemented. It is important to note that the red and green designations were a guide for classroom staff, but the red con- dition did not occur for every occurrence of a red activity, due to staffing issues.
At the start of any scheduled activity in the classroom, the staff instructed David on the condition specific contingency and provided him with the corresponding social story (Appendix B). At any time during the day in which the classroom staff did not run the red condi- tion, they implemented the green condition.
The DRL/RC condition was comprised of three phases. The pur- pose of the phases was to systematically increase the amount of time in which the treatment package was implemented while systematical- ly decreasing the amount of intervals in which David could engage in vocal stereotypy. The first phase was implemented to establish stimu- lus control of vocal stereotypy across 5 min sessions. The purpose of the second and third phases was to systematically increase the dura- tion of time in which he was not engaging in vocal stereotypy. Within the first and second phase, the number of intervals in which David could engage in vocal stereotypy and receive contingent reinforce- ment was systematically decreased (i.e., the DRL). The purpose of the third phase was to evaluate if low levels of vocal stereotypy could be maintained across a longer session duration.
DRL and response cost (DRL/RC). At the start of the condition, the social story (Appendix B) was given to David and he was instructed to read it aloud. Simultaneously, a red wristband or watch was put on David’s wrist to signal the onset of the red condition. Staff presented a token board filled with five tokens and a prize box filled with pre- viously identified preferred items. The timer was set for the phase specific interval and started. If at any point during the interval David engaged in vocal stereotypy, the staff showed him the token board and removed a token. They then prompted a quiet voice by holding a finger up in front of their lips and counting to three, and repeated this procedure until David remained quiet for three consecutive seconds. Instructions were reintroduced regarding the contingency in place. If David engaged in vocal stereotypy at any other time during the same interval, the staff redirected him one time by holding a finger up in front of their lips.
At the end of the condition specific time, if David had kept the criterion specific number of tokens, the green condition was imple- mented and he was given access to the prize box and all tangibles in the box for a predetermined reinforcement time. If David did not
418 LAPRIME and DITTRICH
have the specified number of tokens remaining on the token board, the contingency for reinforcement was described. If the program was going to be run immediately again, the red condition continued; if the program could not be run again, the green condition was imple- mented.
Phase 1. During phase 1 of the DRL/RC, all sessions were com- prised of five 1 min intervals for a total of a 5 min session. Contingent on meeting the criteria, reinforcement time was 1 min. The phase began with a predetermined criterion of 80% or less of intervals with vocal stereotypy (C1). Once this criterion was met, the criterion decreased to 60% or less of intervals with vocal stereotypy (C2). Upon meeting this criterion, the criterion decreased to 40% or less of intervals with vocal stereotypy (C3). The next two criteria changes involved a decrease to 20% or less of intervals with vocal stereotypy (C4) and a reversal back to 40% or less of intervals with vocal stereotypy (C5).
Phase 2. During phase 2 of the DRL/RC, all sessions were com- prised of five 1.5 min intervals for a total of a 7.5 min session. Con- tingent on meeting the criteria, reinforcement time was 1–2 min. The phase began with a predetermined criterion of 20% or less of intervals with vocal stereotypy, and once this was met, the criterion was de- creased to 0% or less of intervals with vocal stereotypy.
Figure 4. Average percentage of intervals where vocal stereotypy occurred across experimental conditions.
0
10
20
30
40
50
60
70
80
90
100
Baseline Criterion 1
Phase 1: Criterion 2
Phase 1: Criterion 3
Phase 1: Criterion 4
Phase 2: Criterion 5
Phase 2: Criterion 6
Phase 2: Criterion 7
Phase 3: Criterion 8
Pe rc
en ta
g e
of I
n te
rv al
s in
w h ic
h V
oc al
S te
re ot
yp y
O cc
u re
d
Experimental Condition
Phase 1:
419VOCAL STEREOTYPY
Phase 3. During phase 3 of the DRL/RC, all sessions were com- prised of five 2 min intervals for a total of a 10 min session. Contingent on meeting the criteria, reinforcement time was 2–3 min. The criterion for this phase was 20% or less of intervals with vocal stereotypy.
No programmed consequence. At the start of the no programmed consequence condition, the social story (Appendix B) was given to David and he was instructed to read it aloud. Simultaneously, a green wristband or watch was put on David’s wrist to signal the onset of the green condition, and David was instructed of the condition spe- cific contingency. There was no programmed consequence for any in- stance of vocal stereotypy.
Interobserver Agreement
Interobserver agreement (IOA) was collected for 25% of sessions. The behavioral consultant simultaneously recorded occurrences of vocal stereotypy independent of the classroom staff. Data were com- pared to identify the number of agreements and disagreements IOA was calculated by dividing the number of agreements by the sum of agreements and disagreements and multiplying by 100 to determine a percentage. Agreement was 97% (range of 60% to 100%). The low data point at 60% occurred during session 16 of baseline, when the behav- ioral consultant recorded repeated occurrences of vocal stereotypy during only one interval, but the classroom staff recorded the repeat- ed occurrences of stereotypy as occurring across consecutive intervals (i.e., although the repeated stereotypy occurred during the same in- terval, the staff recorded each instance in a new interval). Feedback on data collection was given to the staff following the recording error. Recording errors were not observed during any subsequent observa- tions, nor with any other classroom staff; therefore, additional train- ing was not conducted.
Social Validity
Social validity was assessed with a 7-point Likert scale consist- ing of 10 questions and a brief open-ended comments section (Ap- pendix A). The form was adapted from the Treatment Acceptability Rating Form-Revised (TARF-R; Reimers & Wacker, 1988). David’s guardian, teacher, classroom staff, and home-based behavioral con- sultant recorded data on social validity at the end of the study.
Results
Experimental conditions. The results of the baseline condition and DRL/RC are shown in Figures 3 and 4. During baseline, David’s vo- cal stereotypy occurred at moderate levels during recorded sessions,
420 LAPRIME and DITTRICH
averaging 60% of five 1 min partial intervals. When discrimination training and DRL/RC was implemented, vocal stereotypy stabilized with little variability at 56% of five 1min partial intervals on average in criterion 1 of phase 1. Criterion increases resulted in decreases in vocal stereotypy to 37% of five 1 min partial intervals on average in criterion 2 of phase 1, 26% of five 1 min partial intervals on average in criterion 3 of phase 1, and 15% of five 1 min partial intervals on aver- age in criterion 4 of phase 1. When the criterion was decreased vocal stereotypy increased to 28% of five 1 min partial intervals on aver- age in criterion 5 of phase 1.A return to the previous criterion and an increase in the duration of the session resulted in a decrease in vocal stereotypy to 12% of five 1.5 min partial intervals on average in crite- rion 6 of phase 2. The last criterion increase resulted in an additional decrease in vocal stereotypy to 10% of five 1.5 min partial intervals on average in criterion 7 of phase 2. A return to the previous criterion and an increase in the duration of the session resulted in a slight decrease in vocal stereotypy to 8% of five 2 min partial intervals in criteria 8 of phase 3 (see Figure 3 and Figure 4).
Social validity. Results from the social validity assessment (Ap- pendix A) indicated that staff rated the procedures as a socially ac- ceptable way to decrease vocal stereotypy (mean of 5 out of 5 possible points), while being easy to implement (mean of 4.5 out of 5 possible points), and minimally intrusive to the participant (mean of 4.75 out of 5 possible points). The raters also indicated that the procedures resulted in an improved rate of appropriate vocalizations both dur- ing (mean of 5 out of 5 possible points) and outside of session (mean of 4.75 out of 5 possible points respectively), and did not result in adverse side effects for the participant (mean of 5 out of 5 possible points).
Discussion
In the current study, discrimination training paired with a DRL/RC contingency effectively reduced the occurrence of automat- ically reinforced vocal stereotypy in a public school preschool class- room. The results of the intervention were achieved across a time pe- riod of 5 months, demonstrating the efficacy of the intervention. The results of the current study were similar to those of previous studies (e.g., Falcomata et al., 2004; Rapp et al., 2009) and suggested that re- sponse cost and discrimination training effectively reduced vocal stereotypy. In addition, these procedures could be implemented in a setting that was not experimental, with less experimental con- trol, and still result in substantial reductions in interfering problem behaviors.
421VOCAL STEREOTYPY
Although there was some variability in the data, overall, the contingent removal of a token paired with the opportunity to earn a prize box for low rates of vocal stereotypy was effective at reducing vocal stereotypy, even when the intervention was not implemented the same number of times daily. It appeared that the presence of the red bracelet enhanced the discrimination of the contingency when the DRL/RC was in effect; therefore, the red bracelet came to exert SDP control over vocal stereotypy. Rapp and colleagues (2009) suggested that establishing SDP control might be useful in educational settings because it may result in a teacher or trainer needing to provide fewer consequences to maintain low levels of problem behavior. While the current study did not record data on vocal stereotypy in the presence of the red bracelet without the DRL/RC contingency, low levels of re- sponding were maintained across 10 min sessions (Figure 3). These data support that establishing stimulus control of automatically main- tained problem behavior in the presence of arbitrary stimuli and not teachers, may reduce the need for intensive teacher supports to main- tain low levels of responding in educational settings.
Additionally, the red bracelet may have come to exert SD control over other responses that resulted in social positive reinforcement; therefore, in the presence of the red bracelet, positive tangible rein- forcement (i.e., prize box) for behaviors other than vocal stereotypy may have competed with the hypothesized automatic reinforcement derived from the vocal stereotypy. Teacher reports indicated that to- ward the end of the study, David occasionally refused the prize box and asked to engage in the current activity and continue to wear the red bracelet (i.e., restart the response cost). One hypothesis is that by reducing the interference of vocal stereotypy, other activities in the classroom became conditioned reinforcers. Alternatively, the tokens themselves may have become conditioned reinforcers and keeping them may have been as reinforcing as earning the prize box, thus sup- porting the role of tokens as mediating reinforcement to the ultimate reinforcer.
Teacher reports also indicated that there were days when David engaged in vocal stereotypy and then asked the teachers for the red bracelet. The response product of vocal stereotypy may have become aversive due to being paired with the response cost contingency dur- ing the DRL/RC condition. Previous researchers (i.e., Rappet al., 2009; Lerman & Vorndran, 2002) demonstrated that verbal reprimands be- came effective punishers after being paired with a response cost. A similar effect may have occurred in the current study, except that the response product of the vocal stereotypy may have become an effec- tive punisher after the systematic pairing with the response cost.
422 LAPRIME and DITTRICH
The results of the current study contributed to the literature re- garding effective treatment of vocal stereotypy in several ways. First, outcomes replicated and extended the results of previous research (i.e., Falcomata et al., 2004; Rapp et al., 2009) which demonstrated the effectiveness of establishing SDP control via a negative punish- ment procedure. Second, the current data demonstrated the utility of employing differential reinforcement of low rates of behavior (DRL) paired with mild punishment procedures to decrease automatically maintained problem behavior that was not susceptible to other dif- ferential procedures due to initial high rates of response. Lastly, the current procedures were implemented throughout the day in a public school preschool classroom. The classroom was integrated two days per week and staffed by individuals with minimal experience with be- havior analysis, and no experience conducting research. The obtained data suggested that the current procedures were not only effective but could be implemented in a natural environment without exten- sive training in behavior analysis or research methods.
Anecdotal reports suggested that David’s vocal verbal behav- ior increased in the classroom. Though no formal data were collected, during observations David consistently used full sentences to tact the occurrence of classroom events, intraverbals to answer questions, and mands for attention. The DRL/RC contingency was also implemented by David’s guardians in the home and community, and they reported similar effects as the current study.
Several limitations to the current study should be noted. First, a formal functional analysis was not conducted; therefore, other vari- ables may have maintained David’s vocal stereotypy. Future studies should ensure a function-based treatment prior to employing punish- ment procedures. Second, criterion changes did not occur immedi- ately after David met criterion for the phase. The experimenter was only available to visit the classroom one to two times per week. Based on this schedule, the phase was changed when the experimenter was able to go into the classroom, score the data, and then inform the class- room staff of the next phase. This resulted in some phases where Da- vid met criterion and remained in the phase for many consecutive sessions (e.g., criterions 5 and 8), or met criterion but then remained in the phase and variable responding occurred (e.g., criterions 1, 2, 3, 4, 6, and 7). While it is unclear if the delayed criterion change influenced response variability, substantial changes were evident, demonstrating that the procedures were efficacious in a less controlled environment.
An additional limitation to the current study was that two students with developmental disabilities joined the classroom dur- ing the second phase of the study. Variability in those data may be
423VOCAL STEREOTYPY
attributed to less sessions daily as well as increases in staff to stu- dent ratios. Regardless, responding continued to remain at low lev- els, suggesting that the intervention was one that would not require constant staff attention to be successful. Another limitation was that the green condition was implemented when David earned the prize box and items that were hypothesized to evoke vocal stereotypy such as Thomas the Tank Engine© were put into the prize box and not available throughout the day. Previous studies (i.e., Charlop, Kurtz, & Casey, 1990) have evaluated the effects of using automatically rein- forced problem behavior as a reinforcer in differential reinforcement programs. The hypothesis that vocal stereotypy may also have been a reinforcer for low rates during the red condition was not evaluated in this study but may warrant further investigation as an independent variable. Additionally, the number of red and green sessions were not held constant daily. While red sessions did occur two to four times daily, there was no specific schedule; rather, session implementation was based on the classroom staff having the opportunity to run the program. Therefore it is unclear as to how many sessions were neces- sary to establish and maintain SDP control of vocal stereotypy. Lastly, vocal stereotypy was not measured in the green condition; therefore, it is unclear if vocal stereotypy remained the same when not under SDP control of the red bracelet or decreased across all conditions. If data were recorded during the green condition and there was a de- crease in responding compared to baseline, it would suggest that the classroom itself came to exert some SDP control of vocal stereotypy. If vocal stereotypy remained at the same levels in the green condi- tion as in baseline, reductions during the red condition may also be attributed to an abolishing operation in effect (Laraway, Syncerski, Michael, & Poling, 2003). David was able to engage in behaviors that produced automatic reinforcement at other times before and after the onset of the red condition (i.e., during any time he wore the green bracelet); therefore, he may have temporarily satiated on the response product of vocal stereotypy, resulting in a momentary decrease in the value of such reinforcement and reducing the occurrence of vocal stereotypy.
An additional limitation was that David’s guardians were im- plementing the system in the home and community. Reports from the guardian and home-based behavioral consultant suggested that, within the home and community settings, the intervention remained at the beginning phases (i.e., 5 min sessions with 1 min intervals) and that the procedures may have been altered slightly due to the parents’ needs in the home and community. This may have pro- duced a confound in two ways: 1) changing contingencies may have
424 LAPRIME and DITTRICH
mitigated some of the effects in the classroom, increasing variability in vocal stereotypy and requiring that David contact the contingen- cies in the classroom before coming under control of the classroom- specific contingencies in place, and 2) the effects of the intervention may have been artificially increased, and generalization across dif- ferent staff and settings in the school may have been due to expo- sure to the contingencies in locations and with persons outside of the school, suggesting that some programming for generalization was in fact occurring.
The results of the current study provided many areas that war- rant further investigation. First, future research should evaluate if changes in responding would remain suppressed without the token system and only the presence of the red bracelet. Second, the occur- rence of appropriate vocalizations should be measured. Previous re- search (i.e., Ahearn et al., 2007; Rapp et al., 2009) found that there were increases in appropriate vocalizations along with decreases in vocal stereotypy when implementing punishment procedures for vocal ste- reotypy. Future researchers should also compare the effects of estab- lishing stimulus control with arbitrary stimuli (i.e., the red bracelet) or stimuli in the environment that may naturally come to exert stimulus control (i.e., the classroom activity or location). For example, when one walks into the library, there is SDP control over the volume of one’s voice due to a history of punishment from the librarian in the library. One research question would be to evaluate generalization and sup- pression of vocal stereotypy over time with naturally occurring stim- uli being paired with the DRL/RC contingency as compared to the red bracelet or some other arbitrary stimulus. Lastly, it should be evalu- ated if the green condition is necessary for the development of strong SDP control. The results of the current study have many implications for practitioners, in terms of assessing and treating vocal stereotypy in an applied setting, with classroom staff implementing the procedures. The treatment package was developed based on the literature but also addressed desired outcomes reported by the guardian and teacher and the available resources of the behavioral consultant, the teacher, and the classroom staff.
In the current study, David’s vocal stereotypy was brought under control of the colored bracelets, with the addition of a cor- responding response cost system and social story. While effective, the intervention was implemented at various times per day with different staff, and across multiple settings within the school. Re- sults demonstrated that the effects of the intervention were general- ized without direct training and that the effects were maintained throughout vacations, during days with fewer sessions, and without
425VOCAL STEREOTYPY
constant oversight by a graduate-level behavioral clinician. Further- more, results suggested that a DRL with response cost procedure may be an effective way to reduce behaviors maintained by auto- matic reinforcement, without the use of extinction-based procedures or the side effects typically associated with punishment procedures. It remains imperative that behavioral consultants work with guard- ians, teachers, and classroom staff to develop efficacious interven- tions that can be successfully implemented with intermittent over- sight from the consultant.
References
Ahearn, W. H., Clark, K. M., MacDonald, R. P. F., & Chung, B. I. (2007). Assessing and treating vocal stereotypy in children with autism. Journal of Applied BehaviorAnalysis, 40, 263–275. doi:10.1901/jaba.2007.30-06
Alderman, N., & Knight, C. (1997). The effectiveness of DRL in the management and treatment of severe behaviour disorders following brain injury. Brain Injury, 11, 79–101.
Ali, S., & Frederickson, N. (2006). Investigating the evidence base of Social Stories. Educational Psychology in Practice, 22, 355–377.
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Ameri- can Psychiatric Association.
Austin, J. L., & Bevan, D. (2011). Using differential reinforcement of low rates to reduce children’s requests for teacher attention. Journal of Applied Behavior Analysis, 44, 451–461.
Bijou, S. W & Baer, D. M. (1963). Some methodological contributions from a functionalanalysis of child development. In L. P. Lip- sitt & C. S. Spiker (Eds.), Advances in child development and behavior (pp. 197–231) New York, NY: Academic Press.
Buckley, S. C., & Newchock, D. K. (2005). An evaluation of simultane- ous presentationand differential reinforcement with response cost to reduce packing. Journal of Applied Behavior Analysis, 38, 405–409.
Catania, C.A. (1997). Learning (4th ed.). Upper Saddle River, NJ: Pren- tice Hall.
Charlop, M. H., Kurtz, P. F., & Casey, F.G. (1990). Using aberrant be- haviors as reinforcers for autistic children. Journal of Applied Behavior Analysis, 23, 163–181. doi:10.1901/jaba.1990.23-163
Cooper, J. O., Heron, T. E., & Heward, W. L. (2007). Applied behavior analysis (2nd Ed). San Antonio, TX: Pearson.
426 LAPRIME and DITTRICH
Deitz, S. M. (1977). An analysis of programming DRL schedules in ed- ucational settings. Behaviour Research and Therapy, 15, 103–111.
Falcomata, T. S., Roane, H. S., Hovanetz, A. N., Kettering, T. L., & Kee- ney, K. M. (2004). An evaluation of response cost in the treat- ment of inappropriate vocalizations maintained by automatic reinforcement. Journal of Applied Behavior Analysis, 37, 83–87. doi:10.1901/jaba.2004.37-83
Fester, C. B., & Skinner, B. F. (1957). Schedules of reinforcement. East Norwalk, CT: Appleton-Century-Crofts.
Grate, C. A., & Garand, J. D. (1993) Social stories: Improving respons- es of students with autism with accurate social information. Focus on Autistic Behaviors, 8, 1–10.
Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. (1994). Toward a functional analysis of self-injury. Journal of Applied Behavior Analysis, 27, 197–209. (Reprinted from Analy- sis and Intervention in Developmental Disabilities, 2, 3–20, 1982).
Johnston, J., & Pennypacker, H. (1980). Strategies and tactics of human behavioral research. Mahwah, NJ: Lawrence Erlbaum.
Kazdin, A. E. (1982). Single-case research designs: Methods for clinical and applied settings. New York, NY: Oxford University Press.
Kokina, A., & Kern, L. (2010). Social StoryTM Interventions for students with autism spectrum disorders: A meta-analysis. Journal of Autism and Developmental Disorders, 40, 812–826. doi:10.1007/ s10803-009-0931-0
Kostinas, G., Scandlen, A., & Luiselli, J. K. (2001). Effects of drl and drl combined with response cost on perseverative verbal behav- ior of an adult with mental retardation and obsessive compul- sive disorder. Behavioral Interventions, 16, 27–37.
Kuttler, S., Myles, B. S., & Carlson, J. K. (1998). The use of social stories to reduce precursors to tantrum behavior in a student with autism. Focus on Autism and Other Developmental Disabilities, 13, 176-182. doi:10.1177/108835769801300306
Laraway, S., Snycerski, S., Michael, J., & Poling, A. (2003). Motivat- ing operations and terms to describe them: some further re- finements. Journal of Applied Behavior Analysis, 36, 407–414. doi:10.1901/jaba.2003.36-407
Lerman, D. C., & Vorndran, C. M. (2002). On the status of knowl- edge for using punishment: Implications for treating behav- ior disorders. Journal of Applied Behavior Analysis, 35, 431–464. doi:10.1901/jaba.2002.35-431
427VOCAL STEREOTYPY
Lloyd, D. M., & Dunn, D. M. (2007). Peabody Picture Vocabulary Test, Fourth Edition. San Antonio, TX: Pearson Educational Press Inc.
Martin, G., & Pear, J. (2010). Behavior modification: What it is and how to do it (9th ed.). Upper Saddle River, NJ: Prentice Hall.
Quill, K. A. (1997). Instructional considerations for young children with autism: The rationale for visually cued instruction. Jour- nal of Autism and Developmental Disorders, 27, 697–713.
Rapp, J. T., & Vollmer, T. R (2005). Stereotypy I: A review of behav- ioral assessment and treatment. Research in Developmental Dis- abilities, 26, 527–547. doi:10.1016/j.ridd.2004.11.005
Rapp, J. T., Patel, M. R., Ghezzi, P. M., O’Flaherty, C. H., & Titter- ington, C. J. (2009). Establishing stimulus control of vocal ste- reotypy displayed by young children with autism. Behavioral Interventions, 24, 85-105.
Reimers, T. M., & Wacker, D. P. (1988). Parents’ ratings of the accept- ability of behavioral treatment recommendations made in an outpatient clinic: A preliminary analysis of the influence of treatment effectiveness. Behavioral Disorders, 14, 7–15.
Rincover, A. (1978). Sensory extinction: A procedure for eliminat- ing self-stimulatory behavior in developmentally disabled children. Journal of Abnormal Child Psychology. 6, 299–310. doi:10.1007/BF00924733
Rogers, S. J., Donovan, C. M., D’Eugenio, D., Brown, S. L., Lynch, E. W., Moersch, M. S., & Schafer, D. S. (1981). Developmental pro- gramming for infants and young children: Early intervention devel- opmental profiles. Revised. (Vol. 2). Ann Arbor, MI: University of Michigan Press.
Rust, J., & Smith, A. (2006). How should the effectiveness of Social Stories to modify the behavior of children of the autistic spec- trum be tested? Autism, 10, 125–138.
Semel, E., Wiig, E. H., & Secord, W. A. (2003). Clinical evaluation of lan- guage fundamentals: Second edition. San Antonio, TX: Pearson Education Inc.
Singh, N. N., Dawson, M. J., & Manning, P. (1981). Effects of spaced responding DRL on the stereotyped behavior of profoundly retarded persons. Journal of Applied Behavior Analysis, 14, 521– 526.
Skinner, B. F. (1957). Verbal behavior. Englewood Cliffs, NJ: Prentice Hall.
428 LAPRIME and DITTRICH
Smith, C. (2001). Using Social Stories to enhance behavior in children with autistic spectrum difficulties. Educational Psychology in Practice, 17, 337–345.
Truchlicka, M., McLaughlin, T. F., & Swain, J. C. (1998). Effects of token reinforcement and response cost on the accuracy of spelling performance with middle-school special education students with behavior disorders. Behavioral Interventions, 13, 1–10.
Vaughn, M. E., & Michael, J. L. (1982). Automatic reinforcement: An important but ignored concept. Behaviorism, 10, 217–227.
Weis, L., & Hall, R. V. (1971). Modification of cigarette smoking through avoidance of punishment. In R.V. Hall (Ed.), Man- aging behavior modification applications in school and home (pp. 54–55) Lawrence, KS: H& H Enterprises.
429VOCAL STEREOTYPY
Appendix A
Social Validity Scale Token System + Red Bracelet
After implementing the procedures or watching the implemen- tation of the procedures in the classroom, please answer all questions honestly and to the best of your ability. His will help to interpret the overall social validity of the study for future reference.
1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree
1. I like the token system + red bracelet procedure. 1 2 3 4 5
2. I think this procedure was easy to implement. 1 2 3 4 5
3. I am satisfied with my students/child’s response to the procedure.
1 2 3 4 5
4. The token system + red bracelet reduced my stu- dent/child’s vocal stereotypy during the session.
1 2 3 4 5
5. I thought the token system + red bracelet proce- dure was intrusive.
1 2 3 4 5
6. The token system + red bracelet program was a socially appropriate intervention for my student/ child.
1 2 3 4 5
7. I think token system + red bracelet training re- sulted in an improved rate of appropriate vocal- izations during the session.
1 2 3 4 5
8. I think token system + red bracelet resulted in an improved rate of appropriate vocalizations out- side of the sessions.
1 2 3 4 5
9. My child/student will benefit in the long run from the procedures implemented in this study.
1 2 3 4 5
10. I think token system + red bracelet training produced negative emotional reactions in my stu- dent/child.
1 2 3 4 5
Please feel free to write in any additional comments regarding the study here:
430 LAPRIME and DITTRICH
Appendix B
Red Condition Social Story:
Green Condition Social Story:
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
