- Email: [email protected]
Establishing deictic repertoires in autism
G Model
RASD-1035; No. of Pages 11 Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Research in Autism Spectrum Disorders Journal homepage: http://ees.elsevier.com/RASD/default.asp
Establishing deictic repertoires in autism Shawn P. Gilroy a,*, Elizabeth R. Lorah b, Jessica Dodge a, Catherine Fiorello a a b
Temple University, Philadelphia, Pennsylvania, USA University of Arkansas, Fayetteville, Arkansas, USA
A R T I C L E I N F O
A B S T R A C T
Keywords: Relational responding Perspective-taking Autism Relational Frame Theory
Derived relational responding is an increasingly researched topic in the behavioral sciences. Deictic frames, among other forms of relational responding, have been found to underlie complex behavior, including perspective-taking. Researchers have developed procedures for training children, with and without disabilities, to demonstrate relational responding. Recent extensions of these teaching procedures incorporated more naturalistic contexts as an avenue for establishing these repertoires, but the efficacy of these new procedures for children with disabilities and the degree to which these improvements generalize have not yet been explored. The purpose of this study was to determine if relational repertoires could be established using newer, more naturalistic teaching procedures in children diagnosed with an autism spectrum disorder and to what degree improvements would generalize to another adult. Results indicated that all participants met criteria for mastery on all levels of relational complexity. Additionally, all three children demonstrated generalization of these repertoires to a novel adult following training to mastery. The results suggest that an intervention approach utilizing a BarnesHolmes protocol in a story reading context was effective for establishing deictic repertoires in children diagnosed with an autism spectrum disorder. ß 2015 Elsevier Ltd. All rights reserved.
1. Introduction Among areas of intervention for children with developmental disabilities, there are relatively few interventions designed to remediate perspective-taking deficits (Baron-Cohen, 1995; Howlin, Baron-Cohen, & Hadwin, 1999; Heagle & Rehfeldt, 2006; LeBlanc et al., 2003). This limited range of intervention impacts a wide range of individuals, including children with a development delay (Heagle & Rehfeldt, 2006; LeBlanc et al., 2003; Rehfeldt, Dillen, Ziomek, & Kowalchuck, 2007) and even those still acquiring perspective-taking abilities beyond early childhood (McHugh, Barnes-Holmes, & Barnes-Holmes, 2004). Deficits in these repertoires limits one’s ability to respond to emotional and affective stimuli (Baron-Cohen, 2005; Perner, 1988), infer the beliefs of others (McHugh, Barnes-Holmes, Barnes-Holmes, & Stewart, 2006), form and sustain friendships (Klin, Schultz, & Cohen, 2000) and respond appropriately in complex social environments (Perner, 1991). While deficits in perspective-taking and social behavior are frequently researched in autism (Lord & McGee, 2001; Matson, 2008; Volkmar, Lord, Bailey, Schultz, & Klin, 2004), similar challenges are observed to a lesser degree in many other populations. Efforts to remediate perspective-taking deficits would benefit those with a developmental disability (Rehfeldt et al., 2007), various neuro-behavioral disorders (Villatte, Moneste`s, McHugh, Freixa i Baque´, & Loas, 2010), depression and anhedonia (Villatte, Moneste`s, McHugh, Freixa i Baque´, & Loas, 2008), and even typically-developing individuals (Heagle & Rehfeldt, 2006;
* Corresponding author at: Department of School Psychology, Temple University, Philadelphia, PA, USA. Tel.: +1 2152046254. E-mail address: [email protected] (S.P. Gilroy). http://dx.doi.org/10.1016/j.rasd.2015.04.004 1750-9467/ß 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 2
S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
McHugh et al., 2004; McHugh, Barnes-Holmes, Barnes-Holmes, Stewart, & Dymond, 2007; Weil, Hayes, & Capurro, 2011), an increasingly prevalent consumer of behavior analytic interventions and supports. Prior to the emergence of Relational Frame Theory (RFT: Hayes, Barnes-Holmes, & Roche, 2001a), research into perspective-taking was predominantly cast into the theory of mind framework (Baron-Cohen, 1995, 2005; Baron-Cohen, Tager-Flusberg, & Cohen, 2000; Howlin et al., 1999; Volkmar et al., 2004). Within a theory of mind framework, individuals incorrectly identifying the informational states (e.g., beliefs, motivations, etc.) of others were thought to be at less advanced stage of development. Within this framework, individuals were thought to progress through several stages before they could correctly assume the true or false beliefs of others (Baron-Cohen, 1995, 2005; Baron-Cohen et al., 2000; Howlin et al., 1999). The development of perspective-taking was thought to progress from basic visual perspective-taking, to inferring an informational state (e.g., beliefs, motivation, etc.) and to the development of predictions of other’s behavior based on those states—be they accurate or not (Baron-Cohen, 1995, 2005; Baron-Cohen et al., 2000; Howlin et al., 1999). A typically developing child would be expected to demonstrate these capabilities no later than the age of six-years, as biological maturation has been cited as the driving factor in the development of these repertoires historically (Baron-Cohen, 1995, 2005; Baron-Cohen et al., 2000). More recent research, into how perspective-taking emerges, has utilized a Relational Frame Theory (RFT) approach (Hayes et al., 2001a,b; McHugh et al., 2004; Rehfeldt & Barnes-Holmes, 2009). Rather than assuming that biological maturation drives the development of these abilities, a RFT approach highlights how the arrangement of stimuli in the environment continually shapes how individuals come to derive and respond to stimulus relations in new contexts and situations (e.g., demonstrate relational responding) (Hayes et al., 2001a,b; McHugh et al., 2004). Using this new approach, researchers have since re-explored a range of complex social behavior (e.g., perspective-taking) by analyzing these repertoires in terms of the relational responding that supports them (Barnes-Holmes, Barnes-Holmes, & Cullinan, 2001; Barnes-Holmes, McHugh, & Barnes-Holmes, 2004; Hayes et al., 2001a,b). These researchers have conducted research on relational frames (e.g., deictic frames) as a way to both analyze and teach said repertoires. Among the range of relational frames identified, deictic frames have been identified as a class of relational responding relevant to perspective-taking (Hayes et al., 2001a,b). This family of relational frames (e.g., class of relational responding), includes identity (I vs. You), spatial (Here vs. There) and temporal relations (Now vs. Then) (Barnes-Holmes et al., 2001; Dymond & Barnes, 1995; McHugh et al., 2004). Within a Relational Frame Theory approach to perspective-taking, the relational responding that supports taking the perspective of others can be both assessed (McHugh et al., 2004, 2007) and taught through the arrangement of stimuli in the environment (Davlin, Rehfeldt, & Lovett, 2011; McHugh et al., 2007; Weil et al., 2011). In the studies referenced, relational repertoires (e.g., deictic frames) were assessed in terms of identity relations (I vs. You), spatial relations (Here vs. There) and temporal relations (Now vs. Then). As participants were taught to demonstrate these relations, higher degrees of relational complexity were introduced by ‘‘switching’’ one or more of those relations (e.g., ‘‘If I were you and you were me. . .’’). By switching one or more of these relations (e.g., switching I with You) participants had to subsequently derive the novel relations between stimuli in those contexts. For example, if a participant was asked to respond as if they were another person (e.g., switch identity) they should subsequently derive new spatial relations and temporal locations in order to accurately respond as if they were that individual, in that place and time. In cases of deictic frames with a single switch (e.g., Reversed Relations), if the participant was told ‘‘You are in the kitchen and I am in the living room’’ and was then asked ‘‘If I were you and you were I, where would you be’’ the participant would then have to derive new stimulus relations not originally specified in the initial frame. Through systematically increasing the number of switches and varying the frames switched, participants acquired relational responding through differential reinforcement and error correction. Several studies have successfully demonstrated that deictic repertoires can be established using PC-based tools (Heagle & Rehfeldt, 2006; Rehfeldt et al., 2007), instructional table-top (e.g., adult reading from cue cards) activities (Weil et al., 2011) and even story reading activities (Davlin et al., 2011). While earlier studies provide support that relational responding can be assessed and taught using this framework, only one study has utilized this approach with children diagnosed with an autism spectrum disorder (Rehfeldt et al., 2007). The methods used in this study utilized a computer-based approach to delivering instructional trials and feedback to children with autism, the results of which suggested that children diagnosed with an autism spectrum disorder could be taught to derive deictic relations. While encouraging, the methods utilized in this study required specially-developed computer software. More recent expansions upon protocols have since eliminated the need for computer software and have used storybook materials, resources commonly available to young children (Davlin et al., 2011). However, these newer methods have not been evaluated with children diagnosed with an autism spectrum disorder. Thus, the focus of the current study was to evaluate the effectiveness of a teaching protocol to establish relational responding in children diagnosed with an autism spectrum disorder using the more naturalistic story reading approach from Davlin et al. (2011). The goal of this study was to determine (a) if children with autism would acquire a deictic relational repertoire following the completion of these newer teaching procedures and (b) if those improvements in relational responding would generalize to a novel adult. 2. Method 2.1. Participants Four children were screened for participation and three were included in the study. The ages for Andrew, Brian and Charles were 8, 9 and 11, respectively, as shown in Table 1. Enrollment was advertised through local academic consultants in
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
3
Table 1 Participant information. Participant
Age
Sex
Autism diagnosis
Educational services
Andrew Brian Charles
8 9 13
Male Male Male
Yes Yes Yes
Reading and social skills programming Reading, mathematics and social skills programming Social skills programming
the suburbs of a northeastern city in the United States. The criteria for inclusion included a previous diagnosis of autism, the ability to spontaneously and accurately ask and answer questions (e.g., who, what, where, when) and low observed accuracy on deictic relational responding tasks, as measured by a deictic framing probe. All children had a history of early intervention and academic remediation and were enrolled in general education settings with varying degrees of support. All participants demonstrated marked difficulties with social interaction, including perspective-taking. 2.2. Setting and materials All training was conducted in an unused room in the individual homes of the participants. Training was conducted at a small table and the child and the experimenter were seated at the table across from one another. The children’s stories used were derived from a large compilation book of Disneyß stories. Children’s stories selected for use in probes were based on popular movies (e.g., the Lion Kingß, Aladdinß, etc.), similar to the original study (Davlin et al., 2011). The duration of each probe administration ranged from 30 to 45 min, with shorter sessions being instances where participants declined one or more full breaks. Each active participant completed up to three probes per week (e.g., one every other day) while in consecutive baseline or training sessions. Participants receiving only probes (e.g., non-consecutive baseline or post-training) were administered one probe per week, commensurate with the performance of participants in active baseline or training phases. The total length of the intervention was planned to be no longer than two months, as all training was carried out over the participants’ summer break. 2.3. Dependent measures The dependent variable across all conditions was the percentage of accuracy of relational responding, calculated by dividing the number of correct responses over correct and incorrect responses. Identical to that of the Davlin et al. (2011), trials without switching (e.g., Simple frames), trials with a single switch (e.g., Reversed frames) and trials with two switches (e.g., Double Reversed frames) were included in each administration. All levels of difficulty were presented in each session, sequenced in the order of Simple, Reversed and then Double Reversed. Accuracy was measured both in terms of specific relational complexity (e.g., accuracy for Simple, Reversed, etc.) as well as total overall accuracy among complexities. All items in each probe were attempted at every administration. 2.4. Research design A multiple-probe design across participants was selected for this study. A multiple probe design was selected in order to minimize threats to internal validity and unnecessarily lengthy baseline measures (Gast & Ledford, 2009). Additionally, given the high degree of relational complexity inherent in each administration, unnecessarily long baseline and maintenance conditions would have substantially increased the potential for participant frustration and undesired behavior. 2.5. General procedures 2.5.1. Preference assessment Preference assessment interviews were conducted with all participants and caregivers. An initial preference assessment interview was conducted with a wide range of items and activities. Assessments probed participant preferences across a range of arts and craft materials, story time and reading activities, free access to a tablet device or access to a console video game system. Preferences were assessed by reviewing all available items and allowing the child to select the most desired item. Participants then selected from the remaining possible alternatives, without replacing prior selections, until all items were selected or the participants indicated that the remaining options were non-preferred. All participants endorsed access to electronic activities (e.g., tablet device or console video games) as the most highly preferred activity prior to initial probes. Brief informal preference assessments were conducted prior to beginning each training session to confirm that access to items previously indicated to be preferred (e.g., video games) were indeed desirable at the time. The specific items utilized in the preference assessment included the Nintendo Wiiß, a 7’’ or 10’’ Google Nexusß device, an Apple iPadß, basic arts and craft supplies and access to preferred television programs. 2.5.2. Baseline Following an initial probe to determine inclusion, participants were exposed to baseline conditions. Participants began baseline and training procedures in order based on their performances, with the initial participant demonstrating the lowest
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
4 Table 2 Deictic framing level. Framing level
# Frames # Switches Example
Simple
15
0
Reversed
11
1
Double Reversed 11
2
Buzz Lightyear is flying in the air. You are in the dining room. Where is Buzz? Where are you? (No frame switch) You are reading a book. Pongo is playing with the puppies. If you were Pongo and Pongo were you, what would you be doing? If Pongo were you and you were Pongo, what would Pongo be doing? (One I and You frame switch) Right now you are reading, you were playing on the computer before. Goofy is celebrating a birthday right now, he was cleaning the house before. If you were Goofy and Goofy were you and now was then and then was now, what would you be doing now? If Goofy were you and back then was now, what would Goofy be doing? (One I and You and one Now and Then frame switches)
overall performance in baseline and the final participant demonstrating the highest overall performance in baseline. Procedures in baseline resembled that of the training phase, with the exception that no prompting, reinforcement or error correction procedures were in place. Baseline conditions followed the initial probe and continued until responding stabilized for three consecutive observations. 2.5.3. Deictic framing training protocol Following stable baseline responding, training procedures were introduced in the form of reinforcement and error correction procedures. Participants continued in the training phase until three consecutive probes with 80% or greater accuracy were observed across all levels of complexity. This criterion was designed to ensure that students demonstrated mastery on all levels of difficulty prior to concluding the training phase. All training methods were as close to that of Davlin et al. (2011) as possible. The approach utilized in Davlin et al. (2011) was an extension of the Barnes-Holmes deictic framing protocol (Barnes-Holmes et al., 2004; McHugh et al., 2004, 2006; Rehfeldt et al., 2007) utilizing a story reading context. The procedures used in Davlin et al. (2011) consisted of delivering deictic frame trials derived from the content of popular children’s stories. Within the story reading portions of each session, the researcher embedded 15 opportunities to respond to Simple frames (e.g., no switches), 11 opportunities to respond to Reversed frames (e.g., a single switch) and 11 opportunities to respond to Double Reversed frames (e.g., two switches). The specific contents of each probe (e.g., stories referenced) were kept consistent across participants, commensurate with their position in each phase (e.g., first, second and third probes were consistent across baseline, training, etc.). Each probe consisted of a total of thirty-seven deictic frames, as shown in Table 2. Each instance of framing consisted of two queries, with both forming the basis of the relation, see Fig. 1. Participants were required to correctly respond to both queries, with or without a switch, for the response to be considered correct, as shown in Fig. 2. It is important to note that while I-YOU/NOW-THEN Double Reversed relations were included in the protocol, as utilized in the Davlin et al. (2011) protocol, these relations were not present in the original Barnes-Holmes protocol. Additional departures from earlier methods were necessary to ensure that participants with autism remained engaged and completed all trials on the probe without unnecessary frustration, as original demonstrations were lengthy and not developed for children with a developmental disability (Davlin et al., 2011). These deviations took the form of
[(Fig._1)TD$IG]
Probe: “You are here at home, Hercules is by the pillars. If you were Hercules and Hercules was you…” Question 1: Where would Hercules be? Question 2: Where would you be?
[(Fig._2)TD$IG]
Fig. 1. Deictic frame of Here vs. There relation reversed. This figure depicts the structure of a Here vs. There relation with an I vs. You switch.
Aer I & You Switch
Before Switch Parcipant
You Current Locaon Reading Now
Parcipant
You Book Seng Playing earlier
Character
Pongo Book Seng Playing earlier
Character
P on g o Current Locaon Reading Now
Fig. 2. Relations derived following a switch. This figure depicts how reversing the I vs. You relation requires the derivation of new Here vs. There and Now vs. Then relations.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
5
scheduled breaks of at least three minutes offered following each level of complexity throughout the session (e.g., following Simple frames, Reversed frames, etc.). Additionally, a programmed end point was designated to prevent undue frustration for participants not readily acquiring relational responding (e.g., treatment terminated after two sessions without improvement). 2.5.4. Training In the training phase, all participants were granted access to preferred items and activities contingent on equal or greater accuracy from earlier sessions. For example, participants correctly responding to 11 trials in a previous session would be granted access to items identified as desirable in an earlier preference assessment (e.g., preferred electronics, video games, etc.) if they answered 11 or more trials correctly. Participants who did not meet this criterion were offered access to items identified as less desirable, as identified in the initial preference assessment (e.g., items endorsed as less preferred or not at all). The error correction procedure was implemented as in Davlin et al. (2011). In response to an error the experimenter stated ‘‘No, say (correct answer)’’ and re-presented the trial. If the participant provided the correct response after error correction was provided, verbal praise was delivered. The error correction procedure was implemented until correct responding was achieved. 2.5.5. Post-training Participants meeting mastery criteria in the training phase progressed to a post-training phase. Conditions in the posttraining phase were identical to that of baseline. No prompting, error correction and reinforcement contingencies were in place. The first and second participant completed a post-training phase while the remaining participant(s) were still in training. The final participant progressed directly from the training phase to the generalization phase. A post-training phase was not necessary since all others had completed the training phase and were available to begin the generalization probe. 2.5.6. Generalization Conditions in the generalization phase were nearly identical to that of baseline and post-training. The only difference between baseline and generalization conditions was that a novel adult experimenter, with whom the participants had no previous exposure to, ran a session. The novel adult experimenter delivered all trials. No prompting, error correction or reinforcement contingencies were in place during these trials. 2.6. Inter-observer agreement and procedural fidelity Inter-observer agreement (IOA) was assessed for a total of 34% of all sessions. IOA data were calculated by taking the number of agreements and dividing that by the total number of agreements plus disagreements and multiplying by 100. The overall agreement across all sessions and participants was 98.9% (range: 94.5–100%). In addition to agreement, on-going procedural fidelity was assessed using the procedural fidelity checklist utilized in Davlin et al. (2011). The checklist included whether the administrator followed the assigned order, read trails as written, allowed for breaks, provided or withheld praise or error correction as indicated, followed reinforcement procedures and recorded participant responses. The checklist was expanded from that of Davlin et al. (2011) to include departures related to the delivery of additional structured breaks. Procedural fidelity was completed following every session, indicating 100% procedural fidelity across 100% of sessions. 2.7. Experimenters Each experimenter had at least three years of applied experience with behavior analytic instruction. Two experimenters were doctoral students in School Psychology and one experimenter was a doctoral student in Educational Psychology. All experimenters were certified as a Board Certified Behavior Analyst TM. The novel adult utilized in the generalization probes was a doctoral student in school psychology as well, enrolled in behavior analytic coursework. All experimenters received training on delivering relational framing content and had demonstrated these skills prior to beginning the study. 3. Results 3.1. All participants The results for all participants are depicted as overall averages in Fig. 3 and performances on specific relations are listed in Table 3. Consistent, substantial improvements in overall accuracy were observed across all participants. Following the introduction of the training procedures, substantial improvements in overall accuracy were observed by the initial session for all participants. Performances in training sharply contrasted that of baseline conditions, with consistent upward trending observed immediately following the introduction of the training procedures that contrasted with consistently low, stable performances in baseline conditions for all participants. Performances in generalization conditions were markedly different from baseline as well, indicating that improvements from baseline performances generalized to a novel adult in all participants.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11
[(Fig._3)TD$IG]
S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
6
Fig. 3. Rates of overall acquisition across participants. This figure depicts overall accuracy of responding across phases for all participants. Generalization.
3.2. Andrew Results for Andrew are depicted in Fig. 3 and Fig. 4. Andrew demonstrated low accuracy across all levels of relational responding during baseline. His average performances on Simple, Reversed and Double Reversed relations were 30% (range: [(Fig._4)TD$IG]
Fig. 4. Acquisition of relational repertoire for Andrew. This figure illustrates Andrew’s performances across levels of complexity and across phases.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
7
Table 3 Deictic framing accuracy by frame. Andrew
Simple
Reversed
D. Reversed
IC
HT
NT
IC
HT
NT
IC/HT
IC/NT
Baseline
20 80 40 60
50 25 25 50
16.67 0 16.67 0
0 0 0 0
0 0 0 0
0 0 0 0
0 0 0 0
0 0 0 0
Training
20 20 40 60 60 80 100 80
50 75 100 100 125 100 125 100
83.33 83.33 66.67 100 66.67 83.33 83.33 100
33.33 33.33 66.67 100 66.67 100 100 100
33.33 0 33.33 100 100 100 100 66.67
0 80 100 60 60 80 100 100
25 75 50 75 100 100 100 75
0 0 0 14.29 85.71 100 85.71 100
Post-Training
80 100
100 100
100 83.33
66.67 100
66.67 100
80 100
75 75
85.71 100
Generalization
80
100
100
100
100
60
75
71.43
Brian Baseline
IC 40 40 40 20 0
HT 100 100 100 100 100
NT 100 66.67 33.33 50 83.33
IC 0 0 0 0 0
HT 0 0 0 0 0
NT 0 0 0 0 0
IC/HT 0 0 0 0 0
IC/NT 0 0 0 0 0
Training
40 60 80 60 60 100 100 60
75 100 100 100 100 75 100 100
83.33 100 100 100 100 100 100 100
66.67 100 100 100 100 66.67 100 100
66.67 66.67 66.67 100 100 100 100 100
60 100 100 100 100 100 100 100
25 50 75 50 50 100 100 100
0 28.57 42.86 71.43 100 100 100 100
Post-Training Generalization
60 0
100 100
83.33 100
100 100
100 100
100 80
100 100
100 71.43
Charles Baseline
IC 40 80 40 40 40 0
HT 100 100 75 50 75 100
NT 33.33 33.33 16.67 83.33 83.33 66.67
IC 0 0 0 0 66.67 0
HT 0 0 33.33 33.33 66.67 0
NT 0 0 0 0 0 20
IC/HT 0 25 0 0 25 0
IC/NT 0 0 0 0 14.29 0
Training
80 100 100 100 100
100 100 100 100 100
100 100 100 100 83.33
33.33 100 100 100 100
66.67 66.67 100 100 100
80 100 80 100 100
25 25 100 100 100
14.29 42.86 100 85.71 100
Generalization
80
100
100
100
100
80
100
100
27–30%), 0% and 0% respectively, as shown in Fig. 4. Eight training sessions were necessary to establish criteria for mastery on all frame levels. Improvements continued throughout post-training probes, with average accuracies for Simple, Reversed and Double Reversed relations at 93% (range: 93–93%), 85% (range: 73–100%) and 82% (range: 73–91%). The magnitude of difference between baseline and training/post-training phases was high, evidenced by high level change from baseline (12.16% accuracy) to training/post-training (70.27% average). The Percentage of All Non-overlapping Data (PAND) was 100%, indicating complete separation between baseline and training/post-training performances (Parker, Hagan-Burke, & Vannest, 2007). In the presence of a novel adult, Andrew demonstrated relational responding at 93%, 81% and 72% for Simple, Reversed and Double Reversed relations respectively. 3.3. Brian Brian was the second participant. Brian demonstrated low accuracy on all levels of relational responding in baseline, as shown in Fig. 3 and Fig. 5. His average performances on Simple, Reversed and Double Reversed relations were 63%
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11
[(Fig._5)TD$IG]
S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
8
Fig. 5. Acquisition of relational repertoire for Brian. This figure illustrates Brian’s performances across levels of complexity and across phases.
[(Fig._6)TD$IG]
Fig. 6. Acquisition of relational repertoire for Charles. This figure illustrates Charles’ performances across levels of complexity and across phases.
(range: 53–80%), 0% and 0% respectively in baseline, as illustrated in Fig. 5. Eight training sessions were necessary for Brian to demonstrate mastery on all levels of deictic framing. These improvements persisted through one post-training probe, with accuracies for Simple, Reversed and Double Reversed relations at 80%, 100% and 100% respectively. The magnitude of difference between baseline and training/post-training phases was high, evidenced by high level change from baseline (25.4% accuracy) to training/post-training (84.08% accuracy). The Percentage of All Non-overlapping Data (PAND) was 100%, indicating complete separation between baseline and training/post-training performances (Parker et al., 2007). In the presence of a novel adult, Brian demonstrated relational responding at 66%, 90% and 81% for Simple, Reversed and Double Reversed relations. 3.4. Charles Charles was the final participant. Charles demonstrated low overall accuracy in baseline, with accuracy for Simple, Reversed and Double Reversed relations at 57% (range: 40–67%), 11% (range: 0%–36%) and 5% (range: 0%–18%) respectively, as illustrated in Fig. 3 and Fig. 6. Five training sessions were necessary for Charles to meet criteria for mastery on all levels of relational responding, as depicted in Fig. 3. Charles progressed directly to the generalization phase from training and did not have a post-training phase, as all participants were available to begin the generalization phase. The magnitude of difference between baseline and training phases was high, evidenced by high level change from baseline (25.57% accuracy) to training (86.48% accuracy). The Percentage of All Non-overlapping Data (PAND) was 100%, indicating complete separation between baseline and training performances (Parker et al., 2007). In the presence of a novel adult, Charles demonstrated relational responding at 93%, 90% and 100% for simple, reversed and double reversed relations. 4. Discussion The focus of this study was to determine (a) if children with autism would acquire a deictic relational repertoire using the more naturalistic story reading procedures from Davlin et al. (2011) and (b) if improvements in relational responding would generalize to a novel adult. To answer these questions, data were collected on relational responding prior to and following training on a Deictic Frame teaching protocol (Davlin et al., 2011). Following training all participants were presented with additional deictic frame trials delivered by a novel adult. All participants demonstrated improved overall accuracy following the introduction of the teaching protocol and these improvements were observed in the presence of a novel individual.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
9
With respect to the initial research question, improvements in relational responding were observed in all participants. Consistent with the results of earlier studies on derived relational responding, the teaching procedures within deictic frame teaching protocols produced improved overall accuracies of derived relational responding across multiple levels of complexity in all participants (Davlin et al., 2011; Rehfeldt et al., 2007; Rehfeldt & Barnes-Holmes, 2009; Weil et al., 2011). These results extended earlier research suggesting that individuals with an autism spectrum disorder can be taught to demonstrate this form of derived relational responding (Rehfeldt et al., 2007) and that training with a story book teaching protocol can be used to do so (Davlin et al., 2011). More specifically, this approach was effective in teaching children with an autism spectrum disorder to derive purely novel deictic relations—no trial contents were repeated at any time and every trial presentation was derived from a completely different story context. The implications of acquiring an additional avenue for remedying perspective-taking deficits in autism, among other populations, are great. An improved set of teaching procedures to enhance perspective-taking abilities holds promise as a method for more effectively establishing skills that are prerequisite to a range of social behavior, as the ability to assume to perspective of peers, characters in a story and others in the environment is paramount to succeeding in many social, academic and occupational situations. Looking to the issue of generalization in relational responding, the current authors could find no other study exploring the issue of generalization across individuals using the Barnes-Holmes protocol. To clarify, improvements in the accuracy of relational responding using a Barnes-Holmes protocol suggests that relational responding generalized to new arrangements of stimuli (e.g., new trials following training). The results of this study indicated that improvements in derived relational responding were observed in all participants at all levels of complexity in the presence of a completely novel adult. These findings are highly encouraging, with the potential to offer a method for teaching perspective-taking in a way that can be systematically expanded to new contexts (e.g., clinic, home, school) and individuals (e.g., instructors, parents, peers) over time. Such an approach could lead to benefits beyond the teaching situation, something infrequently observed in theory of mind-based approaches to remediating perspective-taking deficits (Beeger et al., 2011; Gevers, Clifford, Mager, & Boer, 2006; Golan & Baron-Cohen, 2006; Silver & Oakes, 2001; Steerneman, Jackson, Pelzer, & Muris, 1996). However, substantial research needs to be conducted regarding the generalization of these skills to real-world social contexts and same-age peers. This most recent iteration of the Barnes-Holmes protocol (e.g., the Deictic Frame Teaching Protocol) expanded upon earlier forms by including more natural contexts. While previous forms have successfully used PC-based (McHugh et al., 2004; Rehfeldt et al., 2007) tools and table-top teaching methods (e.g., flash cards and vocal presentation) to assess and teach relational responding (Weil et al., 2011), these approaches are something not likely available to parents and early childcare providers. Furthermore, PC-based and purely vocal methods may not be developmentally appropriate and engaging to young children. The procedures developed in Davlin et al. (2011) capitalize on the familiar story reading context, a typically preferred activity that occurs throughout early childhood. As such, many of the previous requirements (e.g., PC, computer program, cue cards, etc.) are not required when story reading material and a child’s imagination are enlisted. 5. Limitations Several issues limit the generalizability of these findings to other populations and individuals diagnosed with autism. First and most notably, a high degree of accuracy on double reversed deictic frames should not be equated to having a fully established arbitrarily applicable deictic repertoire. Additional study is warranted to better explore the if accuracy with double reversed relations, as taught and measured using such a protocol, indeed become applicable to a wide range of real-world social situations. As such, while improvements were observed in all participants there is insufficient evidence to suggest that relational responding improved beyond the specific forms of behavior assessed and trained in the protocol. In addition to this, no follow up data is available and it is unclear to what degree these improvements persisted over time. Post-training follow up data, for participants who received it, was very short (i.e., one to two weeks or none at all). This is a substantial limitation, as it is unclear whether improvements would have persisted at such levels over greater periods of time. In terms of generalization, the measures used were narrow in scope and were not assessed in baseline conditions with both novel and familiar adults. Given the arbitrary nature of relational responding, true assessments of generalization might require demonstrations of these skills across varying individuals (e.g., new, novel, story character, adult, peer, etc.), environments (e.g., home, school, story setting, etc.) and times/situations (e.g., current, previous and future contexts, real and imagined contexts, etc.). As such, the assessment of generalization for arbitrarily applicable relational responding would likely require a much more robust and systematic assessment before relational responding could be considered to be truly arbitrarily applicable and likely to impact complex social behavior. Lastly, it is unclear to what degree that familiarity with certain reading material and earlier exposure to the content of children’s stories might have had on the acquisition of derived relational responding. Some participants maybe have had greater familiarity to certain story material (e.g., through books and movies) and engaged in more frequent shared story reading with their caregivers. While this does not detract from the evidence suggesting that the teaching protocol improved these skills, these issues could have been a factor in their acquisition of this form of relational responding. 6. Future research New approaches emerging from behavioral methodologies are maturing to a level that can measure, and intervene on, complex forms of verbal behavior. On-going research into these newer methodologies has led to greater understanding of
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 10
S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
how logic, humor, empathy and responding to social cues might emerge from an environmental perspective (Barnes-Holmes et al., 2004; Hayes et al., 2001a,b; Villatte et al., 2008, 2010; Weil et al., 2011). While these approaches grant a novel perspective on how complex repertoires might develop, additional research needs to explore the applicability of these procedures to exceptional populations (e.g., autism, developmental disabilities, etc.). Additional research is needed to more clearly understand how these repertoires might be taught, how they might be generalized to novel contexts and what other types of repertoires or characteristics might mediate their acquisition. Despite several limitations, the use of a Relational Frame Theory approach (e.g., deictic relational responding) to teach complex repertoires is one that warrants continued research. An environmental understanding of complex forms of human behavior (e.g., perspective-taking, reasoning, etc.) could potentially lead to additional, and potentially superior, interventions applicable to a range of skill deficits and clinical populations (American Psychiatric Association, 2013; Baron-Cohen, 1995, 2005; Baron-Cohen et al., 2000; Howlin et al., 1999; Rehfeldt et al., 2007; Villatte et al., 2008, 2010). Future interventions should continue to explore these new approaches, potentially incorporating additional theory of mind measures to facilitate comparisons between competing approaches, as well as incorporating other age-appropriate avenues for delivering instructional content (e.g., games, mobile technology, etc.). Acknowledgments The authors would like to acknowledge Dr. Ruth Anne Rehfeldt for assistance with replicating the original study and Ms. Mina Kim for her assistance with data collection and assessing generalization. References American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing. Barnes-Holmes, Y., McHugh, L., & Barnes-Holmes, D. (2004). Perspective-taking and theory of mind: A relational frame account. The Behavior Analyst Today, 5(1), 15–25. Barnes-Holmes, Y., Barnes-Holmes, D., & Cullinan, V. (2001). In S. C. Hayes, D. Barnes-Holmes, & B. Roche (Eds.), Relational frame theory: A post-Skinnerian account of human language and cognition (pp. 181–196). New York, NY: Plenum. Baron-Cohen, S. (1995). Mindblindness: An essay on autism and theory of mind. Cambridge, M.A. MIT Press. (Bradford Books). Baron-Cohen, S., Tager-Flusberg, H., & Cohen, D. (2000). Understanding other minds: Perspectives from developmental cognitive neuroscience (2nd ed.). Oxford: Oxford University Press. Baron-Cohen, S. (2005). Autism and the origins of social neuroscience. In A. Easton & N. Emery (Eds.), The cognitive neuroscience of social behaviour (pp. 239–255). New York, NY: Psychology Press. Beeger, S., Gevers, C., Clifford, P., Verhoeve, M., Kat, K., Hoddenbach, E., et al. (2011). Theory of mind training in children with autism: A randomized controlled trial. Journal of Autism and Developmental Disorders, 41, 997–1006. Davlin, N. L., Rehfeldt, R. A., & Lovett, S. (2011). A relational frame theory approach to understanding perspective-taking using children’s stories in typically developing children. European Journal of Applied Behavior Analysis, 12, 403–430. Dymond, S., & Barnes, D. (1995). A transformation of self-discrimination response functions through the arbitrarily applicable relations of sameness, more-than, and less-than. Journal of the Experimental Analysis of Behavior, 64, 163–184. Gast, D. L., & Ledford, J. R. (2009). Single subject research methodology in behavioral sciences. Mahwah, NJ: Routledge. Gevers, C., Clifford, P., Mager, M., & Boer, F. (2006). Brief report: A ToM-based social cognition training program for school-aged children with PDD: An open study of its effectiveness. Journal of Autism and Developmental Disorders, 36, 567–571. Golan, O., & Baron-Cohen, S. (2006). Systemizing empathy: Teaching adults with Asperger syndrome or high-functioning autism to recognize complex emotions using interactive multimedia. Development and Psychopathology, 18, 591–617. Hayes, S., Barnes-Holmes, C., & Roche, D. B. (Eds.). (2001a). Relational frame theory: A post-Skinnerian account of language and cognition. New York: Plenum Press. Hayes, S. C., Fox, E., Gifford, E. V., Wilson, K. G., Barnes-Holmes, D., & Healy, O. (2001). Derived relational responding as learned behavior. In S. C. Hayes, D. Barnes-Holmes, & B. Roche (Eds.), Relational frame theory: A post-Skinnerian account of language and cognition. New York: Plenum Press. Heagle, A., & Rehfeldt, R. A. (2006). Teaching perspective-taking skills to typically developing children through derived relational responding. Journal of Early and Intensive Behavior Intervention, 3, 1–34. Howlin, P., Baron-Cohen, S., & Hadwin, J. (1999). Teaching children with autism to mind-read: A practical guide. Chichester, England: Wiley. Klin, A., Schultz, R., & Cohen, D. (2000). Theory of mind in action: Developmental perspectives on social neuroscience. In S. Baron-Cohen, H. Tager-Flusberg, & D. Cohen (Eds.), Understanding other minds: Perspectives from developmental cognitive neuroscience (2nd ed., pp. 357–390). Oxford, UK: Oxford University Press. 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 Behavior Analysis, 36, 253–257. Lord, C., & McGee, J. P. (Eds.). (2001). Educating children with autism. Committee on educational interventions for children with autism, division of behavioral and social sciences and education national research council. Washington DC: National Academy Press. Matson, J. L. (2008). Clinical assessment and intervention for autism spectrum disorders. Burlington, MA: Academic Press. McHugh, L., Barnes-Holmes, Y., & Barnes-Holmes, D. (2004). Perspective-taking as relational responding: A developmental profile. The Psychological Record, 54, 115–144. McHugh, L., Barnes-Holmes, Y., Barnes-Holmes, D., & Stewart, I. (2006). Understanding false belief as generalized operant behavior. The Psychological Record, 56, 341–364. McHugh, L., Barnes-Holmes, Y., Barnes-Holmes, D., Stewart, I., & Dymond, S. (2007). Deictic relational complexity and the developmental of deception. The Psychological Record, 57, 517–531. Parker, R. I., Hagan-Burke, S., & Vannest, K. (2007). Percent of all non-overlapping data (PAND): An alternative to PND. Journal of Special Education, 40, 194–204. Perner, J. (1988). Higher-order beliefs and intention in children’s understanding of social interaction. In J. W. Astington, P. L. Harris, & D. R. Olson (Eds.), Developing theories of mind (pp. 271–294). New York, NY: Cambridge University Press. Perner, J. (1991). Understanding the representational mind. Cambridge, MA: MIT Press. Rehfeldt, R. A., & Barnes-Holmes, Y. (2009). Derived relational responding: Applications for learners with autism and other developmental disabilities. Oakland, CA: New Harbinger Publications. Rehfeldt, R. A., Dillen, J. E., Ziomek, M. M., & Kowalchuck, R. K. (2007). Assessing relational learning deficits in perspective-taking in children with high-functioning autism spectrum disorder. The Psychological Record, 57, 23–47.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
11
Silver, M., & Oakes, P. (2001). Evaluation of a new computer intervention to teach people with autism or Asperger syndrome to recognize and predict emotions in others. Autism, 5, 299–316. Steerneman, P., Jackson, S., Pelzer, H., & Muris, P. (1996). Children with social handicaps: An intervention program using a ToM approach. Clinical Child Psychology & Psychiatry, 1, 251–263. Villatte, M., Moneste`s, J. L., McHugh, L., Freixa i Baque´, E., & Loas, G. (2008). Assessing deictic relational responding in social anhedonia: A functional approach to the development of theory of mind impairments. International Journal of Behavioral Consultation and Therapy, 4, 360–373. Villatte, M., Moneste`s, J. L., McHugh, L., Freixa i Baque´, E., & Loas, G. (2010). Assessing perspective taking in schizophrenia using relational frame theory. The Psychological Record, 60, 413–436. Volkmar, F. R., Lord, C., Bailey, A., Schultz, R. T., & Klin, A. (2004). Autism and pervasive developmental disorders. Journal of Child Psychology and Psychiatry, 45, 135–170. Weil, T. M., Hayes, S. C., & Capurro, P. (2011). Establishing a deictic repertoire in young children. The Psychological Record, 61, 371–390.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
RASD-1035; No. of Pages 11 Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Research in Autism Spectrum Disorders Journal homepage: http://ees.elsevier.com/RASD/default.asp
Establishing deictic repertoires in autism Shawn P. Gilroy a,*, Elizabeth R. Lorah b, Jessica Dodge a, Catherine Fiorello a a b
Temple University, Philadelphia, Pennsylvania, USA University of Arkansas, Fayetteville, Arkansas, USA
A R T I C L E I N F O
A B S T R A C T
Keywords: Relational responding Perspective-taking Autism Relational Frame Theory
Derived relational responding is an increasingly researched topic in the behavioral sciences. Deictic frames, among other forms of relational responding, have been found to underlie complex behavior, including perspective-taking. Researchers have developed procedures for training children, with and without disabilities, to demonstrate relational responding. Recent extensions of these teaching procedures incorporated more naturalistic contexts as an avenue for establishing these repertoires, but the efficacy of these new procedures for children with disabilities and the degree to which these improvements generalize have not yet been explored. The purpose of this study was to determine if relational repertoires could be established using newer, more naturalistic teaching procedures in children diagnosed with an autism spectrum disorder and to what degree improvements would generalize to another adult. Results indicated that all participants met criteria for mastery on all levels of relational complexity. Additionally, all three children demonstrated generalization of these repertoires to a novel adult following training to mastery. The results suggest that an intervention approach utilizing a BarnesHolmes protocol in a story reading context was effective for establishing deictic repertoires in children diagnosed with an autism spectrum disorder. ß 2015 Elsevier Ltd. All rights reserved.
1. Introduction Among areas of intervention for children with developmental disabilities, there are relatively few interventions designed to remediate perspective-taking deficits (Baron-Cohen, 1995; Howlin, Baron-Cohen, & Hadwin, 1999; Heagle & Rehfeldt, 2006; LeBlanc et al., 2003). This limited range of intervention impacts a wide range of individuals, including children with a development delay (Heagle & Rehfeldt, 2006; LeBlanc et al., 2003; Rehfeldt, Dillen, Ziomek, & Kowalchuck, 2007) and even those still acquiring perspective-taking abilities beyond early childhood (McHugh, Barnes-Holmes, & Barnes-Holmes, 2004). Deficits in these repertoires limits one’s ability to respond to emotional and affective stimuli (Baron-Cohen, 2005; Perner, 1988), infer the beliefs of others (McHugh, Barnes-Holmes, Barnes-Holmes, & Stewart, 2006), form and sustain friendships (Klin, Schultz, & Cohen, 2000) and respond appropriately in complex social environments (Perner, 1991). While deficits in perspective-taking and social behavior are frequently researched in autism (Lord & McGee, 2001; Matson, 2008; Volkmar, Lord, Bailey, Schultz, & Klin, 2004), similar challenges are observed to a lesser degree in many other populations. Efforts to remediate perspective-taking deficits would benefit those with a developmental disability (Rehfeldt et al., 2007), various neuro-behavioral disorders (Villatte, Moneste`s, McHugh, Freixa i Baque´, & Loas, 2010), depression and anhedonia (Villatte, Moneste`s, McHugh, Freixa i Baque´, & Loas, 2008), and even typically-developing individuals (Heagle & Rehfeldt, 2006;
* Corresponding author at: Department of School Psychology, Temple University, Philadelphia, PA, USA. Tel.: +1 2152046254. E-mail address: [email protected] (S.P. Gilroy). http://dx.doi.org/10.1016/j.rasd.2015.04.004 1750-9467/ß 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 2
S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
McHugh et al., 2004; McHugh, Barnes-Holmes, Barnes-Holmes, Stewart, & Dymond, 2007; Weil, Hayes, & Capurro, 2011), an increasingly prevalent consumer of behavior analytic interventions and supports. Prior to the emergence of Relational Frame Theory (RFT: Hayes, Barnes-Holmes, & Roche, 2001a), research into perspective-taking was predominantly cast into the theory of mind framework (Baron-Cohen, 1995, 2005; Baron-Cohen, Tager-Flusberg, & Cohen, 2000; Howlin et al., 1999; Volkmar et al., 2004). Within a theory of mind framework, individuals incorrectly identifying the informational states (e.g., beliefs, motivations, etc.) of others were thought to be at less advanced stage of development. Within this framework, individuals were thought to progress through several stages before they could correctly assume the true or false beliefs of others (Baron-Cohen, 1995, 2005; Baron-Cohen et al., 2000; Howlin et al., 1999). The development of perspective-taking was thought to progress from basic visual perspective-taking, to inferring an informational state (e.g., beliefs, motivation, etc.) and to the development of predictions of other’s behavior based on those states—be they accurate or not (Baron-Cohen, 1995, 2005; Baron-Cohen et al., 2000; Howlin et al., 1999). A typically developing child would be expected to demonstrate these capabilities no later than the age of six-years, as biological maturation has been cited as the driving factor in the development of these repertoires historically (Baron-Cohen, 1995, 2005; Baron-Cohen et al., 2000). More recent research, into how perspective-taking emerges, has utilized a Relational Frame Theory (RFT) approach (Hayes et al., 2001a,b; McHugh et al., 2004; Rehfeldt & Barnes-Holmes, 2009). Rather than assuming that biological maturation drives the development of these abilities, a RFT approach highlights how the arrangement of stimuli in the environment continually shapes how individuals come to derive and respond to stimulus relations in new contexts and situations (e.g., demonstrate relational responding) (Hayes et al., 2001a,b; McHugh et al., 2004). Using this new approach, researchers have since re-explored a range of complex social behavior (e.g., perspective-taking) by analyzing these repertoires in terms of the relational responding that supports them (Barnes-Holmes, Barnes-Holmes, & Cullinan, 2001; Barnes-Holmes, McHugh, & Barnes-Holmes, 2004; Hayes et al., 2001a,b). These researchers have conducted research on relational frames (e.g., deictic frames) as a way to both analyze and teach said repertoires. Among the range of relational frames identified, deictic frames have been identified as a class of relational responding relevant to perspective-taking (Hayes et al., 2001a,b). This family of relational frames (e.g., class of relational responding), includes identity (I vs. You), spatial (Here vs. There) and temporal relations (Now vs. Then) (Barnes-Holmes et al., 2001; Dymond & Barnes, 1995; McHugh et al., 2004). Within a Relational Frame Theory approach to perspective-taking, the relational responding that supports taking the perspective of others can be both assessed (McHugh et al., 2004, 2007) and taught through the arrangement of stimuli in the environment (Davlin, Rehfeldt, & Lovett, 2011; McHugh et al., 2007; Weil et al., 2011). In the studies referenced, relational repertoires (e.g., deictic frames) were assessed in terms of identity relations (I vs. You), spatial relations (Here vs. There) and temporal relations (Now vs. Then). As participants were taught to demonstrate these relations, higher degrees of relational complexity were introduced by ‘‘switching’’ one or more of those relations (e.g., ‘‘If I were you and you were me. . .’’). By switching one or more of these relations (e.g., switching I with You) participants had to subsequently derive the novel relations between stimuli in those contexts. For example, if a participant was asked to respond as if they were another person (e.g., switch identity) they should subsequently derive new spatial relations and temporal locations in order to accurately respond as if they were that individual, in that place and time. In cases of deictic frames with a single switch (e.g., Reversed Relations), if the participant was told ‘‘You are in the kitchen and I am in the living room’’ and was then asked ‘‘If I were you and you were I, where would you be’’ the participant would then have to derive new stimulus relations not originally specified in the initial frame. Through systematically increasing the number of switches and varying the frames switched, participants acquired relational responding through differential reinforcement and error correction. Several studies have successfully demonstrated that deictic repertoires can be established using PC-based tools (Heagle & Rehfeldt, 2006; Rehfeldt et al., 2007), instructional table-top (e.g., adult reading from cue cards) activities (Weil et al., 2011) and even story reading activities (Davlin et al., 2011). While earlier studies provide support that relational responding can be assessed and taught using this framework, only one study has utilized this approach with children diagnosed with an autism spectrum disorder (Rehfeldt et al., 2007). The methods used in this study utilized a computer-based approach to delivering instructional trials and feedback to children with autism, the results of which suggested that children diagnosed with an autism spectrum disorder could be taught to derive deictic relations. While encouraging, the methods utilized in this study required specially-developed computer software. More recent expansions upon protocols have since eliminated the need for computer software and have used storybook materials, resources commonly available to young children (Davlin et al., 2011). However, these newer methods have not been evaluated with children diagnosed with an autism spectrum disorder. Thus, the focus of the current study was to evaluate the effectiveness of a teaching protocol to establish relational responding in children diagnosed with an autism spectrum disorder using the more naturalistic story reading approach from Davlin et al. (2011). The goal of this study was to determine (a) if children with autism would acquire a deictic relational repertoire following the completion of these newer teaching procedures and (b) if those improvements in relational responding would generalize to a novel adult. 2. Method 2.1. Participants Four children were screened for participation and three were included in the study. The ages for Andrew, Brian and Charles were 8, 9 and 11, respectively, as shown in Table 1. Enrollment was advertised through local academic consultants in
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
3
Table 1 Participant information. Participant
Age
Sex
Autism diagnosis
Educational services
Andrew Brian Charles
8 9 13
Male Male Male
Yes Yes Yes
Reading and social skills programming Reading, mathematics and social skills programming Social skills programming
the suburbs of a northeastern city in the United States. The criteria for inclusion included a previous diagnosis of autism, the ability to spontaneously and accurately ask and answer questions (e.g., who, what, where, when) and low observed accuracy on deictic relational responding tasks, as measured by a deictic framing probe. All children had a history of early intervention and academic remediation and were enrolled in general education settings with varying degrees of support. All participants demonstrated marked difficulties with social interaction, including perspective-taking. 2.2. Setting and materials All training was conducted in an unused room in the individual homes of the participants. Training was conducted at a small table and the child and the experimenter were seated at the table across from one another. The children’s stories used were derived from a large compilation book of Disneyß stories. Children’s stories selected for use in probes were based on popular movies (e.g., the Lion Kingß, Aladdinß, etc.), similar to the original study (Davlin et al., 2011). The duration of each probe administration ranged from 30 to 45 min, with shorter sessions being instances where participants declined one or more full breaks. Each active participant completed up to three probes per week (e.g., one every other day) while in consecutive baseline or training sessions. Participants receiving only probes (e.g., non-consecutive baseline or post-training) were administered one probe per week, commensurate with the performance of participants in active baseline or training phases. The total length of the intervention was planned to be no longer than two months, as all training was carried out over the participants’ summer break. 2.3. Dependent measures The dependent variable across all conditions was the percentage of accuracy of relational responding, calculated by dividing the number of correct responses over correct and incorrect responses. Identical to that of the Davlin et al. (2011), trials without switching (e.g., Simple frames), trials with a single switch (e.g., Reversed frames) and trials with two switches (e.g., Double Reversed frames) were included in each administration. All levels of difficulty were presented in each session, sequenced in the order of Simple, Reversed and then Double Reversed. Accuracy was measured both in terms of specific relational complexity (e.g., accuracy for Simple, Reversed, etc.) as well as total overall accuracy among complexities. All items in each probe were attempted at every administration. 2.4. Research design A multiple-probe design across participants was selected for this study. A multiple probe design was selected in order to minimize threats to internal validity and unnecessarily lengthy baseline measures (Gast & Ledford, 2009). Additionally, given the high degree of relational complexity inherent in each administration, unnecessarily long baseline and maintenance conditions would have substantially increased the potential for participant frustration and undesired behavior. 2.5. General procedures 2.5.1. Preference assessment Preference assessment interviews were conducted with all participants and caregivers. An initial preference assessment interview was conducted with a wide range of items and activities. Assessments probed participant preferences across a range of arts and craft materials, story time and reading activities, free access to a tablet device or access to a console video game system. Preferences were assessed by reviewing all available items and allowing the child to select the most desired item. Participants then selected from the remaining possible alternatives, without replacing prior selections, until all items were selected or the participants indicated that the remaining options were non-preferred. All participants endorsed access to electronic activities (e.g., tablet device or console video games) as the most highly preferred activity prior to initial probes. Brief informal preference assessments were conducted prior to beginning each training session to confirm that access to items previously indicated to be preferred (e.g., video games) were indeed desirable at the time. The specific items utilized in the preference assessment included the Nintendo Wiiß, a 7’’ or 10’’ Google Nexusß device, an Apple iPadß, basic arts and craft supplies and access to preferred television programs. 2.5.2. Baseline Following an initial probe to determine inclusion, participants were exposed to baseline conditions. Participants began baseline and training procedures in order based on their performances, with the initial participant demonstrating the lowest
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
4 Table 2 Deictic framing level. Framing level
# Frames # Switches Example
Simple
15
0
Reversed
11
1
Double Reversed 11
2
Buzz Lightyear is flying in the air. You are in the dining room. Where is Buzz? Where are you? (No frame switch) You are reading a book. Pongo is playing with the puppies. If you were Pongo and Pongo were you, what would you be doing? If Pongo were you and you were Pongo, what would Pongo be doing? (One I and You frame switch) Right now you are reading, you were playing on the computer before. Goofy is celebrating a birthday right now, he was cleaning the house before. If you were Goofy and Goofy were you and now was then and then was now, what would you be doing now? If Goofy were you and back then was now, what would Goofy be doing? (One I and You and one Now and Then frame switches)
overall performance in baseline and the final participant demonstrating the highest overall performance in baseline. Procedures in baseline resembled that of the training phase, with the exception that no prompting, reinforcement or error correction procedures were in place. Baseline conditions followed the initial probe and continued until responding stabilized for three consecutive observations. 2.5.3. Deictic framing training protocol Following stable baseline responding, training procedures were introduced in the form of reinforcement and error correction procedures. Participants continued in the training phase until three consecutive probes with 80% or greater accuracy were observed across all levels of complexity. This criterion was designed to ensure that students demonstrated mastery on all levels of difficulty prior to concluding the training phase. All training methods were as close to that of Davlin et al. (2011) as possible. The approach utilized in Davlin et al. (2011) was an extension of the Barnes-Holmes deictic framing protocol (Barnes-Holmes et al., 2004; McHugh et al., 2004, 2006; Rehfeldt et al., 2007) utilizing a story reading context. The procedures used in Davlin et al. (2011) consisted of delivering deictic frame trials derived from the content of popular children’s stories. Within the story reading portions of each session, the researcher embedded 15 opportunities to respond to Simple frames (e.g., no switches), 11 opportunities to respond to Reversed frames (e.g., a single switch) and 11 opportunities to respond to Double Reversed frames (e.g., two switches). The specific contents of each probe (e.g., stories referenced) were kept consistent across participants, commensurate with their position in each phase (e.g., first, second and third probes were consistent across baseline, training, etc.). Each probe consisted of a total of thirty-seven deictic frames, as shown in Table 2. Each instance of framing consisted of two queries, with both forming the basis of the relation, see Fig. 1. Participants were required to correctly respond to both queries, with or without a switch, for the response to be considered correct, as shown in Fig. 2. It is important to note that while I-YOU/NOW-THEN Double Reversed relations were included in the protocol, as utilized in the Davlin et al. (2011) protocol, these relations were not present in the original Barnes-Holmes protocol. Additional departures from earlier methods were necessary to ensure that participants with autism remained engaged and completed all trials on the probe without unnecessary frustration, as original demonstrations were lengthy and not developed for children with a developmental disability (Davlin et al., 2011). These deviations took the form of
[(Fig._1)TD$IG]
Probe: “You are here at home, Hercules is by the pillars. If you were Hercules and Hercules was you…” Question 1: Where would Hercules be? Question 2: Where would you be?
[(Fig._2)TD$IG]
Fig. 1. Deictic frame of Here vs. There relation reversed. This figure depicts the structure of a Here vs. There relation with an I vs. You switch.
Aer I & You Switch
Before Switch Parcipant
You Current Locaon Reading Now
Parcipant
You Book Seng Playing earlier
Character
Pongo Book Seng Playing earlier
Character
P on g o Current Locaon Reading Now
Fig. 2. Relations derived following a switch. This figure depicts how reversing the I vs. You relation requires the derivation of new Here vs. There and Now vs. Then relations.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
5
scheduled breaks of at least three minutes offered following each level of complexity throughout the session (e.g., following Simple frames, Reversed frames, etc.). Additionally, a programmed end point was designated to prevent undue frustration for participants not readily acquiring relational responding (e.g., treatment terminated after two sessions without improvement). 2.5.4. Training In the training phase, all participants were granted access to preferred items and activities contingent on equal or greater accuracy from earlier sessions. For example, participants correctly responding to 11 trials in a previous session would be granted access to items identified as desirable in an earlier preference assessment (e.g., preferred electronics, video games, etc.) if they answered 11 or more trials correctly. Participants who did not meet this criterion were offered access to items identified as less desirable, as identified in the initial preference assessment (e.g., items endorsed as less preferred or not at all). The error correction procedure was implemented as in Davlin et al. (2011). In response to an error the experimenter stated ‘‘No, say (correct answer)’’ and re-presented the trial. If the participant provided the correct response after error correction was provided, verbal praise was delivered. The error correction procedure was implemented until correct responding was achieved. 2.5.5. Post-training Participants meeting mastery criteria in the training phase progressed to a post-training phase. Conditions in the posttraining phase were identical to that of baseline. No prompting, error correction and reinforcement contingencies were in place. The first and second participant completed a post-training phase while the remaining participant(s) were still in training. The final participant progressed directly from the training phase to the generalization phase. A post-training phase was not necessary since all others had completed the training phase and were available to begin the generalization probe. 2.5.6. Generalization Conditions in the generalization phase were nearly identical to that of baseline and post-training. The only difference between baseline and generalization conditions was that a novel adult experimenter, with whom the participants had no previous exposure to, ran a session. The novel adult experimenter delivered all trials. No prompting, error correction or reinforcement contingencies were in place during these trials. 2.6. Inter-observer agreement and procedural fidelity Inter-observer agreement (IOA) was assessed for a total of 34% of all sessions. IOA data were calculated by taking the number of agreements and dividing that by the total number of agreements plus disagreements and multiplying by 100. The overall agreement across all sessions and participants was 98.9% (range: 94.5–100%). In addition to agreement, on-going procedural fidelity was assessed using the procedural fidelity checklist utilized in Davlin et al. (2011). The checklist included whether the administrator followed the assigned order, read trails as written, allowed for breaks, provided or withheld praise or error correction as indicated, followed reinforcement procedures and recorded participant responses. The checklist was expanded from that of Davlin et al. (2011) to include departures related to the delivery of additional structured breaks. Procedural fidelity was completed following every session, indicating 100% procedural fidelity across 100% of sessions. 2.7. Experimenters Each experimenter had at least three years of applied experience with behavior analytic instruction. Two experimenters were doctoral students in School Psychology and one experimenter was a doctoral student in Educational Psychology. All experimenters were certified as a Board Certified Behavior Analyst TM. The novel adult utilized in the generalization probes was a doctoral student in school psychology as well, enrolled in behavior analytic coursework. All experimenters received training on delivering relational framing content and had demonstrated these skills prior to beginning the study. 3. Results 3.1. All participants The results for all participants are depicted as overall averages in Fig. 3 and performances on specific relations are listed in Table 3. Consistent, substantial improvements in overall accuracy were observed across all participants. Following the introduction of the training procedures, substantial improvements in overall accuracy were observed by the initial session for all participants. Performances in training sharply contrasted that of baseline conditions, with consistent upward trending observed immediately following the introduction of the training procedures that contrasted with consistently low, stable performances in baseline conditions for all participants. Performances in generalization conditions were markedly different from baseline as well, indicating that improvements from baseline performances generalized to a novel adult in all participants.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11
[(Fig._3)TD$IG]
S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
6
Fig. 3. Rates of overall acquisition across participants. This figure depicts overall accuracy of responding across phases for all participants. Generalization.
3.2. Andrew Results for Andrew are depicted in Fig. 3 and Fig. 4. Andrew demonstrated low accuracy across all levels of relational responding during baseline. His average performances on Simple, Reversed and Double Reversed relations were 30% (range: [(Fig._4)TD$IG]
Fig. 4. Acquisition of relational repertoire for Andrew. This figure illustrates Andrew’s performances across levels of complexity and across phases.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
7
Table 3 Deictic framing accuracy by frame. Andrew
Simple
Reversed
D. Reversed
IC
HT
NT
IC
HT
NT
IC/HT
IC/NT
Baseline
20 80 40 60
50 25 25 50
16.67 0 16.67 0
0 0 0 0
0 0 0 0
0 0 0 0
0 0 0 0
0 0 0 0
Training
20 20 40 60 60 80 100 80
50 75 100 100 125 100 125 100
83.33 83.33 66.67 100 66.67 83.33 83.33 100
33.33 33.33 66.67 100 66.67 100 100 100
33.33 0 33.33 100 100 100 100 66.67
0 80 100 60 60 80 100 100
25 75 50 75 100 100 100 75
0 0 0 14.29 85.71 100 85.71 100
Post-Training
80 100
100 100
100 83.33
66.67 100
66.67 100
80 100
75 75
85.71 100
Generalization
80
100
100
100
100
60
75
71.43
Brian Baseline
IC 40 40 40 20 0
HT 100 100 100 100 100
NT 100 66.67 33.33 50 83.33
IC 0 0 0 0 0
HT 0 0 0 0 0
NT 0 0 0 0 0
IC/HT 0 0 0 0 0
IC/NT 0 0 0 0 0
Training
40 60 80 60 60 100 100 60
75 100 100 100 100 75 100 100
83.33 100 100 100 100 100 100 100
66.67 100 100 100 100 66.67 100 100
66.67 66.67 66.67 100 100 100 100 100
60 100 100 100 100 100 100 100
25 50 75 50 50 100 100 100
0 28.57 42.86 71.43 100 100 100 100
Post-Training Generalization
60 0
100 100
83.33 100
100 100
100 100
100 80
100 100
100 71.43
Charles Baseline
IC 40 80 40 40 40 0
HT 100 100 75 50 75 100
NT 33.33 33.33 16.67 83.33 83.33 66.67
IC 0 0 0 0 66.67 0
HT 0 0 33.33 33.33 66.67 0
NT 0 0 0 0 0 20
IC/HT 0 25 0 0 25 0
IC/NT 0 0 0 0 14.29 0
Training
80 100 100 100 100
100 100 100 100 100
100 100 100 100 83.33
33.33 100 100 100 100
66.67 66.67 100 100 100
80 100 80 100 100
25 25 100 100 100
14.29 42.86 100 85.71 100
Generalization
80
100
100
100
100
80
100
100
27–30%), 0% and 0% respectively, as shown in Fig. 4. Eight training sessions were necessary to establish criteria for mastery on all frame levels. Improvements continued throughout post-training probes, with average accuracies for Simple, Reversed and Double Reversed relations at 93% (range: 93–93%), 85% (range: 73–100%) and 82% (range: 73–91%). The magnitude of difference between baseline and training/post-training phases was high, evidenced by high level change from baseline (12.16% accuracy) to training/post-training (70.27% average). The Percentage of All Non-overlapping Data (PAND) was 100%, indicating complete separation between baseline and training/post-training performances (Parker, Hagan-Burke, & Vannest, 2007). In the presence of a novel adult, Andrew demonstrated relational responding at 93%, 81% and 72% for Simple, Reversed and Double Reversed relations respectively. 3.3. Brian Brian was the second participant. Brian demonstrated low accuracy on all levels of relational responding in baseline, as shown in Fig. 3 and Fig. 5. His average performances on Simple, Reversed and Double Reversed relations were 63%
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11
[(Fig._5)TD$IG]
S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
8
Fig. 5. Acquisition of relational repertoire for Brian. This figure illustrates Brian’s performances across levels of complexity and across phases.
[(Fig._6)TD$IG]
Fig. 6. Acquisition of relational repertoire for Charles. This figure illustrates Charles’ performances across levels of complexity and across phases.
(range: 53–80%), 0% and 0% respectively in baseline, as illustrated in Fig. 5. Eight training sessions were necessary for Brian to demonstrate mastery on all levels of deictic framing. These improvements persisted through one post-training probe, with accuracies for Simple, Reversed and Double Reversed relations at 80%, 100% and 100% respectively. The magnitude of difference between baseline and training/post-training phases was high, evidenced by high level change from baseline (25.4% accuracy) to training/post-training (84.08% accuracy). The Percentage of All Non-overlapping Data (PAND) was 100%, indicating complete separation between baseline and training/post-training performances (Parker et al., 2007). In the presence of a novel adult, Brian demonstrated relational responding at 66%, 90% and 81% for Simple, Reversed and Double Reversed relations. 3.4. Charles Charles was the final participant. Charles demonstrated low overall accuracy in baseline, with accuracy for Simple, Reversed and Double Reversed relations at 57% (range: 40–67%), 11% (range: 0%–36%) and 5% (range: 0%–18%) respectively, as illustrated in Fig. 3 and Fig. 6. Five training sessions were necessary for Charles to meet criteria for mastery on all levels of relational responding, as depicted in Fig. 3. Charles progressed directly to the generalization phase from training and did not have a post-training phase, as all participants were available to begin the generalization phase. The magnitude of difference between baseline and training phases was high, evidenced by high level change from baseline (25.57% accuracy) to training (86.48% accuracy). The Percentage of All Non-overlapping Data (PAND) was 100%, indicating complete separation between baseline and training performances (Parker et al., 2007). In the presence of a novel adult, Charles demonstrated relational responding at 93%, 90% and 100% for simple, reversed and double reversed relations. 4. Discussion The focus of this study was to determine (a) if children with autism would acquire a deictic relational repertoire using the more naturalistic story reading procedures from Davlin et al. (2011) and (b) if improvements in relational responding would generalize to a novel adult. To answer these questions, data were collected on relational responding prior to and following training on a Deictic Frame teaching protocol (Davlin et al., 2011). Following training all participants were presented with additional deictic frame trials delivered by a novel adult. All participants demonstrated improved overall accuracy following the introduction of the teaching protocol and these improvements were observed in the presence of a novel individual.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
9
With respect to the initial research question, improvements in relational responding were observed in all participants. Consistent with the results of earlier studies on derived relational responding, the teaching procedures within deictic frame teaching protocols produced improved overall accuracies of derived relational responding across multiple levels of complexity in all participants (Davlin et al., 2011; Rehfeldt et al., 2007; Rehfeldt & Barnes-Holmes, 2009; Weil et al., 2011). These results extended earlier research suggesting that individuals with an autism spectrum disorder can be taught to demonstrate this form of derived relational responding (Rehfeldt et al., 2007) and that training with a story book teaching protocol can be used to do so (Davlin et al., 2011). More specifically, this approach was effective in teaching children with an autism spectrum disorder to derive purely novel deictic relations—no trial contents were repeated at any time and every trial presentation was derived from a completely different story context. The implications of acquiring an additional avenue for remedying perspective-taking deficits in autism, among other populations, are great. An improved set of teaching procedures to enhance perspective-taking abilities holds promise as a method for more effectively establishing skills that are prerequisite to a range of social behavior, as the ability to assume to perspective of peers, characters in a story and others in the environment is paramount to succeeding in many social, academic and occupational situations. Looking to the issue of generalization in relational responding, the current authors could find no other study exploring the issue of generalization across individuals using the Barnes-Holmes protocol. To clarify, improvements in the accuracy of relational responding using a Barnes-Holmes protocol suggests that relational responding generalized to new arrangements of stimuli (e.g., new trials following training). The results of this study indicated that improvements in derived relational responding were observed in all participants at all levels of complexity in the presence of a completely novel adult. These findings are highly encouraging, with the potential to offer a method for teaching perspective-taking in a way that can be systematically expanded to new contexts (e.g., clinic, home, school) and individuals (e.g., instructors, parents, peers) over time. Such an approach could lead to benefits beyond the teaching situation, something infrequently observed in theory of mind-based approaches to remediating perspective-taking deficits (Beeger et al., 2011; Gevers, Clifford, Mager, & Boer, 2006; Golan & Baron-Cohen, 2006; Silver & Oakes, 2001; Steerneman, Jackson, Pelzer, & Muris, 1996). However, substantial research needs to be conducted regarding the generalization of these skills to real-world social contexts and same-age peers. This most recent iteration of the Barnes-Holmes protocol (e.g., the Deictic Frame Teaching Protocol) expanded upon earlier forms by including more natural contexts. While previous forms have successfully used PC-based (McHugh et al., 2004; Rehfeldt et al., 2007) tools and table-top teaching methods (e.g., flash cards and vocal presentation) to assess and teach relational responding (Weil et al., 2011), these approaches are something not likely available to parents and early childcare providers. Furthermore, PC-based and purely vocal methods may not be developmentally appropriate and engaging to young children. The procedures developed in Davlin et al. (2011) capitalize on the familiar story reading context, a typically preferred activity that occurs throughout early childhood. As such, many of the previous requirements (e.g., PC, computer program, cue cards, etc.) are not required when story reading material and a child’s imagination are enlisted. 5. Limitations Several issues limit the generalizability of these findings to other populations and individuals diagnosed with autism. First and most notably, a high degree of accuracy on double reversed deictic frames should not be equated to having a fully established arbitrarily applicable deictic repertoire. Additional study is warranted to better explore the if accuracy with double reversed relations, as taught and measured using such a protocol, indeed become applicable to a wide range of real-world social situations. As such, while improvements were observed in all participants there is insufficient evidence to suggest that relational responding improved beyond the specific forms of behavior assessed and trained in the protocol. In addition to this, no follow up data is available and it is unclear to what degree these improvements persisted over time. Post-training follow up data, for participants who received it, was very short (i.e., one to two weeks or none at all). This is a substantial limitation, as it is unclear whether improvements would have persisted at such levels over greater periods of time. In terms of generalization, the measures used were narrow in scope and were not assessed in baseline conditions with both novel and familiar adults. Given the arbitrary nature of relational responding, true assessments of generalization might require demonstrations of these skills across varying individuals (e.g., new, novel, story character, adult, peer, etc.), environments (e.g., home, school, story setting, etc.) and times/situations (e.g., current, previous and future contexts, real and imagined contexts, etc.). As such, the assessment of generalization for arbitrarily applicable relational responding would likely require a much more robust and systematic assessment before relational responding could be considered to be truly arbitrarily applicable and likely to impact complex social behavior. Lastly, it is unclear to what degree that familiarity with certain reading material and earlier exposure to the content of children’s stories might have had on the acquisition of derived relational responding. Some participants maybe have had greater familiarity to certain story material (e.g., through books and movies) and engaged in more frequent shared story reading with their caregivers. While this does not detract from the evidence suggesting that the teaching protocol improved these skills, these issues could have been a factor in their acquisition of this form of relational responding. 6. Future research New approaches emerging from behavioral methodologies are maturing to a level that can measure, and intervene on, complex forms of verbal behavior. On-going research into these newer methodologies has led to greater understanding of
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 10
S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
how logic, humor, empathy and responding to social cues might emerge from an environmental perspective (Barnes-Holmes et al., 2004; Hayes et al., 2001a,b; Villatte et al., 2008, 2010; Weil et al., 2011). While these approaches grant a novel perspective on how complex repertoires might develop, additional research needs to explore the applicability of these procedures to exceptional populations (e.g., autism, developmental disabilities, etc.). Additional research is needed to more clearly understand how these repertoires might be taught, how they might be generalized to novel contexts and what other types of repertoires or characteristics might mediate their acquisition. Despite several limitations, the use of a Relational Frame Theory approach (e.g., deictic relational responding) to teach complex repertoires is one that warrants continued research. An environmental understanding of complex forms of human behavior (e.g., perspective-taking, reasoning, etc.) could potentially lead to additional, and potentially superior, interventions applicable to a range of skill deficits and clinical populations (American Psychiatric Association, 2013; Baron-Cohen, 1995, 2005; Baron-Cohen et al., 2000; Howlin et al., 1999; Rehfeldt et al., 2007; Villatte et al., 2008, 2010). Future interventions should continue to explore these new approaches, potentially incorporating additional theory of mind measures to facilitate comparisons between competing approaches, as well as incorporating other age-appropriate avenues for delivering instructional content (e.g., games, mobile technology, etc.). Acknowledgments The authors would like to acknowledge Dr. Ruth Anne Rehfeldt for assistance with replicating the original study and Ms. Mina Kim for her assistance with data collection and assessing generalization. References American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing. Barnes-Holmes, Y., McHugh, L., & Barnes-Holmes, D. (2004). Perspective-taking and theory of mind: A relational frame account. The Behavior Analyst Today, 5(1), 15–25. Barnes-Holmes, Y., Barnes-Holmes, D., & Cullinan, V. (2001). In S. C. Hayes, D. Barnes-Holmes, & B. Roche (Eds.), Relational frame theory: A post-Skinnerian account of human language and cognition (pp. 181–196). New York, NY: Plenum. Baron-Cohen, S. (1995). Mindblindness: An essay on autism and theory of mind. Cambridge, M.A. MIT Press. (Bradford Books). Baron-Cohen, S., Tager-Flusberg, H., & Cohen, D. (2000). Understanding other minds: Perspectives from developmental cognitive neuroscience (2nd ed.). Oxford: Oxford University Press. Baron-Cohen, S. (2005). Autism and the origins of social neuroscience. In A. Easton & N. Emery (Eds.), The cognitive neuroscience of social behaviour (pp. 239–255). New York, NY: Psychology Press. Beeger, S., Gevers, C., Clifford, P., Verhoeve, M., Kat, K., Hoddenbach, E., et al. (2011). Theory of mind training in children with autism: A randomized controlled trial. Journal of Autism and Developmental Disorders, 41, 997–1006. Davlin, N. L., Rehfeldt, R. A., & Lovett, S. (2011). A relational frame theory approach to understanding perspective-taking using children’s stories in typically developing children. European Journal of Applied Behavior Analysis, 12, 403–430. Dymond, S., & Barnes, D. (1995). A transformation of self-discrimination response functions through the arbitrarily applicable relations of sameness, more-than, and less-than. Journal of the Experimental Analysis of Behavior, 64, 163–184. Gast, D. L., & Ledford, J. R. (2009). Single subject research methodology in behavioral sciences. Mahwah, NJ: Routledge. Gevers, C., Clifford, P., Mager, M., & Boer, F. (2006). Brief report: A ToM-based social cognition training program for school-aged children with PDD: An open study of its effectiveness. Journal of Autism and Developmental Disorders, 36, 567–571. Golan, O., & Baron-Cohen, S. (2006). Systemizing empathy: Teaching adults with Asperger syndrome or high-functioning autism to recognize complex emotions using interactive multimedia. Development and Psychopathology, 18, 591–617. Hayes, S., Barnes-Holmes, C., & Roche, D. B. (Eds.). (2001a). Relational frame theory: A post-Skinnerian account of language and cognition. New York: Plenum Press. Hayes, S. C., Fox, E., Gifford, E. V., Wilson, K. G., Barnes-Holmes, D., & Healy, O. (2001). Derived relational responding as learned behavior. In S. C. Hayes, D. Barnes-Holmes, & B. Roche (Eds.), Relational frame theory: A post-Skinnerian account of language and cognition. New York: Plenum Press. Heagle, A., & Rehfeldt, R. A. (2006). Teaching perspective-taking skills to typically developing children through derived relational responding. Journal of Early and Intensive Behavior Intervention, 3, 1–34. Howlin, P., Baron-Cohen, S., & Hadwin, J. (1999). Teaching children with autism to mind-read: A practical guide. Chichester, England: Wiley. Klin, A., Schultz, R., & Cohen, D. (2000). Theory of mind in action: Developmental perspectives on social neuroscience. In S. Baron-Cohen, H. Tager-Flusberg, & D. Cohen (Eds.), Understanding other minds: Perspectives from developmental cognitive neuroscience (2nd ed., pp. 357–390). Oxford, UK: Oxford University Press. 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 Behavior Analysis, 36, 253–257. Lord, C., & McGee, J. P. (Eds.). (2001). Educating children with autism. Committee on educational interventions for children with autism, division of behavioral and social sciences and education national research council. Washington DC: National Academy Press. Matson, J. L. (2008). Clinical assessment and intervention for autism spectrum disorders. Burlington, MA: Academic Press. McHugh, L., Barnes-Holmes, Y., & Barnes-Holmes, D. (2004). Perspective-taking as relational responding: A developmental profile. The Psychological Record, 54, 115–144. McHugh, L., Barnes-Holmes, Y., Barnes-Holmes, D., & Stewart, I. (2006). Understanding false belief as generalized operant behavior. The Psychological Record, 56, 341–364. McHugh, L., Barnes-Holmes, Y., Barnes-Holmes, D., Stewart, I., & Dymond, S. (2007). Deictic relational complexity and the developmental of deception. The Psychological Record, 57, 517–531. Parker, R. I., Hagan-Burke, S., & Vannest, K. (2007). Percent of all non-overlapping data (PAND): An alternative to PND. Journal of Special Education, 40, 194–204. Perner, J. (1988). Higher-order beliefs and intention in children’s understanding of social interaction. In J. W. Astington, P. L. Harris, & D. R. Olson (Eds.), Developing theories of mind (pp. 271–294). New York, NY: Cambridge University Press. Perner, J. (1991). Understanding the representational mind. Cambridge, MA: MIT Press. Rehfeldt, R. A., & Barnes-Holmes, Y. (2009). Derived relational responding: Applications for learners with autism and other developmental disabilities. Oakland, CA: New Harbinger Publications. Rehfeldt, R. A., Dillen, J. E., Ziomek, M. M., & Kowalchuck, R. K. (2007). Assessing relational learning deficits in perspective-taking in children with high-functioning autism spectrum disorder. The Psychological Record, 57, 23–47.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004
G Model
RASD-1035; No. of Pages 11 S.P. Gilroy et al. / Research in Autism Spectrum Disorders xxx (2015) xxx–xxx
11
Silver, M., & Oakes, P. (2001). Evaluation of a new computer intervention to teach people with autism or Asperger syndrome to recognize and predict emotions in others. Autism, 5, 299–316. Steerneman, P., Jackson, S., Pelzer, H., & Muris, P. (1996). Children with social handicaps: An intervention program using a ToM approach. Clinical Child Psychology & Psychiatry, 1, 251–263. Villatte, M., Moneste`s, J. L., McHugh, L., Freixa i Baque´, E., & Loas, G. (2008). Assessing deictic relational responding in social anhedonia: A functional approach to the development of theory of mind impairments. International Journal of Behavioral Consultation and Therapy, 4, 360–373. Villatte, M., Moneste`s, J. L., McHugh, L., Freixa i Baque´, E., & Loas, G. (2010). Assessing perspective taking in schizophrenia using relational frame theory. The Psychological Record, 60, 413–436. Volkmar, F. R., Lord, C., Bailey, A., Schultz, R. T., & Klin, A. (2004). Autism and pervasive developmental disorders. Journal of Child Psychology and Psychiatry, 45, 135–170. Weil, T. M., Hayes, S. C., & Capurro, P. (2011). Establishing a deictic repertoire in young children. The Psychological Record, 61, 371–390.
Please cite this article in press as: Gilroy, S. P., et al. Establishing deictic repertoires in autism. Research in Autism Spectrum Disorders (2015), http://dx.doi.org/10.1016/j.rasd.2015.04.004