Introduction

Restricted, repetitive behaviors (RRB) are one of the key features of autism spectrum disorder (ASD; American Psychiatric Association 2013). This domain of symptoms refers to a broad range of atypical behaviors that have been characterized as rigid, perseverative and limited in scope (Klinger et al. 2003; Scahill et al. 2015; Turner 1999). Higher-order RRB (H-RRB) have been used to describe cognitively mediated symptoms that include restricted interests and adherence to idiosyncratic routines to distinguish them from motor-sensory repetitive behaviors (Bishop et al. 2006; Carcani-Rathwell et al. 2006; Cuccaro et al. 2003; Richler et al. 2007; Turner-Brown et al. 2011). Impairments with flexibly adjusting behaviors to situational demands is a salient feature of H-RRB (South et al. 2005) that negatively impact the lives of people with ASD, contributing to overall impairments in social, communication, and adaptive functioning (Green et al. 2006; Mercier et al. 2000; Militerni et al. 2002; South et al. 2005; Troyb et al. 2016). Given the persistent nature of these symptoms and the negative implications on functioning, effective interventions to treat this core set of ASD symptoms are needed; however, there is a paucity of research examining interventions that specifically target H-RRB at this time (Harrop 2015).

Impact of Higher-Order RRB

It is important to note that H-RRB can be a source of motivation that can be tapped into to enhance the learning and engagement of youth with ASD (Baker 2000). On the other hand, they also can adversely impact the overall functioning of youth with ASD. A preference for routines and specific interests is a common, healthy, adaptive aspect of childhood—typically developing children often exhibit preferences for particular interests and routines in their play and daily activities. However, they can flexibly vary from or shift their play interests and activities with ease (Evans et al. 1997). In contrast, situational demands to vary from such preferences are often met with significant emotional and behavioral challenges among youth with ASD (Attwood 2003; Klin et al. 2007; Zandt et al. 2007) that is exhibited to an even greater degree than youth with other developmental disorders (Green et al. 2006). Also, Klin and colleagues (2007) found that engagement in restricted interests interfered with children’s self-directed learning and contributed to interpersonal difficulties beginning as early as the preschool-age. Therefore, although H-RRB can be used to make positive changes in the lives of youth with ASD in certain contexts, the impact of H-RRB also can be significantly interfering to youth with ASD, thus requiring intervention.

These behaviors also interfere with children’s interactions with family members, thus having implications for overall family functioning (Mercier et al. 2000; Klin et al. 2007; South et al. 2005). H-RRB are one of the primary factors that contribute to problem behaviors in youth with ASD (Lecavalier et al. 2006). Children’s engagement in such behaviors has been correlated with increased levels of parent stress (Gabriels et al. 2005). Moreover, South and colleagues (2005) found that children’s higher-order, rather than lower-order RRB, were particularly challenging for families. Specifically, parents reported experiencing the most difficulty with their children’s repetitive conversations related to restricted interests and challenging behaviors in response to changes to preferred routines. Therefore, interventions targeting this feature of ASD would benefit from including an emphasis on parent training and parent–child interactions.

Self-Management Intervention and Pivotal Response Treatment

Self-management interventions have a long history as an effective behavioral treatment to promote and maintain positive behavior change for social, communication, and behavioral impairments in youth with ASD (e.g., Lee et al. 2007). The intervention is a multi-component behavioral treatment that emphasizes the monitoring and regulation of behaviors and reinforcements. The core elements of self-management interventions involve (a) self-observation of a targeted behavior, (b) self-recording, and (c) delivery of a reinforcer (Kazdin 1974; Korotitsch and Nelson-Gray 1999; Mahoney 1972). One of the reasons this intervention has been used widely as a treatment for individuals with ASD is the intervention’s emphasis on the active and central role of the participant in monitoring and altering behaviors (Mahoney 1972). It has been proposed that this element of treatment may enhance generalization of skills beyond treatment parameters (Harrower and Dunlap 2001; Koegel et al. 1992). Some targeted behaviors using self-management intervention have included social interactions, conversations, play skills, daily living activities and repetitive play (Koegel et al. 1992; Mancina et al. 2000; Newman et al. 2000; Palmen et al. 2008; Pierce and Schreibman 1994).

Pivotal response treatment (PRT) is considered a naturalistic developmental behavior intervention that is rooted in the principles of applied behavior analysis and has been shown to be effective in treating the symptoms of ASD (Schreibman et al. 2015). It emphasizes the pivotal area of child motivation by using strategies to engage and maintain a child’s interest and engagement through a variety of strategies such as following the child’s interests, varying tasks, interspersing difficult and easy tasks, providing contingent reinforcement, and using rewards directly related to the activity when feasible (Koegel et al. 2001). It also incorporates the role of parent-training in treatment in order to facilitate gains and generalization (Koegel et al. 1996).

Current Study

The purpose of the current pilot study was to examine whether a parent-implemented self-management intervention could improve children’s behavioral flexibility using single-case experimental design. The intervention targeted H-RRB symptoms—restricted, repetitive play interests and insistence on idiosyncratic routines/rituals. The treatment incorporated PRT strategies to enhance child motivation and engagement (e.g., following the child’s interest, varying tasks, mixing up easy and hard tasks). The impact of the intervention on child, parent, and family variables was assessed. Changes in behavioral flexibility, child affect, parent and child engagement, parent affect, overall presence of RRB, and completion of family activities were measured. Given the pervasive nature of these symptoms and the impact they have on the overall outcomes of the child and family functioning, a parent training intervention approach was selected as it has been associated with greater treatment effectiveness and long-term maintenance of behavioral gains in children with ASD (Koegel et al. 1996; McConachie and Diggle 2007; Stahmer and Gist 2001).

Method

Participants

Three children, aged 4 years and 11 months, 4 years and 9 months, and 6 years and 8 months of age respectively, diagnosed with an autism spectrum disorder and their parents were sequentially enrolled and participated in the study. The children were diagnosed according to DSM-IV criteria (American Psychiatric Association 2000) by a psychologist from an outside state agency. Diagnosis was confirmed by review of records, direct observation, and review of child history from a faculty level clinician with more than 20 years of experience with children with autism spectrum disorder. Participants were referred and selected based on parent report of the presence of a long history of H-RRB (e.g., from early childhood) and direct observation of emotional and behavioral difficulties associated with situations in which the children had to vary from their preferred interests or routines. Children had to exhibit problem behaviors, such as crying, yelling, avoidance, or mild aggression (e.g., pushing, throwing items) in the context of their H-RRB in order to be eligible for the study. The children’s parents, two mothers and one father, were selected to participate in the study because they were the primary caregivers. All three parents had completed at least a high school education. The children received behavioral intervention services from a community agency that did not target H-RRB symptoms prior to and during their participation in the study. In addition, no changes to outside interventions, particularly addressing H-RRBs was a requirement for the study. The research study was reviewed and approved by the University’s Institutional Review Board for research with human participants, and informed consent was obtained from the participants. Families did not receive monetary compensation for their participation in the study, but received a gift of a game or toy upon completion.

Receptive and expressive communicative functioning was evaluated with the Vineland Adaptive Behavioral Scales-II (VABS-II; Sparrow et al. 2005). Table 1 summarizes the age, gender, ethnicity, and communicative functioning for each child.

Table 1 Participant characteristics
Full size table

Child 1

Child 1 was diagnosed with Asperger’s Disorder and lived with both her biological parents. She was fully included in her local preschool with the assistance of a classroom aide. Her mother was the primary parent to participate in the study. According to parent report, during the current year, Child 1 demonstrated a circumscribed preference for play interests and routines related to arts-and-crafts activities and certain cartoon characters (e.g., Curious George). She became markedly upset when opportunities to vary from her preferred play interests and routines were introduced or when she was unable to engage in these activities.

Child 2

Child 2 lived with his adoptive father and was enrolled in a local preschool program in which he attended a half-day full-inclusion preschool and a special education classroom. His father was the primary parent in the study. In the past year, Child 2 demonstrated a restricted interest and insistence on routines with toy trains and action figures from Disney movies (e.g., Wall-E). He exhibited problem behaviors when peers or caregivers varied from his preferred play routines and interests. As a result, Child 2 often played alone.

Child 3

Child 3 lived with his parents and two younger siblings. He was fully included in a regular education classroom with the assistance of a classroom aide. His mother was the primary parent to participate in the study. Child 3 demonstrated highly preferred play interests and routines with Lego blocks and repeatedly drawing characters from his favorite cartoons (e.g., Batman). He displayed problem behaviors by protesting, crying, hitting others, or engaging in self-injurious behaviors (i.e., hitting his head with closed fists) when situations related to varying from his preferred play interests and routines arose. As a result, he often played alone.

Experimental Design.

A non-concurrent multiple baseline across participants single case experimental design was implemented to evaluate the effects of the intervention (Barlow and Hersen 1984). This design was selected because it accounts for potential confounding variables related to maturation, history, and habituation (Campbell and Stanley 1963). The baseline phases were systematically staggered from 3, 5, and 7 biweekly data collection probes for Child 1, Child 2, and Child 3, respectively.

Procedures

The study took place primarily within the natural environment of the children’s homes. Due to logistical reasons, some sessions for Child 3 were also conducted in the clinic setting. Materials for the activities used in the study consisted of a variety of age-appropriate games, toys, and activities that the children enjoyed. Toys and activities used also included those that were specifically related to the children’s H-RRB (e.g., Disney figurines), particularly to serve as a context for children to vary from these interests and activities. In addition, toys unrelated to children’s H-RRB were also included for the purposes of expanding children’s engagement to other age-appropriate play interests (e.g., Bingo game). Parents identified activities in which children demonstrated behavioral challenges varying from H-RRB prior to beginning the study. For the self-management intervention, children tracked their points on a sheet of paper with small boxes drawn in. Children were allowed to select from a variety of materials to self-record points such as pens, markers, crayons, and stickers. Prior to starting the study, reinforcers that were unrelated to the children’s H-RRB were selected. Across the study conditions, parents provided opportunities or bids to the children to vary their play interests and routines during H-RRB related activities. For the purposes of data analyses, 10-min video-recordings of parents interacting with their children during these contexts were collected one-time at the beginning of each session.

Baseline

Within the context of children’s highly preferred play interests and routines, parents interacted with their children during H-RRB activities and provided bids, as they typically would in these situations, for their children to vary their play interests and routines by offering suggestions to incorporate other age-appropriate interests or toys, engage in additional developmentally-appropriate activities, or integrate alternative routines.

Intervention

The intervention primarily consisted of a parent-implemented, manualized self-management intervention and incorporated the naturalistic motivational principles of pivotal response treatment (PRT; Koegel et al., 2010; e.g., interspersing easy and difficult goals, reinforcing attempts, providing child choice, task variation, etc.). Self-management intervention procedures were reviewed and parents received instruction and feedback on using them based on a manual that was provided to parents (Koegel et al. 1992). Parents received direct instruction in self-management intervention, incorporating motivational PRT strategies, and in vivo clinician feedback while interacting with their children throughout the treatment phase. Feedback was faded as parents demonstrated mastery of the intervention procedures during the sessions.

Parents implemented the intervention while providing opportunities for their children to vary from H-RRB play interests and routines. Children were encouraged to earn points for being “flexible” during these activities. The number of points children needed to obtain a designated reward was gradually increased across sessions (e.g., starting with 2 points then up to 5 points). The number of parent-provided opportunities varied depending on the child’s responses to the opportunities. For example, if a parent provided an opportunity and the child demonstrated behavioral flexibility, the child accrued the point and the parent provided another opportunity for the child to demonstrate behavioral flexibility to obtain a subsequent point. If the child did not demonstrate behavioral flexibility for the opportunity provided for a point, additional opportunities were provided until the child was able to obtain that point. The average number of parent-provided opportunities per session ranged from 6 to 8 (Child 1: 8 opportunities and Child 1 and 2: 6 opportunities).

Children were presented with a child-friendly definition of behavioral flexibility: “Try something different while keeping your cool.” This concept and behaviors were modeled to children during the initial instruction on self-management. Parents were encouraged to implement the intervention throughout the week and their attempts to practice were discussed in session; however, they were not asked to record their attempts to practice.

Self-management procedures consisted of parents: (a) providing explicit definitions and a behavior model of the targeted behavior (i.e., behavioral flexibility by varying from H-RRB behaviors), (b) having the child select a reward from choices provided by the parent, (c) providing a self-management sheet and instructing children to track and self-record points for targeted behaviors, (d) providing an opportunity for the child to vary from H-RRB, and (e) facilitating children to obtain immediate rewards for accruing a specific number of points.

The following is an example of the treatment session. Parents informed their children that it was time to practice being flexible with a specified activity to earn a treat. Parents asked their children to select from at least two rewards (e.g., “Do you want M&Ms or a popsicle?”). Then they provided the children with a blank self-management point sheet with empty boxes drawn in for the children to mark off. A child-friendly definition of flexibility was provided to the children (e.g., “Keep your cool when we do something a different way. That means staying calm and playing, and not crying or yelling.”). Then they were told that they would be given a chance during the activity to earn points. While engaging in the activity together, parents would present the opportunity for the children. (e.g., “Here’s your chance to get your point, I’m going to add a different train to the railroad track.”) Children would then be praised if the target behaviors were demonstrated (e.g., child allowed parent to add a different train to the activity without demonstrating problem behaviors) and encouraged to mark off a point on their sheet. This would be repeated based on the number of points the children were expected to earn. Once the expected number of points were completed, children were encouraged to obtain the selected reward.

As part of the intervention, a reward system of the children’s reinforcers was used on a contingent basis. The initial set of rewards was determined based on an informal assessment via parent interview during which parents identified their children’s highly preferred reinforcers that was gathered during the baseline phase of the study. Rewards were also determined during the treatment phase as parents became aware of other highly-preferred activities or items that the child demonstrated during the week. Reinforcers that were unrelated to the children’s higher-order RRB were identified and selected. The reinforcers included child-preferred activities, toys, or other tangible items (e.g., favorite snacks, stickers). The identified reinforcers were reserved for intervention sessions and when parents provided the intervention outside of the parent-training sessions.

Sessions were provided approximately two times per week for each family. Each session was about 60 min in duration. The treatment phase was provided for 10 weeks for Child 1 and for 12 weeks for Child 2 and 3. The duration of the treatment phase was not pre-determined but assessed upon observing stability (i.e., less variability) in the data points collected from the treatment sessions, specifically related to behavioral flexibility. Treatment sessions were provided at least two days apart.

The interventionist was a doctoral graduate student in a Clinical Psychology Program in the 4th year of training (4 years training in PRT and self-management intervention) with 5 years of experience of providing one-to-one behavioral intervention to youth with ASD and 4 years of direct parent training to families with ASD in the home setting. This interventionist provided parent-training for all participants in the study.

To determine parent fidelity of treatment implementation, video clips were reviewed in random order from the treatment phase (33% for each parent). Implementation of intervention was coded as present or not present for the following components: (1) parent provides opportunity for child to demonstrate flexibility, (2) parent provides self-management sheet, and (3) child obtains reward. The percentage of components completed by the parent was calculated for each video clip. The total average for all the video clips was obtained for each parent by dividing the total percentage for all video clips with the total number of clips reviewed. Overall, the average percentage of fidelity of treatment implementation ranged between 96 and 100 percent for the parents in the study (parent 1: 100%, parent 2: 96%, and parent 3: 100%).

Follow-Up

Generalization of behavior change was assessed approximately two weeks after the intervention was completed. In this condition, children were presented with opportunities to vary from H-RRB separately by parents and a clinician blinded to the intervention (i.e., uninvolved in the delivery of the treatment and strategies). Parents did not receive feedback and self-management materials were not used, and children were presented with play activities that had not been targeted during intervention.

Dependent Measures

Data were collected across all study conditions in the form of 10-min video-taped probes of dyadic interactions during activities related to H-RRB. The dependent measures consisted of: (1) the percentage of parent-provided opportunities in which children exhibited behavioral flexibility; (2) composite ratings of observed child affect; (3) the percentage of 10-s intervals in which synchronous engagement was demonstrated by the parent and child; (4) composite ratings of observed parent affect during parent–child interactions; and (5) parent ratings of the number of family activities in which they experienced a high degree of difficulty attempting with their child.

Behavioral Flexibility

Behavioral flexibility was defined as children appropriately varying from H-RRB when presented with an opportunity from parents to do so. Parent-provided opportunities were defined as parents providing bids to their children to vary their play interests and routines by offering suggestions to incorporate other age-appropriate interests or toys, engage in additional developmentally-appropriate activities, or integrate alternative routines. Specifically, the criteria for behavioral flexibility were met if children appropriately engaged in an activity that the parents provided a bid for that varied from their H-RRB without demonstrating problem behaviors (e.g., yelling, hitting, throwing items, pushing, snatching items from the hands of parents, avoidance), upset emotions (e.g., crying), or verbal protests for a reasonable amount of time for the activity. Additionally, children had to demonstrate behavioral flexibility within 10-seconds of the opportunity. For each data collection point, the percentage of parent-provided opportunities in which children demonstrated behavioral flexibility was calculated by recording behavioral flexibility as occurring or non-occurring for each opportunity. The sum of opportunities in which behavioral flexibility occurred was then divided by the total number of opportunities and multiplied by 100 to obtain a percentage.

Observed Child Affect

An adapted 6-point Likert rating scale was used to obtain a composite rating of child affect (Koegel and Egel 1979). Video-taped interactions of the parent and child dyad were viewed and a global rating of observed child affect was assigned based on specific criteria (see Table 2 for details). For child affect ratings, interest and happiness were measured separately and then divided by 2 in order to obtain a composite rating for each video clip. Composite affect rating scores between 0 and 1 denoted negative affect, between 2 and 3 reflected neutral affect, and between 4 and 5 indicated positive affect.

Table 2 Global rating of observed child affect
Full size table

Synchronous Engagement

Synchronous engagement was defined as the parent and child appropriately engaging in the same activity in a reciprocal and mutually enjoyed manner (e.g., neither the parent nor the child demonstrated negative affect). These included the parent and child conversing about the activity (e.g., “The monkey in the book is acting silly!”), playing with the items together, or taking turns with the items. Criteria for synchronous engagement were not met if either the parent or child were involved in unrelated activities or if the child engaged in disruptive behaviors (e.g., protesting).

An occurrence of synchronous engagement was scored continuously if the parents and children exhibited synchronous engagement for the majority of a 10-s interval (e.g., more than 70%). A percentage was calculated by dividing the number of intervals in which synchronous engagement occurred by the total number of intervals, and multiplying by 100.

Observed Parent Affect

Observed parent affect ratings were obtained using an adapted 6-point Likert scale from Schreibman et al. (1991). Video-taped interactions of the parent and child dyad were viewed and a global rating of observed parent affect was assigned (see Table 3 for criteria). Low affect was denoted by scores ranging from 0 to 1, neutral affect between 2 and 3, and high affect between 4 and 5.

Table 3 Global rating of observed parent affect
Full size table

Repetitive Behavior Questionnaire (RBQ)

Parents were administered the Repetitive Behavior Questionnaire (RBQ; Turner 1999) at pre- and post-treatment. This parent-rating consists of 26-items assessing for RRB symptoms. The items assess for the presence of sameness behaviors, repetitive movements, repetitive language, and restricted interests on a scale of 0 to 3 or 4 with higher values indicating a greater level of symptoms. A total score was obtained for each child by obtaining a sum of all the scores.

Family Activities Attempted

A parent questionnaire assessing the number of common, daily family activities attempted by the parents and children was completed. Parents were asked to indicate which activities they experienced a high degree of difficulty attempting together with their children at pre- and post-treatment. The questionnaire was developed for the purposes of this study and based on common family activities that were identified in the existing literature (Fenstermaker 1996; Fiese et al. 2002; Hofferth and Sandberg 2001; Sytsma et al. 2001; Zaborskis et al. 2007).

Parents were asked to determine the degree of difficulty they experienced for each item on the questionnaire. Significant difficulty was defined as parents indicating they were almost always unable to begin or complete the indicated activity together, moderate difficulty was defined as parents indicating they were often unable to begin or complete the activity together, mild difficulty was described as parents occasionally being unable to begin or complete the activity together, and no difficulty was defined as parents never or rarely being unable to begin or complete an activity together with their child. The items in which parents reported experiencing significant difficulty with their children were summed to obtain a total for the purposes of this study as this was of primary interest.

Reliability

Inter-rater reliability calculations were obtained for the dependent measures for behavioral flexibility, observed child affect, synchronous engagement, and observed parent affect. For this purpose, two observers independently recorded the dependent variables from the video-taped data probes which were provided in random order to control for possible observer drift and potential experimenter bias. Reliability was computed for approximately 33% of probes across all phases of the study for all the participants. Two undergraduate assistants unfamiliar with the study design and participants were trained by the main author for inter-rater reliability.

Inter-observer reliability was calculated by obtaining the number of agreements, dividing by the total number of agreements plus disagreements, and then multiplying by 100 to obtain a percentage. Agreement was defined as both raters similarly indicating that a behavior was present or absent while a disagreement was defined as raters indicating differences in whether a behavior was present or absent. Disagreements were discussed with the main author and consensus reached if possible. To account for potential chance agreement, Cohen’s kappa coefficient (κ) was calculated (Cohen 1960). For all the dependent measures, kappa ranged between 0.70 and 1.0, indicating strong inter-rater agreement (Landis and Koch 1977; Please see Table 4 for details).

Table 4 Reliability and Cohen’s kappa (κ) values
Full size table

Results

Behavioral Flexibility

Overall, the data across the three children demonstrate that with intervention, behavioral flexibility was exhibited for a greater percentage of parent-provided opportunities relative to baseline levels (Fig. 1). These behavioral gains maintained in the follow-up condition.

Fig. 1
figure 1

Percentage of opportunities to vary from H-RRB in which behavioral flexibility was demonstrated

Full size image

In the baseline condition, Child 1 demonstrated behavioral flexibility for less than 30% of opportunities (range = 0–22%, M = 10%). In comparison, during the intervention condition behavioral flexibility initially was variable, but still above baseline levels, increasing across sessions until consistently staying at 100% (range = 56–100%, M = 91%). In the follow-up condition, Child 1 maintained and generalized these gains at above 90% when treatment materials were removed (range = 92–100, M = 96%).

Similarly, Children 2 and 3 also exhibited gains in behavioral flexibility across conditions. Behavioral flexibility during the treatment condition exceeded baseline levels. Child 2 demonstrated behavioral flexibility for less than 15% of parent-provided opportunities during baseline (range = 5–14%, M = 10%). Behavioral flexibility was variable during the initial portion of treatment, before consistently staying at approximately 100% for subsequent intervention sessions (range = 25–100%, M = 77%). Across all follow-up sessions, Child 2 exhibited flexibility for more than 85% of the opportunities provided (range = 86–100, M = 93%).

Child 3 showed behavioral flexibility for less than 25% of the opportunities during baseline (range = 0–25%, M = 13%). This immediately increased to 100% in the first intervention probe and stayed above baseline levels for most of the intervention condition with periodic variability (range = 33–100%, M = 92%). In the follow-up probes, behavioral flexibility was observed for 60% and 100% of opportunities (M = 80%).

Observed Child Affect

Figure 2 shows the composite ratings of observed child affect ranging from negative, neutral, and positive across conditions. All three children showed gains in their observed affect with the introduction of the intervention. Child 1 demonstrated primarily negative ratings during baseline (range = 0.5–1.5, M = 1.0). With treatment, affect ratings improved to the positive ranges and steadily remained in that range across sessions (range = 2.5-5, M = 4) and continued to be high during the follow-up probes (4.5).

Fig. 2
figure 2

Composite ratings of observed child affect

Full size image

Child 2 demonstrated affect in the negative to neutral ranges during baseline (range = 0.5–2.5, M = 1.5). Affect ratings immediately rose to the positive ranges, becoming more stable in that range across time (range = 1–5, M = 4.0). At follow-up, the composite ratings remained in the positive range and above baseline levels (range = 4.5-5, M = 4.5).

Similarly, affect ratings for Child 3 were mostly in the negative range during baseline (range = 0.5–1.5, M = 0.8) and improved to mostly neutral and positive ratings in the treatment condition (range = 1-4.5, M = 3.5). The ratings remained higher than the baseline range during follow-up (range = 3.5–4.5, M = 4).

Synchronous Engagement

For all the participants, synchronous engagement increased across time with the implementation of treatment and maintained at follow-up (Fig. 3). Child-Parent 1 demonstrated synchronous engagement for less than 50% of the intervals during baseline (range = 23–45%, M = 36%). With treatment, this percentage increased to above baseline levels (range = 71–100%, M = 88%), steadily reaching levels of 100% across data points and maintained at follow-up (M = 100%).

Fig. 3
figure 3

Percentage of 10-s intervals that synchronous engagement was demonstrated

Full size image

Child-Parent 2 exhibited synchronous engagement below 40% during baseline (range = 9–39%, M = 28%). During treatment, the percentage was above baseline levels reaching 100% (range = 60–100%, M = 80) and maintained at follow-up (range = 86–91%, M = 89%).

Last, Child-Parent 3 demonstrated synchronous engagement below 42% across baseline (range = 5–42%, M = 28%). In the treatment condition, the percentage remained above baseline levels reaching 100% at several points (range = 46–100%, M = 85%) and remained well above baseline levels during follow-up (range = 85–88%, M = 87%).

Observed Parent Affect

Composite ratings of observed parent affect changed from the low and neutral ranges during baseline to the high ranges during intervention (Fig. 4). For Parent 1, ratings of parent affect were in the low and neutral ranges during baseline (range = 1–3, M = 2), rose to the high range, and remained at this level (range = 4–5, M = 4.6) across all the treatment sessions. Affect ratings maintained at follow-up (rating = 4). The data for Parents 2 and 3 similarly indicate that at baseline, affect ratings were between low and neutral ranges (Parent 2: range = 0–2, M = 1; Parent 3: range = 1–2, M = 1.60). For Parent 2, in the treatment condition, affect ratings increased to above baseline levels ranging between neutral to high, then steadily stayed within the high ranges toward the latter portion of intervention (range = 2–5, M = 4) that maintained at follow-up (rating = 5). For Parent 3, the introduction of intervention corresponded with affect ratings that were above the levels at baseline with ratings varying between the upper level of the neutral range to the high range (range = 3–5, M = 4) and continuing in the high range at follow-up (rating = 5).

Fig. 4
figure 4

Ratings of observed parent affect

Full size image

Repetitive Behavior Questionnaire

The children showed a decrease in RBQ total scores with the completion of treatment. For Child 1, her RBQ total score was 12 at pre-intervention and 7 at the conclusion of intervention, which reflects a 42% decrease in the scores. Similarly, Child 2’s RBQ total score was 30 at baseline compared to 19 at post-intervention, indicating a 37% decrease in the scores. Last, Child 3 had an RBQ total score of 29 at pre-treatment, which decreased to 16 once intervention was completed, reflecting a 45% change.

Family Activities Attempted

Overall, across all participants, the number of family activities attempted in which the parents reported a high degree of difficulty with their children decreased from pre- to post-treatment. Child 1’s parent reported that 16 family activities were attempted with a high degree of difficulty prior to intervention that then reduced to 5 with a high degree of difficulty after intervention (69% decrease). Child 2’s parent reported that prior to intervention 19 family activities were attempted with a high degree of difficulty; this decreased to 6 after intervention (68% decrease). Child 3’s parent reported that 11 family activities were attempted with a high degree of difficulty before intervention, and this number decreased to 2 after intervention (82% decrease).

Discussion

A parent-implemented self-management intervention incorporating PRT motivational principles resulted in positive gains in behavioral flexibility. Children more easily engaged in a variety of play and daily activities that were unrelated to their H-RRB for a greater percentage of opportunities presented to them. The current study used a combination of self-management intervention and naturalistic developmental behavioral strategies (Koegel et al. 2001) to facilitate the development of behavioral flexibility. The findings show that behavioral treatments for behavioral flexibility are feasible. Interventions treating this area are much needed given that impairments in behavioral flexibility can compromise the capacity to learn new reinforcer relationships by limiting the opportunities to acquire more adaptive behaviors. Adaptive performance involves recognizing and appreciating new contingency relationships (Mischel 1973). Untreated, impairments in behavioral flexibility can have a detrimental effect on children’s adaptability to their environment, limiting their capacity to partake in other age-appropriate activities and social opportunities (Green et al. 2006; Klin et al. 2007). Therefore, it seems that behavioral flexibility is a pivotal area to target in an intervention program.

The beneficial use of H-RRB to facilitate learning and improve social interactions in children with ASD also indicate that these interests and behaviors can be tapped into in order to create positive gains (e.g., Baker 2000). In the future, it seems that a dual-pronged approach in addressing H-RRB would be beneficial by both utilizing children’s interests and activities to enhance skills while also facilitating the development of skills in varying from H-RRB in order to treat this core symptom domain in a multi-faceted manner. Indeed, there has been increasing research examining potential effective treatments to target this area (e.g., Boyd et al. 2011; Ventola et al. 2016), but continued efforts are needed. The current study differed from that of other studies as it specifically targeted varying from H-RRB interests and behaviors using a parent-implemented self-management intervention that incorporated PRT principles. Therefore the focus of the treatment was targeted to a specific group of RRB and treatment was not exclusive to PRT, but incorporated the naturalistic strategies of PRT in order to attain and engage children’s interest in the treatment.

Treatment also was associated with increases in the quality of parent–child interactions, child affect, and parent affect. Before treatment, when children were presented with opportunities to vary from their H-RRB they demonstrated emotional reactivity, parents demonstrated negative affect, and there was a poor quality of parent–child interactions. The improvements observed across these areas highlight the contributing role of improved behavioral flexibility in increasing positive experiences between children and their parents. Behaviors are affected by the interactions between people and their environment (Bandura 1978). Not only do H-RRB limit children’s opportunities for developing adaptive behaviors, but it also compromises their interactions with family members (Green et al. 2006; Klin et al. 2007). Interventions targeting H-RRB are promising given that the asocial nature of H-RRB has been cited as one of the definitive features that differentiates these behaviors and interests from that of typically developing children (Attwood 2003; Klin et al. 2007; Mercier et al. 2000). Parents were able to implement the intervention to fidelity and they made attempts to incorporate the treatment outside of treatment sessions that were then discussed during the intervention sessions.

The findings also show that the intervention was associated with gains in decreasing the overall number of family activities parents experienced difficulty attempting with their children. Such gains were not entirely expected because the intervention primarily targeted a specific subset of H-RRB symptoms. These changes suggest that gains in behavioral flexibility may be associated with broader improvements in daily family functioning. This is important because routines in families of typically developing children have been considered to be an important aspect of both child and family well-being, providing a sense of predictability for children, increasing positive child and family interactions, and by enhancing children’s feelings of security (Sytsma et al. 2001). In contrast, impairments in behavioral flexibility contribute to excessively rigid routines that can have negative effects on child, parent, and family functioning in youth with ASD, thereby limiting their potential benefits.

Improvements in behavioral flexibility maintained once the intervention phase was completed. When parents presented their children with novel activities that had not been targeted during the intervention, the children in the study demonstrated behavioral flexibility at similar levels observed in the intervention phase. Gains also were maintained when the children interacted with a novel adult who was unfamiliar with the intervention. A possible explanation for this finding is that parent training may have contributed to the generalization of skills by providing children with learning opportunities across a variety of different contexts within their natural settings (Lovaas et al. 1973) and multiple exemplars for learning (Albin and Horner 1988).

The findings from the study are promising, but there are limitations. The external validity of the findings of this study is limited to the characteristics of the participants. Future replication of the intervention will provide more robust evidence about the effect of the treatment with additional participants. The follow-up period occurred only 2 weeks after the end of treatment; therefore, the sustained effects of the intervention are unknown. With regard to parent-report measures (e.g., RBQ and family activities questionnaires), the findings could have been biased as parents were interventionists and significantly involved in targeting these symptoms. Therefore, in the future it will be important to include ratings of external change in functioning by raters blind to the intervention to obtain a more objective measure of change. In addition, the generalizability of the study across larger samples of youth with ASD and their families is yet unknown; therefore, the intervention should be further examined using a large randomized group design in the future. Furthermore, the children in this study were responsive to a variety of reinforcers that extended beyond their H-RRB. Therefore, the treatment effects may be limited to children who are responsive to a wider variety of reinforcers. For children responsive to a narrow range of reinforcers or solely to their H-RRB, the effects of this intervention are unclear. Another consideration is that the findings and this type of treatment may be limited to families who prefer a parent-implemented treatment program, rather than primarily a clinician-delivered program. Moreover, the children in the study were verbal. Further research is needed to determine whether this treatment would be appropriate for minimally verbal children. Furthermore, the study did not separately measure treatment integrity for the incorporation of PRT strategies in the self-management intervention as it was part of the intervention package, rather than the sole focus. In the future, it would be worthwhile to separately measure the fidelity of implementation of PRT strategies that were embedded into the self-management intervention. At this time, the field is still unclear as to what are the most effective treatment mechanisms in treating this symptom domain, further exploration into interventions and distilling treatment mechanisms for H-RRB are needed to effectively address the varying needs and presentation of this area for youth with ASD. We anticipate that future research in this area will be very exciting.