Scoping Review: Caregiver Training to Reduce Challenging Behaviors Displayed by Children on the Autism Spectrum

Abstract

Effective intervention for children on the autism spectrum who engage in challenging behavior is critical. To ensure meaningful behavior change, caregiver training to support intervention implementation in the natural environment is often required. It is fortunate that both behavioral interventions and caregiver training are considered evidence-based for the reduction of challenging behavior. However, behavioral interventions may be comprised of an idiosyncratic combination of behavior-change strategies that require caregivers to accurately implement several strategies simultaneously or sequentially. The complexity of these interventions may affect parents’ ability to implement interventions accurately and consistently in the natural environment, which may in turn affect child outcomes. However, no review has synthesized the evidence from single-case design studies that evaluate the effects of caregiver training on a variety of caregiver-mediated interventions for autistic children who engage in challenging behavior. Given the complexity of caregiver training and caregiver-mediated interventions, and their increasing popularity, a greater understanding of the related evidence is warranted. We identified trends in participant demographics, caregiver training approaches implemented, child challenging behavior addressed, and behavioral intervention approaches employed. We propose recommendations for interpreting and applying results in practice, and avenues for future research directions.

Challenging behaviors are highly prevalent among children on the autism spectrum (Lecavalier, 2006; Matson et al., 2009). Furthermore, the incidence of challenging behaviors is higher in children on the autism spectrum compared to both typically developing children (Bauminger et al., 2010; Nicholas et al., 2008) and children with other diagnoses such as attention-deficit/hyperactivity disorder, anxiety disorder, Tourette’s syndrome, and others (Dominick et al., 2007; Matson et al., 2009). These challenging behaviors can compromise children’s safety, interfere with learning, and disrupt family life (Hartley et al., 2008). Research suggests that caregivers of autistic children experience more stress than caregivers of either typically developing children or children with another disability. Further, these challenging behaviors are often correlated with caregiver stress (Hayes & Watson, 2013; Lecavalier et al., 2005).

Interventions based on the principles of behavior analysis (i.e., applied behavior analysis [ABA] interventions) have been shown to effectively reduce the occurrence of challenging behavior displayed by children on the autism spectrum. A variety of behavior-change strategies have been identified as evidence-based with respect to behavior reduction (e.g., antecedent-based intervention, reinforcement, response interruption and redirection, extinction; Steinbrenner et al., 2020). Although interventions that utilize one behavior-change strategy can sometimes be effective, ABA-based interventions commonly involve idiosyncratic combinations of several strategies tailored to the needs of the individual (e.g., choice with task interspersal and reinforcement, antecedent interventions followed by functional communication training with extinction and delay and denial training, noncontingent reinforcement with extinction; National Autism Center [NAC], 2015; Nuta et al., 2021; Saini et al., 2017). Some research shows that a combination of several behavior-change strategies (e.g., antecedent, consequent, and/or function-based skill replacement strategies) may be required to meaningfully reduce challenging behavior (e.g., Smith et al., 2019; Waters et al., 2009).

Caregiver training and caregiver-implemented interventions are also considered evidence-based for the reduction of challenging behavior displayed by autistic children and youth (e.g., NAC, 2015; Steinbrenner et al., 2020) and research indicates that caregivers can effectively implement ABA-based strategies (e.g., Bearss et al., 2015; Miles & Wilder, 2009; Nuta et al., 2021; Symon, 2005). Further, evidence suggests child outcomes may be improved when caregivers are trained to implement ABA-based interventions (Reid & Fitch, 2011; Wainer & Ingersoll, 2013). In fact, there is a substantial body of evidence suggesting that caregiver involvement in interventions which are designed to match each family’s context and preferences and are integrated into salient family routines result in reductions in child challenging behavior (e.g., Dunlap & Fox, 1999; Moes & Frea, 2000, 2002). Further, caregivers report that these interventions are acceptable and sustainable. It should be noted that a considerable number of these early studies demonstrated positive child outcomes following caregiver-implemented functional assessment and functional communication (e.g., Wacker et al., 1998, 2005), as well as more complex multicomponent interventions (e.g., long-term supports, prevention and consequent strategies, replacement skills; Dunlap & Fox, 1999). Taken together, there is substantial evidence to suggest that child outcomes may be improved when caregivers are trained to implement ABA-based interventions (e.g., Reid & Fitch, 2011; Wainer & Ingersoll, 2013), including complex multicomponent interventions.

However, the complexity of these interventions may pose a challenge to meaningful caregiver involvement if caregivers are unable to implement the interventions with sufficient accuracy (i.e., treatment integrity) and consistency (i.e., adherence; Allen & Warzak, 2000). Further, reduced treatment integrity may negatively affect intervention effectiveness (e.g., Fryling et al., 2012; Wilder et al., 2006). Although several of the aforementioned studies reflect pioneering advances in caregiver-mediated intervention (e.g., Dunlap & Fox, 1999; Moes & Frea, 2000, 2002: Wacker et al., 1998, 2005), only one of these studies reported caregiver treatment integrity (i.e., Moes & Frea (2002) reported that caregivers were trained to implement contextualized functional communication training [FCT] with > 80% fidelity). Although reporting treatment integrity was not common practice at that time, the importance of sharing this information has recently been highlighted. Falakfarsa et al. (2022) recommended reporting treatment integrity to increase confidence that manipulations of the independent variable were responsible for outcomes, to rule out the impact of extraneous variables, and to reduce the likelihood that researchers publish results about treatment effectiveness that cannot be replicated by clinicians. Further, they noted the importance of reporting treatment integrity to reduce the possibility that effective treatments may be deemed ineffective (and vice versa) and to avoid the impact of this on treatment decision making and funding. Recent studies have begun to report on caregiver treatment integrity (e.g., Gerow et al., 2019a, 2019b; Nuta et al., 2021). However, there have been no systematic reviews of the literature related to caregiver-implemented intervention for reduction of challenging behavior with a particular focus on caregiver treatment integrity.

Researchers have started to conduct reviews to better understand caregiver training in, and caregiver implementation of, ABA-based interventions (e.g., Becraft et al., 2024; Gerow, Hagan-Burke et al., 2018a; McConachie & Diggle, 2007; McGinnis et al., 2023; Nevill et al., 2018; Oono et al., 2013; Patterson et al., 2011; Postorino et al., 2017; Shalev et al., 2019; Unholz-Bowden et al., 2020). However, only seven of the existing reviews included studies that evaluated the impact of intervention on challenging or maladaptive behavior (Becraft et al., 2024; Gerow, Hagan-Burke et al., 2018a; McConachie & Diggle, 2007; McGinnis et al., 2023; Oono et al., 2013; Postorino et al., 2017; Unholz-Bowden et al., 2020). Of those reviews, only four included studies that used single case design (SCD) methodology (Becraft et al., 2024; Gerow, Hagan-Burke et al., 2018a; McGinnis et al., 2023; Unholz-Bowden et al., 2020), despite the majority of ABA literature using this research design. Of the remaining four studies, one was limited to interventions that utilized FCT and excluded other intervention approaches (Gerow, Hagan-Burke et al., 2018a), one only evaluated the efficacy of caregiver-implemented intervention for autistic adolescents and young adults and excluded studies focused on child participants (McGinnis et al., 2023), one focused solely on caregiver-training delivered via telehealth and excluded studies involving in-person intervention (Unholz-Bowden et al., 2020), and one focused on whether caregiver involvement was broadly categorized as passive or active without analyzing other components of caregiver training and involvement (Becraft et al., 2024).

Because there is currently no comprehensive review of SCD literature related to caregiver implemented ABA-based interventions for the treatment of challenging behavior displayed by autistic children––and no reviews with a focus on studies that report caregiver treatment integrity––the purpose of this scoping review is to address this gap. In particular, this review was designed to enhance understanding of participant demographics, the methods of caregiver training, caregiver-implemented interventions, and the topography and function of targeted challenging behaviors. The presence of follow-up and social validity information was also explored. Given the importance of treatment integrity for establishing a functional relationship between intervention and outcomes, only studies that specifically included caregiver treatment integrity data were included.

The majority of behavior analysts work with autistic children (i.e., 72.2% of Behavior Analyst Certification Board [BACB] certificants; BACB, n.d.) and all have an ethical imperative to involve and maximize outcomes for clients (i.e., BACB Ethical Codes 2.09 and 2.01, respectively; BACB, 2020). For this reason, a greater understanding of variables related to caregiver training and caregiver-implemented ABA interventions in studies that include direct caregiver treatment integrity data seems warranted. Results may be used to inform future directions in caregiver training research. Further, this information may help clinicians better understand how the existing body of evidence may be applied to clinical practice.

Method

The methods for conducting and reporting the results of this review were based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) guidelines (Tricco et al., 2018). A comprehensive search strategy that aligned with PRISMA guidelines was developed in collaboration with two master’s level librarians.

Inclusion Criteria

Title and Abstract Screening

Inclusion criteria during the title and abstract screening phase were as follows: (1) published in a peer-reviewed journal; (2) published in English; (3) utilized a SCD; (4) involved one or more children on the autism spectrum; (5) involved caregiver training and/or caregiver-mediated intervention; (6) targeted reduction of challenging behaviors; and (7) data were collected via direct observation of behavior.

Full Text Screening

Additional criteria were added during the full text screening phase. Inclusion criteria at this phase were the following: (1) published in English; (2) published in 1987 or later1; (3) published in a peer-reviewed journal; (4) utilized a SCD; (5) presented data on caregiver treatment integrity obtained through direct observation2; (6) involved one or more children on the autism spectrum; (7) utilized an ABA-based intervention; and (8) child challenging behavior reduction was the primary goal of the caregiver-mediated intervention (determined based on the purpose of the study and dependent variable definitions).

To create operational definitions for ABA-based intervention approaches we reviewed behavior analytic texts (i.e., Cooper et al., 2007; Miltenberger, 2016) and comprehensive reviews (i.e., NAC, 2015; Wong et al., 2015) to identify existing operational definitions, and consulted four doctoral-level board certified behavior analysts (BCBAs) with expertise in challenging behavior for their recommendations. We integrated this information with the seven dimensions of ABA to establish the inclusion criteria.

Search Strategy

Database Searches

We searched eight databases to identify studies for inclusion: CINAHL Complete, Education Source, Embase, ERIC, Medline, Nursing and Allied Health Source, PsycINFO, and Web of Science Complete. Database searches were conducted on January 15, 2021 (except for Web of Science Complete, which was searched on January 19, 2021). We collaborated with master’s level university librarians to adapt search terms related to challenging behavior, autism, and caregiver treatment integrity for each database (see Appendix for PsycINFO search strategy. Full search strings for all databases available from authors upon request).

Ancillary Searches

We searched reference lists of related reviews and included studies to identify additional relevant studies that were not retrieved by the database search (i.e., backward search; Lefebvre et al., 2019). In addition, we searched journal issues from 2000 (the year of publication of the oldest study identified for inclusion) to January 2021, inclusive, for four publications: Journal of Applied Behavior Analysis, Behavioral Interventions, Journal of Autism and Developmental Disorders, and Research in Developmental Disabilities (i.e., hand search; Lefebvre et al., 2019; Page et al., 2021). The Journal of Applied Behavior Analysis was selected because it is the flagship ABA journal; the remaining three were selected because the included studies were most commonly published in these journals.

Data Collection and Analysis

Selection of Studies

We imported all records retrieved by the database search to Covidence (https://www.covidence.org). For the backward search, we examined the reference lists in each publication. For the hand search, we screened records through journal volumes and issues on the publisher’s website. We independently screened records first by title and abstract and then by full text review. Records were included if they met all inclusion criteria and contained no exclusion criteria.

Data Extraction and Management

The following data were coded and analyzed via descriptive analysis. Full operational definitions for all data items are available from the authors upon request.

Demographic Information

Caregiver demographic information collected included the relationship of the caregiver to the child, level of education, employment status, age, and income. Child demographic information collected included gender, age, diagnosis, age at diagnosis, school status, ABA service or other service status, and hours per week of ABA services, if applicable. Information about the race and ethnicity of participants and the individuals living in the home was also collected. Demographic categories and definitions were adapted from those used by Jones et al. (2020), Gerow, Hagan-Burke et al. (2018a), and the Canadian census (Statistics Canada, 2021).

Format and Setting of Intervention

We coded whether training and intervention were in person and/or online, and whether training and intervention were conducted in the clinic, home, school, university, and/or other setting. These data were coded and analyzed at the study level (as opposed to reported for each individual participant) because this variable was generally consistent across participants within a study.

Caregiver Trainer Experience

When this information was provided, we coded for the experience and/or qualifications of the caregiver trainers. We extracted and reported direct quotations from each text. These data were coded and analyzed at the study level.

Method of Caregiver Training

Behavior skills training (BST; Miltenberger, 2008), which consists of instructions, modeling, rehearsal, and feedback was selected as the framework for defining and coding caregiver training components. This decision was made as BST is current best practice for mediator training broadly, and for training caregivers of children on the autism spectrum in particular (e.g., Erhard et al., 2019; Schaefer & Andzik, 2021). We coded the specific training approaches described in the study instructions (e.g., written and/or vocal instructions), modeling (e.g., in-vivo, online, prerecorded,), rehearsal (e.g., caregivers provided an opportunity to practice the skill in-vivo, online), and feedback (e.g., in-vivo and recorded feedback provided either immediately or following a delay). We also coded if the study reported using BST. We only indicated that a study utilized BST if the authors explicitly reported its use but coded for the presence of individual BST components based on the description of the training procedures. These five categories were not mutually exclusive, and we coded if any combination of these strategies was present (e.g., a study that described using BST and technologically described their caregiver-training procedures would be coded as including all five categories). These data were coded and analyzed at the study level as this variable was generally consistent across participants within a study.

Caregiver-Implemented Intervention

We coded the behavior-change strategies implemented by caregivers using the terms employed by the authors and their technological descriptions of the strategies. For example, although differential reinforcement of alternative behavior (DRA) by definition includes reinforcement delivered following a desired behavior and withheld following the behavior targeted for reduction (Cooper et al., 2020), if authors named DRA as a strategy, technologically described their use of reinforcement, but did not describe withholding reinforcement or the use of extinction, we coded the study as using DRA and reinforcement, but not extinction. This conservative approach was used to accurately categorize the technological use of various intervention approaches. Operational definitions for all intervention strategies are available from the authors upon request.

Caregiver Treatment Integrity

We coded caregiver treatment integrity into one of two categories: 80% or greater on average, or less than 80% on average (NAC, 2015). We employed this dichotomous approach to coding to manage the heterogeneous nature and level of detail of the available treatment integrity data (e.g., visual display, numerical values in tables or text). We extracted scores based on the overall average across all participants and intervention phases, excluding assessment and baseline.

Functional Assessment of Child Behavior

We coded if studies reported the use of functional behavioral assessment (i.e., indirect and/or direct functional behavioral assessments and functional analyses) and if caregivers were involved in these assessments. We also indicated if the authors reported the hypothesized function of child behavior, if applicable.

Topography of Child Behavior

We coded each child’s topography(ies) of challenging behavior (e.g., aggression, property destruction, self-injury). The topography categories selected were based on items included in an initial screener of problem behaviors (i.e., the autism spectrum disorders-behavior problems for children; Matson et al., 2009).

Generalization and Maintenance

We coded if studies presented generalization data (defined as measurement of challenging behavior across novel settings or people, or as the treatment effect spread across untargeted behaviors) and/or maintenance data (defined as direct measures of behavior collected at least 1 week after termination of researcher-mediated intervention or support) for both child and caregiver behavior. When applicable, we also coded the domain(s) across which generalization data were collected (i.e., settings/environments, behaviors/routines, people).

Social Validity

We indicated if social validity data were presented and, if applicable, whether the intervention was rated positively by caregivers. A positive rating was scored if the authors reported that overall the caregivers rated the intervention positively, above neutral, etc., or if the majority of individual items were rated positively. These data were coded and analyzed at the study level as this variable was generally reported for participants as a group.

Interrater Reliability

The first author and a second rater (a BCBA-D or graduate students in an ABA program, all trained to minimum 80% reliability) independently screened articles for inclusion and coded data items for interrater reliability (IRR) purposes. Total count IRR was calculated as the number of agreements divided by the number of agreements plus the number of disagreements, multiplied by 100%. A second independent rater screened at least 33% of records identified through the database and ancillary searches (determined using an online random number generator) for inclusion at both the title and abstract and full text levels. For screening, an agreement was defined as (1) both raters voting “no”; (2) both raters voting “yes”; or (3) one rater voting “yes” and another rater voting “maybe” (as both ratings resulted in the study being retained for full review by the Covidence platform). A disagreement was defined as (1) one rater voting “yes” and another rater voting “no” or (2) one rater voting “maybe” and another rater voting “no”. For the database search, IRR was 97% and 82% for title and abstract and full text screening, respectively. For the ancillary search, IRR was 98% and 94% for title and abstract and full text screening, respectively. All disagreements were resolved by consensus. A second independent rater screened at least 33% of studies (determined using an online random number generator) for all data extracted from included studies. For all the data extraction items (with the exception of some demographic items and caregiver trainer experience), raters indicated either “yes” or “no” to all the variables of interest (e.g., for format of the intervention, raters indicated “yes” or “no” to the intervention being conducted online and “yes” or “no” to the intervention being conducted in-person). An agreement was defined as both raters indicating “yes” or both raters indicating “no” to a single item and a disagreement was defined as one rater indicating “yes” and the other indicating “no” to a single item. Some demographic items and caregiver trainer experience required fillable responses (e.g., child and caregiver age, household income, description of trainer credentials and experience). For these items, the raters indicated exactly what was reported in the study. An agreement was defined as both raters indicating identical information and a disagreement was defined as the raters recording nonidentical information. Mean IRR was 95% (range: 84%–100%) across all categories of data (i.e., caregiver and child demographic information, setting and format of intervention, caregiver trainer experience, method of caregiver training, caregiver-implemented intervention, caregiver treatment integrity, assessment of child behavior, topography and function of child behavior, generalization and maintenance, and social validity). All disagreements were resolved by consensus.

Results

Search Strategy

Database and Ancillary Search

The database search identified 17,278 unique records. Following screening, 32 unique records remained. One record contained two studies, resulting in a total of 33 studies included for analysis. One additional study was added following the ancillary search, resulting in a total of 34 studies included for analysis (see Fig. 1 for PRISMA flow diagram).

Fig. 1.

PRISMA Flow Diagram

Data Extraction

Demographic Information

A total of 85 families participated across 34 studies, with a mean of 2.5 families per study (range: 1–12). Demographic information is presented in Table 1.

Table 1.

Participant Demographics

Characteristic	Participants for Whom Information is Provided
	n	%
Child Race/Ethnicitya(reported for 48% of children; n = 41)
Asian	3	7
Black	4	10
Hispanic	6	15
White	24	58
Biracial/Multiple	4	10
Caregiver Race/Ethnicitya(reported for 24% of caregivers; n = 20)
Asian	4	20
Black	2	10
White	13	65
Biracial/Multiple	1	5
Child Languageb(reported for 22% of participants, n = 19)
English only	2	10
Other language only	15	79
English and other language	2	10
Caregiver Languageb (reported for 36% of participants, n = 31)
English only	10	32
Other language only	16	52
English and other language	5	16
Child Gender/Sex (reported for 96% of participants, n = 82)
Male	65	79
Female	17	21
Child Diagnosis (reported for 66% of participants, n = 56)
Autism only	18	32
Autism + other(s)	38	68
Caregiver Relationship with Child (reported for 100% of participants, n = 85)
Mother	69	81
Father	10	12
Other caregiver	8	9
Caregiver Education (reported for 50% of participants, n = 43)
High school	10	23
College	4	9
Bachelor’s	15	35
Master’s	9	21
Other	5	12
Caregiver Employment (reported for 32% of participants, n = 27)
Unemployed/Not employed outside the home	13	48
Part Time	4	15
Full Time	6	22
Employment hours undefined	4	15
Individuals Living in the Home (reported for 33% of participants, n = 28)
Mother only	1	4
Mother and father	7	25
Mother and sibling(s)	1	4
Mother, father, and sibling(s)	19	68

^aRace/ethnicity was presumed for three children and their caregivers (in Blair et al., 2011); the authors reported that the study was conducted in Korea but did not explicitly describe the race/ethnicity of participants. Race/ethnicity was coded as “Asian” for these three dyads

^bLanguage was presumed for three children and their caregivers (in Blair et al., 2011); the authors reported that the study was conducted in Korea but did not explicitly describe the language of participants. Language was coded as “other language” for these three dyads. Language was also presumed for 12 children (in Tsami et al., 2019); the authors reported the country each dyad lived in and noted that could not understand or speak English as an inclusion criterion but did not describe the children’s known languages. Based on the information provided, language was coded as “other language” for these 12 children

Note. Languages other than English that were endorsed were Arabic, Japanese, Lebanese, and Spanish. For caregiver relationship with child, these values add up to greater than 100%, as two children had multiple caregivers participate

The following information is based on the number of participants for which data were reported. Of the 41 children (48%) for whom race/ethnicity was reported, 24 were described as White (58%), six were Hispanic (15%), four were Black (10%), three were Asian (7%), and four were described as biracial or reported multiple races/ethnicities (10%). For the 20 caregivers for whom race/ethnicity was reported, 13 were described as White (65%), four were Asian (20%), two were Black (10%), and one reported multiple races/ethnicities (5%). Language was reported for 19 children (22%); 15 were described as speaking a language other than English (79%; e.g., Spanish, Turkish, Greek, Arabic), two were described as speaking English only (10%), and two were described as speaking both English and another language (10%). Language was reported for 31 caregivers (36%); 16 were described as speaking a language other than English (52%; e.g., Spanish, Turkish, Greek, Arabic), 10 were described as speaking English only (32%), and 5 were described as speaking English and another language (16%). Sixty-eight of the caregivers were mothers (80%). Caregiver education level was reported for 41 participants (48%); 15 had completed a bachelor’s degree (36%), 10 had completed high school (24%), and nine had a master’s degree (22%). Caregiver employment status was reported for 27 participants (32%); 13 were unemployed/did not work outside the home (48%). Caregiver age was reported for 30 caregivers (35%), and caregivers were about 36 years old on average (range: 27–47). Household income was reported for three families (4%; all participated in the same study) and was greater than $105,000 for each. Finally, the individuals living in the home was reported for 28 families (33%); nineteen families had a mother, father, and at least two children (including the child participant who was on the autism spectrum) living in the home (68%).

Sixty-five of the 85 children who participated in the studies were males (76%), and the mean age of all children was approximately 6 years old (range 20 months–16 years). Thirty-eight children were diagnosed with autism spectrum disorder only (67%), but many had additional diagnoses (e.g., intellectual disability, attention deficit hyperactivity disorder, language disorders). The mean age at diagnosis was reported for 11 participants (13%) and was approximately 3 years of age (range: 1.5 years–5 years). Children’s attendance at school was reported for 34 participants (40%) and 32 of these children were in school (94%). Participation in ABA and non-ABA services was reported for 19 (22%) and 16 (19%) participants, respectively. Thirteen children were receiving ABA services (68%) and 14 were receiving non-ABA services (88%). Of the 13 children who were receiving ABA services, service hours were reported for 6 (46%) and ranged from 2 hr–40 hr per week.

Format and Setting of Intervention

Overall, 23 studies (68%) reported that intervention took place only in-person and six (18%) reported that intervention took place only online. Five studies (15%) reported that intervention took place in a combination of in-person and online formats. Fifteen studies (44%) reported that intervention took place in the home only and 16 (47%) reported that intervention took place in both the home and at least one other setting (i.e., a total of 31 studies [91%] conducted intervention at least in-part in the client home). Only three studies (9%) did not involve any implementation in the home. Eight studies (24%) implemented at least part of the intervention at a university, four (12%) implemented part of the intervention in the clinic, three (9%) involved the school setting, and eight (24%) implemented part of the intervention in some other setting (e.g., playground, shopping, walking).

Caregiver Trainer Experience

Eleven studies (32%) explicitly reported the experience and/or credentials of the caregiver trainer. Overall, caregiver trainers had various backgrounds, but all had graduate degrees or were completing graduate degrees/coursework at the time of the study (e.g., graduate students, BCBAs and doctoral-level BCBAs (BCBA-Ds), master’s and doctoral level clinicians in behavior analysis and related fields). Years of experience working with caregivers, children with disabilities, and/or providing early intervention or behavior analytic services ranged from 2 to 10 years when reported.

Method of Caregiver Training

Seven studies (20%) explicitly reported using BST in their training strategy, and five of those studies (71%) described in technological detail how each component of BST was provided. Eleven studies (32%) described using all four BST components. Altogether, 18 studies (53%) reported using BST and/or described using the four components of BST. Two studies (6%) reported using one of instructions, modeling, rehearsal, and feedback, 13 studies (38%) reported using some combination of the four, and 1 study (3%) did not report any of the training strategies used.

Caregiver Implemented Intervention

Almost all caregivers were taught to use reinforcement (99%), and most were taught to use prompting (90%), manipulate motivating operations (72%), implement extinction (67%), and use FCT (56%). Less commonly, caregivers were trained to use punishment (33%), stimulus control (24%), differential reinforcement (23%), chaining/fading (22%), and choice (12%). Rarely were caregivers taught to use noncontingent reinforcement (5%) or high probability sequences (1%), and none were trained to use imitation or shaping.

An average of five (range: 2–8) strategies were used in combination during intervention (not necessarily simultaneously or for the entire duration of treatment). Several trends in combinations of strategies were identified (see Fig. 2). Reinforcement, extinction, and FCT were presented together (sometimes in combination with additional components) in 40 cases (49%). Further, these components were also often presented in combination with prompting (N = 37 cases; 45%), manipulation of motivating operations, (N = 32; 39%), and chaining/fading (N = 16; 20%). These common combinations of treatment components are in line with clinical practice standards and suggest that authors are providing technological descriptions of intervention procedures that often implicate relevant antecedent and consequent manipulations. For example, these data suggest that authors are thoroughly describing FCT procedures in which relevant motivating operations are addressed and reinforcement is both withheld contingent on challenging behavior and provided contingent on a functional communication response. It is important to note that although extinction was not explicitly described in 53% of cases of differential reinforcement, it is likely that extinction was in place for challenging behavior but simply not described in the article, given that this procedure involves allocating reinforcement based on specific, appropriate responses (Cooper et al., 2020).

Fig. 2.

Caregiver-Implemented Behavior-Change Strategies

Caregiver Treatment Integrity

Caregiver treatment integrity was 80% or greater on average for 28 studies (82%) and below 80% on average for 3 studies (9%). An average caregiver treatment integrity score could not be determined for three studies (9%). For the three studies that reported caregiver treatment integrity below 80%, Gerow, Hagan-Burke et al. (2018a), Gerow, Rispoli et al. (2018b) reported that Michael’s father implemented treatment with 70% integrity on average (range: 0%–100%) and demonstrated a steadily increasing trend throughout, with the last four data points being above 80%. Data from Lerman et al. (2000) were not provided in textual format, and data were extracted from the graphs using online software (WebPlotDigitizer). Average integrity across all treatment phases was 79% (range: 64%–100%); this increased to 86% (range: 64%–100%) when only accounting for data obtained in the terminal treatment phase for all skills and participants. Finally, one caregiver participant in Lanovaz et al. (2016) did implement treatment with an average of 86% integrity (range: 65%–100%), whereas the other demonstrated lower and more variable levels (M = 71%, range: 48%–83%).

Functional Assessment of Child Behavior

Sixty-three caregivers (74%) participated in indirect behavioral assessments such as researcher-developed interviews (e.g., Boyd et al., 2011), open-ended functional assessment interview (Hanley, 2012; e.g., Hoffmann et al., 2019), the Functional Assessment Interview (O’Neill et al., 1997; e.g., Fettig et al., 2015), Functional Assessment Screening Tool (Iwata, 2002; e.g., Crone & Mehta, 2016), and Questions About Behavior Function (Matson & Vollmer, 1995; e.g., Raulston et al., 2019). Twenty caregivers (24%) participated in direct behavioral assessments such as structured descriptive assessments (e.g., Lequia et al., 2013) and antecedent-behavior-consequence observations (e.g., Crone & Mehta, 2016). However, direct behavioral assessments were conducted for an additional seven children (8%) without any caregiver involvement. Forty-four caregivers (52%) participated in a functional analysis, and functional analyses were conducted for an additional nine children (10%) without caregiver involvement. Functional analysis procedures were commonly based on those by Iwata et al. (1982/1994), with modifications such as adding test conditions to assess tangible functions and modifying session length (e.g., Hoffmann et al., 2019; Tsami & Lerman, 2020). Authors also reported conducting trial-based functional analyses (e.g., Gerow, Rispoli et al., 2019b) and brief functional analyses (e.g., Moes & Frea, 2002). For three children (4%), no type of assessment was reported.

A hypothesized function was presented for 66 children (78%). Of those for whom a hypothesized function was indicated, challenging behaviors were most commonly multiply controlled (n = 32; 48%). Less commonly reported were challenging behaviors maintained by a single function, e.g., access to tangible items (n = 15; 23%), escape (n = 10; 15%), automatic reinforcement (n = 6; 9%), and attention (n = 3; 5%). For 19 children (22%), no hypothesized function was presented.

Topography of Child Behavior

The topography of challenging behaviors targeted for intervention varied widely. Children were reported to demonstrate a mean of about 2.5 different topographies of challenging behavior (range: 1–6). Forty-six children engaged in disruptive behavior (which included behaviors such as crying, whining, screaming, swearing, flopping to the floor, and “tantrums”), and aggression (54% each), 32 engaged in property destruction (38%), 24 engaged in self-injury (28%), and 17 engaged in stereotypy (20%). Thirteen children engaged in noncompliance (15%), six engaged in elopement (7%), three had feeding issues (4%), three mouthed inedible objects (4% each), two engaged in disrobing (2%), one engaged in inappropriate sexual behaviors (1%), and seven engaged in other behaviors that did not fit into predetermined categories (8%). No children engaged in fecal smearing.

Generalization and Maintenance

Of the 34 studies included in the review, 32 (94%) collected data on child behavior as a dependent variable whereas 8 (24%) collected data on caregiver treatment integrity as a dependent variable. Given this important difference, we conducted two analyses. First, the 32 studies that reported child behavior as a dependent variable were reviewed for the presence of child generalization and/or maintenance data. Next, the eight studies that reported caregiver treatment integrity as a dependent variable were reviewed for the presence of caregiver generalization and/or maintenance data. Results of data extraction regarding generalization and maintenance data for both dependent variables are summarized in Fig. 3.

Fig. 3.

Studies Presenting Generalization and Maintenance Data for Child Behavior and Caregiver Integrity

Six studies (18% of all studies, 19% of the 32 studies with child behavior as a dependent variable) collected child generalization data. One of these studies investigated generalization across locations only; four investigated generalization across both locations and challenging behaviors or routines; and one investigated generalization across locations, challenging behaviors or routines, and people. Seven studies (20% of all studies, 22% of studies with child behavior as a dependent variable) collected maintenance measures on child behaviors.

Four studies (12% of total studies, 50% of the eight studies with caregiver treatment integrity as a dependent variable) collected caregiver generalization data; three of these investigated generalization across location, three investigated generalization across challenging behaviors or routines, and one investigated generalization across people (for one of three participants only). Four studies (12% of total studies, 50% of studies with caregiver treatment integrity as a dependent variable) collected maintenance measures on caregiver behaviors.

Social Validity

Nineteen studies (56%) provided information on social validity as reported by the caregiver participants. In eighteen studies (95%), caregivers completed a questionnaire or interview. In eight of these (44%), caregivers completed an adapted version of the Treatment Acceptability Rating Form (original or revised; Reimers & Wacker, 1988; Reimers et al., 1992). Two of these 19 studies (11%) included caregiver self-ratings and naïve observer ratings of session videos. All but one of the 19 studies (95%) reported that caregivers rated the programs positively; one study (i.e., Cavell et al., 2018) reported administering the Parenting Stress Index Short Form (Abidin, 1995) as a measure of social validity. These researchers found that two of three caregivers’ Total Stress scores decreased slightly following intervention, but the Total Stress score increased for one caregiver .

Discussion

Caregiver training is considered evidence based for the treatment of challenging behaviors displayed by autistic children (Wong et al., 2015). However, little is known about the current state of the literature with respect to SCDs that directly evaluate caregiver training and treatment integrity while implementing ABA-based interventions to reduce challenging behavior. The purpose of this scoping review was to highlight variables that have been evaluated in SCDs of caregiver training and intervention for autistic children who engage in challenging behavior and to summarize gaps that still exist in the literature. The present review included 34 studies involving a total of 85 families. How authors reported caregiver treatment integrity data varied widely, and although a full synthesis and meta-analysis of the results are beyond the scope of this article, most studies reported an average treatment integrity of 80% or greater. Caregivers were trained to implement a diversity of behavior change strategies, typically involving positive reinforcement contingent on appropriate responses and extinction contingent on inappropriate responses, among other treatment components. Interventions were designed to address an array of common topographies of child challenging behavior (e.g., disruptive behavior, aggression, property destruction, self-injury, stereotypy). Most child participants engaged in several topographies of challenging behavior and these behaviors were often multiply controlled. These results are promising as they indicate that caregivers can accurately implement a variety of behavior change strategies, both individually and in combination, to address a variety of topographies and functions of child challenging behavior. The results also highlight some important considerations for practitioners when training caregivers to address challenging behaviors.

Across studies, demographic information was relatively limited, especially for caregiver demographic variables, most of which were reported in less than half of studies. A relatively homogenous sample of participants were included; many caregivers were White English-speaking mothers. Half were unemployed or did not work outside the home, and many homes were composed of two caregivers. Family demographics and structure are crucial variables to consider in the context of caregiver training. Cultural and family norms, values, and language may affect caregivers’ willingness and ability to engage in training and implement interventions (Dennison et al., 2019; Martinez & Mahoney, 2022; Sivaraman & Fahmie, 2020). Socioeconomic status, particularly caregiver employment status, may place time constraints on caregivers’ abilities to attend training sessions or consistently implement interventions in the home. Caregivers who work outside the home may find it difficult to attend trainings during the day when providers are most available or may have limited time to implement interventions accurately. These caregivers may not be well represented in the current literature, and it is unknown how these and other demographic variables may affect caregiver-mediated interventions and associated child outcomes. Further, more than half of the children in the included studies were reported to have autism diagnoses alone. Child disability status, including the number and nature of diagnoses, may exacerbate autism symptoms, warrant the need for additional interventions, including medical or pharmacological treatments, and affect child outcomes (Koudys et al., 2021). Clinicians should consider how caregiver and child demographics vary from those reported in the literature; reflect on how those variations may affect intervention feasibility, appropriateness, and effectiveness; and take these factors into consideration when designing training and intervention.

Almost all studies in the present review conducted caregiver training and/or implementation at least partially in the home setting, and most conducted at least some sessions in-person. Training caregivers to implement treatments in the natural environment affords numerous benefits, such as the inclusion of relevant discriminative stimuli, opportunities to train for generalization across stimuli and settings, and the opportunity to engage additional family members (Allen & Cowan, 2008; Weiss, 2001). However, there is a growing body of evidence that indicates telehealth caregiver training is effective, acceptable, and confers benefits similar to in-person intervention (e.g., access to natural environmental contingencies, behavior-change agents, settings, and stimuli (e.g., Shawler et al., 2023) as well as benefits unique to telehealth (e.g., providing services to families in rural and remote locations, improving service efficiency and cost-effectiveness; Ferguson et al., 2019; Wacker et al., 2013; Unholz-Bowden et al., 2020). Clinicians for whom in-person or in-home training is not feasible should identify strategies to address the loss of the benefits associated with training in the natural environment, such as incorporating multiple exemplar or general case training (e.g., Treszl et al., 2022) and incorporating salient stimuli in the training setting that can be present in the nontraining setting (e.g., providing checklists of intervention steps; Allen & Warzak, 2000). Given the established evidence for in-person intervention along with increased evidence for the use of telehealth, clinicians should continue using a combination of in-person and virtual modalities for caregiver training and future researchers should continue to explore potential variables that affect the effectiveness and acceptability of in-person versus telehealth caregiver trainings. Related to this, caregiver trainers had a range of experience and credentials, but most were graduate students, or masters or doctoral level clinicians with multiple years’ experience in the field. If planning to provide caregiver training, clinicians should consider the experience of the caregiver trainers; the present study presents little evidence for the effectiveness of training provided by individuals who are new to the field and/or have not completed graduate coursework in behavior analysis or related fields.

About half of studies reported using BST and/or all four components of BST for caregiver training. About one third reported using some combination of instructions, modeling, rehearsal, and feedback, but not all four. As noted earlier, BST was selected as the framework for coding caregiver training components given the evidence supporting its effectiveness in mediator training broadly, and in particular, with caregivers of children on the autism spectrum (e.g., Erhard et al., 2019; Schaefer & Andzik, 2021).We coded training components that included various forms of instructions, modeling, rehearsal, and feedback subsumed under these broad categories, but other recent reviews have reported more specific details of the strategies used (e.g., Germansky et al., 2020; Pellecchia et al., 2023; Unholz-Bowden et al., 2020). Related to this, some studies have included additional training and coaching components, such as collaborating with caregivers to select intervention goals and techniques, integrating training into daily routines or activities, providing caregiver homework assignments, and engaging caregivers in goal-setting and self-assessment or reflection on their performance (Pellecchia et al., 2023; Unholz-Bowden et al., 2020). Clinicians should continue to provide comprehensive training following the BST model when possible, but should be aware that positive outcomes can be obtained even when not all components of BST are used.

Results from the present review indicate that caregivers were most frequently taught to use strategies such as reinforcement, prompting, manipulation of motivating operations, and extinction. Punishment was implemented less frequently than positive reinforcement. It is important to note that all studies that used extinction or punishment also used reinforcement and other skill development and/or behavior-mitigation strategies. These results suggest that the interventions caregivers are trained to implement are consistent with the professional standards of the field that require the use of reinforcement and teaching strategies in advance of and/or alongside more intrusive approaches (i.e., BACB, 2020). Given the present results and ethical requirements, practicing behavior analysts should continue to design caregiver interventions that focus on preventing the occurrence of child challenging behaviors (e.g., manipulating motivating operations) while supporting child skill development (e.g., through reinforcement and prompting). If additional strategies are warranted (e.g., extinction, punishment) these should be combined with an intervention that includes positive reinforcement.

Although some studies did present data on generalization and maintenance, the percentage was relatively low; 50% of studies that reported caregiver outcomes as a dependent variable evaluated generalization and maintenance, and approximately 20% of studies that reported child outcomes as a dependent variable conducted these evaluations (similar limitations were noted in prior reviews; e.g., Gerow, Hagan-Burke et al., 2018a). Given the limited amount of follow-up data, the durability of caregiver behavior change and child outcomes is relatively unknown. To support meaningful outcomes, clinicians should continue to incorporate strategies implicated in included studies to enhance child and caregiver skill generalization and maintenance (e.g., demand fading, stimulus control manipulation, and training in natural settings). Additional strategies could be incorporated as well; for example, Allen and Warzak (2000) recommend that clinicians train caregivers to apply strategies to a broad array of target behaviors (i.e., train sufficient exemplars); arrange a variety of training conditions with a variety of people, including those involved in everyday conditions; and include relevant stimuli that can be present in training and nontraining settings.

Finally, about half of studies included social validity measures. Of these studies, almost all reported that caregivers were generally happy with the intervention, consistent with results reported in other caregiver training reviews (e.g., Gerow, Hagan-Burke et al., 2018a; Kemmerer et al., 2023). These findings are encouraging, and clinicians should continue to provide high quality caregiver training and take social validity data on their mediator training and interventions, including data specific to the goals, procedures, and outcomes of caregiver training and intervention. Further, given the increasing awareness of the impact of process characteristics on treatment outcomes (e.g., therapist characteristics including warmth, empathy, acceptance, and compassion; Chadwell et al., 2018; Taylor et al., 2019) and emphasis on collaborative approaches to determining treatment goals and intervention strategies (e.g., Jimenez-Gomez & Beaulieu, 2022), practitioners should continue to collaborate with caregivers to develop goals for treatment, determine appropriate intervention strategies, and evaluate the outcomes of caregiver training and caregiver-implemented interventions (Jimenez-Gomez & Beaulieu, 2022; Kemmerer et al., 2023; O’Neill et al., 2023).

Limitations and Future Research Directions

The present review has several limitations. As with any scoping review, some relevant studies may have been excluded if there was not enough information to indicate that the study met the inclusion criteria. Related to this, only studies that contained direct data on caregiver treatment integrity were included. Although direct data increases our confidence in the feasibility of caregiver implementation of these interventions, this stringent criterion also resulted in the exclusion of important research related to caregiver-implemented interventions for issues such as food refusal (e.g., Ahearn et al., 1996; Najdowski et al., 2003) and sleep challenges (e.g., McLay et al., 2019a, 2019b), as well as studies involving approaches such as mindfulness (e.g., Singh et al., 2006, 2007) and telehealth (e.g., Wacker et al., 2013; Suess et al., 2016), among others. In addition, seminal caregiver training research, including articles exploring the impact of support provided within the context of family-centered early intervention (e.g., Dunlap & Fox, 1999) and functional communication training (e.g., Wacker et al., 2005, 2008), were excluded for the same reason. Because the purpose of any structured review is to synthesize the best available evidence, the exclusion of these studies is a significant limitation and results should be interpreted accordingly. Caregiver training strategies were broadly defined and categorized in the present review. Although we believe that the definitions accurately capture relevant training strategies, future research may benefit from a more precise analysis of training procedures, such as further exploring the training components used (e.g., timing and mode of instruction delivery, type of model presented, whether training was provided with the child present). Related to this, the present study used the components of BST as a framework for review. Given the breadth and diversity of training and coaching models present in the caregiver training literature, this may have limited the scope of analysis in the present study. Future research should explore additional models of caregiver training and coaching––individually and in combination––in addition to BST. Finally, the present scoping review did not systematically explore factors associated with variations in caregiver treatment integrity (e.g., method of caregiver training, caregiver-implemented interventions, topography and/or function of challenging behavior). For this reason, we cannot draw any firm conclusions about how these factors affect caregiver treatment integrity and child outcomes. Future researchers are encouraged to extend the present review by conducting a formal systematic review or meta-analysis.

Conclusions and Implications for Practice

The purpose of this study was to provide a comprehensive review of the literature with respect to SCDs that directly evaluate caregiver training and treatment integrity in ABA-based interventions to reduce challenging behavior displayed by autistic children. Caregivers were trained to address an array of common topographies of challenging behavior (e.g., disruptive behavior, aggression, property destruction, self-injury, stereotypy) and empirically supported approaches to caregiver training, such as BST or components of BST, were often employed. The results of the present review are promising, because they indicate that caregivers can implement many behavior change strategies, often in various combinations, with high treatment integrity (i.e., 80% or greater on average). The results also highlight some important considerations for practitioners when designing caregiver training and caregiver-mediated interventions. When possible, clinicians should involve caregivers in the functional behavior assessment process using indirect, direct, and/or experimental approaches as appropriate. Clinicians should provide caregiver training in natural settings, in-person or via telehealth technology, to promote meaningful and sustained outcomes for both caregivers and their children. To support caregiver skill acquisition, clinicians should employ caregiver training approaches with empirical support, such as BST. Interventions should emphasize the use of prevention strategies, skill development approaches, and reinforcement.

Appendix

PsycINFO Search Strategy

((Index Terms: (Autism Spectrum Disorders)) OR(Any Field: (autis*) OR Any Field: (ASD) OR Any Field: (PDD*) OR Any Field: ("pervasive developmental disorder*"))) AND ((Index Terms: (Parent Training)) OR (Any Field: ("parent training") OR Any Field: ("parent-training")) OR(Any Field: (parent*) OR Any Field: (caregiver*) OR Any Field: (guardian*) OR Any Field: (mother*) OR Any Field: (father*)) AND (Any Field: (train*) OR Any Field: (guid*) OR Any Field: (mediat*) ORAny Field: (implement*)))) AND ((IndexTerms:("Behavior Problems") OR IndexTerms:("Self-Injurious Behavior") OR IndexTerms:("Self-Stimulation") OR IndexTerms:("Stereotyped Behavior") OR IndexTerms:(Tantrums)) OR (AnyField:(behav*) AND (AnyField:(problem*) OR AnyField:(challeng*) OR AnyField:(interfering) OR AnyField:(disrupt*))) OR (AnyField:(aggress*) OR AnyField:(hit*) OR AnyField:(kick*) OR AnyField:(punch*) OR AnyField:(throw*) OR AnyField:(swing*) OR AnyField:(slap*) OR AnyField:(bit*) OR AnyField:(yell*) OR AnyField:(scream*) OR AnyField:(swear*) OR AnyField:(curs*) OR AnyField:(threat*) OR AnyField:(abus*) OR AnyField:(butt*) OR AnyField:(hairpull*) OR AnyField:("hair-pull*") OR AnyField:("hair pull*") OR AnyField:(spit*) OR AnyField:(push*) OR AnyField:(bang*) OR AnyField:(pinch*) OR AnyField:(squeez*)) OR (AnyField:(disrob*) OR AnyField:(naked) OR AnyField:(bare) OR AnyField:(nud*)) OR (AnyField:(cloth*) AND (AnyField:(remov*) OR AnyField:(strip*))) OR (AnyField:(public) AND (AnyField:(nud*) OR AnyField:(naked) OR AnyField:(remov*) OR AnyField:(indecen*))) OR (AnyField:(elop*) OR AnyField:(run*) OR AnyField:(escap*) OR AnyField:(leav*)) OR (AnyField:("inappropriate sexual behav*")) OR (AnyField:(inappropriate*) AND (AnyField:(touch*) OR AnyField:(hold*) OR AnyField:(hug*) OR AnyField:(kiss*) OR AnyField:(sex*) OR AnyField:(intercourse) OR AnyField:(grab*) OR AnyField:(display*))) OR (AnyField:(public*) AND (AnyField:(masturbat*) OR AnyField:(touch*) OR AnyField:(fondl*) OR AnyField:(sex*) OR AnyField:(intercourse) OR AnyField:(display*))) OR (AnyField:(touch*) AND (AnyField:(other*) OR AnyField:(genital*) OR AnyField:(self*) OR AnyField:(penis*) OR AnyField:(vagina*) OR AnyField:(breast*) OR AnyField:(bum*) OR AnyField:(butt*))) OR (AnyField:(mouth*) OR AnyField:(swallow*) OR AnyField:(ingest*)) AND (AnyField:(object*) OR AnyField:(inedible) OR AnyField:("non-edible") OR AnyField:("non edible")) OR (AnyField:(saliva) AND (AnyField:(play*) OR AnyField:(smear*) OR AnyField:(touch*) OR AnyField:(spit*))) OR (AnyField:(pica)) OR (AnyField:("noncompliance") OR AnyField:("non compliance") OR AnyField:(noncompliance) OR AnyField:(disobe*) OR AnyField:(defy) OR AnyField:(defi*) OR AnyField:(refus*) OR AnyField:(protest*) OR AnyField:(ignor*) OR AnyField:(whin*) OR AnyField:(turn*) OR AnyField:(drop*) OR AnyField:(avoid*)) OR AnyField:("destruct*) OR (AnyField:(property) AND (AnyField:(destruct*) OR AnyField:(damag*))) OR (AnyField:("self-injur*") OR AnyField:("self injur*") OR AnyField:("self harm*") OR AnyField:("self-harm*") OR AnyField:(vomit*) OR AnyField:(ruminat*)) OR (AnyField:(bit*) AND (AnyField:(finger) OR AnyField:(hand) OR AnyField:(self) OR AnyField:(arm) OR AnyField:(body))) OR (AnyField:(head) AND (AnyField:(hit*) OR AnyField:(bang*) OR AnyField:(smash*) OR AnyField:(butt*) OR AnyField:(strik*) OR AnyField:(slap*))) OR (AnyField:(eye*) AND (AnyField:(goug*) OR AnyField:(pok*) OR AnyField:(scratch*))) OR (AnyField:(skin) AND (AnyField:(cut*) OR AnyField:(burn*) OR AnyField:(pick*) OR AnyField:(scratch*) OR AnyField:(pinch*) OR AnyField:(pull*) OR AnyField:(slap*))) OR (AnyField:(hair) AND (AnyField:(pull*) OR AnyField:(rip*))) OR (AnyField:(smear*)) OR (AnyField:(fec*) AND (AnyField:(smear*) OR AnyField:(play*) OR AnyField:(throw*))) OR (AnyField:(stereotyp*) OR AnyField:("self-stim*") OR AnyField:("self stim*")) OR (AnyField:(stereotyp*) AND (AnyField:(behav*) OR AnyField:(vocal*) OR AnyField:(movement*) OR AnyField:(acti*) OR AnyField:(play*) OR AnyField:(engag*)) OR AnyField:(tap*) OR AnyField:(jump*) OR AnyField:(rock*) OR AnyField:(hum*) OR AnyField:(tens*) OR AnyField:(spin*) OR AnyField:(insist*) OR AnyField:(chang*) OR AnyField:(flap*) OR AnyField:(script*) OR AnyField:(repeat*) OR AnyField:(repetitive) OR AnyField:(plac*) OR AnyField:(line) OR AnyField:(lining) OR AnyField:(touch*)) OR (AnyField:(tantrum*) OR AnyField:(disrupt*) OR AnyField:(cry*) OR AnyField:(scream*) OR AnyField:(yell*) OR AnyField:(stomp*) OR AnyField:(flop*) OR AnyField:(drop*) OR AnyField:(swip*) OR AnyField:(knock*) OR AnyField:(break*) OR AnyField:(tear*) OR AnyField:(shout*) OR AnyField:(pull*)))

Funding

Council for Research in the Social Sciences, Brock University, Faculty of Social Sciences

Data Availability

The data that support the findings of this study are available from the corresponding author.

Declarations

Conflicts of Interest

The authors have no conflicts of interest to declare.

Ethics Approval

No ethical approval was required for this study.