Physical activity is crucial for the health and wellbeing of all individuals, yet most children with ASD (representing 1 in 66 children in Canada; Ofner, 2018) do not meet the physical activity guideline of 60 minutes of moderate-to-vigorous daily physical activity to accrue these health and developmental benefits (Liang et al., 2020; McCoy et al., 2016; Pan et al., 2015). Factors contributing to this reduced physical activity participation include parental safety concerns, differences in how children with ASD interact and communicate their feelings, and an overall lack of understanding of ASD in the community (Ayvazoglu et al., 2015, Simpson et al., 2021). Children with ASD may use different body language and communication cues than their neurotypical (NT) peers, leading to communication barriers between them and their NT peers and instructors. This may result in inadvertent exclusion from physical activity opportunities (Ayvazoglu et al., 2015; Kinnear et al., 2016). The double empathy problem, a term coined by Milton (2012) as applied to ASD, suggests that two individuals with ASD or two NT individuals will understand each other better than an NT individual and one with ASD. Given these expressive and interpretive differences between individuals with ASD and NT individuals, it is crucial that instructors of physical activity programs become aware of the meaning of the behaviors and body language cues that children with ASD exhibit during physical activity to help better communicate and support their participation (Welch et al., 2021). To illustrate this, Appendix A provides an example of a fictional boy with ASD named Jason. It is important to emphasize that this story does not reflect all children with ASD, rather, it is a composite of anecdotal data the first author acquired through conversations with parents of children with ASD during the tool development process.
At its most rudimentary, body language is defined in dictionary.com (2021) as gestures and facial expressions that do not involve verbal cues, but that include non-verbal aspects of speech (e.g., accent, tone of voice, and speed of speaking). Behavioral studies have explored how one can characterize another person’s emotions by decoding their facial expressions alone (Smith et al., 2005). However, research suggests that facial expressions and body cues should be interpreted together to get the best picture of an emotion without words (Martinez et al., 2016). Thus, in our body language measurement research project, body language we defined more specifically as the communication of emotions expressed through the use of body cues (posture, speed of movements and proximity to the instructor), facial expressions (smiling, and eye gaze), and tone of voice in verbal responses and vocalizations (flat versus excited tones, speed and volume of speaking; Aviezer et al., 2012; Conway et al., 2023 De Gelder & Vroomen, 2000; Martinez et al., 2016; Meeren et al., 2005; Van den Stock et al., 2007; Wurm et al., 2001).
An initial literature review of training and assessment tools that describe and identify the body language of children with ASD led us to conclude that such a measure does not exist. For example, the Sensory Processing Assessment evaluates the behavioral responses of children with ASD when interacting with novel sensory toys (Kirby et al., 2015), but it does not quantify or identify specific body language cues. The Repetitive and Restricted Behaviour Scale (Bourreau et al., 2009) and the Autism Treatment Evaluation Checklist (Magiati et al., 2011) focus primarily on the stereotypic repetitive behaviors exhibited by children with ASD, based on self- or parent/caregiver-report (Bodfish et al., 2000; Silva & Schalock, 2012), but does not capture the engagement aspect of body language and is not an observational measure. The Emotion Understanding Assessment (Barbosa-Leiker et al., 2014), the Affect Knowledge Test (Sette et al., 2015) and the EmoBio (Egeland, et al., 2019) asks individuals with emotion differences (e.g., ASD, schizophrenia) to score the emotions of NT individuals but do not identify the differences in body language between the two populations, nor are they designed to allow NT individuals to interpret the body language of individuals with the emotion differences.
The absence of published measures or training tools that capture the diverse aspects of body language or identify specific behaviors of individuals with ASD through body language is problematic as such a tool could improve the inclusivity, enjoyment, and engagement of children with ASD in physical activity. In sports, the positive relationship between use of non-verbal communication and team wins (Durdubas et al., 2018) and athlete satisfaction (Sullivan & Gee, 2007) demonstrates the importance of accurate body language interpretation between players and coaches. It is therefore hypothesized that awareness of the body language communication differences of children with ASD by physical activity instructors may be an important aspect of understanding the breadth of a child’s reactions and interactions with their environment. Thus, learning about these behaviors could be the difference between responding appropriately or inadvertently reducing a child’s motivation, confidence, and participation. In thinking about the underlying expression of emotions, basic emotions (i.e., the intrinsic, evolutionary, and universal emotions such as, happiness, surprise, fear, sadness, anger and disgust/contempt; Ekman, 1992) are more accurately recognized than non-basic emotions (i.e., the socially developed and experience derived emotions that include all other emotions including frustration, boredom, and engagement; Ekman, 1992; Fridenson-Hayo et al., 2016; Laukka et al., 2016; Saarimäki et al., 2018) in both NT individuals and those with ASD. It is hypothesized in the current paper that one key aspect of body language that may be different in children with ASD is the display of non-basic emotions. This is particularly salient when we think about the context of physical activity, the motor challenges and related self-efficacy issues that these children often face. For example, the occurrence of associated reactions (i.e., disruptive behaviours) are likely to happen in the case of misunderstanding of more complex non-basic emotions.
In response to putative body language awareness and recognition issues and their potential consequences, a tool for Exploring the EXPRESSions of Autism through Body Language (EXPRESS) was developed by [AUTHOR INITIALS] for physical activity instructors working with children with ASD to: (a) sensitize and educate instructors about body language that might be seen in children with ASD (EXPRESS-Train); and (b) provide them with a way to quantify and interpret the body language of children with ASD in their program (EXPRESS-Code). Specifically, the video-based EXPRESS-Train was designed as a body language awareness building tool that would assist instructors with supporting and engaging children with ASD in their physical activity programs. By referring to the EXPRESS-Train materials, an instructor might be able to initiate and structure conversations with parents about their child’s body language at the outset of a program. The EXPRESS-Train has an optional module as its final section that is the instructional guide for learning to use the EXPRESS-Code. It also comes with a reference manual (Appendix B) that trained users can refer to while scoring. This coding tool allows an instructor to systematically score and then interpret a child’s body language expressions from a video that is created during a child’s performance of a 30-minute advanced motor skills test known as the Ignite Challenge (Wright et al., 2023). Use of the Ignite Challenge (Wright et al., 2023) provides a standardized context within which to observe and score a child’s body language during physical activity. While this more time-intensive EXPRESS-Code was designed for research purposes, it may also be viable for use by instructors as a pre-program onboarding assessment with children with ASD.
As the basis for future use of the EXPRESS-Train and EXPRESS-Code, it was necessary first to establish the reliability of scoring body language expression via the EXPRESS-Code. It is essential to be sure that the descriptive names and their respective categories for each of the body language items measured are clear enough to be observed accurately and interpreted correctly by different instructors. This paper provides the results of the content validity and reliability evaluation of the EXPRESS-Code.
The first version of the EXPRESS-Code (unpublished) consisted of 19 body language expression items split into three categories according to their anticipated behavioral valence: positive (10 items), negative (7 items), and neutral (no perceived associated positive or negative meaning; 2 items). These arbitrary categorizations were a useful place to start as a structural framework to help create a logical flow of item order. However, exploration of the content validity of these categories in the interpretation of body language of children with ASD was identified for further study in the next phase of the tool’s refinement following its initial evaluation of performance. Our initial study with this first version of the EXPRESS-Code (REB# 210) provided evidence of excellent intra-rater reliability but only moderate inter-rater reliability when it was used to assess the body language of 28 children with ASD as scored from videos of their Ignite Challenge performance. Based on these results and the feedback from the study’s four physiotherapy student raters, revisions were made to the EXPRESS-Code’s item structure and coding sheet and to the accompanying EXPRESS-Train manual with a main goal of clarifying scoring criteria and enhancing the training materials to bring inter-rater reliability up to the level of excellent.
The updated version of the EXPRESS-Code (described below in the Methods Instrumentation section) consists of 14 items split into two interpretive ‘valence’ categories: positive cues (7 items) and negative cues (7 items). The main objectives of this study were to: (1) assess the content validity of the EXPRESS-Code items to verify if each item was correctly placed within its respective body language valence category and identified with an accurate cue description (i.e., cue name); (2a) evaluate the intra- and inter-rater reliability of these category scores, and (2b) evaluate the intra- and inter-rater reliability of the individual item scores.