Keywords

1 Introduction

Autism spectrum disorder (ASD) is a neurodevelopmental diagnosis defined by persistent deficits in social interaction and communication and by restricted, and repetitive patterns of behavior, interests, or activities [1]. An overall prevalence of ASD estimate it to be around 1% [2]. Psychosocial deficits and challenges are common in ASD. Various aspects such as less friendships with peers [3], and increased loneliness and isolation in their teens [4] have been reported by adolescents with ASD. Young people with ASD display more school refusal behavior compared to others [5], and children report having experienced anxiety in various settings [6] in addition to bullying [7].

Enhancing social skills is a key target for interventions for this group [8], and there has been a durable trend in empirical research on social skills interventions in ASD [9]. Virtual Reality (VR) as a tool in research on ASD has gained more attention in recent years. In fact, the number of published research articles on using Virtual Reality or other computer-based assessments and interventions on ASD in 2018 alone exceeded all publications produced in the consecutive years of 2001–2011 (see Fig. 1).

Fig. 1.
figure 1

Number of peer-reviewed publications on ASD and the use of VR or other computer-based tools, each year since 2001–2018.

Full size image

The majority of interventions using VR aims at improving social or emotional skills [10, 11], and using VR show promising treatment effects [12]. Several researchers describe VR as having advantages over other forms of tools, such as combining ecological validity provided by real-world simulation [11], with a less stressful environment [13].

There are various forms of VR equipment that enable different levels of immersions [12]. For instance, Ip et al. [14] used a four-side CAVE VR system, in which big screens and projectors halfway surround the participant. Despite some promising treatment effects reported [14, 15], implementation of CAVE VR systems in ASD interventions suffer from practical requirements/challenges related to space, equipment, and its immobility [16].

The highest level of immersion provided by mobile equipment is realized in VR Head-Mounted Displays (HMD), where the goggles provide a sense of a virtual environment completely surrounding the user. An alternative to VR HMD is glasses with Augmented Reality (AR) technology. AR is an enhanced version of reality created by the use of technology to overlay digital information on an image of something being viewed through a camera device [17], and AR-glasses differ from VR HMD as visual information is superimposed on the actual environment [18]. Henceforth, we include both VR and AR-glasses when referring to the abbreviation HMD. Other computer-based interventions use computers, laptops, tablets, mobile phones, Kinect®, or other technology such as for example social robots [19].

In regard to the usability and acceptability of HMDs, previous research suggests that sensory oversensitivity among participants with ASD may cause irritation and thus low acceptability [20,21,22]. This criticism might lead to some reluctance towards the use of HMDs. However, there is little empirical research that support these claims and the references used in the aforementioned studies were published before 2004 e.g., [23, 24] and do not account for recent developments in usability, comfort and improved motion sickness prevention. Findings from a more recent study indicate that in tests for acceptability and sensitivity of participants using HMDs, participants express both acceptance and enjoyment of HMDs [25]. These results where further corroborated in a literature review by Bozgeyikli et al. [26], which also suggested guidelines in design of VR as a training tool for participants with ASD. (We argue that) the question of acceptability is of central importance for the future development of AR/VR-supported intervention research in ASD. Lacking acceptability of an intervention method can affect common factors central for intervention effects [27]. Participants’ outcome expectations, collaboration, affirmation and motivational factors contributing to a sustainable treatment alliance are known to be central to increase effects of psychological interventions [27]. Although the usability and acceptability of HMD is reported in a number of studies [15, 25], a systematization of data across studies reporting acceptability (i.e., participant reports on usability, enjoyment, likeability, tolerability and so on) amongst participants with ASD is lacking. This chapter addresses the gap in knowledge to provide a more contemporary and comprehensive conclusion regarding the acceptability of HMD-supported AR/VR interventions in ASD.

All interventions represent an active and targeted manipulation of the individual’s cognitive and emotional status with the goal to reach sustainable change in how one perceives, interpret, experiences and reacts to the environment. These intended psychosocial changes are to be reached via empirically validated methods. As all intervention can have unintended side effects, which also can apply to VR-interventions, it is important for researchers, clinicians, and practitioners to be aware and account for these. The use of VR and computer-based tools has undergone numerous iterations during its relatively short history and there is a need to maintain a focus on social validity when developing and using new technology.

In this chapter, we provide a number of normative considerations of how the VR interventions can relate to the non-epistemic values of the relevant party (i.e., considerations of how the participant experiences the purpose of and the VR intervention itself). Using these considerations, and report these judgements and decisions in research should contribute to the focus on social validity in research, and ensures that individual consideration is always accounted for.

2 Method

2.1 Literature Search and Screening

We searched PsycInfo, Pubmed, ERIC, Education Research Complete, Web of Science, and IEEE Xplore, between August 23rd 2019 and August 27th 2019, using the thoroughly selected Boolean operators for the search strings: (Pervasive development disorder OR pdd OR pdd-nos OR pervasive developmental disorder not otherwise specified OR autism OR autistic OR Autism Spectrum Disorder OR autism spectrum disorders OR Asperger OR asd OR autism spectrum condition* OR asc) AND (Virtual Reality OR vr OR hmd OR Head-mounted display OR Immersive Virtual Environment OR Augmented Reality OR Artificial Reality OR Oculus OR Immersive Technolog* OR Mixed Reality OR Hybrid Reality OR Immersive Virtual Reality System OR 3D Environment* OR htc vive OR cave OR Virtual Reality Exposure OR vre).

After removing duplicates, we manually screened the results according to pre-defined inclusion criteria: The articles are peer-reviewed and written in English, include participants with ASD, and include the use of virtual reality/environment equipment. We excluded meta-analyses, reviews, conceptual articles, and design articles that were not original research articles providing new empirical data.

2.2 Coding Acceptability Data

In order to discover acceptability data in all the included publications, we searched for the following terms in each publication: accept*, question*, usability*, evaluat*, ask*, enjoy*, valid*, and tolerab*. As the publications does not use a general standard of assessing and reporting acceptability data, we extracted qualitative statements from participants, data provided by questionnaires or other forms for reporting acceptability. If the articles included acceptability data, we coded the data as either; 0 = negative sentiment towards the VR, 1 = inconclusive, and 2 = positive sentiment towards VR. If different participants within a study reported acceptability data that contradicts each other, the data would be coded as inconclusive.

3 Results

The initial screening and removing of duplicates resulted in 226 publications. Figure 2 (flow chart) shows that the number of publications we included according to our inclusion and exclusion criteria where 155. Sixty-three of these publications included acceptability data.

Fig. 2.
figure 2

Flow chart of the search and screening process.

Full size image

A descriptive analysis of the data shows that 63 (40.6%) publications reported data on the acceptability from participants with ASD. See Table 1 for the summary of the evaluation from the participants. Out of the total number of publications, 22 (14.2%) used HMD or AR-glasses, and the remaining used other types of computer-based apparatus (e.g., CAVE, laptop, tablet etc.). Sixteen (72.7%) of the 22 publications using HMD reported acceptability data, three of which was inconclusive and the rest was positive. An independent coder that was unfamiliar with the project coded a randomly selected sample of 16 publications (10.3%) to check for inter-rater reliability (IRR) with the first author. The IRR agreement in this sample was 100%.

Table 1. The table shows descriptive statistics on the number of publications (n) reporting acceptability data. The lower section in the table shows the similar data in the publications using HMDs. The numbers inside the parenthesis is the percent of n its respective unit.
Full size table

4 Discussion

The aim of this research was to provide an indication of whether VR and HMDs are accepted amongst participants with ASD across studies. Our findings indicates that VR interventions are welcomed by participants with ASD as over 80% of the publications reporting evaluation data report positive evaluations. This is in line with suggestions from Newbutt et al. [25]. The majority of the overall studies including both HMD and other computer-based apparatus does not report evaluation data on acceptability. Most of the publications using HMD, however, have reported acceptability data. We encourage researchers to include these kinds of evaluations when doing research and clinical work. As the positive results are shown both across studies and within studies such as Newbutt et al. [25] it seems that there is no evidence suggesting to avoid HMDs in ASD research and clinical practice. It is a vast amount of studies not reporting the evaluation so there might be some unrecorded negative evaluations. However, individual variation will always apply to specific situations and necessary adaptions is needed whenever using VR apparatus in clinical research.

4.1 Future and Normative Considerations

Where the clinical assessment concludes the need for an intervention with the goal of sustainable behavior change, normative considerations are important to determine the appropriateness of the additional burden for the client. Decisions based on evidence-based conclusions regarding the effectiveness and efficacy of an intended intervention should therefore always differentiate between descriptive and normative premises and undergo an additional normative consideration to ensure the participant’s interests.

As shown amongst others by Gillespie-Lynch et al. [28], people with ASD may often be considered experts on autism, and there is an increased and pronounced expectation amongst persons within the autism spectrum to be heard in respect to societal issues, and whether one should change people’s personality, or the construct of society. The increased awareness towards this, emphasize the importance of cooperating with people with ASD in both research and clinical practice.

Normative Considerations According to Løkke and Salthe’s [29] Checklist.

The first step in any clinical interventions is to collect the facts about the participant’s behavior, situation, and somatic and psychological health. Before continuing, one should verify the participant’s values, interests and opinion regarding the topic [29]. Then, after this assessment and before anything else, one should ask the question: is there any reason to do anything? If no, then do not. However, if the answer is yes, one can go on checking evidence, effects and possible side effects of various interventions before collecting data.

It is also important to do normative considerations of the specific intervention, and our paper provides some evidence that VR and HMDs are a tolerable tool in interventions. Even though our results suggest that HMDs and other computer-based tools are accepted amongst the majority of participants with ASD, it is not a justification towards using it against someone’s will. One should principally make sure that the intervention does not contain force/compulsion, reduction of choices, or other unwanted elements for the participant.

In addition to the acceptability towards the intervention method, the normative issues of the possible outcomes of the interventions for the individual is important. According to the philosopher James Griffin [30], one can categorize values that make life worth living into five categories of wishes for our lives: Good experiences, freedom, knowledge, relations, and to have accomplished something. When setting up an interventions, the researcher/clinician should think carefully about if the interventions itself provides a good experience, and if the possible outcome expands the individual’s opportunities later in life for more good experiences. These two considerations may contradict one another; however, the latter may oust the first in many cases. Actually, in regards to important measurements and outcomes, happiness is ranked, by parents of children with ASD, as the most important outcome that researchers should measure [31].

Freedom, to both make one’s own decisions and be able to withstand unwanted decision made unto by others, may also be important. This wish should be considered both prior to and during the interventions, as well as when evaluating outcomes. User involvement is an example which is important prior and during, but one should try as an outcome to give participants enhanced opportunities to increase their own sense of freedom in life. Knowledge in itself is of value, and also the notion of having accomplished something. That might be on a smaller or bigger scale, e.g., from learning how to read or raising happy children. The fifth wish, relations, can often seem to be underrated by the surrounding of people with ASD. Maybe because children with ASD tends to orient themselves more towards non-social stimuli rather than social [32], this might be mistaken to be interpreted as people with ASD not wanting any social relations. A lot of research has refuted this [7]. Again, it is important to emphasize that research does not show that all persons with ASD want social relationships either, just as with people “outside” the autism spectrum.

4.2 Limitations

The analysis of the existing study laid out some systematic knowledge gaps to be addressed by future research, as they are not uncommon in new research fields such as VR interventions and research. There is no coherent way of how to report data on acceptability even though attempts have been made to develop questionnaires for researchers and teachers [33]. The various ways of assessing and reporting evaluation data, may have led to variations and/or errors in our appraisal of the evaluations provided in the different studies. However, the results from IRR suggests agreement and thus reduce the margin of errors. The results would have been strengthened by coding more or all the data with two independent coders.

Many of the studies do not report acceptability data. There are also instances where parents, caregivers, or teachers are asked to evaluate the participants’ acceptance, which may increase the risk of socially desired answers. The fact that participation in studies is based on ethically approved informed consent, may lead to a self-selection effect limiting the generalization of conclusions beyond settings where participants with ASD receive detailed information prior to HMD exposure.

5 Conclusion

This study shows that almost half of the studies on VR and ASD report acceptability data. Of the publications that reported such data, the majority overall and studies using HMD report positive sentiments on acceptability. These data suggest that persons with ASD can readily use computer-based and virtual reality technology, HMDs included. However, over half of the studies we have checked has not reported evaluation data. This indicates that there might be a number of unrecorded cases of negative evaluations. Either way, we emphasize that individual considerations must be made in all cases. Further, we have suggested normative considerations that is eligible to use prior and during interventions in regards to aims and outcome.