Overview of Issues in Autism Intervention Research: Research Design and Reporting

Abstract

Purpose: Neurodiversity-affirming practices for people with autism need to be backed by rigorous evidence of effectiveness. We discuss aspects of research design and reporting in autism intervention research that should be considered by intervention providers when providing evidence-based practices to support people with autism. We describe risks of bias, which are features of the study design that make it more likely that ineffective interventions will show evidence of effectiveness; social validation, which are procedures that determine community members' views on the intervention goals, procedures, and outcomes; the design of intervention outcome variables, including the extent to which they index lasting, meaningful change; adverse event monitoring and reporting, which is the extent to which researchers track and transparently communicate negative events that occur during or after participation in the intervention; and conflicts of interest (COIs), which are relationships held by the researcher that could result in positive researcher benefits for showing particular outcomes in their studies. We discuss several systematic reviews and evaluations of autism intervention literature that have explored the extent to which researchers have adhered to ethical guidelines around each of these issues. Conclusions: Risks of bias are prevalent in this literature, social validation is rarely rigorously conducted, and positive intervention outcomes tend to index narrow, specific change. Additionally, researchers rarely monitor or report adverse events despite evidence that they occur, and they often fail to disclose COIs or falsely claim not to have them. Clinicians and researchers should be aware of these features of autism intervention studies, so that they can make informed decisions as new, and hopefully improved, research is produced.