In The Politics of Autism, I discuss the incentive structure facing academic researchers:
This diversity of research agendas is partially a result of uncertainty. Amid the darkness, it might make sense to shine searchlights in all directions. Some of it may also stem from the availability of autism research money at a time of tight science budgets. To put it bluntly, publication-hungry scientists may have an incentive to rebrand marginally-relevant work as autism-related. Describing her study of how experts on sex differences have landed on the “biomedical platform” of autism, science historian Sarah Richardson says they “have begun to link their very basic research -- even if it’s on nematodes [roundworms] -- to frame it as a contribution to autism.”
As an exercise, I reviewed the content of three journals with the highest impact factors among journals focusing exclusively on ASD research: Autism Research, Journal of Autism and Developmental Disorders, and Autism: International Journal of Research and Practice. According to the ISI Web of Knowledge, the 5-year impact factor for all three journals was above 3.8, and they were in the top 10 Developmental Psychology journals in 2014. From 2010 to 2015, they published a total of 171 treatment studies. To put things into perspective, 2,358 papers were published in these three journals over that 6-year period. This means that about 7% of articles were devoted to treatment studies.[ emphasis added] This 7% sounds less impressive than 171 studies, an absolute number that is far larger than what was published in a 6-year span 25 years ago. Of course, many treatment studies are published in other high-profile journals because of their potential impact.
Why relatively so few treatment studies? Because they are labor-intensive and expensive. Study teams (e.g. therapists, physicians, blind evaluators, research coordinators, parents, and children) take part in repeated visits and assessments. Randomized controlled trials (RCTs) do not mesh well with academic environments that reward speed and quantity. Involved treatment studies are not the untenured scholar’s friend. [emphasis added] For instance, both our recent parent training trial (Bearss et al., 2015) and atomoxetine and parent training study (Handen et al., 2015) took about 6 years to execute. This number approaches 10years when including grant and manuscript writing.