Designing research studies on psychosocial interventions in autism.

Smith et al. (2007) drew a road map. They said autism psychosocial research should move in four clear steps. First, show the idea works in a small lab test. Second, write a manual so others can copy it. Third, run a big randomized trial. Fourth, prove it still works in messy real-world settings.

The paper is a position piece, not an experiment. It gives check-lists and warnings for each step. The goal is to stop studies from jumping straight to big trials before the basics are solid.

There are no numbers in this paper. Instead, the authors show that most autism studies at the time skipped steps. They argue this wastes money and hurts kids. The four-step pipe-line is their fix.

Ouyang et al. (2024) and Patterson et al. (2012) prove the pipe-line works. These reviews pool parent-training studies that followed the stages. They show big child gains and high parent fidelity when steps are done in order.

Schaaf et al. (2015) and John-Bathelt et al. (2019) extend the model. They add real-world tweaks like stress-reduction coaching and a stakeholder checklist. These tools help you move from step three to step four without losing quality.

Ault et al. (2024) show the model is still in use today. Their rural tele-health protocol starts with a tiny feasibility test—exactly step one of the 2007 map.

Before you pilot a new autism program, line up your steps. Run a small proof first, write the manual, then scale. Use the John-Joe checklist to pick a package that already did the early work. This keeps your clinic from wasting time on shiny ideas that crumble in the real world.