What does it mean for an autism intervention to be evidence-based?

Vivanti (2022) wrote a think-piece, not an experiment.

He traced how the words “evidence-based” became popular in autism work.

He looked at old policy papers, meta-analyses, and the fights researchers still have over the label.

The phrase “evidence-based” has no single meaning.

One team may require five single-case studies.

Another may demand two big RCTs.

These hidden rules decide which programs get funded, published, or called “best practice.”

Mammarella et al. (2022) counted 55 school FBA studies.

Only seven checked if the plan still worked in a real classroom.

Giacomo’s point: poor ecological validity lets weak studies claim the “evidence-based” badge.

Lord et al. (2005) already begged for more RCTs and shared outcome measures.

Vivanti (2022) agrees, but adds that even perfect RCTs can ignore culture, choice, and daily fit.

So the 2005 plea is still right, yet no longer enough.

When you pick an “evidence-based” package, ask two questions.

Did researchers test it in a setting that looks like yours?

Did they define “work” the same way the family does?

If either answer is no, keep looking or adapt the plan before you sign the IEP.