...comparisons of common intervention approaches with children who are not yet speaking to test this question and I know for sure we have very few of those. A recently completed dissertation found that only 14/252 studies in our dataset compare theoretically distinct interventions.

And they would need to then do a moderator analysis to show the subgroup of “kids at risk of being nonverbal” (not sure how that group overlaps with all autistic kids who aren’t yet speaking) specifically were more likely to talk after ABA