Assessing change in early intervention programs for children with autism.

Kasari (2002) read every big early-intervention study for preschoolers with autism. She did not run new kids; she told a story about what was missing.

She asked one question: Do we really know which parts of these thick, pricey programs make kids better?

The answer was no. Most papers treated the whole package as a black box. They counted IQ or language at the end but never tracked which teaching move caused the jump.

Without that link, we can’t tell parents why therapy works or trim the parts that don’t.

Egli et al. (2002) looked at the same pile of studies the same year. They used a stricter checklist and reached the same bleak view: most tools were weak and no one measured the middle steps.

Matson (2007) picked up the baton five years later. The message got sharper: pick the wrong ruler and even a so-so program can look like a cure.

Eckes et al. (2023) finally crunched the numbers Connie wanted. Eleven solid trials showed medium gains in IQ and daily skills, but language and parent stress stayed flat. The box opened a crack: ABA works for some things, not everything.

Next time you praise a comprehensive program, pause. Ask which slice you can credit: discrete trials, peer play, parent coaching? Track that slice with a quick daily probe. If it climbs, keep it; if not, swap it out. Connie’s plea is now your Monday-morning scalpel.