Assessing and Improving Early Social Engagement in Infants.

Waterhouse et al. (2014) wrote a how-to guide for scientists who watch babies grow.

They listed the traps that ruin infant-sibling studies: small samples, shaky measures, no plan for sick kids.

The paper tells teams to pick a comparison group early and to have a clinic ready before the first baby walks in.

The team did not run a new experiment.

Instead they showed that most past studies broke basic rules, so their results may be false alarms.

Clear rules—like using the same coder for every video—can fix the mess.

Vanvuchelen et al. (2011) proved a five-minute imitation test spots autism in preschoolers. Lynn says add that test to infant visits so you have a cheap, solid marker from day one.

Uljarević et al. (2017) linked late sitting and toe-walking to worse repetitive behaviors later. Lynn urged teams to log motor milestones; Mirko shows why that data matters.

Yu et al. (2022) re-checked antibiotic studies and found risks vanish when siblings are compared. Lynn warned that failing to control family genetics can fake risk factors—Hai-Ying’s numbers back her up.

If you screen babies who have an older sibling with autism, copy Lynn’s checklist. Build a control group of low-risk infants, film every visit, and decide now which clinic gets the red-flag cases. These steps turn a shaky pilot into solid data you can trust—and help families faster.