And at MIT: @kosakowski.bsky.social, @fkamps.bsky.social, @halieolson.bsky.social, @emilychen.bsky.social, @bmhdeen.bsky.social, Haoyu Du, Camille Osumah, and @rebeccasaxe.bsky.social

Thank you also to the wonderful families and babies who make this work possible!

This exciting collaboration was only possible because of decades of effort from an incredible team of researchers both at Yale: @tristansyates.bsky.social, @jetrach.bsky.social, @camerontellis.bsky.social , and Nick Turk-Browne

We found that movie-based experiments had higher success rates than block or event-related designs, and stimuli featuring social content and faces were more successful than those that did not.

Based on these and other findings, we outline recommendations for future infant fMRI study designs!

Despite substantial methodological differences across labs, we retain a similar amount of usable data per session (about 9 minutes). And while older infant scans can be harder to get started, sessions with older infants, up to ~24 months, ultimately produce the most usable data.

Studies targeting domains like object recognition or social cognition often require infants to encode and later recall complex information. We therefore argue that these tasks tap into episodic-like memory and that evidence for infant memory is more prevalent than previously recognized!

In this fun collaboration with @levelsof.bsky.social and Nick Turk-Browne, we first review the classic tasks used to test infant memory -- but we don’t stop there. We also highlight tasks from other cognitive domains that may place hidden demands on episodic-like memory.