❗New Paper❗Is children's attention more like a spotlight that darts across time, or one that diffuses across many things at once? How might children's immature attention help their learning? Our Dev Sci Paper has answers! 🧵🎯 pubmed.ncbi.nlm.nih.gov/41549519/

We found that kids' immature selective attention allowed them to learn more—even stuff we explicitly told them to ignore! Their "leaky" attention mediated better memory for task-irrelevant information

But kids' broad attention didn't always"diffuse" over everything. Instead, it appeared to "dart" between relevant and irrelevant information across time - narrowing in on one one or the other in each moment

Does darting attention have adaptive benefits over diffusing? Hypothetically it would allow children to retain a narrowly focused attentional scope in each moment, regardless of the content of their focus. Darting could lead to really good and broad learning when averaged across many events.

We also found kids were less likely than adults to learn associations between relevant and irrelevant info—their associative systems "stick" the stuff they encounter together less... keeping things more open or disparate..

This darting pattern makes me think about explore-exploit patterns across development. Higher exploration (darting) early in training/development enabling broader learning

This also made me wonder about neural network models of attention and learning - would attention 'darting' early in training (rapid sequential sampling across features) produce better learning than 'diffuse' attention? Is this immature attention a feature, not a bug 🤔?