Note the posting says 12 months, but we prefer a 2-year commitment. We're looking forward to reviewing applications!

@lisik.bsky.social reports she is experiencing the same thing

Me too! Stay tuned; I'm writing an opinion piece on this as we speak (spoiler: there's no answer yet, but there could be if we allow evidence to build up from multiple smaller-scale studies, rather than dismiss any new brain-behavior prediction finding based on sample size alone)

Oh, also, just to be extra clear, @micahgallen.com and I were referring to using fMRI data from different *in-scanner* tasks as predictor variables, not necessarily using multiple phenotype scores as targets of prediction -- though choosing the right target variables is a whole 🐻 unto itself!

Yeah, it does get complicated, but I think a lot of the Big-Data-brain-phenotype-prediction work these days is more concerned with practical utility than interpretability per se. Not a knock, just different goals

Both great points — completely agree!

Interesting results. Wish we had the numbers to test what happens if you use something other (read: more sensitive) than resting state for the functional acquisitions... 🤔

...and welcome to Bluesky to the piece's author, @ajdinahalilovic.bsky.social!

Original paper: doi.org/10.1093/cerc...@thelablab.bsky.social for the data, provided via the fabulous Naturalistic Neuroimaging Database!

Thanks for your response @jvoigts.bsky.social (and sorry for not tagging you in before!). Bc other subfields of neuro can sometimes be disparaging of (f)MRI, comments like this are a bit triggering ;) but I appreciate your clarification, and also your larger point -- science needs more specialists!