Between-movie variability severely limits generalizability of "naturalistic" neuroimaging
Leipold, S.; Ravi Rao, R.; Schoffelen, J.-M.; Bogels, S.; Toni, I.
Show abstract
"Naturalistic" neuroimaging paradigms, where subjects watch movies during fMRI, have become increasingly popular. Movie fMRI data is often analyzed using inter-subject correlation (ISC), which measures the similarity of neural time series across individuals. Differences in ISC during movie-watching have been linked to individual traits and states. However, movies are complex cultural artifacts that differ in content, structure, and style, raising concerns about the generalizability of ISC findings across movies. Using fMRI data from 112 subjects watching eight animated movies, we quantified between-movie variability in ISC and examined its implications for trait- or state-like effects. ISC varied substantially across movies and brain regions, with the strongest ISC observed in regions showing the greatest variability. Consequently, associations between behavior and ISC differed markedly across movies. These findings suggest that movie-based paradigms should be treated like task-based designs, requiring clear specification of movie features. Broad claims about "movie watching" are not warranted.
Matching journals
The top 4 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.