PT  - JOURNAL ARTICLE
AU  - Maël Lebreton
AU  - Stefano Palminteri
TI  - Assessing inter-individual variability in brain-behavior relationship with functional neuroimaging
AID  - 10.1101/036772
DP  - 2016 Jan 01
TA  - bioRxiv
PG  - 036772
4099  - http://biorxiv.org/content/early/2016/01/18/036772.short
4100  - http://biorxiv.org/content/early/2016/01/18/036772.full
AB  - Investigating inter-individual differences in brain-behavior relationships is fundamental to decipher the neural substrate of cognition, and to realize the full potential of neuroimaging applications. In this context, accurately assessing the statistical dependencies between inter-individual differences in behavior and inter-individual differences in neural activity is essential. In the present perspective we consider two hypotheses: 1) BOLD signal scales linearly with behavioral variables across individuals and 2) BOLD signal encodes behavioral variables on a similar scale across individuals. We formally show that these two hypotheses produce opposite brain-behavior correlational results in group-level analyses. We empirically explore these hypotheses in four fMRI studies, and find that, regarding the representation of values in the prefrontal cortex, the normalization hypothesis dominates. Independently from the generalizability of these findings, our results illustrate the importance of explicitly testing the scaling law between brain signals and behavioral variables before engaging in the study of functional inter-individual differences.