Nicholas Branigan

@michaelcbdavid Thanks! We used nilearn plot_surf_stat_map: #nilearn.plotting.plot_surf_stat_map" target="_blank" rel="nofollow noopener">nilearn.github.io/stable/modules…

@NKBranigan Very cool work! What did you use to make those beautiful brain surface plots btw?

What are the neural correlates of human behavior? Much research has approached versions of this question between individuals but not within individuals. Are results from these 2 approaches similar? No! In over 4000 individuals, the neural correlates of cognitive control diverged between vs. within individuals -- a case of non-ergodicity and Simpson’s paradox. We think that non-ergodicity may be an important principle in human neuroscience. [1/12]

We think this work illustrates the need to distinguish inferences from between and within individuals. And since within-person questions are fundamental but underexplored, with individual-level approaches there could be much left to discover... [12/12]

But non-ergodicity was not coming from our model, since we also saw it with directly observed reaction times. For example, in default mode brain areas, individuals with slower reaction times had more brain activity but when an individual had slower reaction times they had less brain activity. [11/12]

@StanfordBrain We found widespread nonergodicity in brain-behavior relationships. Results underscore the need to distinguish between-subjects and within-subjects inferences in human neuroscience. Preprint: biorxiv.org/content/10.110… Work with @mystrypercy @Stanford_CoSyNe

In his Nanosymposium presentation at 3:45 PM, Nicholas Branigan will address the challenge of nonergodicity and how a new study reveals how mechanisms of inhibitory control differ between and within subjects. #SfN24