Jeremy Hogeveen
@jeremyhogeveen.bsky.social
Brain imager (and occasional tinkerer) @ The University of New Mexico. https://www.hogeveen-lab.com/
No plans to eat yet, but the way it was bullying around the other animals and trying to chase down our kids…. That plan may change. Turkeys seem like real A-holes!
November 24, 2025 at 9:02 PM
No plans to eat yet, but the way it was bullying around the other animals and trying to chase down our kids…. That plan may change. Turkeys seem like real A-holes!
Reposted by Jeremy Hogeveen
If your ultimate inference target is the group level, then what matters is the joint contribution of trial number per condition and participant sample size, not either one in isolation. We explored this point in detail here:
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
Hyperbolic trade-off: The importance of balancing trial and subject sample sizes in neuroimaging
Here we investigate the crucial role of trials in task-based neuroimaging from the perspectives of statistical efficiency and condition-level generali…
www.sciencedirect.com
November 21, 2025 at 3:37 AM
If your ultimate inference target is the group level, then what matters is the joint contribution of trial number per condition and participant sample size, not either one in isolation. We explored this point in detail here:
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
I think @mandymejia.bsky.social and @gangchen6.bsky.social approaches will, hopefully, enable us to move towards Bayesian estimation from the first-level up (eg we’re trying to set up our new study’s analysis with BayesfMRI—https://www.statmindlab.com/software)
November 20, 2025 at 11:03 PM
I think @mandymejia.bsky.social and @gangchen6.bsky.social approaches will, hopefully, enable us to move towards Bayesian estimation from the first-level up (eg we’re trying to set up our new study’s analysis with BayesfMRI—https://www.statmindlab.com/software)
Tremendously kind of you, thank you!
I imagine within-subjects power would be the same needs as always for fMRI. In this case, we modeled first level data using standard FLAME1 in FSL, then Bayesian multilevel model on ROI-level estimates to make group-level inferences using brms in R.
I imagine within-subjects power would be the same needs as always for fMRI. In this case, we modeled first level data using standard FLAME1 in FSL, then Bayesian multilevel model on ROI-level estimates to make group-level inferences using brms in R.
November 20, 2025 at 10:59 PM
Tremendously kind of you, thank you!
I imagine within-subjects power would be the same needs as always for fMRI. In this case, we modeled first level data using standard FLAME1 in FSL, then Bayesian multilevel model on ROI-level estimates to make group-level inferences using brms in R.
I imagine within-subjects power would be the same needs as always for fMRI. In this case, we modeled first level data using standard FLAME1 in FSL, then Bayesian multilevel model on ROI-level estimates to make group-level inferences using brms in R.
Reposted by Jeremy Hogeveen
And the next step? Full voxel-level modeling.
Recent numerical advances cracked the scalability barrier. Voxel-level hierarchical modeling is now feasible, revealing just how punishing traditional multiple-comparison adjustments really are.
arxiv.org/abs/2511.12825
Recent numerical advances cracked the scalability barrier. Voxel-level hierarchical modeling is now feasible, revealing just how punishing traditional multiple-comparison adjustments really are.
arxiv.org/abs/2511.12825
SIMBA: Scalable Image Modeling using a Bayesian Approach, A Consistent Framework for Including Spatial Dependencies in fMRI Studies
Bayesian spatial modeling provides a flexible framework for whole-brain fMRI analysis by explicitly incorporating spatial dependencies, overcoming the limitations of traditional massive univariate app...
arxiv.org
November 18, 2025 at 10:13 PM
And the next step? Full voxel-level modeling.
Recent numerical advances cracked the scalability barrier. Voxel-level hierarchical modeling is now feasible, revealing just how punishing traditional multiple-comparison adjustments really are.
arxiv.org/abs/2511.12825
Recent numerical advances cracked the scalability barrier. Voxel-level hierarchical modeling is now feasible, revealing just how punishing traditional multiple-comparison adjustments really are.
arxiv.org/abs/2511.12825
Thanks Julie Bugg, @kimberlychiew.bsky.social, @tobiasegner.bsky.social, and @evdbussc.bsky.social for organizing--and giving Cidney such a great venue to build her community & share her work!
November 20, 2025 at 4:15 PM
Thanks Julie Bugg, @kimberlychiew.bsky.social, @tobiasegner.bsky.social, and @evdbussc.bsky.social for organizing--and giving Cidney such a great venue to build her community & share her work!
100% agree—though perhaps a bit too perfectionist www.bbc.com/news/enterta...
Thom Yorke 'nearly walked off stage' during Radiohead's 1997 Glastonbury performance
The singer was on the verge of walking off stage during Radiohead's legendary 1997 performance.
www.bbc.com
November 20, 2025 at 3:22 PM
100% agree—though perhaps a bit too perfectionist www.bbc.com/news/enterta...