Education
The curriculum, published
Journal club, an eight-week methods sequence, and workshops. All of it is open to any UT student, whether or not they are a Fellow, and none of it costs anything.
Journal club
Every other week during the semester, one paper, one member presenting, thirty minutes of discussion. Papers chosen for having something wrong with them as often as for being landmarks.
| Week | Focus | Presenter |
|---|---|---|
| 1 | A landmark paper in the current project area | TBA |
| 3 | A direct challenge to that landmark | TBA |
| 5 | A methods paper underlying both | TBA |
| 7 | A failed replication, read closely | TBA |
| 9 | A preprint from the last six months | TBA |
| 11 | Member's choice | TBA |
Methods sequence
Eight sessions, run once per year, ninety minutes each. Published as an outline so it can be argued with rather than taken on trust.
Reading a paper against itself
Figures before abstract. What the data would look like if the claim were false.
Study design and what it can support
Cross-sectional, longitudinal, and within-subject designs, and the claims each cannot license.
Effect sizes, intervals, and power
Why an underpowered significant result is weaker evidence than a null one, not stronger.
Pre-registration and the garden of forking paths
Reading a registration against the paper that came out of it.
Neuroimaging data, concretely
What a BIDS dataset contains, what a preprocessing pipeline does to it, and where the choices are.
Diffusion MRI and tractography
Reconstruction, its ambiguities, and why a tract image is a model and not a photograph.
Systematic search and extraction
Building a search, an inclusion log, and an extraction sheet that someone else could rerun.
Writing the thing
Structure, figures, authorship, and how to submit.
Workshops
Workshop dates for the coming term are published on the events page as they are set.
The tractography on the home page was reconstructed from a chapter member's own diffusion MRI, using the pipeline the sixth session above would cover. The chapter can teach this because it has already done it.