How to get one beta estimate per condition per run? (3dDeconvolve)

AFNI Message Board
1

Hi all,

What’s the best way to get one beta estimate per condition per run in 3dDeconvolve?

When I have one repeat of each condition per run, I just use -stim_times_IM which works great.
But what if I have 2 or more repeats of a condition per run, and I want to estimate a single beta for that condition for every run?

I can think of some workarounds, like running 3dDeconvolve multiple times, but it doesn’t seem ideal…

[Having a single beta per condition per run is often desired for multivariate/decoding analysis, which is what I’d like it for.]

thanks in advance!

Susan

SLaP, Laboratory of Brain & Cognition

2

Hi Susan,

Choices like that are controlled with the timing files, with IM being a special case of that. There is no related convenience for doing multiple runs.

Supposing there are 3 runs, that makes breaking a single timing file into 3 timing files, one per run. For example:

t11 t12 t13 t14
t21
t31 t32 t33

might be broken into:

t11 t12 t13 t14

t21

t31 t32 t33

And then you would specify these as individual timing files.

Does that seem reasonable?

  • rick