Comparing TENT and CSPLIN

Hi AFNI experts,

I want to compare the results after using TENT and CSPLIN to decide which function I should use for the final analysis. I’ve run 3dDeconvolve using TENT(0,16,9), and later, run another 3dDeconvolve using CSPLIN(0,16,9), for both instances, I used the same data (of 1 participant) and the same stim_times 1D files. I expected to see smoother HRFs when using CSPLIN, but surprisingly, when comparing, I got the exact same parameter estimates after running 3dROIstats. Is there an explanation for why I got the same betas after using 2 different functions? The experiment is the Monetary Incentive Delay task with a even-related design.

Thanks,
Tien

Hi all,

I just want to follow-up with my previous question to ask if someone can kindly respond and let me know if it is possible to have the exact same results after running 2 separate 3dDeconvolve with TENT and CSPLIN.

Thank you,
Tien

Hi Tien,

Sorry for being slow.

Are your event times TR-locked? The difference
between them is in the interpolation between time
points.

  • rick

Hi Rick,

Yes, our event times are TR-locked. Now I understand why we’re getting the same results using TENT and CSPLIN.

Thank you for your response!
Tien