Unequal variance assumption/uneven group size in DPABI

Submitted by LukeN on

Dear experts,

Thank you for creating this wonderful software. I have been finding it to be very useful!

I have a study with uneven group sizes (40 patients, 70 controls), which has caused a reviewer some concerns. My understanding is that other software packages (SPM, FSL) allow for the assumption of unequal variances to be specififed as a way to deal with this. Is this also true for DPABI, and if so which methods of correction would it apply to (GRF, permutation etc.)?

Thank you in advance,

Luke

YAN Chao-Gan

Thu, 10/29/2020 - 04:28

Hi Luke,

I think 40 vs. 70 can be taken as normal two-sample t-test. And you may perform supplementary analysis to select 40 vs. 40.

You can apply GRF or permutation test as normal.