Hi there,

I'd like to clarify the use of the -Abcontr option in 3dANOVA3 type 4 [a 2-way repeated-measures, 2*2 ANOVA, A (2) *B (2) *subject].

I'm analyzing the simple effects of B nested in A as well as the simple effects of A nested in B (4 simple effects in total). I am looking at these data with a specific p-value threshold, which corresponds to a specific T value threshold. To my understanding, this T-p correspondence should be the SAME for the simple effects of A nested in B, and for the simple effects of B nested in A (with a 2*2, the df are the same in both cases).

The strange error is that the same T value corresponds to a different p value for Abcontr and for aBcontr. (e.g. a T=3 corresponds to p=.01 in aBcontr, but to p=.001 in Abconr simple effects).

Just to make sure it is not my data that are causing the problem, I switched the A and B levels and re-ran the analysis. The data which were analyzed with aBcontr are now those analyzed with Abcontr, so the T-p correspondence should be the same for these simple effects in the 2 ANOVAs, but this is not the case (what happens is that aBcontr always has the same T-p correspondence, and Abcontr always has the same correspondence, in both ANOVAs, but those are not identical, as I think they should be).

My guess is that my syntax is wrong, so I'm attaching it here.... Any help will be greatly appreciated.

Thanks,
Debbie

3dANOVA3 -type 4 -alevels 2 -blevels 2 -clevels 5 \
-dset 1 1 1 '../sept23_2005/afdir/Sirf_detn.lag23+tlrc' \
-dset 1 2 1 '../sept23_2005/afdir/Sirf_detx.lag23+tlrc' \
-dset 2 1 1 '../sept23_2005/afdir/Sirf_attn.lag23+tlrc' \
-dset 2 2 1 '../sept23_2005/afdir/Sirf_attx.lag23+tlrc' \
-dset 1 1 2 '../sept21_2005/afdir/Sirf_detn.lag23+tlrc' \
-dset 1 2 2 '../sept21_2005/afdir/Sirf_detx.lag23+tlrc' \
-dset 2 1 2 '../sept21_2005/afdir/Sirf_attn.lag23+tlrc' \
-dset 2 2 2 '../sept21_2005/afdir/Sirf_attx.lag23+tlrc' \
-dset 1 1 3 '../oct11_2005/afdir/Sirf_detn.lag23+tlrc' \
-dset 1 2 3 '../oct11_2005/afdir/Sirf_detx.lag23+tlrc' \
-dset 2 1 3 '../oct11_2005/afdir/Sirf_attn.lag23+tlrc' \
-dset 2 2 3 '../oct11_2005/afdir/Sirf_attx.lag23+tlrc' \
-dset 1 1 4 '../may31_2005/afdir/Sirf_detn.lag23+tlrc' \
-dset 1 2 4 '../may31_2005/afdir/Sirf_detx.lag23+tlrc' \
-dset 2 1 4 '../may31_2005/afdir/Sirf_attn.lag23+tlrc' \
-dset 2 2 4 '../may31_2005/afdir/Sirf_attx.lag23+tlrc' \
-dset 1 1 5 '../nov2_2005/afdir/Sirf_detn.lag23+tlrc' \
-dset 1 2 5 '../nov2_2005/afdir/Sirf_detx.lag23+tlrc' \
-dset 2 1 5 '../nov2_2005/afdir/Sirf_attn.lag23+tlrc' \
-dset 2 2 5 '../nov2_2005/afdir/Sirf_attx.lag23+tlrc' \
-amean 1 det \
-amean 2 att \
-bmean 1 neut \
-bmean 2 emot \
-fab interact \
-aBcontr -1 1 : 1 att_det_n \
-aBcontr -1 1 : 2 att_det_x \
-Abcontr 1 : -1 1 x_n_det \
-Abcontr 2 : -1 1 x_n_att \
-adiff 2 1 att_det \
-bdiff 2 1 x_n \
-bucket anov3