Hi JMP Community,
When building contrasts in a Regular Least Squares Fit Model analysis, the number of columns (i.e., discrete contrasts) is limited to the number of conditions (i.e., number of items produced by the combinations of all parameters). In many cases, I'd like to create more contrasts than conditions (i.e., comparisons of multiple pairs of factor combinations)
Example:
VAR1 = A, B, C and VAR2 = X, Y, Z
COMBINATIONS VAR1 * VAR2: AX, AY, AZ, BX, BY, BZ, CX, CY, CZ
CONTRASTS: AX vs AY, AX vs AZ, AY vs AZ, BX vs BY, BX vs BZ, BY vs BZ, CX vs CY, CX vs CZ, and CY vs CZ (==> one too many comparisons to input in the contrast matrix builder)
I have resorted to an ugly workaround by generating the complete pairwise t-test comparison (LSMeans student t-test) and extracting only the 9 contrasts that interest me.
Would you suggest a more efficient approach to this issue?
Thank you.
Best regards,
TS
Thierry R. Sornasse
