help
help > RE: 3 groups - multi-center test
May 1, 2020 01:05 AM | Andrew Zalesky
RE: 3 groups - multi-center test
Hi Riccardo,
1) Yes - this is correct.
2) Yes this is correct. The residuals are permuted and then added back into the model. This approach is described in Freedman and Lane (1983). See Anderson Winkler's work for fantastic explanations of permutation testing and the GLM:
https://www.ncbi.nlm.nih.gov/pmc/article...
3) Yes. But you may also want to try the contrast [0 1 0 0 0]. Also this model will also be somewhat sensitive to effects that are not necessarily linear. For example, g1=g2>g3.
Andrew
Originally posted by David Andre Barriere:
1) Yes - this is correct.
2) Yes this is correct. The residuals are permuted and then added back into the model. This approach is described in Freedman and Lane (1983). See Anderson Winkler's work for fantastic explanations of permutation testing and the GLM:
https://www.ncbi.nlm.nih.gov/pmc/article...
3) Yes. But you may also want to try the contrast [0 1 0 0 0]. Also this model will also be somewhat sensitive to effects that are not necessarily linear. For example, g1=g2>g3.
Andrew
Originally posted by David Andre Barriere:
Hello,
Just a small follow up on this question to clarify:
1) In general terms if we have an ANOVA design with 3 groups, want to test for any possible difference between the groups and we want to control the analysis for different covariates (sex, age, etc) we just need to add those collumns in front of the design as such:
1 0 0 -1 -2.4 1.3
1 0 0 1 1.1 0.1
0 1 0 1 0.3 -0.2
0 1 0 -1 -3.2 -2.1
0 0 1 1 -1.2 -3.1
0 0 1 1 4.1 4.5
and then use with F-contrast:
[1 1 1 0 0 0]
2) I explored NBS code a little bit and if I understood correctly what NBS does is a regression using the co-variables (at 0 in the contrast) to obtain the residuals and then use those residuals to do the ANOVA, is this correct?
3) If I want to explore linear effects (g1>g2>g3), I should replace the design with a more GLM kind format?
Such as
1 1 -1 -2.4 1.3
1 1 1 1.1 0.1
1 2 1 0.3 -0.2
1 2 -1 3.2 -2.1
1 3 1 -1.2 -3.1
1 3 1 4.1 4.5
and then use t-test contrast [ 0 -1 0 0 0] to test.
Best regards, Ricardo
Just a small follow up on this question to clarify:
1) In general terms if we have an ANOVA design with 3 groups, want to test for any possible difference between the groups and we want to control the analysis for different covariates (sex, age, etc) we just need to add those collumns in front of the design as such:
1 0 0 -1 -2.4 1.3
1 0 0 1 1.1 0.1
0 1 0 1 0.3 -0.2
0 1 0 -1 -3.2 -2.1
0 0 1 1 -1.2 -3.1
0 0 1 1 4.1 4.5
and then use with F-contrast:
[1 1 1 0 0 0]
2) I explored NBS code a little bit and if I understood correctly what NBS does is a regression using the co-variables (at 0 in the contrast) to obtain the residuals and then use those residuals to do the ANOVA, is this correct?
3) If I want to explore linear effects (g1>g2>g3), I should replace the design with a more GLM kind format?
Such as
1 1 -1 -2.4 1.3
1 1 1 1.1 0.1
1 2 1 0.3 -0.2
1 2 -1 3.2 -2.1
1 3 1 -1.2 -3.1
1 3 1 4.1 4.5
and then use t-test contrast [ 0 -1 0 0 0] to test.
Best regards, Ricardo
Threaded View
| Title | Author | Date |
|---|---|---|
| Giulia Forcellini | Mar 22, 2019 | |
| David Andre Barriere | Apr 30, 2020 | |
| Andrew Zalesky | May 1, 2020 | |
| Andrew Zalesky | Mar 22, 2019 | |