You are here

Correlated factors ok to use even when phentypic varaiance not significant?

3 posts / 0 new
Last post
K Ram's picture
Offline
Joined: 08/21/2012 - 02:31
Correlated factors ok to use even when phentypic varaiance not significant?

We have 7 variables, of which one of the variable is significantly heritable - modeled using univariate analyses. Is it theoretically congruent to test for shared genetics between all 7 variables using the correlated factors model despite the fact that 6 of the variables are not significantly heritable?

tbates's picture
Offline
Joined: 07/31/2009 - 14:25
sounds fine

Sounds fine. It is unlikely the other 6 variables are genuinely h2 == zero, but the multivariate model will only confirm this if it is the case.

The benefit of going multivariate is that the modest evidence each variable provides for inheritance mode on its own can be pooled by the multivariate model: you may find that the multivariate likelihood indicates influences that were not significant in the univariate models.

best, tim

K Ram's picture
Offline
Joined: 08/21/2012 - 02:31
Thank you for your quick

Thank you for your quick reply Tim!