About the overlap on the genetic and shared environmental factors

Posted on
No user picture. Qiuzhi Xie Joined: 12/27/2018
I am conducting the Cholesky model and wondering how to judge if two phenotypes significantly overlap on a genetic or environmental factor. It should be judged by the loading factor and the 95% CI, right? In the following example, only the third phenotype (P3) does not significantly overlap with the others on the A1 (genetic) factor, as the lower boundary of the 95% CI is smaller than .00. Am I right?

A1
P1 .45 (.20, .81)
P2 .76 (.06, .82)
P3 .39 (-.21, .69)
P4 .30 (.14, .71)
95% confidence interval is presented in the parentheses

Replied on Fri, 12/06/2019 - 09:42
Picture of user. AdminNeale Joined: 03/01/2013

Hi

If the numbers you provide are the Cholesky factor loadings, I think this is not a good approach. In a Cholesky, the covariance between two phenotypes, say 3 and 4, depends not only on the first column of the Cholesky, but on columns 2 and 3 as well. So it is best to answer your question, as Rob says, obtain the confidence intervals of the lower triangle of the A matrix where A = L*L' the result of the algebra to obtain the covariance associated with A.