Test Moderation of Standardized Variance Components

Posted on
No user picture. benruk Joined: 12/09/2019
Hello,
since in the conventional moderated ACE model proposed by Purcell it is only possible to test whether there is a significant moderation of the unstandardized genetic, shared and unshared variance components in a phenotype, I was wondering whether it is also possible to test if there is a significant moderation of the standardized variance components.
Replied on Mon, 03/16/2020 - 15:30
Picture of user. AdminRobK Joined: 01/24/2014

I was wondering whether it is also possible to test if there is a significant moderation of the standardized variance components.

Yes, that can be done via DeFries-Fulker regression with double-entered data. Relevant references:

Kohler, H.P., & Rodgers, J.L. (2001). DF-analyses of heritability with double-entry twin data:
Asymptotic standard errors and efficient estimation. Behavior Genetics, 31(2), 179-191.

Rodgers, J. L., & Kohler, H. P. (2005). Reformulating and simplifying the DF analysis model.
Behavior Genetics, 35(2), 211-217.

Rodgers, J. L., & McGue, M. (1994). A simple algebraic demonstration of the validity of
Defries-Fulker analysis in unselected samples with multiple kinship levels. Behavior
Genetics, 24
(3), 259-262.

I will warn you: the fact that the Purcell approach estimates moderation of the raw, not standardized, variance components is one of its strengths, because moderation of the standardized components can be misleading. See the figure in the attachment. In the trait's raw metric, the nonshared-environmental component slightly increases as the moderator (SES) increases. But. because the total phenotypic variance also increases with the moderator, the standardized nonshared-environmental component actually *decreases* slightly with the moderator.