Equating covariances as a test for sex-limitation effects?

Hi all,

I have a question about sex-limitation model testing. I came across an article which mentioned the following (http://www.ncbi.nlm.nih.gov/pubmed/15685427) :

"For each of QCST, VIQ and PIQ, covariances of female and male MZ twin pairs could be equated, as could covariances for female and male DZ twin pairs, indicating an equal magnitude of genetic effects for females and males for each measure. As such scalar sex limitation effects were not modelled for any of the variables. Likewise, there were no significant differences for any of the variables between same sex DZ twin pairs and opposite sex DZ twin pairs indicating that for each measure the same genes were acting on females and males (i.e., no evidence for non-scalar sex limitation)."

I'm just wondering how this was done. Can I achieve this by fitting reduced models from a saturated univariate model which first constrained the variances across twin order, and then constrained the variances and covariances across sex (of the same zygosity)?

Thanks and best regards,
Yi Ting