I am working with a common path model and I am having difficulty standardizing my factor loadings. In the past, I have identified the model by constraining the variance of the latent phenotype to 1, then I standardize factor loadings by using a matrix of standard deviations (SDs on diagonal, 0s on off-diagonal) and multiplying that by the matrix of the unstandardized factor loadings (I can attach code for this if necessary)
In the model I am currently working with, I have identified the model by fixing the first factor loading to 1 and I am finding that the method of standardizing factor loadings I've used before doesn't seem to be working properly (I get standardized loadings greater than 1). What method should I be using to standardize loadings when the first loading is fixed to 1?
Thanks!
Stephanie