best model fit based on even chr

[Izel Erdogan]

Hi,

Hope you are having a good day!

I have a question about finding the best factor model on subset of the data(even chromosomes). In my analysis, EFA based on all chromosomes did not lead to a good model, however EFA based on even chromosomes did. (Modelfit results were based on CFA on odd chromosomes as suggested.)

I then validated this modelfit on all chromosomes and it is still a good fit. 

My question is, can we justify using this model that was not initially substentiated from all of the data/chrs? I know that in statistics this is not ideal but wanted to clear it out for genomicsem as well. 

Thank you so much for your time and efforts!

Best regards,

Izel

[Elliot Tucker-Drob]

Hi Izel,
I would recommend that you should follow the standard principles of
training and replication. The goal is generally to determine whether a
model that is trained on one portion of your data replicates in an
independent portion of your data (i.e. displays good fit, produces
consistent estimates, etc...). If it does, then it would make sense to
carry forward in the complete dataset. Or perhaps only carry forward a
version of the model with nonzero parameter estimates that replicated
across independent subsets of the data. I don't see how training on
the full dataset comes into play, as that leaves no remaining
independent sample to replicate in...

> --
> You received this message because you are subscribed to the Google Groups "Genomic SEM Users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to genomic-sem-us...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/genomic-sem-users/b12f958e-11a8-46f3-a617-223aa2eb3f58n%40googlegroups.com.