@SashaGusevPosts @thebirdmaniac @monitoringbias @EvolOdonata @charlesmurray @jacobin understand this example. it's a special case. okay so obviously none of my arguments have been processed by you. that's fine. here is a whole subfield of psychology where
@workercommie @thebirdmaniac @SashaGusevPosts @monitoringbias @EvolOdonata @charlesmurray @jacobin if your estimators are looking at the same latent variables, and only those latent variables, then factor invariance comes with it. meaning you can assign va
@ben_golub @aryehazan @littmath @CaltechEconThry @charlesmurray The classic (though a bit dated) citation is Meredith (1993): https://t.co/pIRlDVsPqT. This earlier paper by Meredith & Millsap (https://t.co/VExXaU3GpD) is also worth reading to see why l
Meredith, W. (1993). Measurement invariance, factor analysis and factorial invariance. Psychometrika, 58(4), 525-543. https://t.co/lIOXcYk2Nk 21/32
@endofanerajc @Whiskie2 Just so it's clear, this doesn't mean everyone in X category will have a high or low IQ, etc., etc. Nor does it mean that we should treat people as being less than or more than others. https://t.co/57LszPEIMZ
@endofanerajc @Whiskie2 The reliability for full-scale IQs is generally >0.9. IQ is the best predictor of success. The military uses IQ tests. It is not 100% predictive. But, it's still a very good predictive tool. 2 links on measurement invariance bel
@AlxEtz The earlier papers, e.g., Horn & McArdle (1992). https://t.co/94uDjfBxvr & Meredith (1993). https://t.co/Qh8FL0WxSK are nice papers on MI
@jd_wilko for a comprehensive intro I'd suggest this book: https://t.co/bZV1XQzBwu These are some classic articles on measurement invariance : https://t.co/owcONRYBDc https://t.co/3uHw3jMo0M