Elina Dale posted on Friday, August 01, 2014 - 2:12 am
I would like to get the polyserial corr estimates (inter-item & item-total) as described in Raykov & Marcoulides (2011). I am following the commands provided in their textbook but for some reason as soon as I add the total scale score, MPlus does not recognize that my items are measured on a categorical scale, even though I specify them under "CATEGORICAL" line.
I can see in the output the items are treated as continuous. If I run the same commands but without the tot1 (which is a sum of all the responses and is continuous), then it's OK: MPlus recognizes that my items are categorical.
Is there an error in my commands? How to get corr matrix for items some of which are categorical, while others are cont?