r - Psych package: Which exploratory factor analysis method to apply -


i trying conduct exploratory factor analysis r using psych package. data consists of items different scales- skewed, differ in number of response catgories (ranging 4 7). factor analysis method should use? minres or wls? after conducting efa want conduct cfas using lavaan package. here estimator mlm seems appropriate data. in mplus there mlm method available exploratory , confirmatory factor analysis. know factor analysis method corresponds in psych mlm estimator? or think transforming data? thank help, kind regards a.

for categorical items ranging 4-7 categories, use cor="poly" option in fa , run either default (minres) or wls option.

when finding correlation matrix cor="poly" option, give warning number of categories not same, can ignored.

for complete data, cor="poly" option produces correlation matrices identical lavann polychoric matrices. have not compared them mplus.

i have found minres (default) option in fa more robust wls, mileage may vary.


Comments

Popular posts from this blog

javascript - AngularJS custom datepicker directive -

javascript - jQuery date picker - Disable dates after the selection from the first date picker -