[R] Exploratory Factor Analysis with Multiply Imputed Data

Heather Kettrey heather.h.kettrey at vanderbilt.edu
Tue Dec 20 05:01:30 CET 2016


I am trying to conduct exploratory factor analysis (EFA) with multiply
imputed data. I used Amelia II to create 40 imputed datasets and have saved
the imputed data in .csv format (I saved the imputations both as a single
file with all imputations and as 40 separate files - each containing one
set of imputations).

I plan to use the psych package to conduct 40 separate factor analyses
(without rotation) and then need to conduct consensus rotation of the 40
factor loading matrices that these separate EFAs produce.

I read that the promin() function in the PCovR package is capable of
consensus rotation of multiple (unrotated) factor loading matrices, but I
cannot figure out how to input multiple matrices into this function.

Does anyone know how to use the promin() function to conduct consensus
rotation - or have any other suggestions of how to conduct EFA with
multiply imputed data in R? Any help would be greatly appreciated.



Heather Hensman Kettrey, PhD
Research Associate, Peabody Research Institute (PRI)
Senior Researcher, Meta-Analysis Center at PRI
Vanderbilt University

	[[alternative HTML version deleted]]

More information about the R-help mailing list