#### Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Supported by

# What matrix reports JASP in EFA?

edited March 2019

Hi!

(not sure if this is the right forum... sorry!)

I am running an exploratory factor analysis using oblique rotation (Promax) and maximum likelihood. I am not sure, though, which matrix appears in JASP: pattern or structure matrix? We know that we should report both matrices when using oblique rotation, so this is a bit of a problem, so I am trying to complement the results form JASP with those of SPSS. However, the factor loadings reported in JASP do not match those I get in SPSS (EFA with ML and Promax). Similarly, the correlations between factors are very different between SPSS and JASP.

Since I would like to report the goodness of fit indices that JASP reports (which is great!), I would like to understand why I get these differences. The results I get in other tests, say, reliability, are identical both in SPSS and JASP, so I am confident I am using exactly the same data set.

• I'll ask the expert, Erik-Jan! (yes, almost the same first name)

• Hi! I had the same experience, but did not find follow-up answers to that question, could somebody maybe indicate the relevant documentation to get answers ? Thanks.

• I'll also ask that this will be made clear in the output or the help file.

E.J.

• Hi all,

sorry for the late response. Here is some more information on EFA in JASP:

1. EFA is estimated in JASP not by ML but by a procedure called "MinRes", which according to the author of the `psych` package in `R` produces "solutions very similar to maximum likelihood even for badly behaved matrices". This may be the source of the difference between JASP and SPSS.
2. The "factor loadings" matrix is the "pattern matrix". We currently do not output the structure matrix, although you may compute it as S = P × Phi, where S is the structure matrix, P is the pattern matrix, and Phi is the factor correlation matrix (check the "factor correlations" box under "output options").
3. A lot of information is included in the `fa` function documentation of the `psych` package which we use internally, you can find that here: https://www.rdocumentation.org/packages/psych/versions/1.8.12/topics/fa