-1
$\begingroup$

I am doing factor analysis with Principal component extraction method.

Below is the Structure matrix

Extraction method: PCA

Rotation method: Promax with kaiser normalization ....................................................................COMPONENT............................................................. ............................................................. 1 ............ 2 ......... 3 .............4 ..........................................
Effeciency class ............................ .893.......... .236........ .139......... .876

Pre sales support......................... .839 ......... -.224 ........ .811.......... .326

I have two variable Eff_class and Pre_sales support and four component..Now variable seems to load alomost equally on two components. How to decide which component to choose.

$\endgroup$
0
$\begingroup$

First, are you doing factor analysis or principal component analysis? You use the word "factor" and also "component". (I admit that the language in this topic is very confusing)

This is really a sign that the PCA/FA isn't working right. You may need more factors or fewer or a different rotation method (there are a lot!) Explore a bunch and try to find one that is substantively useful.

$\endgroup$
  • $\begingroup$ @prof_jack, and another malicious dilemma to add for you is whether select or interpret by pattern matrix or structure matrix. $\endgroup$ – ttnphns Apr 28 '14 at 12:02
  • $\begingroup$ Why would you do that. SPSS has already taken a toll on me. Anyways i guess for loading, i should be using pattern matrix but now i am confused between loading and co-relation as they both seems to depict co-relation but gives different value for oblique rotation $\endgroup$ – prof_jack Apr 28 '14 at 12:28
  • 1
    $\begingroup$ @prof_jack, it's not me to plan any harm, neither SPSS. But the issue of factor analysis is intricate and demanding enough, so one has to read some books or good articles before starting to do it. $\endgroup$ – ttnphns Apr 28 '14 at 12:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.