firts of all thanks for your attention! this problem is really important for me because it is linked with the good exit of my thesis
I am doing a confirmatory factor analysis on a 8item scale to assess its validity,
in particular I am evaluating the two factors represented by the first and the last four items of the scale:
the first four items represents the expertise factor
while the last four items represents the trustworthiness factor
the correlations are expected to be high because we are talking about a scale which is very often filled in an automatic way in the questionnaires by the informants, and the item of the scale are very similar (es. the XYZ company has a great experience; the XYZ has a great deal of expertise; and so on.. the fourth element is negative because it is an inverted question like " I don't think that XYZ company is expert"
my problem is that running the factor analysis in SPSS, with PRINCIPAL AXIS COMPONENT extraction method, and OBLIMIN rotation method (this choice is due to the willingness to use the same analysis of the paper i took as model)
i obtain ONE LOADING IN THE PATTERN MATRIX GREATER THAN ONE
I attach the SPSS output to show you my problem;
i have 2 questions:
- I would like to know why a loading can be grater than 1 (I always thaught that it was a standardized value which can vary from -1 to 1)
- What is the difference between a PATTERN MATIX and a STRUCTURE MATIX?
- What is shown in the pattern matrix? correlation factors or regression factors ?
how can i motivate this results if they are not wrong?
if you can answer me in the simplest way possible, I will be really grateful
thanks for your precious help!
it is really important to me
Advertise on Talk Stats