Factor analysis Flashcards
-Understand what a thematic analysis is and what it is for -Understand how to read the SPSS outputs and what sig values for what section!
what is factor analysis for
- A method to assess the extent to which a
questionnaire is measuring a latent variable. - It explores correlation patterns between items to
establish an underlying structure.
‒ Think of a factor as a ‘subscale’ - A data reduction method
what SPSS OUTputs do we need to look at
. the correlation matrix
. the KMO and Barletts test of sphericity (Factorability of the data)
. table of communalities
. total variance explained table
. scree plot
. pattern matrix
The correlation matrix
All items must correlate (above ..30) with at least ONE other item
None should correlate above .80 with eachother as this would be multicollinearity/singularity (too similar)
Factorability of the data
KMO- accept anything above .5 or .6 (suitability for factor analysis). Write KMO=
Barlett’s test of sphericity= p<.05
x2= chi2 (df), p< sig.
e.g. x2= 555.6 (75)=, p<0.01
suitability for factor analysis
Communalities table
All extraction numbers must ne above .40
>.40
get rid if there is
total variance explained table
Shows us how many factors have been extracted-
there specific eigenvalues are under ‘total’
WHEN ASKED- ‘how much of the total varience does the proposed factor structure explain’- look at the cumulative % and take the last number of the last factor excluded (this is the total)
scree plot
Look for devees in the L
this should match the total number of factors excluded
should be an L
The pattern matrix
Does each item have a factor it loads on to?
To load it must be above .40
are there any cross loading- does one item load onto more than one factor?
UNSTABLE FACTORS- if a factor has less than 3 items, this is unstable and we should consider removing
