Extraction

Question 1

what are the range of factor extraction methods?

Answer 1

PAMIU

1) Principal axis factoring
2) Alpha factoring
3) Maximum likelihood
4) Image factoring
5) Unweighted least squares

Question 2

Communalities

Answer 2

look at communalities (h squared) table in SPSS output

PCA- they will all be 1

FA- shows SMC- one variable correlated with all other variables

Question 3

What is the extraction process for?

Answer 3

Extraction is an iterative process that aims to maximise the variance explained

Question 4

Eigenvalues (Kaiser method of extraction)

Answer 4

Look that the initial Eigenvalues in the TOTAL VARIANCE EXPLAINED box

used to see how many factors/ components are worth keeping

Generally keep the Eigenvalues above 1 (in SPSS- but this is often over-extracted done this way :( )

It tells you how much of the variance (%) the components account for

Question 5

Scree plot

Answer 5

Examine the scree plot to see where the inflexion is- extract the factors at this point

Question 6

Parallel analysis

Answer 6

Creates a random dataset with the same number of cases and variables

Run PCA/FA to generate average Eeigenvalues

compare these to the real Eigenvalues, and retain the ones from the real dataset that are higher than the ones from the random dataset