Experiment with our
Marsh SDQ data set (or another large data set you have available).
Using the results of the EFA from the entire sample, draw small random
samples that represent the following subject to item ratios. Explore
the effect of having an inappropriately small sample on the goodness
of the solution. Do the results of the small samples replicate the
results from the large sample “population”?
2:1
5:1
10:1
20:1
Review EFA analyses
in top journals in your field. What subject to item ratios do you
find in these articles? Are they sufficient, given the results from
this chapter and your experiments in #1?