I just completed a factor analysis with SPSS using this guide. The factor analysis was conducted on a survey that was completed by first year engineering students that completed a semester using Dotfolio. The analysis was roughly done, and I’m not sure if I did it right as I had extracted 10 factors, when really I was looking for something closer to 3 or 4. In fact if I use the scree plot I can reduce it to 4 factors, however if I select initial eigenvalues that have a total greater than 1, then I end up with 10 factors.
Nevertheless, on preliminary inspection it appears that one factor is “Interest in a topic” and another factor is “Taking a shallow approach to learning”, and a third is “E-Portfolios are too much work”. I haven’t interpreted the remaining factors just yet. However, what I am interested in finding out is whether or not students that take a shallow approach to learning perceive e-portfolios to be a useful tool in their learning experience. I suspect that they don’t, which is a good thing, because prior research suggests that e-portfolios encourage deeper learning, which may indicate why they require a lot of work to maintain. Which also appears to be a factor as to why students do not wish to use e-portfolios in other engineering units of study. Anyway the factor analysis was kind of cool, and the results look promising. Just as well, because I needed the factor analysis to complete the last chapter of my thesis.
I can see now how businesses would employ such techniques for market intelligence. I guess those pesky marketers that phone you up during dinner asking you to participate in a survey would run a factor analysis on the responses, compile everything into a report and sell them for a lot of money to large corporations with big marketing budgets.