Hello, guys!
I'm trying to run a Factor A on the entirety of my test and I came up with a doubt:
Since the test has been designed to measure 3 aspects of the same construct, running a Factor A on SPSS shows me that that there is ONE factor that determines 40 percent of the variance in scores, but there are several others that are determining small percentages of the variance.
My assumption is that -hopefully- this one factor reflects de construct which I'm investigating. So, my questions are:
1 - Is this assumption plausible? Are there specific texts or keywords that bring up this kind of assumption?
2 - Lets say it's an intelligence test, so the test would be devided into musical, mathematical and linguistic proficiency. Can I run a Factor A in each section of my test?
This would help me to investigate the factors that are determining variance in each section of the text. Again, is it a recognized strategy of investigation? Are there any keywords that would help me to sustain this strategy?
Thank you very much,
Pedro de Alcântara
A confirmatory factor model might be appropriate. For a lot of tests like this there is a general factor and then each has its own. This can be set up in several ways, for example the bi-factor model. The lavaan manual (freeware written in R) I believe provides an example of this and an alternative.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS