Should I use cluster analysis or latent class analysis for my dissertation?
Hi,
I am testing the validity of model that posits why gifted students underachieve in school. There are three latent constructs: environmental perceptions, self-efficacy, and goal valuation. If students' attitudes are low toward any of these constructs, it is theorized they will lack self-regulation and underachieve in school. I have subscales from two different instruments (very reliable) to measure the constructs, a total of 41 questions.
I think a cluster analysis would be simpler to run, but I know that LCA has many advantages over a traditional cluster analysis. I am just not as familiar with LCA. Which would you recommend I use and why?
Thank you!
|