In my research model the independent variables have questions Likert scale , while the dependent variables are dichotomous questions; with PLS Can I combine these types of questions?
Dear Carlos,
You can, but usually when the dependent variable is dichotomic we allpy logistic regression.
I think you want to aplly PLSbecasuse you have a small sample or you have no teory to support your hypotheses. becasuse if it is not this reasion, it will be better to usae SEm, instead of PLS.
Have a nice time
Helena
Hi Carlos,
Hair et al. (2014) suggest that the PLS-SEM algorithm works well with binary coded data that are used as a means of categorical control variables or moderators in PLS-SEM models. They further suggest that one should be cautious with the use of categorical (i.e., dummy-coded) scaled data in PLS models as they do have certain limitations.
Hope that helps.
Kevin
PLS-SEM can be problematic when endogenous variables are categorical. PLS texts talk about these issues and require that the outcome variable should be of metric data, not categorical. (A Primer on Partial Least Squares Structural Equation Modeling, PLS-SEM; Hair, Hult, Ringle and Sarstedt, 2013, page 16). Categorical as moderator or control variable is not a problem though.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques