I have data with 5 likert scale 1- Strongly Disagree to 5-Strongly Agree and also a Don't Know Option. I want to use Structural Equation Modelling (SEM) to determine causes of corruption from this data but having problem with the don't know option. I would like to seek suggestions.
If you are treating the Likert items or Likert scales as numeric variables, you would have to remove cases with opt-out answers like "Not applicable".
Other than that, I'm not so sure about using SEM on Likert data.
The 'don't know' option is a valid response, and it is neither a 'missing response' or 'not applicable'. and I would report the number of cases, in either a table or a plot. However in an analysis it would depend on your research question - is an SEM really what you wish to do? SEM require a large sample, and maybe there are other analytic or visual methods to answer your question?
Hello Namgay,
I agree with Miranda that the don't know option is data that might be used depending upon your research question.
I am not familiar with SEM but you would need to check the assumptions associated with SEM.
best wishes.
Tom
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials
- Inorganic Materials
- Ceramics