I have the mean scores of a survey that used a 5-point Likert scale and the total scores of a test that had a maximum of 10 points. I would like to check for their correlation but I am unsure of what test to use given that the Likert scale is ordinal and exam scores are continuous.
But given that I obtained the mean scores for the Likert type such that the final values have decimal places, would that already qualify the values to be considered as continuous, allowing me to proceed with the non-parametric Spearman correlation?
Usually when you combine several Likert-type items into a single score that score is considered interval or quasi-continuous. Of course, you are implicitly making some assumptions about the response categories that allow you to sum or average the individual items (questions). But that's usually the way we handle it.
It's not really an interval scale, so taking the mean is a bit dubious. I'd go for Spearmans rho as a measure of association.
https://stackoverflow.com/questions/43059228/correlation-of-two-ordinal-variables-in-r?fbclid=IwAR11kLL7m7A_affmhRS_glQHsZc4D6tDJXJp87LIpruBFSzeXu6stfxKGbw
- Badges
- Science topic
- Similar topics
- Biological Science
- Faunistics
- Ornithology
- birds