İ have a data set with six categorical variables, with responses on a scale of 1-5; the reliability test for the individual variables is very strong but when combined for all variables the reliability test give very low figures. What could be the problem. Also what would be an appropriate regression for this analysis.
Can you clarify ? Are your six variables combined into a single scale, with the sum or mean of the individual questions ? And it's inter-item reliability that you're unhappy with ?
When i run the reliability for the set of questions per variable the test showed about 90% reliability. But when i combine the questions using median/mean to represent the variable as one, the reliability test give very low results. So this is what I don't understand why. (The intervariable reliability)
When you used a single test score variable (i.e., the median or mean score of the items), how did you compute reliability for that single variable (e.g., test-retest correlation?)? And what reliability index did you use at the item level (e.g., Cronbach's alpha?)?
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression