I will be doing ryff's psychological well being scale's psychometric properties. Could I administer alternate form reliability simultaneously? (42-item and 18-item)
You could, but it would annoy the respondents to have to answer the identical questions twice, so I do not recommend it.
If you can get by the problem clearly identify by David Morgan, 82 total items is not extremely problematic unless the items require a great deal of time each to respond. If you wish to go ahead and do it, first, do not put the corresponding items adjacent to each other. My suggestion would be to have two forms of the 82 item scale. For half of sample, put the items on Form 1 first and the items on Form 2 second. Reverse that for the other half and randomly administer the full 82 items. This way, you would remove the order effect or at least be able to empirically verify if there was one.
Since the 18 items are a subset of the 42 items, why can't you just ask all 42, and then extract the 18 for a comparison to the full scale?
As what i understood on alternate forms, the alternate instrument should be adminsitered days after the first scale was administered. So, is it practical to do it if the 18-item test is a subscale of 42?
- Badges
- Science topic
- Similar topics
- Psychology
- Psychometrics