I have T1,T2,T3 values. If I want to run Little's MCAR test with spss, should I put all the values together or run it separately (e.g. run Little's test for T1, then run again for T2)?
I think you should be able to obtain satisfactory results about your missing data when you run the values separately. Since you'll have to impute the missing values (or use a more rudimentary method such as replacement by means), you can try first with using the values separately. Hope this helps :)
Hi Hansika, thanks for helping. When I ran the test for all the time waves at once, it showed non-sig. However, when i ran ran only T2 or only T3, the chi-square doesnt work (Chi-square.000, DF=0,sig=.)
I have about 30% drop out in T2. Do you think 30% attrition rate is the reason why little's test doest work?, thank you very much
Hey Mengyin,
Yes, such a huge attrition in T2 is problematic, but i don't see why MCAR shouldn't run when you only run the T2 data. Have a look at this video: http://www.youtube.com/watch?v=ytQedMywOjQ
It's 45mins long, but really informative, and helps with imputing missing values as well :)
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software