2
$\begingroup$

I have a study where patients were measured pre-operatively (time_pre), post-operatively (time_post) and long-term follow-up (time_lt). I have 60 subjects.

Data consists of: 5 scale variables (1-100) 14 ordinal variables (1-5) and some patient specific characteristics (none missing) like age, gender, co-morbidities (yes / no)

I want to check if there is a significant difference between pre-operative scores and post-operative scores (long- and short-term) and between short and long-term follow-up. Therefore I used a repeated measures ANOVA on the scale variables and with post-hoc Bonferroni correction, and Friedman's test for the ordinal variables with post-hoc Bonferroni correction.

Now I want to repeat these tests, but with the missing values filled in.

Therefore I performed multiple imputations, but I can't figure out how I should perform a repeated measures ANOVA from this point.

I use SPSS v22 statistical software.

$\endgroup$
  • $\begingroup$ Hi @Justin, if you can add some raw data (at least fragment of full dataset) it will be helpful. $\endgroup$ – Ladislav Naďo Jun 20 '14 at 12:51
  • $\begingroup$ Hi @LadislavNado, thank you for your respons. Here you can find a screenshot of the Variable view link and some raw data link $\endgroup$ – Justin Jun 20 '14 at 13:09
  • $\begingroup$ Let me get this straight. You have 60 pacients, each was measured three times. WHAT variable(s) was measured? All 19 variables are dependent variables in your analysis? You also want to [ check if there is a significant difference between pre-operative scores and post-operative scores...] scores of what? All 19 variables? Please re-edit your question for better clarity and add some (even fabricated) dataset. $\endgroup$ – Ladislav Naďo Jun 20 '14 at 13:45
  • $\begingroup$ Correct. We performed measurements ourself (with a constant valueable as outcomes (quotient)) and had patients fill out questionnaires (both ordinal outcomes and nominal outcomes). I have measurements from three time-points (pre-, post- and long-time-post-intervention). Main outcome is the difference in mean of the quotient between pre-intervention and long-term-post-intervention. (if the intervention has a significant effect on the quotient on the long term). Same goes For the other outcomes. Secondary I want to know if there is a sig. effect between short- and long-term-post-intervention $\endgroup$ – Justin Jun 23 '14 at 7:08
  • $\begingroup$ Here is a (fake) data snippet link $\endgroup$ – Justin Jun 23 '14 at 7:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.