Hi All,
Firstly, many thanks for taking the time to address my query. I greatly appreciate your help and insights.
I am currently working on a longitudinal dataset for my research project and I opted linear mixed effect model as my analytical method. Due to missing data problems in several model covariates, I have to resort to multiple imputations to fill in the missing values. At the moment, I am still trying to familiarize myself with the method and therefore I am still a novice with respect to this technique.
I had created 3 imputated datasets since I am still playing around with the technique. After I executed the linear mixed model command, I managed to obtain the pooled estimates for the main model parameters, covariance matrix parameters and estimated marginal means for my main factor interest which is the treatment. Unfortunately when I inspected the output for pairwise comparisons, no pooled estimates were given. The screenshots for the spss output for the pairwise comparisons of such variable (treatment by period) with Bonferroni correction for multiple comparison is given below:
Please also find below the screenshot showing the estimated marginal means for treatment by period as a reference:
In this case I would like to know whether it is feasible to obtain the pooled estimated for each treatment comparison (for instance, the pooled estimate of the mean differences between sham patch + 1 FTU EMLA 30 minutes and MN+1FTU EMLA 30 minutes groups together with its 95% CI and p value) in this case after multiple imputation has been carried out on SPSS using Rubin's rules. Or the pooled estimates could not be produced using Rubin's rules since there are statistical issues that I am not aware of? Or should I calculate the pooled estimate for each pairwise comparisons manually using Rubin's rules? Based on my readings of the IBM SPSS Missing Values 28 manual, it is feasible to obtain the pooled estimate for each pairwise comparisons (the screenshot of the manual is given below, final bullet point):
Again many thanks for your insightful comments and have a great day ahead. All the best and cheers.
Jagjeet
------------------------------
Jagjeet Singh Hullon Data Scientist
------------------------------
#SPSSStatistics