Thanks, I thought it had something to do with that, but everything I saw on Multiple Imputation seemed pretty abstract & i wasn't really sure how to implement it.

I think you are looking for pooled variance.

The above answer is very well done, but I think this source helps provide some context to the question.

The Q value is determined based off the confidence level, the degrees of freedom, and the number of groups. So basically, the CI is based calculated using the Tukeys's Q value, rather than a say a t value like if you were doing a student's t test. See this table for an example. real-statistics.com/statistics-tables/studentized-range-q-ta‌​ble

The mean is just the mean for each group. The CI is the 95% CI determined using the Tukey's Q critical value to compute the width of each groups CI. statsmodels.org/dev/generated/…