Using bal.tab() with Longitudinal Treatments
When using bal.tab() with longitudinal treatments, the output will be different from the case with point treatments, and there are some options that are common across all bal.tab() methods for dealing with longitudinal data. This page outlines the outputs and options in this case.
There are two main components of the output of bal.tab() with longitudinal treatments: the time-point-specific balance summary and across-time-points balance summary. The time-point-specific balance summaries are standard point treatment balance summaries at each time point.
The across-time-points balance summary is, for each variable, the greatest imbalance across all time-point-specific balance summaries. If the greatest observed imbalance is tolerable, then all other imbalances for that variable will be tolerable too, so focusing on reducing the greatest imbalance is sufficient for reducing imbalance overall. The balance summary will not be computed if multi-category treatments or multiply imputed data are used.
There are two additional arguments for each bal.tab() method that can handle longitudinal treatments: which.time and msm.summary.
which.time |
This is a display option that does not affect computation. If |
msm.summary |
This is a display option that does not affect computation. If |
The output is a bal.tab.msm object, which inherits from bal.tab. It has the following elements:
Time.Balance |
For each time point, a regular |
Balance.Across.Times |
The balance summary across time points. This will include the maximum balance statistic(s) for each covariate across all time points. |
Observations |
A table of sample sizes or effective sample sizes for each time point before and after adjustment. |
As with other methods, multiple weights can be specified, and values for all weights will appear in all tables.
Noah Greifer
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.