The treatment is comprised of four banks policies and outcome is risk taking behaviour of banks. How Panel data and multiple treatment should be taken care of at the same? Can we deal with it in Stata or some different software is needed. Any help will be much appreciated.
If there is adequate common overlap (i.e., a range of common values shared by all reference and treatment groups) for all of the variables, then you can use the R Twang package to do inverse propensity weighting with multiple treatments. A tutorial for this can be found at http://www.rand.org/content/dam/rand/pubs/tools/TL100/TL169z1/RAND_TL169z1.pdf.
If there is not adequate common overlap for all groups, then it may be necessary to use multiple sub-analyses. An example of this using panel data can be found at http://www.sciencedirect.com/science/article/pii/S0001457515002262. When this method is used, it can be easily implemented in Stata.
When working with panel data, it is often best to match all observations for entity (banks in your case?) to all observations for another entity. This can be done by matching with the data in wide form, then changing the data to long form and using regression analysis methods that account for panel data. This could include fixed-effects, random-effects, multilevel, or other regression methods.
- Badges
- Science topic