Method workshops for advanced students at the IFS
- News
- Events

The advanced curriculum offers regular workshops on various methods for all research assistants at the IFS.
In February, two workshops were held on methods for estimating the average treatment effect as accurately as possible in an observational study. A biased estimate can occur if relevant characteristics of the study participants in two groups to be compared differ before the start of treatment. By applying propensity score methods or entropy balancing, comparable groups can be formed and an unbiased and precise estimate of the average treatment effect (on the treated) can be obtained.
On February 13, Dr. Lilly-Marlen Bihler, a research assistant at the IFS, introduced the topic of propensity score matching in the first workshop. The second workshop on February 27 offered an introduction to the methods of inverse probability weighting and entropy balancing.
Both workshops included practical applications of the methods to a data set called lalonde implemented in R. Participants assessed the achieved balance and the effective sample sizes and compared the results of the different methods and specifications. In addition, the advantages and disadvantages of each method were discussed.