Weighting in Surveys
Ehsan Karim 
Note
This section contains advanced materials for those who are interested (optional).
Video content
Slides
Summary of topics discussed
Tabular summary
| Step | Basic IPW Process | IPW in Complex Survey |
|---|---|---|
| 1 | Fit PS model (A~L) | Fit PS model (A~L) with survey-weights as design variable (as in Austin et al. (2018))/covariate (as in DuGoff et al. (2014)) |
| 2 | Convert PS to IPW(ATE) | Convert PS to IPW(ATE) |
| \(IPW = 1/PS\), if \(A = 1\) | \(IPW = 1/PS\), if \(A = 1\) | |
| \(IPW = 1/(1-PS)\), if \(A = 0\) | \(IPW = 1/(1-PS)\), if \(A = 0\) | |
| 3 | Check balance using SMD in IPW-weighted data | Check balance using SMD in data weighted by w = IPW * survey-weights |
| 4 | Apply outcome model with weight = IPW | Apply outcome model with weight = IPW * survey-weights |
If ATT is the target parameter, then use
\(IPW(ATT) = 1\), if \(A = 1\)
\(IPW(ATT) = PS/(1-PS)\), if \(A = 0\)
Reference (Optional)
- Ridgeway, G., Kovalchik, S. A., Griffin, B. A., & Kabeto, M. U. (2015). Propensity score analysis with survey weighted data. Journal of causal inference, 3(2), 237-249.
- Austin, Peter C., and Elizabeth A. Stuart. 2015. Moving Towards Best Practice When Using Inverse Probability of Treatment Weighting (IPTW) Using the Propensity Score to Estimate Causal Treatment Effects in Observational Studies. Statistics in Medicine 34 (28): 3661–79.