The finite sample performance of estimators for mediation analysis under sequential conditional independence
Using a comprehensive simulation study based on empirical data, this paper investigates the finite sample properties of different classes of parametric and semi-parametric estimators of (natural or pure) direct and indirect causal effects used in mediation analysis under sequential conditional independence assumptions. The estimators are based on regression, inverse probability weighting, and combinations thereof. Our simulation design uses a large population of Swiss jobseekers and considers variations of several features of the data generating process and the implementation of the estimators that are of practical relevance. We find that no estimator performs uniformly best (in terms of root mean squared error) in all simulations. Overall, so-called ‘g-computation’ dominates. However, differences between estimators are often (but not always) minor in the various setups and the relative performance of the methods often (but not always) varies with the features of the data generating process.
|Date of creation:||Jun 2014|
|Date of revision:||Nov 2014|
|Contact details of provider:|| Postal: Dufourstrasse 50, CH - 9000 St.Gallen|
Phone: +41 71 224 23 25
Fax: +41 71 224 31 35
Web page: http://www.seps.unisg.ch/
More information through EDIRC
References listed on IDEAS
Please report citation or reference errors to , or , if you are the registered author of the cited work, log in to your RePEc Author Service profile, click on "citations" and make appropriate adjustments.:
- Stefanie Behncke & Markus Frölich & Michael Lechner, 2010. "A Caseworker Like Me - Does The Similarity Between The Unemployed and Their Caseworkers Increase Job Placements?," Economic Journal, Royal Economic Society, vol. 120(549), pages 1430-1459, December.
- Arun Advani & Tymon Sloczynski, 2013.
"Mostly harmless simulations? On the internal validity of empirical Monte Carlo studies,"
CeMMAP working papers
CWP64/13, Centre for Microdata Methods and Practice, Institute for Fiscal Studies.
- Advani, Arun & Sloczynski, Tymon, 2013. "Mostly Harmless Simulations? On the Internal Validity of Empirical Monte Carlo Studies," IZA Discussion Papers 7874, Institute for the Study of Labor (IZA).
- Lechner, Michael & Wunsch, Conny, 2013. "Sensitivity of matching-based program evaluations to the availability of control variables," Labour Economics, Elsevier, vol. 21(C), pages 111-121.
- Lechner, Michael & Wunsch, Conny, 2011. "Sensitivity of matching-based program evaluations to the availability of control variables," CEPR Discussion Papers 8294, C.E.P.R. Discussion Papers.
- Michael Lechner & Conny Wunsch, 2011. "Sensitivity of Matching-Based Program Evaluations to the Availability of Control Variables," CESifo Working Paper Series 3381, CESifo Group Munich.
- Lechner, Michael & Wunsch, Conny, 2011. "Sensitivity of matching-based program evaluations to the availability of control variables," Economics Working Paper Series 1105, University of St. Gallen, School of Economics and Political Science.
- Lechner, Michael & Wunsch, Conny, 2011. "Sensitivity of Matching-Based Program Evaluations to the Availability of Control Variables," IZA Discussion Papers 5553, Institute for the Study of Labor (IZA).
- Andrew Gelman & Guido Imbens, 2013. "Why ask Why? Forward Causal Inference and Reverse Causal Questions," NBER Working Papers 19614, National Bureau of Economic Research, Inc.
- Flores, Carlos A. & Flores-Lagunes, Alfonso, 2009. "Identification and Estimation of Causal Mechanisms and Net Effects of a Treatment under Unconfoundedness," IZA Discussion Papers 4237, Institute for the Study of Labor (IZA).
- Newey, Whitney K., 1984. "A method of moments interpretation of sequential estimators," Economics Letters, Elsevier, vol. 14(2-3), pages 201-206.
- van der Laan Mark J. & Rubin Daniel, 2006. "Targeted Maximum Likelihood Learning," The International Journal of Biostatistics, De Gruyter, vol. 2(1), pages 1-40, December.
- Raymond Hicks & Dustin Tingley, 2011. "Causal mediation analysis," Stata Journal, StataCorp LP, vol. 11(4), pages 605-619, December.
- repec:tpr:restat:v:99:y:2017:i:1:p:180-183 is not listed on IDEAS
- Imai, Kosuke & Yamamoto, Teppei, 2013. "Identification and Sensitivity Analysis for Multiple Causal Mechanisms: Revisiting Evidence from Framing Experiments," Political Analysis, Cambridge University Press, vol. 21(02), pages 141-171, March.
- Huber, Martin & Lechner, Michael & Wunsch, Conny, 2013. "The performance of estimators based on the propensity score," Journal of Econometrics, Elsevier, vol. 175(1), pages 1-21.
- Martin Huber & Michael Lechner & Giovanni Mellace, 2017. "Why Do Tougher Caseworkers Increase Employment? The Role of Program Assignment as a Causal Mechanism," The Review of Economics and Statistics, MIT Press, vol. 99(1), pages 180-183, March.
- Huber, Martin & Mellace, Giovanni & Lechner, Michael, 2014. "Why do tougher caseworkers increase employment? The role of programme assignment as a causal mechanism," Economics Working Paper Series 1414, University of St. Gallen, School of Economics and Political Science.