Abstract
In causal randomized experiments or psychological trials, the two-stage least squares (2SLS) model with instrument variables (IVs) is a widely used approach to address the issue of treatment endogeneity. The IVs are used to estimate a part of the causal effect whose estimation is not affected by the violation of the linearity assumption in the causal model, and the causal effect of interest in the 2SLS model becomes the local average treatment effect (LATE). Because practical data usually violate the normality assumption, the LATE estimate from the traditional normal-distribution-based method may be inefficient or even biased. This study proposes a robust Bayesian estimation method using Student’s t distributions to model data with heavy tails or containing outliers and compares the performance of the proposed robust method to that of the traditional normal-distribution-based method. A Monte Carlo simulation study is conducted and shows that the proposed robust method outperforms the traditional method when data are contaminated. The robust method provides more accurate and efficient LATE estimates and better model fits and thus is recommended to be used in general in the 2SLS modeling with IVs.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
D. Angrist, G. Imbens, Two stage least squares estimation of average causal effects in models with variable treatment intensity. J. Am. Stat. Assoc. 90, 431–442 (1995)
D. Angrist, A.B. Krueger, Does compulsory school attendance affect schooling and earnings? Tech. Rep. Natl. Bur. Econ. Res. 10, 1–14 (1990)
D. Angrist, A.B. Krueger, The effect of age at school entry on educational attainment: an application of instrumental variables with moments from two samples. J. Am. Stat. Assoc. 87, 328–336 (1992)
D. Angrist, J. Pischke, Mostly Harmless Econometrics: An Empiricist’s Companion (Princeton University Press, Princeton, 2008)
D. Angrist, J. Pischke, Mastering Metrics: The Path from Cause to Effect (Princeton University Press, Princeton, 2014)
D. Angrist, G. Imbens, D. Rubin, Identification of causal effects using instrumental variables. J. Am. Stat. Assoc. 91, 444–455 (1996)
A. Azzalini, M.G. Genton, Robust likelihood methods based on the sked-t and related distributions. Int. Stat. Rev. 76, 106–129 (2008)
D. Card, Using Geographical Variation in College Proximity to Estimate the Return to Schooling. Aspects of Labor Market Behavior: Essays in Honor of John Vanderkamp (University of Toronto Press, Toronto, 1995)
J. Currie, A. Yelowitz, Are public housing projects good for kids? J. Public Econ. 75, 99–124 (2000)
O.D. Duncan, A.O. Haller, A. Portes, Peer influence on aspirations: a reinterpretation. Am. J. Sociol. 74, 119–137 (1968)
S. Greenland, An introduction to instrumental variables for epidemiologists. Int. J. Epidemiol. 29, 722–729 (2000)
F.R. Hampel, E.M. Ronchetti, P.J. Rousseeuw, W.A. Stahel, Robust Statistcs: The Approach Based on Influence Functions (Wiley, New York, 1986)
P.J. Huber, Robust Statistics (Wiley, New York, 1981)
G. Imbens, D. Rubin, Bayesian inference for causal effects in randomized experiments with noncompliance. Ann. Stat. 29, 305–327 (1997)
K.L. Lange, R.J.A. Little, J.M.G. Taylor, Robust statistical modeling using the t distribution. J. Am. Stat. Assoc. 84 (408), 881–896 (1989)
S.Y. Lee, Y.M. Xia, Maximum likelihood methods in threating outliers and symmetrically heavy-tailed distributions for nonlinear structural equation models with missing data. Psychometrika 71, 565–585 (2006)
J.C. Pinheiro, C. Liu, Y.N. Wu, Efficient algorithms for robust estimation in linear mixed-effects models using the multivariate t distribution. J. Comput. Graph. Stat. 10 (2), 249–276 (2001)
D. Rachman-Moore, R.G. Wolfe, Robust analysis of a nonlinear model for multilevel educational survey data. J. Educ. Stat. 9 (4), 277–293 (1984)
M. Seltzer, K. Choi, Sensitivity analysis for hierarchical models: downweighting and identifying extreme cases using the t distribution. Multilevel Model. Methodol. Adv. Issues Appl. 13, 25–52 (2003)
S. Shoham, Robust clustering by deterministic agglomeration em of mixtures of multivariate t-distributions. Pattern Recogn. 35, 1127–1142 (2002)
P. Song, P. Zhang, A. Qu, Maximum likelihood inference in robust linear mixed-effects models using multivariate t-distribution. Stat. Sin. 17, 929–943 (2007)
D.O. Staiger, J.H. Stock, Instrumental variables regression with weak instruments. Econometrica 65, 557–586 (1997)
J. Steele, R. Murnane, J. Willet, Do financial incentives help low-performing schools attract and keep academically talented teachers? Evidence from California. J. Policy Anal. Manag. 29, 451–478 (2010)
X. Tong, Z. Zhang, Diagnostics of robust growth curve modeling using student’s t distribution. Multivar. Behav. Res. 47, 493–518 (2012)
H.X. Wang, Q.B. Zhang, B. Luo, S. Wei, Robust mixture modelling using multivariate t-distributions with missing information. Pattern Recogn. Lett. 25, 701–710 (2004)
J. Wang, Z. Lu, A.S. Cohen, The sensitivity analysis of two-level hierarchical linear models to outliers, in Quantitative Psychology Research. Springer Proccedings in Mathematics and Statistics (Springer, Cham, 2015), pp. 307–320
K.-H. Yuan, P.M. Bentler, Structural equation modeling with robust covariances. Sociol. Methodol. 28, 363–396 (1998)
K.-H. Yuan, Z. Zhang, Structural equation modeling diagnostics using R package semdiag and EQS. Struct. Equ. Model. 19, 683–702 (2012)
K.-H. Yuan, L.L. Marshall, P.M. Bentler, A unified approach to exploratory factor analysis with missing data, nonnormal data, and in the presence of outliers. Psychometrika 67, 95–111 (2002)
K.-H. Yuan, P.L. Lambert, R.T. Fouladi, Mardia’s multivariate kurtosis with missing data. Multivar. Behav. Res. 39, 413–437 (2004)
Z. Zhang, K. Lai, Z. Lu, X. Tong, Bayesian inference and application of robust growth curve models using student’s t distribution. Struct. Equ. Model. 20, 47–78 (2013)
D. Zimmerman, A note on the influence of outliers on parametric and nonparametric tests. J. Gen. Psychol. 121, 391–401 (1994)
D. Zimmerman, Invalidation of parametric and nonparametric statistical tests by concurrent violation of two assumptions. J. Exp. Educ. 67, 55–68 (1998)
J. Zu, K.-H. Yuan, Local influence and robust procedures for mediation analysis. Multivar. Behav. Res. 45, 1–44 (2010)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Shi, D., Tong, X. (2017). Robust Bayesian Estimation in Causal Two-Stage Least Squares Modeling with Instrumental Variables. In: van der Ark, L.A., Wiberg, M., Culpepper, S.A., Douglas, J.A., Wang, WC. (eds) Quantitative Psychology. IMPS 2016. Springer Proceedings in Mathematics & Statistics, vol 196. Springer, Cham. https://doi.org/10.1007/978-3-319-56294-0_34
Download citation
DOI: https://doi.org/10.1007/978-3-319-56294-0_34
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-56293-3
Online ISBN: 978-3-319-56294-0
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)