- â (1967), âThe Behavior of Maximum Likelihood Estimates under Nonstandard Conditions,â in Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, L.M. LeCam, J. Neyman (Eds.), Berkeley: University of California Press, 221â233.
Paper not yet in RePEc: Add citation now
- â (2016), âRobust Inference in Sample Selection Models,â Journal of the Royal Statistical Society B, 78, 805â827. A Derivation of Proposition 1 We derive the IF of the Rivers-Vuong two-step estimator. The first step is OLS for which the IF is known. Using our notation we have IF{(x1, x2, y1); S, F} = â Z x1x> 1 x1x> x2x> 1 x2x>
Paper not yet in RePEc: Add citation now
- Avella-Medina, M. (2020), âPrivacy-Preserving Parametric Inference: A Case for Robust Statistics,â Journal of the American Statistical Association, to Appear.
Paper not yet in RePEc: Add citation now
Blundell, R. W. and Powell, J. L. (2004), âEndogeneity in Semiparametric Binary Response Models,â Review of Economic Studies, 71, 655â679.
Bonhomme, S. and Weidner, M. (2018), âMinimizing Sensitivity to Model Misspecification,â Working Paper.
Brunello, G., Fabbri, D., and Fort, M. (2013), âThe causal effect of education on body mass: Evidence from Europe,â Journal of Labor Economics, 31, 195â223.
- Cameron, A. C. and Trivedi, P. K. (2005), Microeconometrics: methods and applications, Cambridge university press.
Paper not yet in RePEc: Add citation now
Cantoni, E. and Ronchetti, E. (2001), âRobust Inference for Generalized Linear Models,â Journal of the American Statistical Association, 96, 1022â1030.
Duncan, G. M. (1987), âA Simplified Approach to M-Estimation with Application to TwoStage Estimators,â Journal of Econometrics, 34, 373â389.
- Eicker, F. (1967), âLimit Theorems for Regression with Unequal and Dependent Errors,â in Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, L.M. LeCam, J. Neyman (Eds.), Berkeley: University of California Press, 59â82.
Paper not yet in RePEc: Add citation now
Freue, G. V. C., Ortiz-Molina, H., and Zamar, R. H. (2013), âA natural robustification of the ordinary instrumental variables estimator,â Biometrics, 69, 641â650.
- Hampel, F. (1974), âThe Influence Curve and Its Role in Robust Estimation,â Journal of the American Statistical Association, 69, 383â393.
Paper not yet in RePEc: Add citation now
- Hampel, F., Ronchetti, E. M., Rousseeuw, P. J., and Stahel, W. A. (1986), Robust Statistics: The Approach Based on Influence Functions, New York: John Wiley and Sons.
Paper not yet in RePEc: Add citation now
- Hampel, F., Rousseeuw, P. J., and Ronchetti, E. (1981), âThe Change-of-Variance Curve and Optimal Redescending M-Estimators,â Journal of the American Statistical Association, 76, 643â648.
Paper not yet in RePEc: Add citation now
- Huber, P. J. (1964),âRobust Estimation of a Location Parameter,âThe Annals of Mathematical Statistics, 35, 73â101.
Paper not yet in RePEc: Add citation now
- Ψr 1(z1; γ)> , Ψr 2(z; γ, δ)> . Assume that the following regularity conditions (adapted from Duncan 1987) hold: 1. zi, for i = 1, . . . , N, is a sequence of independent identically distributed random vectors with distribution F defined on a space Z. 2. The parameter space Î is a compact subset of R à R2p1 à Rp2 à [0, 1] à R+ . 3. R Ψr (z; θ)dF = 0 has a unique solution θ0 in the interior of Î. 4. Ψr (z; θ) and â âθ Ψr (z; θ) are measurable for each θ in Î, continuous for each z in Z, and there exist F-integrable functions ξ1 and ξ2 such that for all θ â Î and z â Z |Ψr (z; θ)Ψr (z; θ)T | ⤠ξ1 and | â âθ Ψr (z; θ)| ⤠ξ2. 5. R Ψr (z; θ)Ψr (z; θ)> dF is non-singular for each θ â Î. 6. R âΨr (z; θ0)/âθdF is finite and non-singular.
Paper not yet in RePEc: Add citation now
- Maronna, R. A., Martin, G. R., and Yohai, V. J. (2006), Robust Statistics: Theory and Methods, Chichester: John Wiley and Sons.
Paper not yet in RePEc: Add citation now
Orrenius, P. M. and Zavodny, M. (2015), âDoes immigration affect whether US natives major in science and engineering?â Journal of Labor Economics, 33, S79âS108.
- Proof of Proposition 2. Consistency and asymptotic normality follow from Theorems 1-4 in Duncan (1987). The asymptotic variance reduces to two terms. This is because the error terms 1 in the first step and y2 â xÌ> δ in the second step are independent by construction and thus Î¥1(z)Î¥2(z)> and Î¥2(z)Î¥1(z)> vanish after integration.
Paper not yet in RePEc: Add citation now
Rivers, D. and Vuong, Q. H. (1988), âLimited Information Estimators and Exogeneity Tests for Simultaneous Probit Models,â Journal of Econometrics, 39, 347â366.
- Rousseeuw, P. J. and Van Driessen, K. (1999), âA Fast Algorithm for the Minimum Covariance Determinant Estimator,â Technometrics, 41, 212â223.
Paper not yet in RePEc: Add citation now
- Since MLE is the particular case of an M-estimator, its IF is known to be proportional to the score function (Hampel et al., 1986). Hence the IF is (un)bounded if the score function is (un)bounded. The score functions for different parameters are as follows: âl âγj = ây2 Ï(u) Φ(u) Ïxj/Ï1 (1 â Ï2)1/2 + (1 â y2) Ï(u) {1 â Φ(u)} Ïxj/Ï1 (1 â Ï2)1/2 + Ï2 (y1 â x> 1iγ1 â x> 2iγ2)xj, (21) for j = 1, 2. âl âβ = Ï(u) y2 â Φ(u) Φ(u){1 â Φ(u)} x1 (1 â Ï2)1/2 , (22) âl âα = Ï(u) y2 â Φ(u) Φ(u){1 â Φ(u)} y1 (1 â Ï2)1/2 . (23) The score functions in (22) and (23) are structurally the same as (6). It is clear that all the score functions in (21)-(23) are unbounded. C Assumptions and Proof of Proposition 2 Denote Ψr (z; θ) =
Paper not yet in RePEc: Add citation now
- von Mises, R. (1947), âOn the Asymptotic Distribution of Differentiable Statistical Functions,â The Annals of Mathematical Statistics, 18, 309â348.
Paper not yet in RePEc: Add citation now
White, H. (1980), âA Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity,â Econometrica, 48, 817â838.
Wooldridge, J. M. (2010), Econometric Analysis of Cross Section and Panel Data, Cambridge, MA: The MIT Press, 2nd ed.
- Yohai, V. J. (1987), âHigh Breakdown-Point and High Efficiency Robust Estimates for Regression, â The Annals of Statistics, 15, 642â656.
Paper not yet in RePEc: Add citation now
- Zhelonkin, M. (2013), âRobustness in Sample Selection Models,â PhD Thesis, University of Geneva, Switzerland.
Paper not yet in RePEc: Add citation now
Zhelonkin, M., Genton, M. G., and Ronchetti, E. (2012), âOn the Robustness of Two-Stage Estimators,â Statistics & Probability Letters, 82, 726â732.