Model Selection Tests for Complex Survey Samples

Iraj Rahmani, Jeffrey Wooldridge

Research output: Working paper

Abstract

We extend Vuong’s (1989) model-selection statistic to allow for complex survey samples. As a further extension, we use an M-estimation setting so that the tests apply to general estimation problems – such as linear and nonlinear least squares, Poisson regression and fractional response models, to name just a few – and not only to maximum likelihood settings. With stratified sampling, we show how the difference in objective functions should be weighted in order to obtain a suitable test statistic. Interestingly, the weights are needed in computing the model-selection statistic even in cases where stratification is appropriately exogenous, in which case the usual unweighted estimators for the parameters are consistent. With cluster samples and panel data, we show how to combine the weighted objective function with a cluster-robust variance estimator in order to expand the scope of the model-selection tests. A small simulation study shows that the weighted test is promising.
Original languageEnglish
Pages109-135
Number of pages26
Publication statusPublished - 2019

Publication series

NameAdvances in Econometrics
PublisherEmerald Publishing Limited
Volume39
ISSN (Print)0731-9053

Fingerprint

Sample Survey
Model Selection
Statistic
Objective function
M-estimation
Stratified Sampling
Poisson Regression
Linear Least Squares
Nonlinear Least Squares
Robust Estimators
Variance Estimator
Panel Data
Stratification
Expand
Test Statistic
Maximum Likelihood
Fractional
Simulation Study
Estimator
Computing

Cite this

Rahmani, I., & Wooldridge, J. (2019). Model Selection Tests for Complex Survey Samples. (pp. 109-135). (Advances in Econometrics; Vol. 39).

Model Selection Tests for Complex Survey Samples. / Rahmani, Iraj; Wooldridge, Jeffrey.

2019. p. 109-135 (Advances in Econometrics; Vol. 39).

Research output: Working paper

Rahmani, I & Wooldridge, J 2019 'Model Selection Tests for Complex Survey Samples' Advances in Econometrics, vol. 39, pp. 109-135.
Rahmani I, Wooldridge J. Model Selection Tests for Complex Survey Samples. 2019, p. 109-135. (Advances in Econometrics).
Rahmani, Iraj ; Wooldridge, Jeffrey. / Model Selection Tests for Complex Survey Samples. 2019. pp. 109-135 (Advances in Econometrics).
@techreport{1ccf153ebb6f4660acaea2d2713e163a,
title = "Model Selection Tests for Complex Survey Samples",
abstract = "We extend Vuong’s (1989) model-selection statistic to allow for complex survey samples. As a further extension, we use an M-estimation setting so that the tests apply to general estimation problems – such as linear and nonlinear least squares, Poisson regression and fractional response models, to name just a few – and not only to maximum likelihood settings. With stratified sampling, we show how the difference in objective functions should be weighted in order to obtain a suitable test statistic. Interestingly, the weights are needed in computing the model-selection statistic even in cases where stratification is appropriately exogenous, in which case the usual unweighted estimators for the parameters are consistent. With cluster samples and panel data, we show how to combine the weighted objective function with a cluster-robust variance estimator in order to expand the scope of the model-selection tests. A small simulation study shows that the weighted test is promising.",
author = "Iraj Rahmani and Jeffrey Wooldridge",
year = "2019",
language = "English",
series = "Advances in Econometrics",
publisher = "Emerald Publishing Limited",
pages = "109--135",
type = "WorkingPaper",
institution = "Emerald Publishing Limited",

}

TY - UNPB

T1 - Model Selection Tests for Complex Survey Samples

AU - Rahmani, Iraj

AU - Wooldridge, Jeffrey

PY - 2019

Y1 - 2019

N2 - We extend Vuong’s (1989) model-selection statistic to allow for complex survey samples. As a further extension, we use an M-estimation setting so that the tests apply to general estimation problems – such as linear and nonlinear least squares, Poisson regression and fractional response models, to name just a few – and not only to maximum likelihood settings. With stratified sampling, we show how the difference in objective functions should be weighted in order to obtain a suitable test statistic. Interestingly, the weights are needed in computing the model-selection statistic even in cases where stratification is appropriately exogenous, in which case the usual unweighted estimators for the parameters are consistent. With cluster samples and panel data, we show how to combine the weighted objective function with a cluster-robust variance estimator in order to expand the scope of the model-selection tests. A small simulation study shows that the weighted test is promising.

AB - We extend Vuong’s (1989) model-selection statistic to allow for complex survey samples. As a further extension, we use an M-estimation setting so that the tests apply to general estimation problems – such as linear and nonlinear least squares, Poisson regression and fractional response models, to name just a few – and not only to maximum likelihood settings. With stratified sampling, we show how the difference in objective functions should be weighted in order to obtain a suitable test statistic. Interestingly, the weights are needed in computing the model-selection statistic even in cases where stratification is appropriately exogenous, in which case the usual unweighted estimators for the parameters are consistent. With cluster samples and panel data, we show how to combine the weighted objective function with a cluster-robust variance estimator in order to expand the scope of the model-selection tests. A small simulation study shows that the weighted test is promising.

M3 - Working paper

T3 - Advances in Econometrics

SP - 109

EP - 135

BT - Model Selection Tests for Complex Survey Samples

ER -