------------------------------------------------------------------------------- help forwgttest-------------------------------------------------------------------------------

Test the impact of sampling weights in regression analysis

wgttestdepvar[varlist] [ifexp] [inrange],wgt(wgtvar)[cmd(estimation_command)prefix(string)testopt(test_options)nonoiseestimation_options]

by...:may be used withwgttest; see help by.

DescriptionWhether to use sampling weights (pweights) in regression analysis should be carefully evaluated. Often, the weights do not have great influence on the parameter estimates (see e.g. Winship and Radbill, 1994, to learn when and why). In such cases, unweighted estimates are preferable because they are more efficient than the weighted estimates.

wgttestperforms a test proposed by DuMouchel and Duncan (1983) to evaluate the significance of the impact of sampling weights on estimation results. First, a regression model ofdepvaronvarlistincludingwgtvarand the first order interactions betweenvarlistandwgtvaras additional covariates will be estimated. Second, the coefficients of these covariates (i.e.wgtvarand the interactions) are tested against zero using a standard F test (as provided by the post-estimation commandtest). "If the F test is not significant, then the weighted and unweighted estimates are not significantly different and the analyst can proceed by using unweighted OLS. Weighted and unweighted estimates are significantly different if the F test is significant" (Winship and Radbill, 1994: 248). In the later case, the unweighted estimators are probably biased by sample selection and the weighted estimators are preferable. Be aware, however, that significant differences between weighted and unweighted estimates may also be due to model misspecification.

Options

cmd(estimation_command)allows users to choose a command other thanregressfor model estimation. Technically,wgttestwill work with most estimation commands (if not all). This, however, does not mean that the test is always valid (DuMouchel and Duncan, 1983, who proposed the test, discuss it solely within the framework of linear regression).

nonoisesuppresses the estimation results.

prefix(string)allows users to choose a prefix other than_Ifor the interaction variables. The length of the prefix is restricted to 4 characters. Note that the interaction variables will only be created temporarily.

testopt(test_options)may be used to pass options thru to the post-estimation commandtest.

wgt(wgtvar)specifies the sampling weights (mandatory).

estimation_optionsare passed thru to the estimation command.

ExamplesTest the impact of the sampling weights (variable

pwt) for a linear regression model of wages on education and work experience:

. wgttest wage education experience, wgt(pwt)

ReferencesDuMouchel, W. H. and G. J. Duncan (1983). Using Sample Survey Weights in Multiple Regression Analyses of Stratified Samples. Journal of the American Statistical Association 78: 535-543. Winship, C. and L. Radbill (1994). Sampling Weights and Regression Analysis. Sociological Methods and Research 23: 230-257.

AuthorBen Jann, ETH Zurich, jann@soz.gess.ethz.ch

Also seeManual:

[U] 23 Estimation and post-estimation commands,[U] 29 Overview of Stata estimation commands,[R] test,[R] regress