The Accuracy of Panel-corrected

Standard Errors. OLS

standard errors are accurate in the presence of either panel heteroscedasticity or contemporaneous correlation of the errors if the terms in the error covariance matrix, [omega], are not related to the squares and cross products of the independent variables.

To enable more scientific evaluation of

standard errors in future surveys conducted locally, this research was undertaken with the objective of estimating the value of the design factor in Hong Kong.

Accordingly, we turn to an investigation of how the two estimators and their

standard errors compare in the logit model.

Although they have been shown to be quite useful, multilevel models are susceptible to outliers occurring at each level of the data, leading to parameter estimation bias and inflated

standard errors (e.g.

In Table 2, the regression coefficients and

standard errors obtained from GEE and mixed model analysis applied to MAR, MCAR and complete data sets for the continuous result variable (linoleic acid) are almost same.

where [[gamma].sup.[dagger]] is the relevant "null" value of [gamma] (as in a test of the null hypothesis [H.sub.0]: [gamma] = [[gamma].sup.[dagger]]), and se([??]) is the asymptotic

standard error of equation (1) defined as [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] being a consistent estimator of the asymptotic variance of [??].

As shown in the Table 1, both the classic and semi-parametric models yielded extremely large estimates that are un-interpretable, impossibly large

standard errors, and type I errors close to 1.

Just as I described the differences between standard deviation and

standard error in this editorial, I plan to help readers navigate this increasingly complex range of statistical testing being reported in results sections in JVIB in future Statistical Sidebars.

The

standard errors for monthly natural gas volume and price estimates by State are given in Table C1.

The same ICC calculations were also performed for the

standard errors of disability estimates, and they were used for calculating between the items both before and after imputation.

While the discussion that follows concentrates on the effects of alcohol consumption and results from specification tests, appendix 1 (for binge drinking) shows the IV coefficients and marginal effect

standard errors of all explanatory variables on both absenteeism measures.

Crompton (2000) also pointed out this deficiency and extended the new stochastic approach to derive robust

standard errors for the rate of inflation by relaxing the earlier restriction on the variance of the error term by considering an unknown form of heteroscedasticity.