MSE-improvement of the least squares estimator by dropping variables

It is well known that dropping variables in regression analysis decreases the variance of the least squares (LS) estimator of the remaining parameters. However, after elimination estimates of these parameters are biased, if the full model is correct. In his recent paper, Boscher (1991) showed that the LS-estimator in the special case of a mean shift model (cf. Cook and Weisberg, 1982) which assumes no “outliers” can be considered in the framework of a linear regression model where some variables are deleted. He derived conditions under which this estimator outperforms the LS-estimator of the full model in terms of the mean squared error (MSE)-matrix criterion. We demonstrate that this approach can be extended to the general set-up of dropping variables. Necessary and sufficient conditions for the MSE-matrix superiority of the LS-estimator in the reduced model over that in the full model are derived. We also provide a uniformly most powerful F-statistic for testing the MSE-improvement.     

Air pollution-mortality models: A demonstration of the effects of random measurement error

Errors of measurement have long been recognized as a chronic problem in statistical analysis. Although there is a vast statistical literature of multiple regression models estimating the air pollution-mortality relationship, this problem has been largely ignored. It is well known that pollution measures contain error, but the consequences of this error for regression estimates is not known. We use Lave and Seskin's air pollution model to demonstrate the consequences of random measurement error. We assume a range of 0% to 50% of the variance of the pollution measures is due to error. We find large differences in the estimated effects on mortality of the pollution variables as well as the other explanatory variables once this measurement error is taken into account. These results cast doubt on the usual regression estimates of the mortality effects of air pollution. More generally our results demonstrate the consequences of random measurement error in the explanatory variable of a multiple regression analysis and the misleading conclusions that may result in policy research if this error is ignored.     

The effects of sampling on measures of association between variables based on sample means

Many studies that involve people's perceptions or behaviors focus on aggregate rather than individual responses. For example, variables describing public perceptions for some set of events may be represented as mean scores for each event. Event mean scores then become the unit of analysis for each variable. The variance of these mean scores for a variable is not only a function of the variation among the events themselves, but is also due to the variation among respondents and their possible responses. This is also the case for the covariances between variables based on event mean scores. In many contexts the variance and covariance components attributable to the sampling of respondents and their responses may be large; these components can be described as measurement error. In this paper we show how to estimate variances and covariances of aggregate variables that are free of these sources of measurement error. We also present a measure of reliability for the event means and examine the effect of the number of respondents on these spurious components. To illustrate how these estimates are computed, forty-two respondents were asked to rate forty events on seven risk perception variables. Computing the variances and covariances for these variables based on event means resulted in relatively large components attributable to measurement error. A demonstration is given of how this error is removed and the resulting effect on our estimates.     

Alternative value estimates of owner-occupied housing: Evidence on sample selection bias and systematic errors

Numerous econometric models have used various estimates of housing value as dependent variables. The three most common measures, in order of descending popularity, have been homeowner estimates, sales price, and assessed value. Each of these measures has limitations. The use of sales price can cause sample selection bias, while owner and tax assessor estimates are subject to measurement error. This study investigates the magnitude of the selection bias associated with sales price samples, and whether the errors in owner and assessor estimates are systematically related to independent variables typically included in estimated equations. Our most important conclusion is that the use of owner estimates may cause bias in the estimated coefficients on many independent variables.     

Assessing the importance of measurement error in capital investment models

New developments in the economics of capital investment emphasize the role of financial variables. Econometric evidence on these hypotheses is potentially compromised by measurement error due to accounting conventions. The paper reviews new capital investment models and considers ways in which accounting procedures might lead to measurement error biases. Advances in errors-in-variables econometric models are employed to gauge the impact of measurement error on estimates of financial influences on capital investment. Cash-flow models appear to be especially susceptible to measurement error but q models seem fairly insensitive to measurement problems.     

Estimating structural equation models with non-normal variables by using transformations

We discuss structural equation models for non-normal variables. In this situation the maximum likelihood and the generalized least-squares estimates of the model parameters can give incorrect estimates of the standard errors and the associated goodness-of-fit chi-squared statistics. If the sample size is not large, for instance smaller than about 1000, asymptotic distribution-free estimation methods are also not applicable. This paper assumes that the observed variables are transformed to normally distributed variables. The non-normally distributed variables are transformed with a Box–Cox function. Estimation of the model parameters and the transformation parameters is done by the maximum likelihood method. Furthermore, the test statistics (i.e. standard deviations) of these parameters are derived. This makes it possible to show the importance of the transformations. Finally, an empirical example is presented.     

Corrections for coarsely categorized measures: LISREL's polyserial and polychoric correlations

Versions 5 and 6 of LISREL (Joreskog and Sorbom, 1981) contain procedures that estimate the underlying correlation between continuous variables on the basis of crude rank category measures. The procedures assume that the distribution of the measured variables would have been bivariate normal if they had not been categorized. Using survey data and simulations, the accuracy of these polyserial/polychoric (P/P) based estimates of the underlying correlations are compared with those based on simple equal distance scoring of the categories. The results indicate that under some conditions, e.g., nearly normally distributed variables and moderate to high correlations, the polyserial/polychoric based estimates are better. Under other conditions, e.g., a moderate to high degree of skew and kurtosis and low correlations, the equal distance score based estimates are better. Under all conditions, the amount of error decreases fairly rapidly as the number of categories is increased from two to five.     

Measurement error in causal analysis of panel data: Attenuated versus inflated relationships

The presence of random measurement error is commonly thought to cause attenuation of statistical relationships. While this is an unquestionable truth in bivariate analysis, it cannot be generalized to the multivariate case without qualification. This paper shows that measurement error may give rise to overestimates of parameters in causal analysis whenever there is more than one independent variable and the independent variables are correlated. If the independent variables are not measured with the same amount of reliability, there may also be considerable error in estimates of the relative magnitude of their impact. Both problems are particularly serious when the amount of measurement error is large relative to some of the causal effects such as in panel analysis with lagged dependent variables.     

Return to education in Finland

This study presents estimates of the return to education in Finland using an individual-level data set that also includes ability measures and information on family background. It is found that ability test scores have a strong effect on the choice of education and on subsequent earnings. Estimating the return to education with no information on ability leads to an upward bias in the estimates. However, this bias is more than offset by a downward bias caused by endogeneity or measurement error. Instrumental variables estimates that utilize family background variables as instruments produce estimates of the return to schooling that are approximately 60% higher than the least squares estimates.     

Estimating panel data models in the presence of endogeneity and selection

We consider estimation of panel data models with sample selection when the equation of interest contains endogenous explanatory variables as well as unobserved heterogeneity. Assuming that appropriate instruments are available, we propose several tests for selection bias and two estimation procedures that correct for selection in the presence of endogenous regressors. The tests are based on the fixed effects two-stage least squares estimator, thereby permitting arbitrary correlation between unobserved heterogeneity and explanatory variables. The first correction procedure is parametric and is valid under the assumption that the errors in the selection equation are normally distributed. The second procedure estimates the model parameters semiparametrically using series estimators. In the proposed testing and correction procedures, the error terms may be heterogeneously distributed and serially dependent in both selection and primary equations. Because these methods allow for a rather flexible structure of the error variance and do not impose any nonstandard assumptions on the conditional distributions of explanatory variables, they provide a useful alternative to the existing approaches presented in the literature.     

VARs,common factors and the empirical validation of equilibrium business cycle models

Equilibrium business cycle models have typically less shocks than variables. As pointed out by Altug (1989) International Economic Review 30 (4) 889–920 and Sargent (1989) The Journal of Political Economy 97 (2) 251–287, if variables are measured with error, this characteristic implies that the model solution for measured variables has a factor structure. This paper compares estimation performance for the impulse response coefficients based on a VAR approximation to this class of models and an estimation method that explicitly takes into account the restrictions implied by the factor structure. Bias and mean-squared error for both factor- and VAR-based estimates of impulse response functions are quantified using, as data-generating process, a calibrated standard equilibrium business cycle model. We show that, at short horizons, VAR estimates of impulse response functions are less accurate than factor estimates while the two methods perform similarly at medium and long run horizons.     

Successive Sampling to Estimate Quantiles with P-Auxiliary Variables

The successive sampling is a known technique that can be used in longitudinal surveys to estimate population parameters and measurements of difference or change of a study variable. The paper discusses the estimation of quantiles for the current occasion based on sampling in two successive occasions and using p-auxiliary variables obtained of the previous occasion. A multivariate ratio estimator from the matched portion is used to provide the optimum estimate of a quantile by weighting the estimates inversely to derived optimum weights. Its properties are studied under large–sample approximation and the expressions of the variances are established. The behavior of these asymptotic variances is analyzed on the basis of data from natural populations. A simulation study is also used to measure the precision of the proposed estimator.     

Modelling response error in school effectiveness research

Statistical modelling of school effectiveness in educational research is considered. Variance component models are generally accepted for the analysis of such studies. A shortcoming is that outcome variables are still treated as measured without an error. Unreliable variables produce biases in the estimates of the other model parameters. The variability of the relationships across schools and the effects of schools on students' outcomes differ substantially when taking the measurement error in the dependent variables of the variance component models into account. The random effects model can be extended to handle measurement error using a response model, leading to a random effects item response theory model. This extended random effects model is in particular suitable when subjects are measured repeatedly on the same outcome at several points in time.     

A note on estimation based on record data

In data-processing standpoint, an efficient algorithm for identifying the minimum value among a set of measurements are record statistics. From a sequence of n independent identically distributed continuous random variables only about log(n) records are expected, so we expect to have little data, hence any prior information is welcome (Houchens, Record value theory and inference, Ph.D. thesis, University of California, Riverside, 1984). In this paper, non-Bayesian and Bayesian estimates are derived for the two parameters of the Exponential distribution based on record statistics with respect to the squared error and Linear-Exponential loss functions and then compared with together. The admissibility of some estimators is discussed.     

Estimating the Demand for Housing, Land, and Neighbourhood Characteristics

This paper provides estimates of the structure of demand for individual housing and neighbourhood characteristics and for land in two British cities. We estimate a hedonic price function, and from this obtain the implicit prices of house attributes. These prices are used to estimate a demand system for each city. These perform well, and enable us to calculate price and income elasticities for each of the non-dichotomous characteristics and for land. To counteract criticisms of demand estimates derived within the hedonic framework a method is developed for selecting an appropriate set of instrumental variables. Estimates derived from this method, however, differ only slightly from those obtained using the conventional techniques. Several features of these estimates provide insights into the unusual characteristics of the British housing market, the effects of constraints imposed by land use planning, and the effects of changing income distribution on the structure of demand.     

Regression with errors in variables: estimators based on third order moments

In this paper consistent and, in a well–defined sense, optimal moment–estimators of the regression coefficient in a simple regression model with errors in variables are derived. The asymptotic variance and other asymptotic properties of these estimators are given. As is known for a long time, serious estimation problems exist in this model. There are two ways out of this problem: using either additional assumptions or additional information in the data. A lot of attention has been paid to the use of additional assumptions. However, quite often this leads to rather unrealistic models. In this paper we use additional information in the data. That means here that, besides first and second order moments, third order moments are formulated as functions of the model parameters. Besides theoretical derivations a small study with generated data is discussed. This study shows that for samples larger than 50 the estimates we consider behave nicely.     

Assessing Variability of Complex Descriptive Statistics in Monte Carlo Studies Using Resampling Methods

Good statistical practice dictates that summaries in Monte Carlo studies should always be accompanied by standard errors. Those standard errors are easy to provide for summaries that are sample means over the replications of the Monte Carlo output: for example, bias estimates, power estimates for tests and mean squared error estimates. But often more complex summaries are of interest: medians (often displayed in boxplots), sample variances, ratios of sample variances and non‐normality measures such as skewness and kurtosis. In principle, standard errors for most of these latter summaries may be derived from the Delta Method, but that extra step is often a barrier for standard errors to be provided. Here, we highlight the simplicity of using the jackknife and bootstrap to compute these standard errors, even when the summaries are somewhat complicated. © 2014 The Authors. International Statistical Review © 2014 International Statistical Institute     

Evaluating estimation methods for ordinal data in structural equation modeling

This study examined the performance of two alternative estimation approaches in structural equation modeling for ordinal data under different levels of model misspecification, score skewness, sample size, and model size. Both approaches involve analyzing a polychoric correlation matrix as well as adjusting standard error estimates and model chi-squared, but one estimates model parameters with maximum likelihood and the other with robust weighted least-squared. Relative bias in parameter estimates and standard error estimates, Type I error rate, and empirical power of the model test, where appropriate, were evaluated through Monte Carlo simulations. These alternative approaches generally provided unbiased parameter estimates when the model was correctly specified. They also provided unbiased standard error estimates and adequate Type I error control in general unless sample size was small and the measured variables were moderately skewed. Differences between the methods in convergence problems and the evaluation criteria, especially under small sample and skewed variable conditions, were discussed.     

Adjusting for measurement error in the value added model: evidence from Portugal

The multilevel value added approach to measuring school effectiveness is now widely used. We propose a method to adjust for measurement error to investigate the extent to which this changes school effect estimates. It is applied to longitudinal data collected in the region of Cova da Beira (NUT III) for 1st, 3rd, 5th, 7th and 8th grades. Three different variance component models are considered, depending on the predictor variables included. Assuming measurement error occurs in explanatory and/or response variables, corrections are made for different values of the coefficient of reliability. Moreover, models are fitted under the assumption of either independent or correlated measurement errors.     

A General Method of Deriving the Inefficiencies of Banks from a Profit Function

This article develops a new method of estimating inefficiencies in joint production and shows that unlike the approaches utilized in the previous studies of inefficiency, this method maintains a consistent relationship between the error term of a profit function and the error terms of its price derivatives. A useful by-product of the method is a proof of a Hotelling-like lemma that relates stochastic input demand and output supply functions to stochastic profit functions. While the previous studies fit a single frontier to data on all firms, this paper estimates a frontier unique to every observed firm to allow each one to have a different potential of achieving maximal levels of profit. The new method is applied in the analysis of annual data, 1984–1989, for U.S. commercial banks. Both the analytical and numerical results of the paper show that the residual that the previous studies attribute to inefficiency includes the effects of excluded variables and of inaccuracies in the specified functional forms. Once accurate estimates of these effects are subtracted from the residual, the distortions in the measured inefficiencies should be considerably reduced. Consequently, this article considers how such estimates might be obtained.