A Detailed Evaluation of Adjustment Methods for Multiplicative Measurement Error in Linear Regression with Applications in Occupational Epidemiology

Biometrics ◽  
1997 ◽  
Vol 53 (3) ◽  
pp. 1008 ◽  
Author(s):  
Robert H. Lyles ◽  
Lawrence L. Kupper
Keyword(s):  
Measurement Error ◽  
Linear Regression ◽  
Occupational Epidemiology ◽  
Detailed Evaluation ◽  
Multiplicative Measurement Error

Effects of measurement error on Monte Carlo integration estimators of tree volume: critical height sampling and vertical Monte Carlo methods

2015 ◽  
Vol 45 (4) ◽  
pp. 463-470 ◽  
Author(s):  
Thomas B. Lynch
Keyword(s):  
Monte Carlo ◽  
Measurement Error ◽  
Sampling Error ◽  
Monte Carlo Integration ◽  
Critical Height ◽  
Volume Estimate ◽  
Cross Sectional ◽  
Individual Tree ◽  
Control Variate ◽  
Multiplicative Measurement Error

The effects of measurement error on Monte Carlo (MC) integration estimators of individual-tree volume that sample upper-stem heights at randomly selected cross-sectional areas (termed vertical methods) were studied. These methods included critical height sampling (on an individual-tree basis), vertical importance sampling (VIS), and vertical control variate sampling (VCS). These estimators were unbiased in the presence of two error models: additive measurement error with mean zero and multiplicative measurement error with mean one. Exact mathematical expressions were derived for the variances of VIS and VCS that include additive components for sampling error and measurement error, which together comprise total variance. Previous studies of sampling error for MC integration estimators of tree volume were combined with estimates of upper-stem measurement error obtained from the mensurational literature to compute typical estimates of total standard errors for VIS and VCS. Through examples, it is shown that measurement error can substantially increase the total root mean square error of the volume estimate, especially for small trees.

scholarly journals Occupational epidemiologist's quest to tame measurement error in exposure

2019 ◽  
Author(s):  
Igor Burstyn
Keyword(s):  
Measurement Error ◽  
False Positive ◽  
False Negative ◽  
Epidemiologic Study ◽  
Occupational Epidemiology ◽  
Exposure Misclassification ◽  
Exposure Limits ◽  
Workplace Exposure ◽  
Negative Results ◽  
Differential Exposure

I aimed to assess current practices and opportunities for addressing the problem of errors in exposure in occupational epidemiology. Occupational epidemiologists appreciate that errors in exposure are a concern, but almost none correct for these errors, although there are currently no theoretical and practical barriers for this inertia. The most serious barrier to change is a faulty belief that a well-conducted epidemiologic study suffers only non-differential exposure misclassification and that its sole impact is to attenuate risk gradients, causing a false negative. On the contrary, differential exposure misclassification is the most defensible model in occupational epidemiology, and errors in exposure increase chance of both false positive and negative results. Resistance to mathematical adjustment (correction) for errors in exposure is equivalent to denying the value of more valid exposure estimates and undermines the discipline’s relevance to protection of workers by informing workplace exposure limits.

scholarly journals How measurement error affects inference in linear regression

2020 ◽  
Author(s):  
Erik Meijer ◽  
Edward Oczkowski ◽  
Tom Wansbeek
Keyword(s):  
Measurement Error ◽  
Panel Data ◽  
Linear Regression ◽  
Error Variance ◽  
Standard Errors ◽  
Test Statistics ◽  
Wine Quality ◽  
Measurement Error Variance ◽  
Infeasible Estimator ◽  
Data Context

Abstract Measurement error biases OLS results. When the measurement error variance in absolute or relative (reliability) form is known, adjustment is simple. We link the (known) estimators for these cases to GMM theory and provide simple derivations of their standard errors. Our focus is on the test statistics. We show monotonic relations between the t-statistics and $$R^2$$ R 2 s of the (infeasible) estimator if there was no measurement error, the inconsistent OLS estimator, and the consistent estimator that corrects for measurement error and show the relation between the t-value and the magnitude of the assumed measurement error variance or reliability. We also discuss how standard errors can be computed when the measurement error variance or reliability is estimated, rather than known, and we indicate how the estimators generalize to the panel data context, where we have to deal with dependency among observations. By way of illustration, we estimate a hedonic wine price function for different values of the reliability of the proxy used for the wine quality variable.

Some Variations on Standard Income Measurement Error Models

2004 ◽  
Author(s):  
Brigham B. Frandsen ◽  
James B. McDonald
Keyword(s):  
Measurement Error ◽  
Gini Coefficient ◽  
Measurement Error Models ◽  
The Gini Coefficient ◽  
Parametric Specification ◽  
Income Measurement ◽  
The Difference ◽  
Multiplicative Measurement Error ◽  
The Relationship ◽  
Measures Of Inequality

Measurement error can have a significant impact on measures of inequality. Using a fairly flexible parametric specification of an independent multiplicative measurement error (IMME) model we explore the relationship between changes in the variance of measurement error, for a given mean of measurement error, on the Gini Coefficient. While the measured Gini is greater than the true Gini, the difference decreases as the variance of measurement error decreases. Copulas are used to relax the assumption of independence of measurement error and true income. In this case the measured Gini can be larger or smaller than the true Gini, depending on the correlation between true income and measurement error. Using the same approach with simulations the effect of a different distribution of measurement error is investigated.

Sign in / Sign up

Export Citation Format

Share Document