Measurement Error and Misclassification in Statistics and Epidemiology

Gustafson, Paul

doi:10.1201/9780203502761

Cited by 246 publications

(209 citation statements)

References 0 publications

Supporting

Mentioning

201

Contrasting

Unclassified

Order By: Relevance

“…Another well-known situation in which a similar phenomenon arises is in dichotomization of a miss-measured continuous variable (Gustafson, 2003). Thus, our proposed quasi-Berkson error structure emphasizes the fact that the independence of measurement error and outcome may not be retained on transformation of a mismeasured covariate.…”

Section: Resultsmentioning

confidence: 82%

Bias in the estimation of exposure effects with individual- or group-based exposure assessment

Kim

Richardson

Loomis

et al. 2010

J Expo Sci Environ Epidemiol

View full text Add to dashboard Cite

In this paper, we develop models of bias in estimates of exposure-disease associations for epidemiological studies that use group-and individual-based exposure assessments. In a study that uses a group-based exposure assessment, individuals are grouped according to shared attributes, such as job title or work area, and assigned an exposure score, usually the mean of some concentration measurements made on samples drawn from the group. We considered bias in the estimation of exposure effects in the context of both linear and logistic regression disease models, and the classical measurement error in the exposure model. To understand group-based exposure assessment, we introduced a quasi-Berkson error structure that can be justified with a moderate number of exposure measurements from each group. In the quasi-Berkson error structure, the true value is equal to the observed one plus error, and the error is not independent of the observed value. The bias in estimates with individual-based assessment depends on all variance components in the exposure model and is smaller when the between-group and between-subject variances are large. In group-based exposure assessment, group means can be assumed to be either fixed or random effects. Regardless of this assumption, the behavior of estimates is similar: the estimates of regression coefficients were less attenuated with a large sample size used to estimate group means, when between-subject variability was small and the spread between group means was large. However, if groups are considered to be random effects, bias is present, even with large number of measurements from each group. This does not occur when group effects are treated as fixed. We illustrate these models in analyses of the associations between exposure to magnetic fields and cancer mortality among electric utility workers and respiratory symptoms due to carbon black.

show abstract

Section: Resultsmentioning

confidence: 82%

Bias in the estimation of exposure effects with individual- or group-based exposure assessment

Kim

Richardson

Loomis

et al. 2010

J Expo Sci Environ Epidemiol

View full text Add to dashboard Cite

show abstract

“…Closed form posterior distributions can be found for the naïve analysis and the validation alone analysis, however the posterior for the ME model is not straightforward. A similar problem is outlined by Gustafson (2003) who provides the solution. In more complex cases, Bayesian analysis using Markov Chain Monte Carlo could be undertaken using WinBUGS (Spiegelhalter et al.…”

Section: Combining Low and High Quality Datamentioning

confidence: 98%

The Bayesian conditional independence model for measurement error: applications in ecology

2010

View full text Add to dashboard Cite

The measurement error model is a well established statistical method for regression problems in medical sciences, although rarely used in ecological studies. While the situations in which it is appropriate may be less common in ecology, there are instances in which there may be benefits in its use for prediction and estimation of parameters of interest. We have chosen to explore this topic using a conditional independence model in a Bayesian framework using a Gibbs sampler, as this gives a great deal of flexibility, allowing us to analyse a number of different models without losing generality. Using simulations and two examples, we show how the conditional independence model can be used in ecology, and when it is appropriate.

show abstract

“…However, this does not appear to be the case in our study. To apply Bayesian correction for measurement error, we also need to characterize measurement error in X and derive its distribution (Gustafson, 2003), which also does not appear to be straightforward. Application of the regression calibration approach is hindered in this case by the absence of a gold standard of D through a validation study (Rosner et al, 1989;Spiegelman et al, 1997) and it is not clear how it would be obtained from geocoded data.…”

Section: Discussionmentioning

confidence: 99%

“…A rich statistical body of knowledge exists on the measurement error problem and how to use information on uncertainty in a dependent variable to obtain an unbiased estimate of a regression parameter (Carroll et al, 1995;Gustafson, 2003). In environmental hygiene and exposure modeling, the measurement error problem arises mostly in the context of linear regression of determinants of exposure on the measured levels of air pollution.…”

Section: Introductionmentioning

confidence: 99%

Impact of measurement error on quantifying the importance of proximity to point sources of air pollution

Burstyn

2008

J Expo Sci Environ Epidemiol

View full text Add to dashboard Cite

This project was motivated by the investigation of the impact of primary oil and gas infrastructure on levels of air pollutants in western Canada. In the published models, we assumed that the distances between sources and air monitors were the key determinants of exposure and were measured precisely. These models related the logarithm of air pollutant concentration to a function of separation distance (''distance weight''). We undertook a simulation study to determine the impact on the observed source-pollutant association of uncertainty in the separation distance and the number of relevant sources per monitoring station. We observed that both the number of sources in the vicinity of the monitoring station and the extent of error in the estimate of separation distance influence the estimate of the slope of the source-pollution association. Measurement error tended to attenuate the association and degrade power, whereas the greater number of sources per monitoring station also led to a shallower observed slope. Attempts to correct the estimates of the slope were hampered by the non-standard nature of the frequency distribution of the difference between distance weights based on true and mismeasured distances. Our results revealed unanticipated challenges in the interpretation and estimation of the original analyses.

show abstract

Measurement Error and Misclassification in Statistics and Epidemiology

Cited by 246 publications

References 0 publications

Bias in the estimation of exposure effects with individual- or group-based exposure assessment

Bias in the estimation of exposure effects with individual- or group-based exposure assessment

The Bayesian conditional independence model for measurement error: applications in ecology

Impact of measurement error on quantifying the importance of proximity to point sources of air pollution

Contact Info

Product

Resources

About