mecor

mecor implements measurement-error correction methods for regression models with continuous outcomes to reduce bias in estimates of covariate–outcome associations.

Key Features:

• Measurement Error Correction Methods: Implements regression calibration and maximum likelihood methods to correct measurement error in covariates and outcomes within regression models.
• Study Designs for Parameter Estimation: Uses information from internal validation studies, replicates studies, calibration studies, and external validation studies to estimate parameters of the measurement-error model.
• Correction for Continuous Covariates and Outcomes: Provides procedures to correct measurement error in continuous covariates and methods of moments approaches to correct measurement error in continuous outcomes.
• Variance Estimation: Supplies closed-form solutions and bootstrap techniques for variance estimation of the corrected estimators.

Scientific Applications:

• Epidemiology: Improves estimation of exposure–disease associations in regression analyses by correcting measurement error.
• Clinical research: Adjusts clinical covariate–outcome associations for measurement error to reduce bias in effect estimates.
• Environmental studies: Corrects measurement error in exposure assessment to improve regression-based inference in environmental analyses.

Methodology:

Uses information from internal validation, replicates, calibration, and external validation studies to parameterize measurement-error models; applies regression calibration and maximum likelihood methods to adjust for covariate errors; implements methods of moments to address outcome measurement error; and provides closed-form and bootstrap variance estimation.