CPMCGLM

CPMCGLM performs p-value adjustment for generalized linear models when continuous explanatory variables are transformed or categorized to determine optimal variable transformations.

Key Features:

• P-Value Adjustment for Variable Transformation: Adjusts p-values when selecting optimal transformations of explanatory continuous variables in generalized linear models.
• Multiple Transformation Support: Supports transformations including dichotomization, Box–Cox transformation, and conversion to categorical variables with multiple levels.
• Significance Level Correction: Implements an asymptotic correction method for significance levels in generalized linear models proposed by Liquet and Commenges (2005).
• Resampling-Based Significance Estimation: Applies resampling techniques for estimating significance levels to control type-I error rates.

Scientific Applications:

• Biostatistical Modeling: Evaluates associations between transformed explanatory variables and outcomes using generalized linear models.
• Epidemiological Data Analysis: Identifies optimal cutoff points for continuous variables in epidemiological studies.
• Statistical Methodology Research: Supports development and evaluation of transformation-based modeling strategies in biomedical datasets.

Methodology:

CPMCGLM evaluates transformations of continuous explanatory variables in generalized linear models, adjusts p-values for multiple testing using asymptotic correction methods, and estimates significance levels using resampling techniques.