Estimating standardized parameters from generalized linear models

Area of Science:

• Biostatistics
• Epidemiology
• Statistical Modeling

Background:

• Unrestricted (non-parametric) estimators for standardized rates and rate differences are unbiased but imprecise with sparse data.
• Parametric models offer more precise estimators, addressing the instability of traditional methods.

Purpose of the Study:

• To present a general method for creating smoothed estimators of standardized parameters using generalized linear models.
• To demonstrate the potential for simplified forms of these model-based estimators in specific scenarios.

Main Methods:

• Utilizing generalized linear models (GLMs) to construct estimators for standardized parameters.
• Applying parametric modeling approaches to rate estimation.

Main Results:

• The proposed approach provides a general framework for smoothed estimation.
• In common special cases, the model-based estimators simplify to exceptionally straightforward forms.

Conclusions:

• Generalized linear models offer a robust framework for developing precise, smoothed estimators of standardized rates.
• These smoothed estimators can overcome the precision limitations of traditional methods, particularly in sparse data settings.