Reporting error in weight and its implications for bias in economic models

Area of Science:

• Health Economics
• Biostatistics
• Public Health

Background:

• Economic research on obesity often relies on self-reported weight.
• Self-reported data introduces reporting errors, potentially biasing economic model estimates.
• Accurate weight data is essential for understanding obesity's economic and healthcare consequences.

Purpose of the Study:

• To quantify the extent and characteristics of weight reporting errors.
• To assess the impact of these errors on regression coefficients in healthcare consequence models.
• To compare economic models using self-reported versus measured weight.

Main Methods:

• Analysis of National Health and Nutrition Examination Survey (NHANES) data (2003-2010).
• Utilized datasets containing both self-reported and measured weight and height.
• Compared regression models using self-reported weight against those using measured weight.

Main Results:

• Weight reporting error is non-classical; underweight individuals overreport, while overweight/obese individuals underreport.
• Approximately 1 in 7 obese individuals are misclassified due to reporting errors.
• Reporting errors can lead to upward bias in coefficient estimates, overestimating healthcare utilization for obese men.

Conclusions:

• Weight reporting errors significantly impact economic models of obesity.
• Models should prioritize using measured weight data over self-reported data.
• Future social science datasets should include measured weight to improve accuracy.