This study develops and estimates a model of measurement error in self-reported health conditions. The model allows self-reports of a health condition to differ from a contemporaneous medical examination, prior medical records, or both. The model is estimated using a two-sample strategy, which combines survey data linked medical examination results and survey data linked to prior medical records. The study finds substantial inconsistencies between self-reported health, the medical record, and prior medical records. The study proposes alternative estimators for the prevalence of diagnosed and undiagnosed conditions and estimates the bias that arises when using self-reported health conditions as explanatory variables.

Document Type

Working Paper


Summer 8-2016


Measurement Error, Disease Prevalence, Diabetes, Hypertension, Medical Examination, Self-Reported, Health Condition




Working Papers Series


Health Policy | Medicine and Health | Public Affairs, Public Policy and Public Administration | Social and Behavioral Sciences



Additional Information

Working paper no. 191

The authors would like to thank John Cawley, Gary Engelhardt, Alfonso Flores-Lagunes, Bruce Meyer, Jeffrey Kubik, Yoonseok Lee, and seminar participants at Cornell University for valuable comments.

wp191.pdf (1495 kB)
Accessible PDF version


Local Input

Creative Commons License

Creative Commons Attribution 3.0 License
This work is licensed under a Creative Commons Attribution 3.0 License.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.