Research Magazine 2017

Executive Takeaways

■ Regression Discontinuity allows multiple forcing variables to determine treatment assignment. ■ When the magnitude of discontinuity in receiving treatment is small, the outcome by t- test is unreliable.

■ Our test is robust to magnitude of discontinuity, and is applicable with multiple forcing variables. ■ Whether the hypertension diagnosis decreases fat intake needs further investigation.

Zhaoguo Zhan, Assistant Professor of Economics

Made with FlippingBook - Online catalogs