Measurement error in economic data can impact research findings. By analyzing revised GDP data and GDI estimates, researchers found that most models give the same results when using revised GDP. However, using GDI instead of GDP led to different results in 13% of models.