This is all super-simple; still, it might be useful. In class today a student asked for some intuition as to why, when you’re regressing y on x, measurement error on x biases the coefficient estimate but measurement error on y … Continue reading