Interpreting Effect Sizes

The head of marketing at Bizer, Inc. is trying to decide whether to include a new cholesterol medicine's average treatment effect in the company's advertisements for that medicine. He knows that on average, the medicine reduces cholesterol in respondents by about 1mg/dL and knows that the effect is statistically different from zero. However, he's not sure whether that number is large or small; that is, whether the medicine's average treatment effect is worth advertising.

If cholesterol is highly variable in the population (about 70% of the population has cholesterol between 160mg/dL and 240mg/dL), does an average treatment effect of 1mg/dL seem relatively small or relatively large? In other words, when an average treatment effect is substantially smaller than the standard error of the mean for a parameter, how would we typically interpret the size of that average treatment effect?

Possible Answers

The average treatment effect is large.
The average treatment effect is small.

Exercise

Interpreting Effect Sizes

Instructions

Possible Answers