Exercise

Pitfalls in Regression II

Another problem that can occur in regression is inappropriately interpretting correlation as causation. This means that we think our predictor variable causes our response variable, but actually they are just related to each other. Which of the following is the most likely potential reason why we cannot say that giving our sample money causes them to like us?

Instructions

50 XP

Possible Answers

The more money you give a participant, the more you smile as you hand it over - and people like smiling :)
Our sample may unrepresentatively like money.
We may be unrepresentatively likable.