Session Ready
Exercise

Analyzing A/A data

Just as for a normal A/B experiment, we need to statistically analyze our A/A results to be sure there really was no effect. Here we'll build a logistic regression again, this time hoping for a null effect.

Instructions
100 XP
  • Load the package we've used in the past to clean up modeling results.
  • Set our independent variable to condition.
  • Set the family argument to the expected value for a logistic regression.