📊AP Statistics
3. A streaming platform's analytics team is investigating whether a new machine-learning recommendation algorithm increases user engagement compared to their standard algorithm. A random sample of 1,250 active users was selected for a study. Of these users, 650 were randomly assigned to receive recommendations using the new algorithm, and the remaining 600 users received recommendations using the standard algorithm. The team recorded whether each user clicked on at least one recommended movie within 24 hours.
In the group shown the new algorithm, 182 users clicked on a recommendation. In the group shown the standard algorithm, 138 users clicked on a recommendation.
State the appropriate null and alternative hypotheses for testing whether the proportion of all active users who would click on a recommendation is greater for the new algorithm than for the standard algorithm. Define any parameters used.
Identify the appropriate inference procedure and verify that the conditions for this procedure are met.
Calculate the test statistic and the -value for the procedure identified in Part B.
Based on the -value calculated in Part C, what conclusion should the analytics team make at the significance level?
Describe a Type II error in the context of this study. Identify one potential consequence of making a Type II error for the streaming platform.
