Business Context
Lyft is testing a redesigned home screen in the rider app intended to increase completed ride requests by making destination entry and ride-type selection faster. A 14-day A/B test randomly assigned eligible riders to the current Lyft home screen (control) or the redesigned version (treatment).
Problem Statement
Determine whether the redesigned Lyft home screen increased ride request conversion rate enough to justify rollout.
Given Data
| Group | Exposed Riders | Riders Who Completed a Ride Request | Conversion Rate |
|---|
| Control | 82,400 | 14,008 | 17.0% |
| Treatment | 81,900 | 14,578 | 17.8% |
Use a two-sided hypothesis test at a 5% significance level. Also compute a 95% confidence interval for the lift in conversion rate.
Requirements
- State the null and alternative hypotheses for the conversion-rate difference.
- Compute the sample proportions for control and treatment.
- Calculate the pooled proportion and pooled standard error for a two-proportion z-test.
- Compute the z-statistic and two-sided p-value.
- Construct a 95% confidence interval for the difference in conversion rates.
- Decide whether Lyft should roll out the redesigned home screen based on both statistical and practical significance.
Assumptions
- Rider-level randomization was implemented correctly.
- Each rider is counted once during the analysis window.
- No major outages or pricing changes disproportionately affected one group.
- The normal approximation is appropriate because both groups have large sample sizes and many conversions/non-conversions.
