# Answer key · RQ25

## Did returning visitors buy more often than new visitors?

At this shop, returning visitors accounted for more sessions with recorded revenue. Which group had the higher share of sessions with revenue? Divide each group’s revenue-positive sessions by all of its sessions.

- One record is — one recorded website session
- Period — One-year sampling period; dataset released 2018. Calendar dates unspecified.
- Scope — E-commerce website; country unspecified

## The table

Purchase in session × Visitor type

| Purchase in session | Returning visitor | New visitor |
| --- | ---: | ---: |
| Revenue recorded | 1,470 | 422 |
| No revenue recorded | 9,081 | 1,272 |

## Returning visitors account for most sessions with revenue. Does that mean they bought more often?

1. Yes, because they had more sessions with revenue.
2. Yes, because they made up most of the revenue sessions.
3. No. A smaller share of their sessions had revenue.

Choose one statement. Show the counts and denominators that justify it, then describe one limit of the comparison.

## The figures behind the check

- **1. Not supported.** 1,470 against 422 compares counts, and returning visitors simply had many more sessions: 10,551 against 1,694.
- **2. Not supported.** 77.7% of the 1,892 revenue sessions came from returning visitors. That base is revenue sessions, so it describes who the buyers were, not how often each group bought.
- **3. Supported.** Using each group’s own sessions as the base, 13.9% of returning-visitor sessions had revenue against 24.9% of new-visitor sessions, 11.0 percentage points lower.

- Base = each visitor type’s sessions: Returning visitor 13.9%, New visitor 24.9%.
- Base = the 1,892 revenue sessions: Returning visitor 77.7%, New visitor 22.3%.
- No base (raw counts): Returning visitor 1,470, New visitor 422. The source’s Other visitor class is excluded, leaving 12,245 sessions.

Quote the share with its denominator: of returning-visitor sessions, of new-visitor sessions. The same counts look reversed when divided by revenue sessions.

## Limits

These are selected sessions from one website over one year. A recorded purchase is only yes or no: the file gives no order amount and no experimental assignment, so nothing here says what a change to the site would do. Each user contributes exactly one session in the sample year, so a returning visitor’s earlier visits can lie outside the sample.

## Source

Data: C. Sakar, Yomi Kastro. Online Shoppers Purchasing Intention Dataset. UCI Machine Learning Repository. Licensed under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). Crosstabs selected, recoded and aggregated the records into this table.

- [C. Sakar, Yomi Kastro. Online Shoppers Purchasing Intention Dataset. UCI Machine Learning Repository.](https://archive.ics.uci.edu/dataset/468/online+shoppers+purchasing+intention+dataset)
- [Reproduce it · RQ25](https://www.crosstabs.com/examples/rq25-purchasing-among-new-and-returning-visitors)
- [CSV](https://www.crosstabs.com/examples/rq25-purchasing-among-new-and-returning-visitors/download/csv) · [JSON](https://www.crosstabs.com/examples/rq25-purchasing-among-new-and-returning-visitors/download/json)
