🧪 A/B test of a checkout page
A fictional online shop showed 4,000 visitors either the current checkout page (A) or a redesigned one (B) at random, and recorded whether each visitor bought. 4000 cases, 3 variables. Field: business · level: beginner.
Open in StatLab Download: CSV · Excel (.xlsx, with codebook sheet) · SPSS (.sav, with labels)
What does each variable mean? (codebook)
| Variable | Label | Measure | Value labels | Values |
|---|---|---|---|---|
visitor | Visitor | scale | 1–4000, mean 2000.50 | |
variant | Page variant | nominal | 1 = A (current), 2 = B (new) | 2 categories |
converted | Bought | nominal | 0 = No, 1 = Yes | 2 categories |
Which analyses suit this dataset?
- Chi-square 2×2 — see the case study
- Full case study
Licence
Synthetic data produced by a seeded random process for teaching (no real people or organisations). Released under CC0 1.0 (public domain dedication) by Jupiter2u — free to use, share and adapt.