Test whether average order value differs between A and B

This analysis requires DataCello Pro

Running the test requires DataCello Pro. The dataset needs a group field with exactly two values and a numeric order-value field. Also confirm that each row can be treated as an independent observation.

Configure the t-test

1. Select the A/B dataset, click the Analysis tab, and choose Hypothesis Test.
2. Select t-test from the method menu.
3. Set order value as Target and the A/B field as Group.
4. Resolve the run checklist at the top and click Run.

DataCello uses Welch's t-test here, rather than Student's t-test with an equal-variance assumption.

Read p-value and effect size together

The result includes the test statistic, p-value, Cohen's d, group count, mean, median, standard deviation, and a box plot.

• The p-value relates the observed result to a no-difference assumption.
• Cohen's d standardizes the size of the mean difference.
• Summaries and the box plot show spread and possible outliers.

Decision limits

A test result alone does not prove the causal effect of a treatment. Review random assignment, observation period, repeat users, attrition, and sample size. A paired t-test is not part of this workflow.