Statistical significance report

This report does a statistical test to see whether any differences in response rate are significant, given the size of the samples used. The (±) column shows how different the response rate could have been on another occasion with a similar sized run. For example, a response rate of 30% ± 20% could be anywhere between 10% and 50%. The Z-Score for a given row shows whether this row's result is significant — a value less than 1.96 is not significant, meaning that a different run could well yield the opposite result when compared against the control.

Viewing the report¶

1. Navigate to a Message step which includes a Response step.

2. Select the Statistical Significance Report from the Results view.

   

3. Refresh the Report on run list and pick a run from the drop-down.

   

4. Click Generate.

Example Statistical Significance Report broken down by Brochure:

The (±) column shows how different the response rate could have been on another occasion with a similar sized run. E.g. a response rate of 30% ± 20% could be anywhere between 10% and 50%. The Z-Score for a given row shows whether this row’s result is significant – a value less than 1.96 is not significant, meaning that a different run could well yield the opposite result when compared against the control.

Z-Scores confidence table¶