In this article
The Settings menu - - allows you to apply Significance Testing, select whether you show the results as counts, percentages or both, and show or hide the totals.
Figure 1 - The Settings options
If the survey has a weight model for use with the IA report (the model must be named IAWeight), then the option Show weighted data will be available in the menu (go to Applying Weighting to the Data for more information). If no IA weight model is available, the menu option will not be listed.
Rating Statistics allows you to show only the specified number of elements in the table. For example, if you select Top 2, only the top 2 elements will be displayed, or if you type 5 in Top N, only the top 5 elements will be displayed; the other elements will be removed from the table. You can also type the % character into the field after the figure - in this case the top/bottom x% of the elements will be displayed.
Figure 2 - The Rating Statistics options
You can also perform Significance Testing on the results - select Significance Testing and choose a suitable confidence level - -. The higher the confidence level, the more certain you can be that there really is a difference in the two groups being tested. For example, 90% confidence means that there is a 10% chance that a difference in scores could have been found purely through the effects of sampling.
Figure 3 - Selecting a confidence level
Most of the statistical tests are based on standard t tests that assume that the two samples being compared are independent of each other. When the columns of a table are formed from the categories of a multi question, data from the same case can be present in both of the columns being tested. This is known as overlapping data, and it means that the two samples cannot be considered independent. Overlapping Samples should be used when the columns to be tested contain data that are not independent. A separate algorithm “Choi-Stablein” should then be used to calculate the significance test.
The settings in this menu will be maintained in saved reports and exports. Once the functionality is enabled and a banner is applied to create a table, Significance Testing will automatically highlight applicable table cells in green, and letters will indicate the comparable columns. Note that this feature uses the same column proportional T-test as Reportal, but is limited to a single level of confidence.
Check or uncheck the options as required. Note that you must have at least one of the counts and percents options checked; you cannot uncheck both.