<aside> 💡

This is part of the Statistics Add-On

</aside>

Summary Table Reports are great for analyzing and comparing large sets of data. They can be created using the “New” dropdown at the top of the report list.

From here, you can choose a base type which will represent the unit of each item of the table. For example, if the base type is “Patient” then each item will come from a different patient. Note, if you choose a base type other than patient, then if patients have more than one item of that type, then they can have data on multiple rows.

After selecting a base type, you can choose a default report inclusion filter, feature list, and population list. From the dropdowns, you can select to copy any of these from another report.

Once you hit save, you will see the following configuration options. Features are data points which will be aggregated across one or more sub-populations. Populations are defined in the “Populations” tab and can only be filters for inclusion in the report base type.

Opening the report, you will see the following. Each population shows up as a separate column alongside rows for each feature. Text features show up as a list of the top most common values (full lists are shown in downloaded reports) alongside rates as a percentage and absolute value. Numbers are shown as mean or median based on the feature definition. True/false variables such as boolean or “Yes”-If filters show up as a percentage of items which are true.

One sub-population is helpful for generating summary metrics for a list. Adding multiple populations will automatically add statistical significance p-values in an extra column on the right. For continuous variables, an ANOVA is performed, while for categorical or text variables, a chi-squared test is performed.