Model tab > Statistical Tests - Test for Proportions

Data Fields

• Category Field

Dropdown w/Expression

Represents non-numeric groupings (eg, product type, customer segment) used to compare differences across multiple samples.

• Value Field

Dropdown w/Expression

The numerical data being analyzed (eg, sales, conversion rate) to test whether differences between groups are statistically significant.

• Preview Model Data

Button

Opens a dialog to view the data that will be used by the model. See Model > Model tab.

Model Configuration

• Confidence Score

Number field

95

Represents the range in which the true prediction is expected to fall, based on a chosen confidence level. For example, a 95% confidence level means there’s a 95% chance the actual value lies within that range. Users can adjust the confidence level (eg, 90 to 99), with higher levels widening the interval to reflect greater certainty.

• Report Insight By

Dropdown

Lower Bound

Defines how insights are generated based on key statistical measures:

• Lower Bound - the minimum expected value in the analysis, helping identify underperformance.

• Upper Bound - the maximum expected value, highlighting high-performing data points.

• Mean - the average value, providing a baseline comparison for evaluating trends and differences.

• Report Insights For

  • Top N Percentages

  • Bottom N Percentages

Checkboxes

2

This setting allows users to focus on the Top N or Bottom N categories based on their proportions.

Top N - highlights the categories with the highest proportions (eg, top-performing products or highest conversion rates).

Bottom N - identifies the categories with the lowest proportions (eg, underperforming regions or lowest customer engagement).

Rules for Adding to Insights Feed Page from Scheduler

Radio-buttons

This section allows the user to select if the Quick Insight will be added to the Insights Feed page when generated from a Scheduled Task. See Model > Model tab.