Quick Insights for Statistical Tests
Intended audience: END-USERS DATA SCIENCE DEVELOPERS
AO Easy Answers: 4.3
Overview
Quick Insights can be generated at runtime by the user from Apps on the Easy Answers Results page, or from a scheduled task created by the Solution Developer using the Insight Composer in the AO Platform. This topic covers the Quick Insights relating to Categorical Data.
For Categorical Data
Summary of Quick Insights
| Chi-Squared Test | One-Sample T-Test | Two-Sample T-Test | Binomial Test | Correlation - Categories |
|---|---|---|---|---|---|
Number of Series | Should have 1 series | Should have 1 series | Should have 2 series | Should have 2 series | Should have at least 2 series |
Type of Data | Discrete Data (not float values) | Continuous Data (not integer values) | Continuous or Discrete Data | Discrete Data | Continuous or Discrete Data |
Charts Examples | Bar / Line Chart / Forest Plot | Bar Chart / Forest Plot | Stacked or Grouped Bar / Forest Plot | Stacked or Grouped Bar / Forest Plot | Correlation Heatmap |
Examples | The number of subscriptions from different states of the USA will show the highest and lowest. | Is the average percentage change of the stock different from zero, will compare the average of the series with a given value. | Is the average quarterly revenue different after the increase in the marketing budget. Will compare average quarterly revenues before and after the change in marketing budget. | Number of male subscriptions vs female subscriptions for each state/ month, will count of male subscriptions vs female subscriptions. | Time series of 2 stocks data will give insights into positive or negative correlation. |