Group Comparison Analyses allow you to compare the difference in performance between to groups of data.
|Since data analysis is iterative in nature, you can edit and re-run your Analysis configuration at any time. Alternatively, you can copy the Analysis and edit copy instead, to leave your original Analysis configuration intact. Refer to Duplicating and Editing an Analysis
- Understanding Sisu's Analysis Types
- Interpreting a Group Comparison Analysis
- Tuning an Analyses with Advanced Settings
What is a Group Comparison Analysis?
A Group Comparison Analysis is one of the three analysis types that Sisu performs:
- General Performance
- Time Comparison
- Group Comparison
In this type of analysis, the data between multiple groups are compared against each other. Groups are mutually exclusive subgroups of data. Common examples of this are:
- Sales from TX versus Sales from CA
- Completed orders versus uncompleted orders
- Sessions that converted versus sessions that did not convert
Sisu analyzes all possible combinations in each of the groups, determines the relevant statistics for each subgroup, and then uses our algorithm to determine which subpopulations are statistically significant.
Statistical significance can be driven from changes in metric calculation, subpopulation size, or combination of the two.
|Statistical significance can be driven from changes in metric calculation, subpopulation size, or combination of the two.|
For more details, please refer to Understanding Sisu’s Analysis Types.
Configuring a Group Comparison Analysis
To create and configure a Group Comparison Analysis:
- Ensure that you have either uploaded data to Sisu or have connected to a data source.
Refer to the “Connecting to External Data Sources” articles in the Data Input & Connections section for details on connecting to a data source.
- From the Projects module, select the Project Name you wish to work with.
Refer to Understanding Project & Analysis Hierarchy for more information.
- Select a Metric for your analysis.
Refer to Defining a Metric for details.
- Click + (plus sign) to create a new Analysis.
- Select “Rename” from the dropdown menu on the new Analysis and give it a descriptive title.
- In the Compare menu, select Groups.
In this modal, define the two groups that you want Sisu to compare using the fields provided for both Group A and Group B.
Optional. This will appear in the name of the comparison type for future use.
For example: “Compare Morning vs. Night” where Morning is Group A name and Night is Group B name.
Data Field dropdown
Select the data point (column) in your data set that will define the group.For example: ORDER_DAY_PART
Select the logical operation to determine a condition for the drill down. Options include equal to, less than or equal to, greater than or equal to, less than, greater than, and not equal to.For example: = STORE_NEWYORK
Select Data Value dropdown
For equals to and not equals to, this dropdown will reveal all values in the field and the value selected will create the group.
For example: morning for Group A and night for Group B
Save and Run
Click to save these settings and run the Analysis immediately if all other settings are complete.
Click to save these settings and return to other Analysis settings without running the analysis yet.
Click to clear your current settings in this modal and start over.
Refer to Creating & Managing Queries for details.
- Optional: Select Custom filter to define any filters you would like to apply to the Analysis, and click Save to save the filter and return to the Analysis settings (or Save and run if all settings are complete).
Refer to Setting Custom Filters for more information.
- Optional: Choose any additional settings you would like for your Analysis.
Refer to Understanding Sisu's Statistical Models for more information on available advanced settings.
- If you selected Save in the define Groups modal, click Run.
Sisu will process the Analysis and display the results. This example shows orders made in the morning compared to orders made in the Evening.
If you receive a “Run failed” message, you can click “Show details” to display information about what went wrong. Usually, it’s a problem with the data that can be easily fixed.
Refer to Troubleshooting Common Issues for further assistance.
|NEXT STEP: Refer to Interpreting a Group Comparison Analysis for information on understanding the analysis results.|