How to Temporarily Exclude Outliers from your study data

Overview

This guide explains how to temporarily exclude measurements from your study to manage outliers and ensure accurate analysis. You can exclude individual data points or entire animals for a particular metric.

Identifying Outliers

1. Visual Identification: Navigate to the graphs section to visually identify any significant variations. Outliers will appear as deviations from the group and are evident in spider plots by group color. 

2. Measurement Table Review: Review the table below the graphs to compare data across animals.  

Excluding Outliers

1. Exclude an Entire Animal row:

1. Hover over the animal ID field.
2. An option to exclude all data points on this row will appear. Click to select the row.
3. Click “submit Changes'' to exclude all data points for this measurement 

Note: This must be repeated for each measurement type, this only excludes the visible data points on this page, not all measurements for that animal.

2. Exclude a Specific Data Point:

1.  Hover over the specific data point under the graph.
2.  The cursor will change to a pointer with a line through it. 
3.  Click to select that particular data point.

After selecting the data points to exclude, click the Submit Changes button to apply the exclusions. Excluded data points are shown in Blue font with a line through the value. Multiple points can be excluded and submitted simultaneously to streamline your workflow. This will update the measurements immediately and remove them from both the graphs and subsequent analysis.

Pressing the submit button will also trigger webhook events to sync these changes with any integrations set up, such as data lakes or dashboards, ensuring all systems reflect the updated data.

 Reverting Exclusions

If you need to revert your exclusions and include all previously excluded measurements, simply press the Reset button. This will restore all settings to their original status, including all measurements back into the analysis.

Downloading Reports

Download and review exported files to see how exclusions impact your data representation. For instance:

• Prism Export: Excluded measurements will be reflected in the Prism files, maintaining consistency with the in vivo UI.
• Excel Summary Data: Excluded data points are marked in a different font or with a strike-through, and the analysis will reflect the adjusted dataset counts.

Please contact Benchling support to enable this feature for your organization if it is not already available on your Benchling In Vivo tenant. Please also reach out to us if you need to disable this feature.