Data Studio Statistical Significance Table

gs://data-studio-statistical-significance-table/prod

github.com/phiture/data-studio-statistical-significance-table

Contents

• Usage
  • What you will need
  • Adding the table to your report
  • Column naming
  • Troubleshooting
    • There is no Statistical Significance column
    • The table displayes N/A
• Development
  • Setup
  • Scripts

Usage

What you will need

• Working knowledge of Google Data Studio
• A report in Google Data Studio

Adding the table to your report

1. Add the table to your report by ID, using gs://data-studio-statistical-significance-table/prod as the Manifest path.
2. Enable Community Visualizations in your data source

Column naming

The table needs the following data to calculate statistical significance. In order to specify which data the table should use, please make sure you have one dimension or metric for each category below (named as specified). In the case that multiple column names match the same category, the last matching column will be used.

• Unique Recipients - The number of non control group recipients
  • Unique Recipients
  • Sends
  • Sent
  • Target Sends
  • Variant Entries
• Conversions - The number of non control group conversions
  • Conversions
  • Target Conversions
  • Variant Conversions
• Control Entries - The number of control group recipients
  • Control Entries
  • Control Sends
• Control Conversions - The number of control group conversions
  • Control Conversions

If at least one dimension or metric from each category is present the table will automatically display an additional Statistical Significance column.

Troubleshooting

There is no Statistical Significance column

This is because you are missing required columns. See Column naming.

The Statistical Significance column displays N/A

The data in this row is not valid data for calculating statistical significance. Hover over the cell to see more information.

Development

• NodeJS

This repository is forked from anvilinsights/data-studio-react-starter. The original README is still in this repository in /doc/STARTER_README.md. The statistical significance formulas in statisticalSignificance.ts were modified from www.math.ucla.edu/~tom/distributions/normal.html.

Setup

• Get access to a storage bucket in Google Cloud
• Clone the repository and install dependencies with git clone and npm i
• Edit dsccViz.gcsDevBucket and dsccViz.gcsProdBucket in package.json to point to the folders you would like to deploy to.

Scripts

npm start

Start the visualization locally

npm run deploy:dev

Deploy the visualization to your development folder. Changes will be reflected instantly.

npm run deploy:prod

Deploy the visualization to your production folder. If you do this, devMode in manifest.json is set to false and Data Studio will cache your visualization for 12 hours. Read Google's documentation for more information.