Anscombe's Quartet and the Datasaurus

Anscombe's Quartet and the Datasaurus


Read More: Wikipedia - Anscombe's quartet

For this activity, the following datasets are needed:

If needed, a free, no obligation, 30-day trial of SigmaXL can be downloaded here. Click here for installation instructions.

 Anscombe's quartet - Advanced Multiple Regression Example

Anscombe's quartet:

  1. Open Anscombe's Quartet Data.xlsx

  2. Click Sheet1. Click (1) SigmaXL > (2) Statistical Tools > (3) Advanced Multiple Regression >  (4) Fit Multiple Regression Model:



  3. Click (1) Use Entire Data Table and click (2) Next >>:



  4. Select y1 and click Numeric Response (Y) >>. Select x1 and click Continuous Predictors (X) >>. Click Next >>:



  5. We will leave the Term Generator at its default setting for Main Effects. Click Select All >> and OK >>:



  6. The Multiple Regression Model is generated for response y1 with term x1. Note that in the model summary, R-Square Predicted is 50.14%:



  7. Click sheet MReg1 - Residuals y1. Note the residual charts:



  8. Click Recall SigmaXL Dialog:



  9. Repeat steps 4 and 5 for response y2 with continuous predictor x2:



  10. Response y3 with continuous predictor x3:



  11. And response y4 with continuous predictor x4:



 Datasaurus - Scatter Plots Example

Datasaurus:

  1. Open Datasaurus_data.xlsx

  2. Click (1) SigmaXL > (2) Graphical Tools > (3) Scatter Plots:



  3. Click (1) Use Entire Data Table and click (2) Next >>:



  4. Select y and click Numeric Response (Y) >>. Select x, and click Numeric Predictor (X1) >>. Uncheck Trendline. Click OK >>:



  5. Rawr!