Finding cheating schools in Texas with linear regression

Some schools in Texas had an odd jump in standardized test scores between different grades. Was it cheating? Linear regression is on the case!

linear regression standard deviation education standardized testing

Readings and links


Some schools in Texas scored suspiciously well on some of their standardized tests. Could it have been cheating? The Dallas Morning News is on the case!

This chapter takes a look at using linear regression to find outliers. Taking the expected relationship between standardized scores in math and reading, we'll find schools that did poorly on one yet unexpectedly well on the other. Topics include standard deviation and residuals, along with drawing regression plots.

Notebooks, Assignments, and Walkthroughs

Finding outliers with standard deviation and regression

A full reproduction of the Dallas Morning News's analysis of TAKS standardized test scores. Uses standard deviation and regression to uncover suspicious school scores.

Finding outliers with regression residuals (short version)

A denser reproduction of the Dallas Morning News's analysis.

Reproducing the graphics from The Dallas Morning News piece

If we've had enough with numbers, this one is a walkthrough of how to use seaborn to reproduce the graphics that were published with the Dallas Morning News piece.