Regression

An extensive series into the statistical concept of regression.
We start by developing your intuition around simple regression and make our way through multiple regression, categorical variables, logit models and a detailed inspection of the assumptions underlying regression. Enjoy!

Regression 1: The basics

What is regression | SSE, SSR and SST | R-squared | Errors (ε vs. e)

Regression 2: Adjusted R-squared

Degrees of Freedom ACTUALLY explained! | Adjusted R-squared

Regression 3: Understanding Regression Output

This video looks at EVERYTHING generated in a typical regression output and breaks it down. This is where you might find that “AHA!” moment.

Regression 4: Advanced Regression!

This tutorial contains more advanced concepts for those keen on pushing their understanding of regression beyond the basics.

There’s alot here so it is split into three separate videos:

Part (a) - Non-linear relationships | Logarithms

Part (b) - Categorical X variables | Interaction terms

Part (c) - Categorical Y variables | Logit models

Download the original jaybob dataset here, and the modelling workbook here.

Regression 5: Regression assumptions

EXTRA: Heteroskedasticity

A more in-depth discussion of the regression assumption violation of heteroskedasticity

EXTRA: Multicollinearity

A more in-depth discussion of the regression assumption violation of multicollinearity

Justin’s simulation spreadsheet (click here)

EXTRA: Autocorrelation

A more in-depth discussion of the regression assumption violation of autocorrelation. Durbin-Watson test | Breush-Godfrey test | Cochrane-Orchutt procedure | AR(1) procedure