Correlation and Linear Regression

What this page covers

The unit brings together resources on how two or more variables relate, and how to predict one from another. The mind maps are the largest group. Regression Analysis defines regression as a procedure for analysing associative relationships between a metric dependent variable and one or more independent variables. The first Linear Regression map lists the assumptions: a linear relationship between x and y, independent observations, normally distributed residuals with the same variability, and x measured without error.

Where to start

1. Open the Regression Analysis mind map for the big picture.
2. Move to a Linear Regression map. One covers the OLS estimator, slope and intercept, and measures of fit such as R and SER.
3. Read Multiple Linear Regression, which sets out the model Y = a + b₁x₁ + b₂x₂ and so on, with k explanatory variables.
4. Try the Linear Regression and Prediction Quiz, which uses a television-watching study as its worked scenario.

Using quizzes and maps together

Take a quiz first and note which questions you miss. Then go to the matching map. The chapter quizzes cover nonlinear models, heteroskedasticity and the meaning of a zero slope coefficient, so they suit a second pass once the basics are secure. Flashcard sets such as Regression diagnostics and assumptions and Pearsons and Spearmans population correlation coefficient are handy for short revision sessions. Notes on Spearman's rank and on correlation and causation sit alongside them.

Where it fits

This unit sits in the Correlation, Regression and Modelling strand of Statistics and probability. Most resources are pitched at university level, with a smaller share for school students. All were made by GoConqr users, so styles vary, and it helps to compare two or three on the same idea.

Ready to begin? Pick a map, take a quiz, and then make your own flashcards from what you find tricky.