Regression

This is not a statistics course, so we will not be able to cover regression in full detail. However, we need key parts of regression in order to discuss other methods, such as factor analysis. So will discuss the non-inferential parts of regression in some detail.

We also discuss variants of regression, especially logistic regression, and learn how to use Excel to estimate ad hoc regression-like models that you invent yourself.

QUOTE OF THE DAY

"While it is easy to lie with statistics, it is even easier to lie without them." -- Frederick Mosteller

Topics

- Correlation as the expected number of SD's that Y goes up when X goes up by 1 SD
- Linear models of the form Y = b₀+b₁X1+b₂X2+ε
- When to use various regression-like methods
- Using Excel solver

Datasets

Neogen dataset

Handouts

Introduction to Regression. [html]
Regression line. [html]
Least Squares Approach [html]
Multiple Regression [html]
regression-like models [html]

Worksheets

meanmedian.xls- a spreadsheet for calculating mean and median via Excel Solver
OLS regression with Solver
Logistic regression with Solver

Bibliography

- David Freedman, Robert Pisani & Roger Purves (1998) Statistics, 3^rd Ed. New York: Norton [a superb textbook]