Regression Notes

Link to notes: Notes1a.PDF

To go to specific section, in Adobe Reader do Acrobat Find
i.e. Find: VI.Problems in Non-Standard Statistical Analysis

I. Objective
II. Means, Conditional Means, and Simple Regression

A. Mean
B. The Conditional Mean
C. Simple Regression
D. Regression in Deviation Notation and Moment Notation

III. Multiple Regression
IV. Variance, Covariance, Coefficient of Determination (R²), and correlation (r)
V. Tests of Hypothesis: Significance of Difference in Group Means; Variance in b

A. Tests of Significance of Difference in Group Means
B. Difference in Means as a Dummy Variable Regression
C. F-Statistic, Analysis of Variance, Hypothesis Tests on Several Parameters
D.Tests for Joint Significance of Multiple Regression Coefficient
E. F-Statistic, Analysis of Variance, Hypothesis Tests on Several Parameters
F. Tests for Joint Significance of Multiple Regression Coefficients

VI. Problems in Non-Standard Statistical Analysis

A. Multicolinearity
B. Omitted Variable Bias

VII. Basic Assumptions about the Population Model and Problems Due to their Violation
VIII. Violations of the Basic Assumptions

A. Autocorrelated Disturbances
B. Heteroskedasticity
C. Generalized Least Squares: An Example

IX. Distributed Lags

X. Estimation of Equations in Simultaneous Systems

A. Two examples of Simultaneous Equations bias of OLS
B. Consistent Estimators of Structural Equations
C. The Identification Problem

1. Breakdown of the indirect
2. Breakdown of Two-stage Least Squares
3. Linear Independence and Identification

D. Rules for Identification

1. The Order Condition (Counting Rule)
2. The Rank Condition
3. Identification through Other Restrictictions

XI. Miscellany

A. Partitioning R²
B. Several Alternative Measures of Impact or a Variable, or Group of Variables

Return to Econ 31
Return to Swarthmore Economics Home Page