Lecture 1
Econometrics - definition

Flow chart
  -   econometric models
  -   estimation
  -   structural analysis and forecasting
Illustrations
  -   demand for gasoline (not done this year)
  -   forensic economics: Exxon Valdez
 

Lecture 2
Review of Univariate and Bivariate Statistics

Random Variables
Relative Frequency - Probability
Properties of Probabilities
Probability Functions and Probability Density Functions
  -   Pj = Mj/N
Bivariate Distributions
Marginal Probabilities
Conditional Distributions
 

Lecture 3
Characteristics of Distributions; Normal Distribution

"Moments" of Distributions
  -   Expected Value and Properties
  -   Variance and Properties
  -   Covariance and Joint Distributions
  -   Correlation
Normal Distribution
Standardization of Normal Distribution
 

Lecture 4
Review of Estimation and Inference (univariate case)

Central Limit Theorem and Distribution of x-bar
  -   x-bar is normally distributed with mean mu and s.e. sigma squared/n
t Distribution
Point Estimates
Confidence Intervals (as sets of acceptable hypotheses)
 

Lecture 5
Properties of Estimators; Hypothesis Testing

  -   Linearity
  -   Unbiasedness
  -   Efficiency
  -   Consistency
Hypothesis Testing

Simple Regression Digression on Linearity
Sample Regression Function
4 Maintained Hypotheses for OLS
  -   E(u given X i) = 0
  -   Cov(ui,uj) = 0, for i not equal to j
  -   Var(ui given Xi) = sigma squared (constant)
  -   Cov(ui given Xi) = 0
Rules of Summation Notation
 

Lecture 6
Sample Regression Function (SRF)

SRF Properties
  -   goes through (x-bar,y-bar)
  -   yi hat bar = yi bar
  -   mean ei = 0
  -   ei uncorrelated with Xi with Yi hat
Variance b2
Variance b1
Gauss Markov Theorem: OLS is "BLUE"
Confidence Intervals for B1, B2
 

Lecture 7
Goodness of Fit

Variation in Y
TSS, ResSS, ExSS
R squared = ExSS/TSS
Using Regression for Prediction
 

Lecture 8
From Simple to Multiple Regression

Explanation of t-ratios on SHAZAM output
Multiple Regression Models
Omitted Variables Bias
  -   omitted variable is relevant to explaining Y
  -   omitted variable correlated with included variable
 

Lecture 9
Goodness of Fit

In Multiple Regression Models
  -   general model
  -   restricted model
R squared
Adjusted R squared
Hypothesis Testing in Multiple Regression
Confidence Ellipses
 

Lecture 10
Joint Hypotheses

Joint Confidence Ellipses
  -   F-tests
  -   ANOVA from means
Functional form
  -   Linear vs log-linear vs log-log models
  -   Use of /LOGLOG, /LOGLIN option on OLS
 

Lecture 11
More Flexible (General) Models

Interaction Terms
  -   non-constant derivatives
Dummy Variables
  -   intercept shifter(s)
  -   slope shifter(s)
 

Lecture 12
Categorical Explanatory Variables; Regime Changes

More on Dummy Variables
  -   m categories, m-1 dummies
  -   Changes in Regime
Class Survey
 

Lecture 13
Regression Pathologies: Multicollinearity

definition
evidence of Multicollinearity
  -   lousy t-ratios
  -   high R squared
  -   joint significance - "good" F-test results
  -   point estimates jump around as you add more regressors
What is the source of multicollinearity?
  -   look at pairwise correlations (STAT/PCOR)
  -   auxiliary regressions among RHS variables
  -   check R squared values
  -   high R squared - multicollinearity?
Remedies for Multicollinearity
  -   gather more data
  -   use information from prior studies
  -   drop an offending variable - CAUTION
  -   may just have to give up
 

Lecture 14
Regression Pathologies: Heteroscedasticity

definition - unequal error variances
consequences
  -   point estimates still unbiased
  -   standard error terms b1, b2 incorrect
  -   flawed hypothesis testing, t-ratios, P-values, CI's
Recourse
  -   use sigma i squared if known
  -   switch to weighted least squares (WLS)
  -   wi = 1/sigma i squared
Diagnosis: any relationship between error variances X variable(s)?
  -   OLS Y X/Resid=e
  -   GENR e2=e*e
  -   OLS e2 X
Try logs
Look diligently for any evidence of some relationship explaining error variances
Goldfeld-Quandt Test
  -   rank Observations (SORT) by variable that appears to be "culprit"
  -   omit "c" observations in middle (use SAMPLE command)
  -   examine (error variance 2)/(error variance 1)
 

Lecture 15
Heteroscedasticity, cont'd

  -   Weighted Least Squares
  -   Examples: assumptions about error variances
 

Lecture 16
Pathologies: Serially Correlated Errors

definition - autocorrelation in errors
Time Series
Detection
  -   run OLS, save residuals
  -   create lagged errors, assess correlation
Is correlation significant? - Durbin-Watson test statistics
FGLS Estimation
AUTO Y X gives desired b1, b2 purged of AR(1) error problems
 

Lecture 17
Pathologies: Endogeneity of Regressors

Exogenous Regressors - flip coin to decide participation; X's beyond observation's control
Endogenous Regressors - self selection into program; observation chooses own X values
Examples: debate over capital punishment and fishing avidity
Remedies - Two Stage Least Squares (overview only)
 

Lecture 18
Discrete Dependent Variables

Maximum Likelihood Estimation (MLE) intuition
PROBIT (and LOGIT) discrete choice models (overview only)
Example
 

Lecture 19
Overflow; Recap


COURSE OUTLINE LECTURE OUTLINES PROBLEM SETS PROBLEM SOLUTIONS COMPUTER LABS
SHAZAM EXAMPLES DATA SETS ONLINE QUIZZES GRAPHICS HANDOUTS

Updated: January 9, 1997
Prepared by: Trudy Ann Cameron