STATA and Econometrics: A Bite-sized Overview

 

Course Aims

This half-day intensive course provides a broad refresher of the theory and practice of econometrics. The focus is on hands on use of an econometrics package to estimate the parameters of economic models, and test economic hypotheses.

N.B. We also run a two day 'Intro to Quant Analysis Using STATA' course. Please get in touch for the full course outline. 

 

Course Objectives (Specific Learning Outcomes)

On successful completion, attendees should

  • have refreshed their knowledge and understanding of a large part of core econometric concepts.
  • be able to test statistical assumptions and economic hypotheses.
  • be familiar with the techniques of spotting pitfalls in an econometric model.

 

Course Outline

1. Introduction

2. Bivariate and Multiple Regression Analysis – Led by tutor

  • Classic Linear Regression Model
    • Five assumptions (Unbiasedness and Efficiency)
  • Goodness of Fit (R2)
  • Hypothesis testing (testing a single linear restriction which involves 2 parameters
  • Multiple Linear Regression Model
  • Hypothesis testing (testing a single and multiple restrictions)
  • Failure of the classical regression assumptions
    • Heteroskedasticity
    • Multicollinearity
    • Endogeneity (Ommited variable bias)

3. Running regressions and Interpreting results (group work + feedback from tutor)
Here we aim to implement multiple regression in Stata. Attendees will be expected to run regressions with multiple explanatory variables and interpret the resultant output. This includes an understanding of estimated coefficients, standard errors, confidence intervals, t-stats and p-values, R2, F-tests and recovering the variance-covariance matrix.

We also cover diagnostic commands such as residual plot analysis, multicollinearity, heteroskedasticity and functional form tests. Specific attention will be paid to omitted variable bias and possible ways to fix this.

A running commentary is provided by the tutor on all these matters. The data will be based on a real world problem; estimating the return to education using the Labour Force Survey data. This classic problem will introduce participants to issues of data manipulation, functional form and omitted variable bias.

4. Wrap up

For further information, or to register your interest in future dates, please contact us.

Course Summary: 

  • Duration: Half Day or One Day 
  • Price: Please contact us
  • Dates: 15 Nov 2016
  • Code: BS/STAT

Book Online ButtonDownload Booking Button

 

course-footer