You are here: Department of Management / Prospective students / Postgraduate courses / Postgraduate modules / Statistical Methods for Business Research
Document Actions

Statistical Methods for Business Research

Convenor: Dr Frederick Guy
Assessment – 100% coursework. The assessed work will consist of a series of assignments in which you analyze data (provided), and write up your results.

Aims

This module provides a foundation in methods of descriptive statistics, statistical inference, analysis of variance, linear regression, and regression with qualitative dependent variables. It should allow you to begin using these methods, and also to interpret and assess research by others using these methods. It is geared to researchers who do not have a strong mathematical background.

Topics covered

  • Descriptive statistics
  • Statistical inference
  • Analysis of variance and correlation
  • Ordinary least squares (OLS) regression
    • Model specification and variable transformations
    • Goodness of fit, diagnostics, outliers
    • What does the regression tell us? Sign, size, significance
    • Standardization of variables
    • ‘Short’ regressions and added variables
    • Variable selection: building up (stepwise), testing down, theory
    • Heteroskedasticity, autocorrelation, collinearity
    • Qualitative (dummy) independent variables
    • Interaction terms
    • Omitted variables, errors in variables, endogenous regressors, and instrumental variables
  • Regression with qualitative independent variables – logit and probit estimators

There are many statistical techniques that you might find useful which are not addressed in this module (analysis of time series data, panel data, structural equations, multivariate methods, among others). However, if you want to use or understand any of those, you will be much better off if you already know the material presented here.

Methods

This is a hands-on module: most of the learning and assessment involve using the Stata programme on real data. Throughout, there will be an emphasis on:

  • Good research practice: every step documented in a program file
  • What can your data really tell you? Causation; substantive vs. statistical significance
  • Presentation of data and results, both in tables and in graphics

Background Reading

  • Gujarati, Damodar N., and Dawn C. Porter. 2009. Basic Econometrics. 5th ed: McGraw-Hill. This is the book I’ll assume you have access to. Earlier editions are also OK, and are widely available. (Note: this is not the same as Gujarati’s Essentials of Econometrics.)
  • Kennedy, Peter. 2008. A Guide to Econometrics. 6th ed: Wiley-Blackwell. This book has cornered the market in verbal econometrics – it explains both theory and method in some detail, mostly without equations. It can be a very useful resource, but is not a substitute for Gujarati and Porter or some other another standard text. Try in the library before buying.

Of the more advanced texts (those using matrix algebra – which takes some learning, but simplifies things once you know it), the following two are very clearly written:

  • Johnston, Jack, and John DiNardo. 1996. Econometric Methods. 4th ed: McGraw-Hill.
  • Goldberger, Arthur S. 1991. A Course in Econometrics. Cambridge, Massachusetts: Harvard University Press.