Multivariate Linear to Logistic Regression

Module fact file

Masters Level
Summer Term
10 credits
Contact Hours:
10:00-16:00 (Over two days)
MA Social Research

Contact details

Module lead
Matthew Bennett

Module description

This course aims to serve as a ‘bridging’ course between the Data Collection and Analysis modules and a wider range of short courses dealing with particular data analysis and statistical approaches, e.g. Factor Analysis. The difference between continuous and categorical data will be explained and appropriate modelling strategies introduced. Working with small to large-scale secondary data sets, students will learn to judge when to use what statistical analyses. We will also briefly survey the range of other statistical methods available, to enable informed choices about other techniques (or courses). Software requirements vary depending on the type of analyses, but given the current site licence provision, we will generally use SPSS for this course.

Learning outcomes

On completing this course, students will be able to:

  • Have a sound understanding of the role data analysis in social science research.
  • Develop an appreciation of different statistical approaches to analysing social science data.
  • Understand the principles of some of the most frequently used statistical modelling methods such as ordinary least square (OLS).
  • Understand the key assumptions required for different types of regression model.
  • Understand the importance of looking at a range of diagnostic information and the dangers of over-reliance on some popular summary statistics.
  • Understand how this model may be extended to logistic regression where the. dependent variable is categorical, how the approach differs, and alternative model specifications and extensions to multi-category dependent variables with and without an ‘order’.
  • Develop important data management skills and basic programming skills (writing syntax) using statistical software such as SPSS.
  • Identify and locate appropriate data sources for their research.
  • Critique existing research, and apply statistical modelling methods to their own  research questions.
  • Develop research questions and apply appropriate modelling techniques to address them according to the nature of the outcome variables.
  • Understand and interpret the outputs and findings in both statistical and substantive terms (i.e. one that relates to the research question).
  • Write-up the statistical results and present the findings in a journal-acceptable format (i.e. not merely cutting and pasting SPSS output in your report/paper).


A 2000-word data analysis report using logistic regression and multinomial logit models (fully referenced).  Inclusion of multiple linear regression will be possible but it must include fitting an interaction term.

Related degrees:

The optional modules listed on the website for this programme may unfortunately occasionally be subject to change. As you will appreciate key members of staff may leave the University and this necessitates a review of the modules that are offered. Where the module is no longer available we will let you know as soon as we can and help you make other choices.