Mr Peter Sims SAR.6.07

This course is available on the MPhil/PhD in Economic History, MSc in Economic History, MSc in Economic History (Research) and MSc in Global Economic History (Erasmus Mundus). This course is not available as an outside option.

This course is concerned with how economic historians have used quantitative methods and with how researchers design and structure a research project. In terms of quantitative methods the emphasis is on the applied and practical rather than the theoretical and will range from the use of simple summary descriptive statistics to multiple regression. The course will start with a consideration of broad issues in research design, this might include, for example, models, narrative and case studies. The rest of the course will then be concerned with quantitative issues, the problems of analysing and interpreting quantitative historical evidence. It will consider topics such as sampling and statistical distributions, correlation, simple and multiple regression, specification problems, hypothesis testing, logit and probit analysis, non-parametric tests, and modern time series analysis, although the content may vary slightly from year to year. The course will also provide students with training in using an econometrics software package. An important component of the course is the deconstruction of historical articles that have used quantitative techniques.

9 hours of seminars and 20 hours of computer workshops in the MT.

MT only. Three hours per week.

Students on this course will have a reading week in Week 6, in line with departmental policy.

Students are required to do weekly exercises and to submit one paper during the term.

C H Feinstein and M Thomas, Making History Count (2002); P Hudson, History by Numbers (2002); C H Lee, The Quantitative Approach to Economic History (1977); G Hawthorn, Plausible Words (1991).

Exam (100%, duration: 2 hours) in the January exam period.

Survey questions on feedback to students may be non-informative because assessed work comes later in the term than the survey.