Introduction to MSc in Statistics
The MSc in Statistics provides students with intensive training in statistics applicable to the social sciences, econometrics and finance. The aim of the course is to foster an interest in applied statistics and equip students for work as professional statisticians. The MSc also provides an opportunity to study specialist courses in related disciplines. There are excellent prospects for employment and further study for our graduates. Former MSc in Statistics students have taken up positions in consulting firms, banks, and in the public sector, where there is a shortage of well-qualified statisticians. Many go on to take higher degrees.
Social Statistics related courses:
ST405 Multivariate methods
An introduction to the theory and application of modern multivariate methods used in the Social Sciences: Principal components analysis, factor analysis, latent variable models, multivariate normal distribution, exponential family, and structural equations models.
ST411 Generalised linear modelling and survival analysis
Generalized linear modelling with an emphasis on diagnostics, estimation, and inference. Variables belonging to the exponential family. Survival analysis. Linear regressions. Variable selection and model building. Deletion diagnostics. Analysis of variance. Transformation of the response, constructed variables. Maximum likelihood estimation. Exponential family and generalized linear models. Categorical data, binary variables and logistic regressions. Log-linear models and contingency tables. Exploratory analysis of survivor distributions and hazard rates. Regression modelling for survival data. The use of R for data analysis.
ST416 Multilevel modelling
A practical introduction to multilevel modelling with applications in social research. This course deals with the analysis of data from hierarchically structured populations (e.g., individuals nested within households or geographical areas) and longitudinal data (e.g. repeated measurements of individuals in a panel survey). Multilevel (random-effects) extensions of standard statistical techniques, including multiple linear regression and logistic regression, will be considered. The course will have an applied emphasis with computer sessions using appropriate software (e.g. Stata).
ST421 Developments in statistical methods
Our aim is to teach students important statistical methodologies that reflect the exciting development of the subject over the last ten years, which include empirical likelihood, MCMC, bootstrap, local likelihood and local fitting, model Assessment and selection methods, boosting, support vector machines. These are computationally intensive techniques that are particularly powerful in analysing large-scale data sets with complex structure. A selection from the following topics. Robustness of likelihood approaches: distance between working model and "truth", maximum likelihood under wrong models, quasi-MLE, model selection with AIC, robust estimation. Empirical likelihood: empirical likelihood of mean. Bayesian methods and Markov chain Monte Carlo (MCMC) basic Bayes, Gibbs sampler, Metropolis-Hastings algorithm. Elements of statistical learning: global fitting versus local fitting, linear methods for regression, splines, kernel methods and local likelihood. Model assessment and selection: bias-variance trade-off, effective number of parameters, BIC, cross-validation. Further topics: additive models, varying-coefficient linear models, boosting, neural network, support vector machines. The course will be continuously updated to reflect important new developments in statistics.
ST442 Longitudinal data analysis
A practical introduction to methods for the analysis of repeated measures data, including continuous and binary outcomes. Topics include: longitudinal study designs, (random effects) growth curve models, marginal models, dynamic (autoregressive) models, latent class models, and multiprocess models for multivariate outcomes. The course will have an applied emphasis with weekly computer classes using appropriate software (e.g. Stata).
MPhil/PhD application process
MPhil/PhD programmes of study in social statistics typically include both methodological development and the application of statistical methods to a social science field or to address new developments in social data, such as in sample surveys or social networks. Research topics may be identified in advance by the applicant or may be arrived at through communication with a potential supervisor. The relative emphasis on methodology/theory vs. application may vary. Methodological areas of research include latent variable modelling, multilevel modelling, Longitudinal data analysis, causal modelling, categorical data analysis, measurement error, missing data, survey sampling, model selection and Bayesian methods.
There are three possible start dates; in October, January and March/April. If the final result of your taught postgraduate degree is not published before the end of September your entry date should be the following January or later.
We recommend that you apply online as this is the quickest and cheapest method. Please refer to Graduate Admissions for further information.
When applying, you should provide evidence of your ability to undertake independent research and state your research topic as accurately as possible on a separate sheet. Your research proposal should address the following questions:
What is your general topic?
What questions do you want to answer?
What is the key literature and its limitations?
What are the main hypotheses of the work?
What methodology do you intend to use?
What theoretical/conceptual framework will you adopt?
What are your case studies, if any, and what are your case selection criteria
What previous research have you undertaken in this field?
Your proposal should be approximately 1,500 words in length. MPhil/PhD applications that are received without a research proposal that addresses these questions will not be considered.
We are primarily assessing the potential of the applicant for research and their topic. The following are guidelines of what to emphasise in the proposal.
A research question rather than a very broad research area
A statement of how the proposed research builds upon earlier research on the topic, with reference to 2 or 3 key papers, demonstrating an understanding of the area and the need for further research
Most topics will involve an application of the proposed methods to a substantive research question. Why this particular question? What dataset might be used?
Do you have the training and skills to undertake the proposed research? Be specific. A list of courses attended is not helpful as this can be seen in the CV and transcripts
Further information about PhD research areas can be found here.
In addition, you should submit a personal statement of between 1,000 and 1,500 words, describing your academic interests and your purpose and objectives in undertaking a doctoral research degree.
You should also submit a scanned copy of a marked assignment or a research paper/report, ideally from your most recent programme of study.
Further guidance about completing your application can be found here.