This course examines tools for analysing quantitative data beyond the basic linear model. Students will gain expertise in managing common problems in data analysis, such as violated assumptions and missing data, and complex data formats, such as panel data.  Topics will include binomial, multinomial, proportional, and count data; generalised linear models; fixed and random effects in panel data; and multiple equation models (such as instrumental variables estimators).

Two hours of lecture and two hours of tutorials per week over 11 weeks.

Lectures will run Tuesdays 12-2pm. 1 lecture will take place each week.

Tutorials will run Tuesdays 2-4pm. 1 tutorial will take place each week.

None

SPS5033 or URBAN5127 (which are equivalent courses) must completed before beginning SPS5062. This co-requisite may be waived if the student has previously completed a similar course at Glasgow or at another institution.

Students will complete four problem sets related to the topics of the course for 40% of the summative assessment. The sets will require students to execute statistical models using R and provide concise, written interpretations of the results. Each problem set will be weighted equally.

Students will conduct a quantitative data analysis project related to one of the topics addressed in the course. They will choose a dataset, research question, dependent variable, and independent variables and apply the skills acquired in the course to the data in order to answer the research question. The students will produce a paper of 3,000 words (excluding references and R code), in which they focus on the statistical analysis and present the results.  Each paper must contain an advanced regression model and diagnostic analysis.

Students will learn the assumptions of advanced regression models and be able to explain the situations for which they are useful. They will gain experience in applying these models to a variety of data sets and analysing whether they satisfy criteria such as unbiasedness, consistency, and efficiency. Working with publicly available data they select themselves, students will develop regression models and evaluate their adequacy for explaining social phenomena.

By the end of this course students will be able to access publicly available data sets with complex structures and analyse them using a variety of statistical tools available in the R programming language. They will be able to evaluate problems in the data and apply the appropriate estimators to create new statistical models. They will develop the capacity to evaluate the adequacy of these models.

Students must submit at least 75% by weight of the components (including examinations) of the course's summative assessment.