BASS Philosophy and Criminology

This course will equip students with the knowledge and skills to answer theoretically-driven research questions involving causality in the social sciences. Specifically, student will use modern causal theory to specify and fit linear and binary logistic regression models using the R software platform.

The unit aims to:

(i) Give students an introduction to the causal theory of Directed Acyclic Graphs (DAGs)

(ii) Show students how DAGs can be seen as representations of theories in social science and other domains.

(iii) Show how DAGs and causal theory can be used to guide the specification of quantitative statistical models, specifically linear and binary logistic regression models.

(iv) Give students an introduction in how to use the R software package to specify and fit linear and binary logistic regression models to real-world social data, based upon prior causal analysis of DAGs.

(v) Show students how to interpret the results of the regression models, and make inferences from them to the wider population.

Each week except the first, students will be given homework activities (something to read, and/or watch, and/or do).

During the following 2-hour session we will review and explore those activities to check our understanding. It is imperative that students carry out the homework activities before the session, as the sessions will not be purely lectures as such; they will be a chance for us to ask each other questions to check our understanding of the material.

The sessions will feature presentations/lecturettes, demonstrations using R software, causal analysis and data analysis tasks. Students will need to register with the UK data service, to gain access to real-world datasets that will be used extensively throughout the course.

Student should/will be able to:

• Understand the causal theory of DAGs;
• Distinguish potentially causal relationships from spurious ones. 
• Understand the statistical formulation of regression models. 
• Understand the basis of inference from samples to populations

Student should/will be able to:

• Distinguish between levels of measurement of variables, and use models and variables appropriately.
• Appreciate different types of functional relationship among variables, and use this to specify models appropriately.
• Evaluate correlations and consider to what extent they may represent causal as opposed to spurious, non-causal processes.

Student should/will be able to:

• Use the R software package to fit linear and binary logistic regression models.

Student should/will be able to:

• Use the R software package.
• Critically evaluate claims of causal effects, e.g. those presented in the media and in research papers.

• Formative assignment (0%): 500 words
• Written assessment (40%): 1200 words
• Written Exam (60%): Up to 2000 words

All Social Statistics courses include both formative feedback - which lets you know how you’re getting on and what you could do to improve - and summative feedback - which gives you a mark for your assessed work.

Recommended reading

Agresti, A. (2018). Statistical methods for the social sciences, Global Edition. Pearson/ Prentice Hall.

McShane, B. B., Gal, D., Gelman, A., Robert, C., & Tackett, J. L. (2019). Abandon Statistical Significance. American Statistician, 73(sup1), 235–245. https://doi.org/10.1080/00031305.2018.1527253

Rohrer, J. M. (2018). Thinking Clearly About Correlations and Causation: Graphical Causal Models for Observational Data. Advances in Methods and Practices in Psychological Science, 1(1), 27-42. https://doi.org/10.1177/2515245917745629 

Verzani, J. (2001). SimpleR: Using R for Introductory Statistics. https://cran.r-project.org/doc/contrib/Verzani-SimpleR.pdf

On-line Resources

Absolute basic introduction to R: http://stats.idre.ucla.edu/stat/data/intro_r/intro_r_interactive.html#(1)

Analyze Survey Data for Free: http://asdfree.com/