Workshop: "Using Causal Inference to Express, Explore and Extend your Research"
We will learn about and discuss a tool (Directed Acyclical Graphs, or Graphical Causal Models), which is very useful to describe and discuss implied causal mechanisms, which helps in structuring and writing research papers and even grant applications.
Info about event
Using Causal Inference to Express, Explore and Extend your Research
Date and time: Thursday November 23rd – 10am to 1pm
A) Local collective attendance in Nobelparken, Building 1485-238 (Zoom-link will be send to email at the day for workshop).
B) Remote workshop, teacher on zoom (Zoom-link will be send to email at the day for workshop).
The teacher: Sean Roberts (an associate professor at Cardiff University - https://profiles.cardiff.ac.uk/staff/robertss55) has been working on describing and formalizing the different causal assumptions implicit in studies of the evolution of language to better guide our understanding and the planning of future work (https://correlation-machine.com/CHIELD/).
Why to attend: We will learn about and discuss a tool (Directed Acyclical Graphs, or Graphical Causal Models), which is very useful to describe and discuss implied causal mechanisms, which helps in structuring and writing research papers and even grant applications. In the Spring we will follow up with a more quantitatively oriented workshop (led by a different instructor) on the more statistical and computational implications of the graphs in helping us making causal inference from observational data, but this first workshop is really about the idea of causal graphs which applies equally to conceptual and qualitative work.
Target audience: primarily members of the Linguistics, Cognitive Science and Semiotics research program, from phd students to permanent staff. Additional participants are welcome, in presence if space allows, online otherwise. A mail will follow once sign up is over.
This will be a short crash-course in Causal Inference and using causal graphs to represent your hypotheses. A causal graph is a way to express a hypothesis about how the world works in a graphical form. Their origin is in clinical trials, but they are now used in many disciplines from psychology to anthropology. They can help you be explicit about your hypotheses and communicate them clearly to your collaborators and readers. Together with some causal axioms they can help you identify confounders and alternative explanations in order to choose which variables you need to control for. In this workshop, we'll cover some of the basic theory, and then have a practice building your own causal graphs. There will be plenty of time for Q&A.
- Please watch this 20 minute video introducing some basic concepts of causal graphs: https://youtu.be/S5CS8wL7tjM
- Please come to the workshop prepared to create a causal graph of one of your own hypotheses. If you don't want to use your own, please pick a recent paper with a hypothesis you can use.
- 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.
- Roberts, S. G., Killin, A., Deb, A., Sheard, C., Greenhill, S. J., Sinnemäki, K., ... & Jordan, F. (2020). CHIELD: the causal hypotheses in evolutionary linguistics database. Journal of Language Evolution, 5(2), 101-120.
- Pearl, J., & Mackenzie, D. (2018). The book of why: the new science of cause and effect. Basic Books.