Causal reasoning for TEL research

Causal reasoning for TEL research

Monday 23/05 10:30-12:00h

Outdoor Area A

Abstract

This workshop will introduce you to a causal inference framework using graphical causal models. These models are conceptual tools to help you reason about causal inference in quantitative TEL research. You will learn how to use them to detect bias, facilitate study design, and improve data analysis in ways that allow for robust causal inferences. At the end of the workshop you will be able to craft, inspect, and analyze causal models to appraise the TEL literature and improve your own TEL research.

Needs Analysis

Robust causal inference is a central goal of empirical research. To understand the impact of Educational Technology and TEL on learning experiences, empirical researchers must reason about cause and effect. This has implications for interpreting the state of research, but also directly impacts decisions regarding study design and data analysis. As such, the causal inference framework using Directed Acyclic Graphs (DAGs) provides a principled set of tools through which emerging TEL researchers will be able to conduct more robust research. Thus, this workshop should be of great interest to any TEL researcher who conducts quantitative research.

Learning Objectives

At the end of the workshop, participants will be able to answer the following questions:

1. What are Directed Acyclic Graphs and how do they work?
2. What are the three elemental configurations and sources of bias?
3. How can bias be avoided with the help of the back-door criterion?

Pre-activities

Participants will not have to prepare for this workshop. The workshop will provide a beginner’s introduction to the necessary tools so that participants will be able to craft causal models and reason about causal inference alone or collaboratively.

Session Description

The session will start with an introduction to the topic of about 15-20 mins. Then, participants will draw causal models based on their own area of research. We will then analyze them together, reason about bias in their causal models, and discuss options how bias made be avoided. At the end of the session, participants will have produced several causal models about research questions that interest them.

Post-activities

• 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.
• Pearl, J., & Mackenzie, D. (2018). The Book of Why: The new science of cause and effect.