Software Application Profile: The daggle app-a tool to support learning and teaching the graphical rules of selecting...

Area of Science:

• Epidemiological research
• Causal inference methodology
• Statistical learning

Background:

• Directed Acyclic Graphs (DAGs) are crucial for visualizing causal relationships in epidemiology.
• Identifying appropriate covariate adjustment sets is essential for unbiased estimation of causal effects.
• Existing methods for adjustment set identification can be complex to learn and apply.

Purpose of the Study:

• To introduce the daggle app, a web-based tool designed to facilitate learning and teaching of adjustment set identification using DAGs.
• To provide a practical resource for researchers to understand and apply graphical rules for causal inference.
• To enhance the understanding of causal assumptions and their impact on covariate selection.

Main Methods:

• The daggle app features two modes: a tutorial for guided learning and a random mode for practice.
• Tutorial mode explains common causal structures and graphical rules for identifying minimally sufficient adjustment sets.
• Random mode presents users with randomly generated DAGs (daggles) to test their ability to identify correct adjustment sets.

Main Results:

• The app successfully demonstrates the application of graphical rules for adjustment set identification.
• Users can actively engage with DAGs and practice identifying adjustment sets in a simulated research environment.
• The tool provides immediate feedback through the challenge of solving randomly generated daggles.

Conclusions:

• The daggle app serves as a valuable educational tool for researchers learning causal inference with DAGs.
• It simplifies the process of understanding and applying complex graphical criteria for covariate selection.
• Accessible online with open-source code, the app promotes wider adoption of rigorous causal inference methods.