Skip Navigation LinksHome > March 2013 - Volume 24 - Issue 2 > Causal Pie Bingo!
Epidemiology:
doi: 10.1097/EDE.0b013e318281e1cf
Letters

Causal Pie Bingo!

Johnson, Candice Y.; Howards, Penelope P.

Free Access
Article Outline
Collapse Box

Author Information

Department of Epidemiology, Emory University, Atlanta, GA. cyjohnson@cdc.gov

Back to Top | Article Outline

To the Editor:

The sufficient-cause model, also known as “causal pies,” is a causal model commonly taught in introductory epidemiology courses. Under the sufficient-cause model, outcomes are usually not caused by a single causative factor, but by a combination of “component causes” (exposures) that might occur minutes or years apart.1,2 A person gets the outcome only if he or she has accumulated all component causes. Together, the combination of component causes that are sufficient to cause an outcome is called a “sufficient cause.” Typically, there is more than one sufficient cause for each outcome, meaning that there are different mechanisms by which a person can get the same outcome.

Some students struggle with the sufficient-cause model and other causal models because the concepts can seem abstract, even when examples are given. We have developed a fun and intuitive game called “Causal Pie Bingo!” that introduces students to the sufficient-cause model in a format that makes these examples more concrete.

Causal Pie Bingo! follows the same format as the “bingo” game popular in many countries. Each student receives a game card with one or more sufficient causes (Fig.). The instructor has a box containing each of the component causes and randomly selects a component cause from the box. The students color in the corresponding component cause on their game cards. This process is repeated, simulating accumulated exposures within the population. When all the component causes in a sufficient cause have been colored in, the student gets the outcome. The game continues until all students have the outcome or all component causes have been selected from the box.

FIGURE. Example of a...
FIGURE. Example of a...
Image Tools

Causal Pie Bingo! is suitable for epidemiology students and others, from elementary school (age 10+ years) through graduate school. Instructors can tailor the game to their teaching needs by using the exposures and outcomes of their choice, demonstrating differences among necessary, component, and sufficient causes and calculating risk or odds among the exposed and unexposed. We have developed several game variations, briefly described here.

Back to Top | Article Outline
Infectious Disease:

The infectious organism can be transmitted person-to-person as students get the outcome and become infectious. Vaccination can be introduced to halt the spread of disease, and herd immunity can be produced if an unvaccinated student is surrounded by vaccinated students.

Back to Top | Article Outline
Competing Risks:

Students are given game cards for multiple outcomes; some outcomes prevent others from occurring. Instructors can use this to demonstrate calculation of conditional and unconditional risks.

Back to Top | Article Outline
Counterfactual:

Advanced graduate students can use Causal Pie Bingo! to relate the sufficient-cause model to the potential outcome (counterfactual) model. At various time points in the game, the game is paused and students decide which of the causal types they correspond to at that moment, relative to a given exposure.

Full instructions, a ready-to-play game (using H1N1 influenza as the outcome), and blank templates for creating customized games are available on the Causal Pie Bingo! Web site, https://sites.google.com/site/causalpiebingo.

Causal Pie Bingo! is a fun and effective tool for teaching causal models. We encourage instructors who use the game to contact us and share ideas for improvements or new variations.

Candice Y. Johnson

Penelope P. Howards

Department of Epidemiology

Emory University

Atlanta, GA

cyjohnson@cdc.gov

Back to Top | Article Outline

REFERENCES

1. Rothman KJ. Causes. Am J Epidemiol. 1976;104:587–592

2. Rothman KJ, Greenland S, Poole C, Lash TLRothman KJ, Greenland S, Lash TL. Causation and causal inference. In: Modern Epidemiology. 20083rd ed Philadelphia, Pa Lippincott Williams & Wilkins:5–31

© 2013 Lippincott Williams & Wilkins, Inc.

Twitter  Facebook

Login

Article Tools

Images

Share