Skip to Main Content

Bayes with Beans: Teaching Prototype

2018

In this video, Professor Myriam Hunink introduces the concept of “diagnostic Bayesian thinking” through a simple example and a visual representation using beans! Students consider how an initial probability or belief is influenced by new diagnostic information through the use of Bayes’ theorem.

Access the video. Bayes with Beans: Teaching Prototype (~14 min)

Using the example of a tick bite after a walk through the woods, she asks you to imagine sitting in a doctor’s office with 99 other patients just like yourself, all worried about tick bites and the possibility of Lyme disease. What will the physician do? Treat everybody with antibiotics? Perform a diagnostic test? Wait and monitor without testing or treating?

Students learn how to combine information about the prior probability with diagnostic information that is not perfect, to calculate two “post-test” probabilities of disease – the probability of disease given a positive test and the probability of disease given a negative test. Professor Hunink differentiates conditional probabilities that are test characteristics (e.g., the probability of a positive test given disease positive or sensitivity) from conditional probabilities that provide information helpful for clinical decision making or post-test probabilities (e.g., probability of disease given a positive test, and probability of disease given a negative test).

This video is one of a series developed by Professor Myriam Hunink during an immersion residency at the Center for Health Decision Science (CHDS) Media Hub. The video series reflect experiments to augment brick and mortar teaching with multimedia materials that emphasize visualization of basic concepts.

 

Source:

Bayes with Beans: Teaching Prototype. Teaching Pack: Teaching Prototypes for Decision Analysis. Center for Health Decision Science, Harvard T.H. Chan School of Public Health 2018. https://vimeo.com/236607953/d0592a55a8