Jake Shannin

Jake Shannin


Babette Brumback, Ph.D.


College of Medicine, College of Public Health and Health Professions


Statistics and Mathematics




Pair-a-Dice Gaming (Treasurer), Statistics Club (Treasurer), Phi Beta Kappa, University Math Society, Student Honors Organization

Academic Awards

University Scholars Program (2021), University Research Scholars Program: First Research Award (2020), President’s Honor Roll (2019-2020), Putnam Math Competition: Top 3 at UF (2019), National Merit Scholar (2019)


OCPS Outstanding Youth Volunteer Award for providing geometry tutoring to underrepresented students

Research Interests

Causal Inference, Effect-Measure Modification, Interaction

Hobbies and Interests

Card games, board games, playing trombone

Research Project

Inferring Interaction Causal Types

For many diseases, doctors choose between a variety of treatments to administer or prescribe to patients. They often must decide whether to administer combinations of those treatments. Some patients may respond to one treatment but not others (causal types 4 and 6); giving multiple treatments may decrease an individual’s chance of recovery. However, the combination of treatments may lead to unforeseen interactions.

For example, a patient may not respond to a combination of two treatments even though she would have responded positively to either of the two without the other. This interaction is known as antagonism — causal type 7 in the context of two treatments. On the other hand, a patient may respond positively to a combination of two treatments even though he would have not responded to either treatment separately. This is synergism, causal type 2.

Due to the fundamental problem of causal inference, it is often impossible to directly assess a patient’s causal type. However, our preliminary research shows that, under certain assumptions, we can infer bounds for each of the causal types. We will continue investigating the inference of interaction causal types, developing and proving conjectures given various sets of assumptions.