Teaching and research: less cage match, more separate lanes

Everyone loves a rivalry — sometimes, the nastier the better. Marsh vs. Cope (paleontologists in a vicious knock-down, drag-out fight over dinosaur bones). Edison vs. Tesla (You might be surprised about who won). Michelangelo vs. da Vinci (16th-century contemporaries vying for the same Italian Jobs). William Jennings Bryan vs. Clarence Darrow. Coke vs. Pepsi. Apple vs. Microsoft. Ali vs. Frazier. Real Madrid vs. FC Barcelona (El Clasico). The Duttons vs. Broken Rock. Drake vs. Kendrick Lamar. Nature vs. Nurture.

Teaching vs. Research.

Yeee-ouch. That last can leave a mark from an oftentimes pitched battle in top-tier institutions like ours. It evokes and stirs up faculty beliefs about workloads, fairness, and even careers. Are the underlying tensions resolvable? As with most science-related disputes: Yes, they are. To get to that resolution, let’s start with simplified versions of the propositions driving the debate.

First is the scarcity or fixed capacity theory. If every faculty member has a limited set of resources (e.g., time and energy) from which they draw to do their research (R) and teaching (T) tasks, then resources devoted to T tasks are not available for R tasks and vice versa. In negotiation, this is called a cake-slicing, distributive, or zero-sum mechanism that splits what’s there. This notion broadly predicts that being better at one means being worse at the other; research and teaching performance should be negatively correlated. Other processes that might underlie this antagonism include personality types (e.g., extraverts) that emphasize the R task or T task at the detriment of the other, as well as divergent rather than aligned academic reward systems.

Second is the mutually supportive theory. This notion says research and teaching are intertwined in a complementary way. Teaching yields deeper understandings of what we do and don’t know, which fuels research. Research keeps teaching fresh and advantages students in terms of learning cutting-edge knowledge. In negotiation, this is called an expand-the-pie (Who doesn’t love dueling dessert metaphors?), integrative, or cooperative mechanism that broadens two-party solutions. Research and teaching effectiveness should be positively correlated; being better at one reinforces the other. A well-known result found elsewhere (the “positive manifold” of job performance) might imply that talents such as communication skill and even general intelligence benefit R and T simultaneously, or that harder-working faculty simply devote more resources to both.

What does the evidence say? Do our efforts to generate and test Big Ideas^tm help or hurt our efforts at top-notch instruction? The answer is a bit of a letdown. Neither. All the many dozens of studies and hundreds of thousands of data points on this topic yield a correlation between research and teaching that is essentially zero. Cue the womp-womp trombone.

So, research and teaching are not in a fight or a lovefest. They are not structurally “versus.” They are not elementally “yes, and ….” They are empirically independent. That result should not only inform our discussions about workload and performance; it should give us more freedom to make our own effort-directed choices.

Our “I in AI” column deals with a related interplay of (junior) auditors’ effort and their use of artificial intelligence to do complex tasks. Using the latter might be seen by supervisors as lazy, but making a mistake would be seen as incompetent. Andrew Jones’ research shows that the answer to this tension is different. Even in the face of heightened scrutiny, professional judgment and AI interpretations about what’s on the books can be mutually supportive.