EXCLUSIVE: Conference Buffet Quality Inversely Correlated With Paper Quality

A comprehensive five-year longitudinal study spanning 47 academic conferences across 19 countries has confirmed what attendees have long suspected but been too hungry to quantify: there is a statistically significant inverse correlation between the quality of the conference buffet and the quality of the research presented. The effect size, described by the lead author as “embarrassingly large,” holds across subfields, geographic regions, and funding sources, and is robust to several alternative explanations including “venue cost,” “organizational competence,” and “whether the general chair had a good caterer.”

The data, collected by a team of four graduate students who attended conferences on expenses they have yet to be reimbursed for, assigned buffet quality scores based on a validated 12-point rubric covering food temperature, protein-to-carbohydrate ratio, the presence of a labeled vegetarian option, and whether the dessert table was replenished after the first wave. Paper quality scores were derived from subsequent citation counts, a methodology the authors acknowledge is “imperfect” but “better than asking Reviewer #2.” The correlation coefficient across all observations was -0.71, which the paper describes as “striking” and one anonymous statistician described as “frankly too clean, which is suspicious, but I’m choosing to believe it.”

Among the strongest predictors of low paper quality was what the researchers term the “artisanal cheese board effect”: conferences featuring individually labeled single-origin cheeses, charcuterie arrangements with handwritten cards, or any item described as “locally sourced” on a placard were associated with a 34% reduction in mean citation count per paper. Conversely, conferences featuring industrial trays of unidentifiable pasta and a single vegetarian option that was a second tray of the same pasta produced papers that, on average, cited more prior work, made more modest claims, and were accepted to better venues on resubmission.

The researchers offer three potential explanations. First, conferences with large budgets spend them on catering rather than speaker quality. Second, researchers who attend well-catered conferences spend more time at the buffet and less time at the talks, reducing the peer feedback that improves subsequent work. Third, and the theory the authors personally favor, there is a latent variable they have named “institutional pretension” that drives both expensive food choices and the kind of overconfident, under-validated research that generates abstract buzz but no replicable findings. A preregistered follow-up study is planned. It will be presented at a conference the organizing committee has promised will feature “light refreshments.”