�鶹���

Sweetest of Studies

A VSB professor teamed up with two students to study judge bias in The Great British Bake Off

By Albert Stumm

Contrary to the manufactured drama of typical reality television, The Great British Bake Off charmed Americans with its heartwarming tone as an antidote to pandemic stress.

But while indulging in the series, in which competitors have 10 episodes to rise to the top and be crowned the UK’s best amateur baker, one University economist saw through the sugary pastries and wholesome dialogue to detect something else—a pattern. “As economists, we see the world as economics,” says Maira Emy Reimão, PhD, the Michele and Christopher Iannaccone ’91 Assistant Professor of Economics in the �鶹��� School of Business.

Dr. Reimão noticed that the first dessert judges tasted seemed to come out ahead more often than subsequent samples. It gave her an idea to study judge bias in competitions. The results were published in January in the peer-reviewed journal Kyklos: International Review for Social Sciences, and two undergraduate students earned coauthorship of the academic literature.

In 2021, Dr. Reimão, Rachel Sabbadini ’25 VSB and Eric Rego ’25 CLAS began watching 14 seasons of the show, as well as versions from Australia, Canada, Kenya, New Zealand, South Africa and the United States. After analyzing the data from one segment that involves a blind taste test, they found that the dessert that was tasted first was 14% more likely to be ranked in the top half.

A phenomenon called primacy bias, or that the average person likely remembers the first option of a series best, had long been established in the field of behavioral psychology. But this study determined that bias also extended to a panel of experts–and it went one step further. “Not only do you remember it better, it's actually that you rank it better,” Dr. Reimão says. “You prefer it.”

From a consumer perspective, the findings demonstrate that expert opinion is not necessarily pure, with potential implications for ratings for products such as cars, wine and mutual funds, she says.

The researchers were surprised to find the bias present in every country’s version of the show. Equally surprising was that some of their original assumptions were incorrect, including about a concept called the recency effect. “We thought if judges were trying a dish last, that would lead to a higher score, but we didn’t find that evidence,” says Rachel, who decided to major in Accounting and Economics after the study.

Eric, a Computer Science major, wondered if the opposite would be true, that the last dish would rank lower since judges would be full or have tired taste buds. That theory didn't hold either, but he learned valuable lessons about the research process, critical thinking and scientific collaboration. "The three of us collaborated seamlessly to produce the paper, and being a part of that process is something I am deeply grateful for," he says.