Information Frequency,Value,and Difficulty as Sources of Social Inequality: Competitive Imbalances on Jeopardy!

Abstract

Informational content and topics strategically curated by institutions underpin reward structures in knowledge economies. Using a historical database of 298,879 questions from popular American television trivia game show Jeopardy!, this article presents a case study revealing competitive inequalities rooted in a large, historical information corpus. Historically, women contestants comprise 45.5 percent of Jeopardy! contestants but only 32.5 percent of game winners. This raises questions about the fairness of the game and mechanisms underpinning the gender performance gap. Contestant gender and occupation are predictive of topical strengths and weaknesses. Information frequency, value, and difficulty are identified as knowledge properties that underpin competitive advantages and disadvantages. Questions with female answers on Jeopardy! are less frequent, valuable, and difficult than male and nongendered questions. This deprives women contestants of competitive advantages because contestants exhibit homophilous tendencies with gendered knowledge; women exhibit advantages with female questions, and men are advantaged by male questions. The Jeopardy! gender performance gap can be reduced—but not eliminated—by equalizing the frequency, value, and difficulty of gendered questions. As a microcosm of dominant cultural trends and powerful societal knowledge institutions, Jeopardy! is a broadly applicable case study revealing inequalities in ostensibly meritocratic knowledge evaluation systems. Results reveal specific means to identify and mitigate social inequalities in knowledge institutions and ostensibly meritocratic information-based competitions.