Recency Effect, Expert Judgement versus Public Opinion: the Case of the Masked Singer

Sequential presentations, such as job interviews, personnel decision-making, proposal selections, and competitions in fields like music, art, and sports, often introduce biases into performance evaluations. Among these biases is the serial position effect, characterized by two primary components: the primacy effect, which involves better recall of initial events, and the recency effect, which pertains to the enhanced recall of the most recent occurrences. Consequently, competitors of equal talent may receive unfair evaluations based solely on their position in the lineup. Using a substantial dataset (n = 1462) from performance assessments in the reality singing competition ‘Masked Singer,’ this study presents a persistent case of recency bias over an extended period. The odds significantly favor the second performer, with nearly three times higher winning chances than the first performer. This study also aims to address the long-standing inquiry into the comparative abilities of experts versus the general public in assessing the quality of cultural events. The findings unveil that both the Expert panel and the public (Audience) panel exhibit the recency effect. However, the Expert panel displays a lower sensitivity to the order effect, as the second performer’s winning ratio consistently remains lower compared to that under the Audience panel. The mean vote margin, representing the difference in votes earned by competitors, reveals intriguing insights. A significantly larger mean vote margin occurs when the second performer wins, suggesting a potential positional advantage or premium. Furthermore, regardless of the order of appearance, the mean vote margin is larger under the Expert panel than under the Audience panel. This suggests that the Expert panel tends to make more decisive decisions, possibly indicating greater confidence in their evaluations.