Engagement without approval: comparing YouTube discourse, interpretive framing, and sentiment integrity of Usher and Kendrick Lamar Super Bowl halftime shows

The purpose of this study is to systematically compare large-scale YouTube discourse surrounding two Super Bowl halftime shows by identifying recurring discussion themes, quantifying evaluative tone, and assessing whether high engagement metrics reflect genuinely favorable reception. Specifically, it contrasts an entertainment-centered performance (Usher) with a more symbolic, contested performance (Kendrick Lamar) to examine whether discourse structure and sentiment integrity vary across performance types. We analyzed 138,734 YouTube comments from the official NFL videos of the two halftime shows (Usher: N = 26,982; Kendrick Lamar: N = 111,752). The analysis combined automated topic modeling, human-reviewed theme labeling, frame classification, multilingual sentiment analysis, and a Sentiment Integrity Index (SII) designed to distinguish attention from approval. Usher’s discourse was highly consolidated (88.8% on a single topic) and almost exclusively “hedonic,” resulting in high net positivity. Conversely, Lamar’s discourse was less concentrated across 103 topics and multiple interpretive frames (Hedonic, Symbolic, Hybrid), leading to a more mixed and contested evaluative structure. Crucially, while Lamar had higher reach, Usher’s “sentiment integrity” was higher, demonstrating that massive engagement does not always equal favorable reception. The study contributes to sport management and media research by demonstrating how interpretive frames shape audience evaluation in platform ecosystems. It introduces the Sentiment Integrity Index (SII) as a novel tool for decoupling raw engagement from actual sentiment quality. This approach provides a scalable method for mapping contested symbolic sites and evaluating the public construction of cultural meaning in global sporting spectacles.