Explaining Disparities in Paper Impact across Similar-Level Journals: An Analysis of Sociology Publications

Why do papers of seemingly comparable quality generate widely different levels of scholarly impact? This study addresses this unresolved question by examining internal knowledge characteristics that may explain citation disparities among articles published in three top sociology journals between 1999 and 2022. Drawing on bibliometric techniques, we analyze a dataset of 3,190 articles using Latent Dirichlet allocation (LDA) to extract topic-based representations of each paper. We construct measures of context novelty, content novelty, and knowledge focus, and employ OLS regression models to assess their relationship with five-year citation counts. The analysis was conducted using Python. Contrary to prior findings, novelty measures fail to significantly predict citation outcomes when journal-level prestige is held constant. In contrast, knowledge focus-indicating the thematic concentration of a paper-shows a robust and consistent positive association with citation counts, suggesting that cognitive coherence may enhance scholarly visibility even in structurally comparable settings. This study contributes to ongoing debates about what drives scientific recognition by shifting the focus from external prestige signals to internal cognitive features. It also demonstrates the value of analyzing impact variation within similarly ranked journals-a context often overlooked in large-scale citation studies. The study is limited to the field of sociology, and does not account for factors such as network embeddedness or institutional affiliation. Future research could apply similar designs to other disciplines and incorporate additional contextual variables.