让下一篇重要论文，不再错过。

原文摘要

User engagement is well studied in other domains. Work to date has proposed conceptual models and definitions, approaches to measurement, and factors that influence engagement in digital interactions. However, of the articles in the visualization literature that refer to engagement, few of them make evidence-based, generalizable claims, and research that does is challenging to locate due to inconsistent reporting and often lacks connection to the broader engagement literature. To address these issues and promote the consideration of the broader engagement literature, we conduct a comprehensive scoping review of visualization literature and engagement. Using scoping review methodology, we identify and analyze 243 articles that make claims about engagement with visualizations. We coded the studies conducted, literature cited, definitions provided, and claims made in these articles and analyze the codes to evaluate the quality of the work in the field and to synthesize existing findings. We find that, despite frequent references to engagement and apparent interest in the topic, the overall quality of visualization engagement research is poor, often with weak evidence, inconsistent conceptualizations, and infrequent citation of existing models, methods, or claims. We report the evidence-based claims about what engages people with visualization, including the factors of animation, interaction, and immersion, additionally describing both corroboration and apparent contradictions between these claims. Finally, we propose the COMFI framework and mnemonic for reporting important information in future engagement studies.