Not All Transparency Is Equal: Source Presentation Effects on Attention, Interaction, and Persuasion in Conversational Search

Not All Transparency Is Equal: Source Presentation Effects on Attention, Interaction, and Persuasion in Conversational Search Jiangen He The University of Tennessee Knoxville, TN, USA jiangen@utk.edu Jiqun Liu The University of Oklahoma Norman, OK, USA jiqunliu@ou.edu Abstract Conversational search systems increasingly provide source cita- tions, yet how citation or source presentation formats influence user engagement remains unclear. We conducted a crowdsourc- ing user experiment with 394 participants comparing four source presentation designs that varied citation visibility and accessibil- ity: collapsible lists, hover cards, footer lists, and aligned sidebars. High-visibility interfaces generated substantially more hovering on sources, though clicking remained infrequent across all condi- tions. While interface design showed limited effects on user experi- ence and perception measures, it significantly influenced knowl- edge, interest, and agreement changes. High-visibility interfaces initially reduced knowledge gain and interest, but these positive ef- fects emerged with increasing source usage. The sidebar condition uniquely increased agreement change. Our findings demonstrate that source presentation alone may not enhance engagement and can even reduce it when insufficient sources are provided. CCS Concepts • Information systems →Search interfaces; • Human-centered computing →User studies. Keywords conversational search, large language models, source citation, cred- ibility, transparency, information seeking, attitude change ACM Reference Format: Jiangen He and Jiqun Liu. 2018. Not All Transparency Is Equal: Source Presentation Effects on Attention, Interaction, and Persuasion in Conversa- tional Search. In Proceedings of Make sure to enter the correct conference title from your rights confirmation email (Conference acronym ’XX). ACM, New York, NY, USA, 5 pages. https://doi.org/XXXXXXX.XXXXXXX 1 Introduction Conversational search and large-language-model (LLM)–enabled intelligent assistants are reshaping how people seek, judge, and use information. Within the field of human-computer interaction and retrieval (HCIR), decades of research show that interaction de- sign, sense-making, and credible source use are central to effective Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from permissions@acm.org. Conference acronym ’XX, Woodstock, NY © 2018 Copyright held by the owner/author(s). Publication rights licensed to ACM. ACM ISBN 978-1-4503-XXXX-X/2018/06 https://doi.org/XXXXXXX.XXXXXXX search [13, 17, 20, 27]. In particular, user evaluation studies further demonstrate that the interface is not a neutral conduit: presenta- tion choices modulate attention, implicit feedback, and verification behaviors [2, 8, 10]. In parallel, credibility research in information science establishes that users apply heuristics and strategies sensi- tive to cues exposed by the interface [14, 22, 25, 28]. For LLM-based systems, transparency features, such as explanations, provenance, and citations, are repeatedly recommended to support scrutiny and appropriate trust [6, 16, 19]. Yet, real-world studies of generative search caution that users often do not meaningfully inspect sources even when citations are provided, risking shallow verification, mis- trust, and homogenized exposure [4, 11, 23]. Together, these strands indicate that source-presentation design is a first-order determinant of credible, accountable conversational and generative interactions. Despite significant progress, two gaps persist. First, most work examines transparency at a conceptual or system-level (e.g., “pro- vide citations”) rather than as a precise interface variable whose visibility and accessibility can be manipulated and measured. We lack controlled evidence isolating how concrete presentation for- mats (e.g., inline hover cards vs. footers vs. side panels) affect micro- behaviors such as hovering, clicking, and revisiting sources—the very behaviors through which users enact scrutiny and build cred- ibility judgments [2, 9, 10]. Second, prior studies often stop at global perceptions (e.g., trust, satisfaction, engagement) without connecting interface-induced behaviors to downstream knowledge and attitude change [26], which are outcomes long recognized in information-seeking and persuasion research [5, 7, 22, 25]. Closing these gaps is crucial for both research and practice: it sharpens theories linking interface affordances, attention, and cr

This content is AI-processed based on open access ArXiv data.