Conversational search systems increasingly provide source citations, yet how citation or source presentation formats influence user engagement remains unclear. We conducted a crowdsourcing user experiment with 394 participants comparing four source presentation designs that varied citation visibility and accessibility: collapsible lists, hover cards, footer lists, and aligned sidebars. High-visibility interfaces generated more hovering on sources, though clicking remained infrequent across all conditions. While interface design showed limited effects on user experience and perception measures, it significantly influenced knowledge, interest, and agreement changes. High-visibility interfaces initially reduced knowledge gain and interest, but these positive effects 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.

Characterizing users' judgments and interactions with search engine result pages (SERPs) has been a central theme in Interactive Information Retrieval (IIR) evaluation.In contrast to the perfect rationality assumptions underpinning most existing formal models, people are boundedly rational and are subject to the influence of systematic cognitive biases.To enhance the psychological foundation of user models and better understand the in-situ decisions of boundedly rational users, our between-subject crowdsourcing experiment explored Decoy Effect on users' vulnerability to COVID treatment misinformation, which causes enduring impacts on people's personal health management and wellbeing, and examined the extent to which this effect is moderated by contextual factors and user characteristics.Our results, derived from 540 participants and 2,160 valid SERP evaluation records, indicate that: 1) users' interactions with decoy results may increase their vulnerability to medical misinformation in usefulness and credibility judgments; 2) the size of decoy effect is conditioned by users' prior knowledge and the rank position of decoy results.This research empirically reveals the impact of decoy results on users' context-dependent preferences on ranked search results under varying topics and conditions of medical information evaluation.More broadly, it demonstrates the value of representing and modeling users interacting with information as boundedly rational agents and serves as a step forward towards achieving the goal of truly human-centered IIR.

Abstract Promotions and academic mobility are trajectory‐altering events in a researcher's career. This paper compiles a unique large data set and investigates publication and citation differences between two groups of researchers: the ones who are mobile and their counterparts who stay at a university with a promotion. This paper finds that mobile researchers often have a lesser productivity increase than their post‐promotion counterparts. The difference is largely driven by male professors in physical science and clinical health fields moving from more research‐intensive to less research‐intensive institutions. In contrast, the citational impact differences between the two groups are largely minimal.

Finding relevant publications is a common task. Typically, a researcher browses through a list of publications and traces additional relevant publications. When relevant publications are identified, the list may be expanded by the citation links of the relevant publications. The information needs of researchers may change as they go through such iterative processes. The exploration process quickly becomes cumbersome as the list expands. Most existing academic search systems tend to be limited in terms of the extent to which searchers can adapt their search as they proceed. In this article, we introduce an adaptive visual exploration system named PaperPoles to support exploration of scientific publications in a context‐aware environment. Searchers can express their information needs by intuitively formulating positive and negative queries. The search results are grouped and displayed in a cluster view, which shows aspects and relevance patterns of the results to support navigation and exploration. We conducted an experiment to compare PaperPoles with a list‐based interface in performing two academic search tasks with different complexity. The results show that PaperPoles can improve the accuracy of searching for the simple and complex tasks. It can also reduce the completion time of searching and improve exploration effectiveness in the complex task. PaperPoles demonstrates a potentially effective workflow for adaptive visual search of complex information.

Using visual analytic systems effectively may incur a steep learning curve for users, especially for those who have little prior knowledge of either using the tool or accomplishing analytic tasks. How do users deal with a steep learning curve over time? Are there particularly problematic aspects of an analytic process? In this article we investigate these questions through an integrative study of the use of CiteSpace—a visual analytic tool for finding trends and patterns in scientific literature. In particular, we analyze millions of interactive events in logs generated by users worldwide over a 14‐month period. The key findings are: (i) three levels of proficiency are identified, namely, level 1: low proficiency, level 2: intermediate proficiency, and level 3: high proficiency, and (ii) behavioral patterns at level 3 are resulted from a more engaging interaction with the system, involving a wider variety of events and being characterized by longer state transition paths, whereas behavioral patterns at levels 1 and 2 seem to focus on learning how to use the tool. This study contributes to the development and evaluation of visual analytic systems in realistic settings and provides a valuable addition to the study of interactive visual analytic processes.