Advice perceptions and upvote pragmatics: An experimental analysis of Reddit advice-giving dynamics

This study examines how linguistic formulations influence our perceptions of advice within the computer-mediated context of Reddit, bridging the gap between linguistics and other disciplines. A between-subjects experiment was designed to both explore a selection of advice-giving moves on Reddit and to better understand Reddit's upvote system. Using frameworks from communication and social psychology to examine the underlying effect of linguistic variables, 180 participants were provided with an advice scenario and then invited to emulate the Reddit voting system, rate the advice on a variety of psychometric dimensions, and reflect qualitatively on the advice in the experiment and about advice-giving as a communicative act. This paper reports on how participants perceived the voting system and used it to evaluate advice, along with the quantitative aspects of the experiment. Results show that the manipulation of advice-giving moves has a significant effect on perceptions of the advice-giver's warmth and competence, as well as the advice message's feasibility, limitations, and positive politeness perceptions; all ultimately influencing whether one will take up the advice. There is also evidence that Reddit's upvote system, as a binary affordance, has numerous, graded, pragmatic meanings that are not only obscured by the way the feature is operationalized, but also extend beyond the intention of the site's creators. • Interdisciplinary research on linguistic forms' effect on our evaluations is required. • Why people use paralinguistic digital affordances like Reddit upvotes needs more study. • Experiments help reveal the impact of different linguistic units on our perceptions. • Advice moves shape how advice is perceived across psychological and communicative domains. • Upvotes carry varied pragmatic meanings and influence how we interact online.