Transparency as a moderator of the relationship between web-scraping awareness and privacy-protective behavior: a quasi-experimental onboarding study
Journal
Frontiers in Computer Science
ISSN
2624-9898
Publisher
Frontiers Media SA
Date Issued
2026
Author(s)
Betancourt-Barrita, Cecilia
Cal Y Mayor Peña, Francisco
Type
text::journal::journal article
Abstract
Introduction: Automated data collection practices, such as web scraping, are widely used for personalization on interactive platforms, and they raise recurring questions about privacy, transparency, and user trust. Prior work has examined these constructs separately. Less is known about whether transparency in onboarding disclosures is associated with a weaker relationship between awareness of automated data collection and privacy-protective behavior. This study examines the relationship between privacy-protective behavior and user trust in a simulated social media environment.Methods: A quantitative, between-subjects, scenario-based quasi-experimental survey was conducted with 325 participants. The study employed a simulated Instagram onboarding scenario with two transparency conditions: low and high. Privacy-related outcomes were operationalized as two binary privacy-protective behaviors (adjusting privacy settings and refusing to use a platform), summed into a behavior index that takes integer values of 0, 1, or 2. Additional measures assessed web-scraping awareness, user trust, perceived transparency, and control variables related to social media use and privacy sensitivity. Hypotheses were tested using ordinary least squares regression for the behavior index and continuous outcomes, and logistic regression for each binary behavior, including an interaction term to examine the moderating effect of the transparency condition.Results: Greater awareness of web scraping was associated with a higher likelihood of low-barrier privacy-protective behavior, particularly adjusting privacy settings. The transparency condition was associated with a weaker awareness–behavior relationship, consistent with the proposed moderation model. In contrast, the privacy-protective behavior index did not significantly predict user trust in the platform.Discussion: The findings are consistent with transparency operating as a contextual buffer of the link between awareness of automated data collection and privacy-protective behavior, rather than as a direct antecedent of trust. Because condition assignment was self-selected and baseline awareness differed across conditions, the results are interpreted as conditional associations within a simulated onboarding context rather than as causal effects. The results inform the design of onboarding disclosures in AI-driven platforms. © The authors © Frontiers in Computer Science.
License
Acceso Abierto
How to cite
Betancourt-Barrita C, Berumen-Glinz LA, Peña FCYM and Escobar-Castillejos D (2026) Transparency as a moderator of the relationship between web-scraping awareness and privacy-protective behavior: a quasi-experimental onboarding study. Front. Comput. Sci. 8:1832107. doi: 10.3389/fcomp.2026.1832107