Hello colleagues,

Our team is conducting a meta-analysis examining the effects of AI use disclosure, AI labeling, and related forms of transparency about the use of AI on audience perceptions and behavioral outcomes.

Are we missing your study? If you have a study that may fit this topic—published, in press, under review, a working paper, conference paper, dissertation/thesis, or unpublished study—I would be very grateful if you could send me a citation, manuscript, abstract, or brief description. Unpublished studies and null findings are especially valuable for helping us assess and reduce publication bias.

We are broadly interested in studies in which participants are informed that AI was used in generating, creating, assisting with, editing, or otherwise producing the focal content, message, product, or service. The disclosure or label need not reflect actual AI use; studies that experimentally assign an AI-use or AI-authorship label to otherwise identical content are also eligible. Relevant comparisons may include AI disclosure versus no disclosure, AI versus human-authorship labels, or different forms of AI-use disclosure.

We are particularly interested in outcomes including, but not limited to:

• Trust and trustworthiness
• Credibility
• Perceived accuracy or reliability
• Authenticity
• Perceived quality or persuasiveness
• Attitudes or evaluations
• Behavioral intentions or behaviors (e.g., willingness to use, adopt, purchase, engage, share, comply with, or follow recommendations)

We are primarily interested in experimental or quasi-experimental studies that manipulate or compare the disclosure or labeling of AI use. Studies in which AI disclosure is one factor within a larger factorial design are also very welcome.

If the paper, appendix, OSF page, or other repository already contains sufficient information to calculate effect sizes, a citation or link is sufficient. If additional information is needed, we may follow up regarding relevant summary statistics, such as condition sample sizes, means and standard deviations, proportions, correlations, or test statistics.

Please send any relevant studies or information to S. Mo Jones-Jang (jangsr@bc.edu).