Assessing credibility factors of short-form social media posts: A crowdsourced online experiment
Abstract
People commonly turn to the Internet and social media for their information needs. Most popular social media platforms focus on short-form content that can be consumed rapidly. Given how fast such content spreads online, its trustworthiness and credibility have become important research areas. We investigate how different factors of social media posts influence their perceived credibility. We generated health-themed short-form social media posts, varied specific aspects of those posts, and deployed the variations on three different online crowdsourcing platforms for credibility assessment. Our quantitative data analysis reveals, for instance, how author professions related to healthcare and science increase the perceived credibility of health-themed posts. Moreover, a higher number of likes and shares increased the credibility in two out of the three platforms. Our qualitative results based on questionnaires highlight personal filtering strategies and critical thinking skills as factors that influence post credibility online. Consequently, our results encourage experts to provide information on social media and to be part of correcting any misinformation as they have higher credibility. Our work strengthens the previous body of work on the credibility of online content in general and acts as a starting point for further studies on social media post content by demonstrating a systematic, crowdsourced, and scalable approach.
Study specs
Crowdsourced online credibility assessment using health-themed social media posts with varied content features deployed across three platforms; quantitative and qualitative data collection.
- Authors
- J Li,M Kuutila,E Huusko,N Kariyakarawana
- Institution
- University of Oulu
- Discipline
- Human-Computer Interaction
- Study Type
- Experimental Study
- Year
- 2025
- Human Data Platform
- Prolific
- Source
- View Source DOI Google Scholar
Measured Outcomes
Credibility factors like author profession, engagement metrics (likes/shares), and personal strategies influencing perceived trustworthiness of social media posts.
Peer Review & Critical Discussion
Potential Selection Bias in 2023 Cohort
The participant pool shows a concerning overrepresentation of users from high-income demographics. Looking at Table 3, we can see that 78% of respondents had annual incomes above $75k, which significantly limits the generalizability of these findings to broader populations.
Non-naive Participants Issue
I've noticed a methodological concern regarding participant naivety. Given that Prolific users often complete multiple studies, there's a real risk that participants had prior exposure to similar experimental paradigms, which could confound the results.
RLHF Applicability to This Study Design
The implications for RLHF training pipelines are understated. If we accept the authors' conclusions about preference stability, this has direct consequences for how we should structure reward model training. The temporal decay effect described in Section 4.2 is particularly relevant.
Verify your expertise to join discussion
Create an account and verify your credentials to participate in peer discussions.