People with a firm opinion chose more texts that supported it
People who were confident in their opinion opened more like-minded texts in a simulated search about psychotherapy. In the end, learning was similar.
The key idea
Confidence in an opinion influenced which texts people opened and how they rated them, while source credibility influenced which tags they chose.
The study, in plain language
Participants began by recalling an opinion about psychotherapy or antidepressants, along with a time when they felt confident or doubtful about it. They then browsed a simulated search page with tags, short posts, and a banner describing the community as expert or student-based. The combinations were randomly assigned.
When confidence in an opinion was high, that position guided the choice of like-minded texts more strongly. The community’s credibility mattered at a different point: when choosing tags. When the source seemed less expert, prior opinions guided tag selection more, and also guided which texts people opened.
Confidence also mattered afterward, when participants judged how effective the texts were: with high confidence, they rated texts agreeing with their position more favorably. With low confidence, they rated texts challenging it more favorably. Even so, differences in clicks and ratings did not produce a clear difference in how much participants learned.
What they found
- With greater confidence in a prior opinion, people opened more texts that agreed with it.
- The community’s credibility mattered when people chose browsing tags.
- When rating texts afterward, high confidence favored like-minded texts and low confidence favored texts that challenged the opinion.
How they studied it
One hundred twenty-five people completed a simulated search, with groups randomly assigned across two factors. Researchers varied confidence in the prior opinion and the credibility of the community labeling texts, then measured what participants opened and how much they learned. The simulation offered posts favoring either alternative, but no neutral texts.
What the study cannot show
- The search took place in a simulated environment and covered only one health topic. Available texts favored the options presented, and the study did not measure how much the topic mattered to participants.
Something to think about
When reading about an idea you already hold, ask whether you chose the text for its evidence or because it agrees with you. This is a question for reflection, not a tested technique for reducing bias.
The original study
Attitude confidence and source credibility in information foraging with social tags
Source under CC BY 4.0. English adaptation and explanation by PaperFlow, with AI assistance. The text has been reorganized for a general audience; it is not the complete paper and does not imply endorsement by its authors.
One study alone does not represent a field's consensus. Educational content does not replace professional advice. Report a possible error.
