16 university students used FeedReflect for 3 weeks. Half of the students (the treatment group) were asked credibility questions without the visual nudges on each Tweet, while the other half (the experimental group) was asked credibility questions about the Tweets and given the visual cues. The treatment group scored mainstream sources with higher credibility and non-mainstream sources with lower credibility than the control group did. Further, after qualitative analysis, the following themes emerged among the treatment group:

• Users read news stories that normally would not have interested them
• Users consulted external sources to measure credibility of news sources
• Users actively participated in credibility assessment rather than making subconscious decisions