Traditionally, methods such as opinion polls have been pivotal in shaping media depictions of public sentiment. Journalists have several ways at their disposal to draw on in order to report on public opinion in their coverage. However, the advent of platforms like Twitter (X)1 introduces new layers of complexity to this dynamic. Through a survey-experiment with 143 Flemish political journalists, we examined how different sources (Twitter versus traditional methods) and different types (aggregate versus individual quotes) of public opinion influence journalists’ perceptions of representativeness, as well as their own estimations of public opinion. Our findings reveal that while journalists seem to value traditional opinion polls for their representativeness more than references to public opinion on Twitter, they are “Twitter vox pops”, i.e., a selection of citizen tweets, that have the strongest influence on journalists’ public opinion estimations, especially in soft news topics. Our results underscore the importance of investigating the evolving nature of public opinion representation within journalism and the impact thereof, as traditional methods coexist with and are increasingly complemented by social media.
