No Backfire Effects from Factual Corrections to Misinformation in Taiwan

The next speaker in this IAMCR 2024 session is Victoria Chen, whose interest is in the influence of political misinformation in Taiwan. There are frequent presidential, parliamentary, and mid-term elections in Taiwan, and political misinformation about political parties is common. This manipulates public opinion, and can lead to polarisation and unconscious bias – the key question here is how people believe in and deal with such misinformation.

Fact-checking is one response to this, but can also produce backfire effects or a continued influence effect of misinformation. Backfire effects here mean that correction of misinformation only strengthens the convictions of those who believe in such misinformation, as it creates cognitive dissonance and increases partisan voters’ perceptions of their own difference from opposing world views. Such effects are not universal, however, and may relate to particular political ideologies and issues.

This project explores such backfire effects from fact-checks amongst liberal and conservative voters in Taiwan; it is interested in whether the source of the correction (from one’s own or the opposing party) affects the success of the correction.

Further, the continued influence effect suggests that people will continue to believe in misinformation even when they recall that such misinformation has been corrected; this is because the perceptions they have formed are so strong that even corrections can no longer break through and change people’s attitudes. (Also a reason why people persist in their beliefs in ‘echo chambers’ and ‘filter bubbles’ even in spite of the abundant evidence that debunks these utterly idiotic concepts, just quietly.)

The project examined these effects through experiments with some 500 participants, divided into conservative and liberal voters variously encountering misinformation corrections from liberal, neutral, and conservative sources. It found that liberal voters responded more successfully to corrections than conservative voters, but that the source of the correction statement did not not produce any meaningful effects; participants trusted neutral sources more than partisan sources, however. Even a perception of sources as credible did not necessarily result in a change in misperceptions, however: source credibility did not affect whether or not people changed their misperceptions.

Overall, 60% of participants did not change their views; in only 5%, there was a backfire effect from corrections. Political ideology did not play any meaningful role in any of this. What reduces misperceptions, then? The main factor appears to be relevance – people with greater familiarity with an issue respond more positively to fact-checking corrections.