In an attempt to act responsibly, social platforms now flag content that is certain to be false. However, flagging disputed content has some unintended consequences. False headlines that aren’t flagged are often thought to be true. In fact, according to new research, The Implied Truth Effect, conducted by Gordon Pennycook, Adam Bear, Evan T. Collins, and David G. Ran, false headlines that fail to get tagged are viewed as more accurate. Thus, the research appropriately questions whether the policy of using warning tags to fight misinformation is effective.

False news headlines with flagging

This research includes two studies. In the first, the control group was shown both true and false news headlines without any warning labels. The test group was shown both true and false headlines; the false news headlines included warning labels. Participants were asked how accurate the headlines were and if they would consider sharing the story on social media (such as Facebook or Twitter).

The first study confirms that content with warning labels decreased the belief in items that are flagged (the Warning Effect) but increases belief in items that are untagged (the Implied Truth Effect). In other words, headlines that were not flagged in the test group, were rated as more accurate, by at least one-third, than those in the control group

True news headlines with flagging

The second test included a control group and two test groups. Participants were presented with news headlines and asked whether they would share on social media. They were told that 75% of the headlines had been fact-checked by Snopes.com.

The control group was shown true and false news headlines without any labeling. The first test group was shown false news headlines, half with a “FALSE” stamp and the other half without any stamps. The second test group was shown false news headlines with a “FALSE” stamp and true news headlines with a “TRUE” stamp.

The findings show that participants in the first test group were less likely to consider sharing false headlines tagged with a warning compared to false headlines in the control group. Further, participants in the second test group were more likely to consider sharing true headlines tagged with a verification compared to true headlines in the control group.

This research identifies the consequence of attaching warnings to some inaccurate headlines but not all. It’s safe to assume that a large percentage of false headlines will continue to appear on social platforms and remain untagged. However, it is important to note that it may be even more valuable to tag true headlines.

Labeling truthful and verified headlines helps consumer identify what is true and suggests that everything outside this stamp is potentially false. While this research was is experimental in its design, it’s important to take next steps to explore the impact of full stories. Future work should investigate the impact of warnings on the users’ likelihood of clicking through to read the full articles, and the impact on sharing after reading the article.