Illusory Truth Effect

Media & Technology

Hasher and Goldstein (1977)

If we are told something often enough, will we believe it?


On a day-to-day basis, we  have to make judgments about the truth or falsity of so many statements that we must generally do so without having the required knowledge. To solve this problem, we resort to heuristics, i.e., rules of thumb that are not entirely rational, but nevertheless practical ways to reach an immediate solution. One such rule is that ideas are more likely to be true the more familiar they sound. This phenomenon, called the Illusory Truth Effect, can be explained by both Zajonc (1968)’s Mere Exposure Effect and Kahneman’s (2011) Dual System Theory. According to the former, repeated exposure to a certain stimulus induces a positive attitudinal change, even if its familiarity remains unconscious. Indeed, as explained by Kahneman, most of our thinking is unconscious and the product of a “system 1” that has evolved as an energy-saving mechanism. As such, system 1 positively values speed and cognitive ease, and thus automatically endorses statements the processing of which is facilitated by implicit memory. More generally, system 1 is guided by the following principle: WYSIATI, what you see is all there is. One of its applications is that familiarity can be used as a proxy for more reliable truth criteria, such as peer review or empirical support. If familiarity leads to “truthiness” (the quality of something that feels true), then mere repetition should be a “major access route that plausible statements have into our pool of general knowledge” (Hasher and Goldstein, 1977), and repeated statements should be more likely to be judged as true than similar non-repeated statements.

Participants & Procedures

30 college students from Temple University (Philadephia, PA, USA).

To test this hypothesis, Hasher and Goldstein (1977) exposed their subjects to 140 plausible assertions on three successive occasions separated by two-week intervals. The participants thought that they were helping to validate a new general knowledge test. They were told that the statements might be true or false, and asked to rate each one on a seven-point certainty scale (1 = “Definitely false – 7 = “Definitely true”). For each trial, the researchers used a recording of 60 assertions (30 true, 30 false) covering 10 different topics (history, politics, sport, art, geography, etc) and chosen in such a way as to sound plausible to the participants, but not be known by them to be true or false. Of those statements, 20 had been randomly selected to be repeated on all three trials (the other 40 being new each time).


As can be seen in the table above, the participants gave higher certainty ratings to the statements that were true than to the ones that were false. Without knowing them, they were able to successfully discriminate between them. However, repeated statements saw their ratings increase over time, and this regardless of whether they were true or not. The effect was actually even stronger for false statements. Importantly, the ratings of non-repeated statements did not show the same pattern. They actually diminished slightly, maybe due to an order effect. Likewise, the repeated statements did not receive higher scores during the first sessions, indicating that the effect was indeed due to mere repetition. Thus, results confirmed the Illusory Truth Effect: “if people are told something often enough, they’ll believe it”. This phenomenon is of course even more relevant now than ever, with the recent explosion in communication brought about by digital and social media.

IA Tip

In line with IB guidelines, we recommend that students only compare two conditions in their experiment and obtain a single measurable result for each participant in each condition. Doing otherwise would complicate inferential statistics without any benefit as far as the IA is concerned.

Here, students could measure the difference between participants’ scores on the first and last trial. Properly choosing a sufficient number of statements and using filler tasks can also make it possible to conduct a modified replication of this study in one sitting.


Lynn Hasher, David Goldstein, Thomas Toppino, (1977). Frequency and the conference of referential validity. Journal of Verbal Learning and Verbal Behavior, 16, 107-112.