WASHINGTON — Newly unsealed archival material is underpinning claims from independent scholar Thomas Kelly that the 1956 cult case study “When Prophecy Fails” was misrepresented, as a separate 39-lab replication across 19 countries also failed to reproduce a key lab prediction, Dec. 21, 2025. The double hit arrives as psychologists lean harder on preregistration, shared materials and reopened records to test whether cherished effects actually hold.
Cognitive dissonance is the discomfort that comes from holding conflicting beliefs — or acting against what you say you believe — and then trying to reduce that tension. For decades, the concept has been taught with two familiar supports: a dramatic doomsday narrative in the field and a tidy “free choice” pattern in the lab.
Why cognitive dissonance is suddenly under pressure
Those supports are now being pressed at the same time. One paper is reopening the historical record behind the cult story. Another is testing, at scale, whether a foundational laboratory setup still delivers the result students expect.
Newly unsealed archives reopen the cult story
Kelly’s critique, indexed in PubMed, draws on what he describes as newly unsealed archival material connected to “When Prophecy Fails,” the 1956 book by Leon Festinger, Henry Riecken and Stanley Schachter about a group led by Dorothy Martin that predicted an apocalyptic flood, Dec. 21, 1954. In his abstract, Kelly says the book’s “central claims are false,” arguing the group proselytized well before the prophecy failed and quickly abandoned the specific prediction afterward.
Kelly also alleges serious ethical violations by the researchers, including fabricated psychic messages, covert manipulation and interference in a child welfare investigation. He notes that Riecken posed as a spiritual authority and later admitted he had “precipitated” the climactic events — a detail that, if accurate, blurs the line between observing cognitive dissonance and manufacturing it.
A 39-lab replication finds the “choice” effect missing
In a separate line of scrutiny, researchers ran a preregistered, multilab replication of the induced-compliance paradigm, a staple lab method for studying cognitive dissonance through attitude change after people write a counterattitudinal essay. The registered replication report — 4,898 participants across 39 labs in 19 countries and 14 languages — is published as an open-access replication report.
The replication did find attitude movement after participants wrote the counterattitudinal essay compared with a neutral essay. But it failed to support the signature “free choice” prediction: those in a high-choice condition were no more likely to shift their attitudes than those in a low-choice condition. “In other words, we did not replicate the effect of choice on attitude change,” lead author David C. Vaidis said in an APS summary of the study.
Where the cognitive dissonance debate goes next
This is not the first time the theory has been forced to defend its foundations. The induced-compliance tradition traces back to the classic forced-compliance experiment by Festinger and James M. Carlsmith, described in their 1959 journal article. And the field’s appetite for large replications grew after warnings that flexible research practices could inflate false positives — concerns raised in a 2011 paper by Joseph Simmons, Leif Nelson and Uri Simonsohn — and after large-scale efforts such as the Open Science Collaboration’s 2015 reproducibility project.
For many researchers, the emerging message looks more like revision than rejection. Cognitive dissonance is not anchored to a single cult narrative or a single lab trick. But the archive-driven critique and the 39-lab replication point in the same direction: psychology may need to be clearer about which demonstrations still deserve “bedrock” status — and which belong in the lesson plan as cautionary tales about evidence, methods and mythmaking.
