The Sleeper Effect: Why Bad Sources Persuade You

Your brain keeps messages longer than it keeps sources. The sleeper effect — confirmed across 72 experiments in a meta-analysis by Kumkale and Albarracín — shows that claims from untrustworthy sources become more persuasive over time, not less. Your memory discards who said it faster than what was said, leaving discredited claims floating in your head with no warning label attached.

During World War II, the US Army showed propaganda films to soldiers and measured their opinions five days later, then nine weeks later. The surprise: messages the soldiers had initially dismissed — because they knew the source was biased — became more convincing with time. Carl Hovland and his team at Yale had stumbled onto something that would puzzle psychologists for decades.

They called it the sleeper effect. And it's never been more relevant than right now.

Your Brain Has a Filing Problem

Here's what happens. You scroll past a post claiming a new study found that a common vaccine causes heart inflammation. You check the source — it's a website you don't recognise, with a name designed to sound medical. You dismiss it. Smart move.

Three weeks later, someone mentions vaccine side effects in conversation. A thought surfaces: didn't I read something about heart inflammation? You can't remember where. The sketchy source is gone from memory. The claim remains.

This isn't a failure of intelligence. It's a failure of how your brain encodes information. Memory researchers call it source amnesia — your brain stores what you learned in a different system than where you learned it. The "what" system is robust. The "where" system is fragile.

Kumkale and Albarracín's 2004 meta-analysis across 72 experiments confirmed the pattern: when people received a persuasive message paired with a reason to distrust it — a non-credible source, a disclaimer, a fact-check warning — they were less persuaded immediately. But over time, the discounting cue faded faster than the message. The persuasion increased.

The message won by outlasting the warning.

Why This Is Worse in 2026

Hovland studied soldiers watching a single propaganda film in a theatre. They saw one message, from one source, at one time. Today, the average person encounters thousands of claims per day across dozens of sources. Your brain was never designed to tag each claim with its origin.

Think about how you actually consume information. You open a social media app. A headline catches your eye in someone's repost. You see a screenshot of a tweet embedded in an Instagram story. You hear a podcast guest reference a study. You skim a push notification. Each claim arrives stripped of context, detached from its source, pre-packaged for sleeper-effect delivery.

Social media doesn't just accelerate the sleeper effect. It optimises for it. Every share, repost, and screenshot removes one more layer of source information. By the time a claim reaches you, it might have passed through five intermediaries. The original source — whether it was Reuters or a random blog — is invisible.

A 2023 study published in PubMed examined the sleeper effect specifically in social media contexts. Researchers showed 324 workers Facebook posts containing false safety information. Even when the posts came from obviously unreliable pages, participants who were surveyed weeks later showed increased agreement with the false claims. The discounting cue — the sketchy Facebook page — faded. The claim persisted.

The Perception Gap No One Talks About

At Albis, we track how different regions frame the same stories differently. But the sleeper effect reveals a deeper problem: it doesn't matter how a story is framed if people forget the framing and only remember the claim.

Consider the coverage of the Iran-US conflict. CNN, Al Jazeera, IRNA, and Xinhua all report different versions of the same events. Readers who check multiple sources might notice the contradictions in real time. But what about a month later? The specific framings blend together. Source labels dissolve. What remains is a residue of claims — some from credible reporting, some from state propaganda — all feeling equally plausible because the brain has shed the metadata.

This is the perception gap that no media literacy programme addresses. We teach people to check sources in the moment. We don't teach them that their brains will erase that work weeks later.

The Repetition Multiplier

The sleeper effect gets more dangerous when it meets the illusory truth effect. Repetition makes claims feel true. Source amnesia strips away the reasons to doubt them. Together, they're a one-two punch.

A false claim from an unreliable source gets dismissed. Then you encounter the same claim from a different unreliable source. Dismissed again. Then a third time, in a meme. By now, the claim feels familiar — and familiar feels true. Meanwhile, all three dismissals have faded from memory. You're left with a claim that feels like common knowledge, sourced from nowhere and everywhere.

Propagandists have understood this intuitively for centuries. Modern information warfare just industrialises the process. Flood the zone with a claim from enough low-credibility sources, and time does the rest. You don't need people to believe the messenger. You just need them to remember the message.

What You Can Actually Do

The uncomfortable truth is that awareness alone doesn't fix this. Knowing about the sleeper effect doesn't make your brain suddenly retain source information. The hardware limitation is real.

But there are practical workarounds:

Write it down. When you encounter a claim that surprises you, note the source alongside the claim. "Heard X from Y" in a notes app takes three seconds and defeats the entire mechanism. Your phone's memory doesn't decay. Be suspicious of "common knowledge." When you catch yourself thinking everyone knows that or I read somewhere that, pause. If you can't name the source, the claim deserves fresh scrutiny. The feeling of familiarity isn't evidence. Watch for orphaned claims in your own arguments. If you're making a point and can't remember where the supporting fact came from, that's a red flag. The sleeper effect may have already done its work. Distrust time. The older a memory is, the less reliable your source tag becomes. A claim you evaluated critically last month might need re-evaluation today — not because the facts changed, but because your mental metadata did.

The Claim That Outlasts Its Warning

Every fact-check, every "misinformation" label, every community note operates on an assumption: if we flag the source, people will discount the claim. And they do — immediately. The meta-analysis confirms it.

But the flag fades. The note dissolves. The claim stays.

The sleeper effect means every piece of misinformation you've ever dismissed is still in your head, slowly shedding its warning label. Not all of them will resurface. But some will. And when they do, they'll feel like things you've always known.

Your brain doesn't delete bad information. It just loses the receipt.