Ask anyone why trivia is good for your memory and you will hear the same answer: curiosity. Wanting to know makes the brain pay attention, so the answer sticks. A study published in March 2026 put that tidy story to a stricter test — and it did not survive intact. When curiosity, confidence, and how satisfying the answer felt were measured separately and weighed in the same analysis, curiosity on its own stopped predicting what people remembered a week later. What predicted recall instead was the moment after the reveal: whether the answer felt satisfying, and whether it beat what the person had expected.
That is not a reason to stop being curious. It is a better map of where memory is actually made during a trivia question — and it changes how you should play today's question.
The study, by Michelle Hirsch, William Fisher, and Andrée-Ann Cyr of York University in Toronto, was published on March 13, 2026 in Frontiers in Cognition (full paper). It compared generations directly: 80 younger adults (ages 18–33, average about 20) and 80 older adults (ages 54–94, average about 71) worked through the same online task.
The task used 70 general-knowledge trivia questions from an established set (Kang et al., 2009). One sample question: What animal's milk does not curdle? (Answer: the camel.) For each question, the sequence mattered more than the question itself:
Then the part participants were not warned about: seven days later, they were shown the same questions again and asked to recall the answers. Anyone who admitted looking answers up was excluded, and first-time correct guesses were removed from the analysis, so the results are about learning answers you did not already know.
In the researchers' main model, initial curiosity had essentially no effect on recall a week later (an odds ratio of 0.99 — a coin flip). Confidence did predict recall (odds ratio 1.10), and post-answer satisfaction predicted it more strongly (odds ratio 1.19). The pattern was identical for younger and older adults.
This does not prove curiosity is useless, and the authors do not claim that. Curiosity and confidence were strongly correlated — people feel most curious about questions they almost know — so curiosity may have little separate signal once confidence is in the model. An earlier study the authors cite (Fastrich and colleagues, 1,495 participants) did find pre-answer interest stayed significant. The honest reading is narrower: wanting the answer is not, by itself, what fixes it in memory.
The study's second measure is the clever one. The researchers subtracted each person's curiosity rating from their satisfaction rating to get an information prediction error (IPE) — in plain terms, how much better or worse the answer was than the wanting had promised. A positive IPE means the answer beat the expectation ("oh, that's good"); a negative one means it underwhelmed.
Answers with positive prediction errors were significantly more likely to be recalled a week later (odds ratio 1.09), again in both age groups equally. This lines up with the hypercorrection effect, which the authors also observed: wrong guesses made with high confidence — the ones producing the biggest expectation mismatch — were more likely to be corrected at recall than low-confidence misses. Surprise at being confidently wrong is a strong update signal, at 20 and at 71 alike.
The mechanism, then, looks less like "curiosity opens the door" and more like this: the guess sets an expectation, and memory keeps score of how the answer compares with it.
The older group reported higher state curiosity, confidence, and satisfaction during the task than the younger group, and started with more knowledge — they guessed 12% of answers correctly at first pass, against 7% for the younger group. Younger adults still recalled a higher share of the new answers a week later, and more years of education predicted better recall in both groups. But the emotional machinery — how confidence, satisfaction, and surprise related to memory — worked identically across the 50-year age gap.
A companion lifespan study points the same way. Whatley, Murayama, Sakaki, and Castel, in PLOS ONE (May 7, 2025), separated trait curiosity (how inquisitive you are in general) from state curiosity (the momentary itch one question creates), in a pilot of 193 people and a preregistered main study of 1,218 adults aged 20 to 84. Trait curiosity declined with age. State curiosity did the opposite: it dipped in early adulthood, then rose sharply after middle age and kept rising into old age. As UCLA's summary puts it, senior author Alan Castel's reading is selectivity, not disengagement.
Two honest caveats. First, the York study's older sample was unusually high-functioning — more educated than the younger group (16.4 versus 13.5 years of education) and higher in openness — so the authors warn it may not represent older adults generally. Second, the suggestion that maintained curiosity may help protect against cognitive decline is a hypothesis these data fit, not something either study proved; neither tested an intervention or should be read as medical advice.
No. In this study, curiosity added no predictive value once confidence and satisfaction were in the same model. Treat "curiosity alone is enough" as unproven, not disproven.
The gap between how much you wanted an answer and how satisfying it turned out to be. A positive gap — better than expected — predicted better recall; a letdown predicted forgetting.
They recalled a smaller share of the new answers after a week — but started knowing more (12% versus 7% correct at first guess), reported higher curiosity and satisfaction, and showed the same memory effects as the younger group.
Trait curiosity is your general disposition. State curiosity is situational — the itch created by one specific question. The 2025 lifespan study found trait curiosity declines with age while state curiosity rises after middle age.
This article is the starting point. Copy any of these prompts into your favorite AI assistant to learn more about curiosity, satisfying answers, and trivia memory:
🔎 Run the York study's sequence on yourself for one topic
Quiz me on [your topic, e.g. "world capitals" or "human anatomy"] using this exact protocol, one question at a time: (1) ask the question and take my typed guess; (2) ask me to rate my curiosity 1-10 and my confidence 1-10; (3) reveal the answer; (4) ask me to rate how satisfying it was 1-10; (5) record my information prediction error (satisfaction minus curiosity). After 10 questions, show me a table of all five numbers per question, sorted by prediction error, so I can see which answers beat my expectations.
🧭 Stress-test the "curiosity didn't matter" headline like a skeptic
A March 2026 Frontiers in Cognition study (Hirsch, Fisher and Cyr; 160 adults, 70 trivia questions, recall after 7 days) found that when curiosity, confidence, and post-answer satisfaction were modeled together, curiosity did not predict recall (odds ratio 0.99), but satisfaction (1.19) and information prediction errors (1.09) did. Give me the three strongest reasons this might NOT mean curiosity is unimportant for memory — consider overlapping measures, model choices, and the earlier Fastrich study of 1,495 people — and say what evidence would settle it.
💡 Build a one-week-later retest from your own misses
Here are trivia questions I got wrong this week, with my wrong guess and the correct answer: [paste your list]. Turn them into a 7-day-later retest plan: quiz me cold on day 7 without showing answers until I commit, flag the ones where my original guess was confident (hypercorrection candidates), and for each answer write one sentence explaining why it is satisfying or surprising, so I get a positive prediction error the second time too.
Tip: replace the bracketed parts with your own situation — the more specific your prompt, the more useful the answer.
A March 2026 York University study (160 adults, 70 trivia questions, surprise recall test 7 days later) separated the feelings around a trivia answer and weighed them together. Initial curiosity, on its own, did not predict which answers people remembered (odds ratio 0.99). Confidence did (1.10), post-answer satisfaction did more (1.19), and answers that felt better than expected — positive information prediction errors — were also remembered better (1.09), identically in adults aged 18–33 and 54–94. A 2025 lifespan study of 1,218 adults adds the other half: general trait curiosity fades with age, but the momentary curiosity a specific question creates rises after middle age.
The single most important takeaway: do not just get curious — get an answer that beats your expectation, and linger on it for a few seconds, because that post-answer moment is what your memory actually keeps.
One sharper AI prompt + one new daily trivia topic every week. Free forever, unsubscribe anytime.
We never share your email. One email a week, that's the deal.
Published October 19, 2026