Most Americans Say They Don’t Understand Statistics

Confused by Statistics? So Is Most of America, a New Survey Finds

Most Americans say they have little or no understanding of statistics, including basic ideas like p-values, the kind of number that shows up constantly in news stories about new drugs, diets, and studies. Yet in the same breath, nine out of ten of those surveyed said they would actually use that information to make better decisions if they understood it more. That combination, widespread confusion paired with genuine curiosity, sits at the center of a new nationally representative survey published in the journal PLoS ONE.

Researchers Samuel Anyaso-Samuel of the National Cancer Institute and Mark Louie Ramos of Penn State went looking for a simple answer: how many people actually feel comfortable with statistics, and would they use them more if someone helped them along? To find out, they turned to a survey of 1,000 U.S. adults designed to reflect the country as a whole, matched to the population on things like age, income, education, and where people live. What they found suggests confusion about data is common, but the appetite to fix that is even more common.

Most Adults Never Learned the Basics

Twenty-five percent of adults said they had no idea what statistics or p-values were. Another 37% said they’d heard the terms before but couldn’t explain them if asked. Add those two groups together and more than six in ten adults report knowing almost nothing about a type of information that quietly shapes news, medicine, policy, and now artificial intelligence, given that the algorithms behind AI chatbots and search tools are themselves built on statistical reasoning.

Only 27% said they’d picked up some statistics in school, and just 11% called themselves regular users.

Despite that gap in knowledge, the willingness to learn was clear. Among people who said they knew nothing about statistics, 78% still said they’d use the information to make decisions if they understood it better. That number climbed to 92% among people with a little familiarity, and to 96% among those who’d at least learned some statistics in school. In other words, the less someone knows, the more room there is for that number to grow, and even at the very bottom of the knowledge scale, most people are already leaning toward yes.

Who Reports Knowing More, and Who Doesn’t

Education made the biggest difference. People with at least some college were about twice as likely to say they understood statistics well compared to those with only a high school diploma. More than 38% of adults with a high school education or less said they had no idea what statistics or p-values even meant, compared to just 12% of those with a four-year degree or higher. Even a postgraduate degree wasn’t a guarantee of comfort with numbers: only 22% of that group called themselves regular users of statistics, suggesting formal education helps, but it doesn’t close the gap on its own.

Younger adults consistently reported feeling more comfortable with statistics. Gen Z had the lowest share reporting zero knowledge, at just 14%, and the highest share of regular users, at 17%. Baby Boomers landed at the opposite end on both counts. Researchers suspect this tracks with how much more statistics has crept into school curricula over the past couple decades, along with younger adults simply swimming in data-heavy content online every day, though the survey itself can’t prove why the generations differ.

Men reported slightly more comfort with statistics than women, though the gap was modest. Income mattered too: adults earning between $50,000 and $150,000 a year said they’d be more willing to use statistics than people earning less or considerably more, a wrinkle the researchers say is worth digging into further. Political affiliation and race weren’t linked to how confident people felt about their own statistics knowledge, though independents and Black respondents were somewhat more willing to use statistics than other groups if they understood the subject better, another finding researchers flagged for future study.

Researchers Tie the Findings to the Rise of AI

Researchers frame their findings within a concern that goes beyond math class: artificial intelligence. The large language models now embedded in search engines, medical tools, and everyday apps run on statistical reasoning under the hood, even if the average person never sees the math happening behind the screen. Making sense of what those systems produce, or at least knowing when to be skeptical of it, requires some basic grasp of how statistics work. Without that, people risk taking whatever an algorithm tells them at face value, with no real way to push back or ask whether the answer actually holds up. As AI tools become the default way many people search for information, get medical guidance, or check facts, that gap in understanding starts to matter well beyond the classroom.

As the authors write, “Without statistical literacy, individuals risk becoming passive consumers of AI-generated information rather than critical evaluators.”

The public, it turns out, is ready to meet that challenge. The bigger question is whether schools, workplaces, and policymakers are ready to help them get there.

---

---