(OSV News) — Claims of a Christian revival among the UK’s young adults may be based on data from surveys whose methodology can’t guarantee accurate results — while two major UK studies show a downturn, rather than a resurgence, in faith, said Pew Research Center.
On Jan. 23, Pew published a short analysis on conflicting data among surveys assessing whether Britain’s young adults are embracing Christianity.
A flurry of opt-in surveys — where participants sign up to participate “often in response to website ads or email campaigns” — have touted an upswing in Christian belief, practice and identity among those in the 18-34 age range, wrote Conrad Hackett, associate director of research and senior demographer at Pew, in his analysis.
Sponsoring organizations for such surveys have included the UK charity Bible Society, which announced in April 2025 that “church attendance in England and Wales is on the rise,” with “the most dramatic increase seen among young people.”
Tearfund, another British-based Christian charity, said its 2024 opt-in survey showed a spike in young adult participation in online worship, with its 2025 survey indicating that a greater share of younger rather than older adults planned to attend Christmas services.
Opt-in surveys by the Church of England, polling firm YouGov, the Eternal Wall of Answered Prayer project and the Institute for the Impact of Faith in Life have all found “high levels of religious activity among young adults in recent years,” Hackett wrote.
Data disparity
However, data from major surveys drawing on random population samples “show that Christian identity and practice are not increasing among young adults in Britain,” he noted.
Britain’s Labor Force Survey, which typically interviews more than 50,000 persons per quarter, has measured a continuing decline in Christian identity in the UK across age groups, with 44% of British adults identifying as Christian by the summer of 2025, down from 54% in early 2018, according to Pew.
“Throughout this period, older British adults consistently identified as Christian at higher levels than young adults,” wrote Hackett.
Data from the annual British Social Attitudes survey, which polls a random sample of over 3,000 adults, also “shows no clear evidence of a Christian revival,” he said.
“Among all adult respondents, the share who identify as Christians and who go to church at least once a month was 12% in 2018 and 9% in 2024,” Hackett wrote.
The BSA poll found that the number of young adult Christian churchgoers “has not risen above pre-pandemic levels,” standing in 2024 at 6% for adults ages 18 to 34, down from 8% in 2018.
Trouble with opt-in polling
The data disparity reveals some key issues with the opt-in survey approach.
Hackett told OSV News that “a lot of times, opt-in polls can produce misleading results, especially for young adults.”
In his analysis, he cited earlier observations by Pew that “young, online opt-in respondents are much more likely to answer ‘Yes’ in Yes/No questions, regardless of the truth.”
A 2024 report from Statistics Canada, that nation’s official statistical agency, “found 12% of young online, opt-in respondents claiming they are licensed to pilot a nuclear submarine (versus 1% of those ages 65 and older),” wrote Hackett in his analysis.
Hackett explained to OSV News that his team has found that “there are some people who seem to not be answering questions sincerely” among those who take opt-in polls.
And, he added, “that includes people who might be aware that there’s always a shortage of young adults in these pools.
“So somebody who just wants to be able to take a survey and get a reward might say they’re a young adult, even if they’re not — and whatever their actual identity is, may take this survey in a way that lets them complete it quickly and doesn’t reflect their actual views,” Hackett said. “And so this bogus respondent data gives us information that’s not helpful, and it can pollute what’s actually happening.”
Hackett told OSV News that opt-in surveys on religious identity and practice stand to be further imperiled by some respondents’ use of large language model (LLM) tools, a form of artificial intelligence, “to fill out surveys very quickly.”
“According to a pre-programmed person, they can say, ‘For the purposes of this exercise, I’m a 25-year-old from Kansas and this is my job,'” said Hacket. “And the LLM will look at the individual questions and answer in a way that would be appropriate for that kind of persona.”
“The question increasingly is how much value they have in any kind of research,” David Voas, emeritus professor of social science at University College London, told OSV News in an email. “I used to think that they generally deliver results in the right ballpark, but with the advent and adoption of AI, and some signs that samples are far from representative, I’m having doubts.”
In his analysis, Hackett cited political scientist Sean J. Westwood of Dartmouth College, who in a November 2025 article warned that “the advancement of large language models poses a severe, potentially existential threat to online survey research.”
“We’ll come up with detection methods to beat it, and then the LLMs will figure out a way to evade the detection models,” said political scientist and statistician Ryan Burge of Washington University, author of the Substack newsletter “Graphs about Religion.”
Hackett told OSV News that “the challenge is that online, opt-in polls are inexpensive, relative to doing a survey random sample where — as at Pew — we sample people using a random list of mailing addresses, and then we send mailers and invite people to be part of our panel.
“So this is a time-consuming and expensive process,” Hackett admitted. “Not all organizations can do that.”
But “if you just post something online inviting, say, Catholics to respond, it’s probably going to be worthless,” Voas said. “If any incentives are offered, and possibly even if there aren’t, you’ll have non-Catholics responding, and even if you can avoid that problem, the people who do take the time are likely to be atypical. … You’re more likely to pick up the extremes than the big group in the middle.”
Caution with survey data
Voas stressed that research on religious belief, identity and practice — like any social science research — is to a great extent “all about sampling.”
And random sampling is “the gold standard for a reason,” Burge told OSV News.
He also stressed against uncritically accepting survey data, especially when it doesn’t align with common sense and lived experience.
“Some of those results, like a tripling of belief in God — you would never see that in the real world,” Burge said. “A lot of these results are just eye-poppingly wrong, especially among young people.”
Burge also highlighted the need for self-awareness in reporting on such statistics.
“If you’re a Christian, obviously you want to see Christianity make a revival,” he said. “Even, I think, the media wants it too, because they’ve been telling the same story about religion now for 30 years, which is decline, decline, decline. And anytime you can tell a story that’s a little bit opposite of that, you’re going to get more clicks, likes, traffic and conversation.”
“You’ve got to think about how you’re thinking about it,” said Hackett.
Gina Christian is a multimedia reporter for OSV News. Follow her on X @GinaJesseReina.
