Can we really trust cellphones to tell us how often people go to church?

This article was first published in the State of Faith newsletter. Sign up to receive the newsletter in your inbox each Monday night.

About a month ago, I experienced a crisis of faith.

The issue wasn’t my relationship with God. It was my relationship with research about God.

You see, a new working paper had been released that challenged prevailing assumptions about church attendance.

While traditional surveys, based on phone or online interviews, showed that around 22% of U.S. adults attend worship services at least weekly, the new report, based on location data from cellphones, claimed that only 5% of Americans actually do.

I was shocked, and I was worried.

How was I supposed to know which number to trust?

I comforted myself by thinking of reasons why the cellphone data might not be accurate. My dad often goes to church without his phone, I thought. He can’t be the only one.

I remembered that cops on crime shows always seem to have problems with cellphone data. I assumed the researchers must be overlooking its flaws.

After working through my angst on my own for a while, I recalled that I was, in fact, a journalist and could send an interview request. I got the author of the new working paper on the phone.

Devin G. Pope, a professor of behavioral science and economics at the University of Chicago, told me he got the idea for the new research while working on a different project with cellphone data. His team was exploring whether voters in Black neighborhoods have to wait in longer lines at polling stations than voters in white ones.

Pope realized that the cellphone data could help him circumvent one of the challenges that affects religion surveys — people’s desire to sound a bit more religious than they actually are.

The problem was that research based on cellphone data comes with its own issues. For one thing, people don’t always carry their phones with them when they’re out and about. They might leave it at home or in the car. For another, signals are a bit messy in crowded areas, making it hard for researchers to tell whether someone is at an urban church or the convenience store next door.

Pope’s team tried to avoid the first of those potential problems by eliminating cellphones that rarely changed locations. They winnowed a very large dataset down to a nationally representative sample of about 2.1 million phones that had offered consistent location updates from April 2019 to February 2020.

Then, they stress-tested their approach by checking the data against known attendance numbers, such as the size of the crowd at sporting events.

“We did a bunch of checks with things like baseball games, basketball games, shoppers at Walmart and Target and visitors to Six Flags. We looked at the actual attendance at those places and then looked at what our cellphone data would predict,” Pope said.

For the most part, the predictions were solid. When they weren’t, Pope felt like there were reasonable explanations for the gaps.

“We underreported people who attend basketball games by about 30%, and I could come up with lots of reasons for that. For example, cellphone coverage is poor at crowded arenas,” he said.

Although worship services don’t often take place in crowded arenas, houses of worship pose other, unique challenges for researchers using cellphone data. Churches sometimes share buildings with schools or community centers. Some people of faith avoid all electronics on their Sabbath day.

Still, Pope told me that he stands by the conclusion that notably fewer Americans are in church every week than claim to be.

“Almost certainly the real number is slightly higher than 5%, but I think 5% is better estimate than 22%,” he said.

But Pope also emphasized that religion research is complicated. You have to be thoughtful about using cellphone data, and you have to be thoughtful about traditional surveys, as well.

Although he didn’t say it, my lesson was clear: There’s no need to be in crisis mode when there are still so many interesting questions left to ask.

Fresh off the press

How the Bible became a star in the world of sports

‘Jeopardy!’ just featured a world religion category. Can you answer these 5 clues?

Quiz: Could you qualify for the Swiss Guard?

An attempted shooting appeared on the livestream of a church service. Here’s what happened next

Why Congress’ push to address antisemitism is angering some Christians

Term of the week: Jubilee

In the Catholic context, a jubilee is a Holy Year that takes place every 25 years.

During the Jubilee, the Vatican welcomes more visitors than normal. They visit famous pilgrimage sites and seek out forgiveness for their sins.

“The first Jubilee in the Catholic Church was proclaimed in the year 1300 by Pope Boniface VIII,” according to America magazine.

Jubilee Year 2025 will start on Dec. 24, 2024, and end on Jan. 6, 2026, the article said.

What I’m reading ...

Ben Kaplan got the surprise of his life at his Bar Mitzvah: His dad had arranged for the “Inside the NBA” crew to make a video in his honor. Twenty-four years later, Kaplan put together a beautiful tweet thread about that experience and how it changed his family forever.

The United States is going through something of a closeness crisis, and decreasing interest in organized religion likely plays a role, according to Pew Research Center. Pew found that Americans are less likely than people in other countries to feel close to their compatriots and that religiously unaffiliated Americans are much less likely than religious Americans to feel close to others in the U.S.

This month, Buddhists will celebrate the birthday of the historical Buddha. The holiday, known as Vesak, is associated with a different date and different types of celebrations depending on which country you live in, according to The Associated Press.

Odds and ends

A February survey from Pew Research Center found that U.S. adults are more comfortable speaking with friends and family members about their mental health and emotional well-being than with religious or spiritual leaders.

• Fifty-seven percent of Americans said they’d be “extremely” or “very” comfortable talking about those topics with close friends.

• Fifty-two percent said the same about immediate family members.

• Fifty percent said the same about mental health therapists.

• Just 31% of U.S. adults said they’d be “extremely” or “very” comfortable talking about their mental health and emotional well-being with religious or spiritual leaders.