[excerpts only, to avoid copyright violation]

Researchers find that many poll respondents exaggerate religious involvement

"OWENSBORO, KY. --

[...]

"For decades, sociologists and journalists have repeated public opinion polls reporting 40 percent of Americans attend a worship service once a week. That is far higher than in most other Western nations."

[...]

This figure was challenged by C. Kirk Hadaway and Penny Long Marler, writing in "Christian Century" "last year" (1998). In 1993, they "consulted pastors and did head counts at Protestant religious services in Ashtabula County, Ohio. They also checked attendance figures at several Catholic dioceses around the country."

"What they found was closer to 20 percent worship attendance than 40 percent, writes Andrew Walsh in 'Religion in the News.'"

[...]

"Some reasons found for inflated poll figures include":

The most committed members of religious groups are most likely to exaggerate. If they usually attend, they often claim credit even if they missed the previous week.

Once members have established a "churched identity" over several years, they may not feel the need to affirm it with attendance every week.

Some respondents will claim credit for other things, like attending a wedding, or mowing the church lawn.