Barna Group Surveys and Statistics Challenged

Lies, damned lies, and statistics: How the Evangelical-run Barna Group is unreliable

To start with, was there possibly some confusion among the respondents? An unusually high number were unsure of their response to Evangelicals, answering with a “don’t know.” This number was twice as high as it was for any other category. The reason for this confusion may have been that the question appears to have been worded peculiarly, for it asked about Evangelicals, not evangelical Christians. Perhaps some respondents thought the survey was asking about evangelists—the people who knock at your door when you’re just sitting down for dinner.

 

Christians Are Hate-Filled Hypocrites…and Other Lies You’ve Been Told: A Sociologist Shatters Myths From the Secular and Christian Media, by Bradley R.E. Wright (Bethany House, 2010).

In 2002, the Barna Group conducted a survey of 270 non-Christians. They asked these non-Christians their impressions of actors, lawyers, and nine other different groups in society, including ministers, Evangelicals, and born-again Christians as three of the eleven groups. The Barna Group found that born-again Christians and ministers scored high in respect, but Evangelicals scored rather low. Based on these data, the Barna Group concluded that non-Christians are “dismissive” of Evangelicals. According to the article, this negative opinion has consequences: “One reason why evangelical churches across the nation are not growing is due to the image that non-Christian adults have of evangelical individuals.” Wow, if this is true, it gives us a key to church growth—changing non-Christians’ negative impressions of Christians. But the author is not sure how much credence we should give to the Barna Group’s conclusion, for several reasons, and Select Items readers know my own negative opinion of what Barna reports.

To start with, was there possibly some confusion among the respondents? An unusually high number were unsure of their response to Evangelicals, answering with a “don’t know.” This number was twice as high as it was for any other category. The reason for this confusion may have been that the question appears to have been worded peculiarly, for it asked about Evangelicals, not evangelical Christians. Perhaps some respondents thought the survey was asking about evangelists—the people who knock at your door when you’re just sitting down for dinner.

When the Barna Group asked specifically about born-again Christians, the respondents were much more favorable, ranking them third highest overall. How many of us, Christian or otherwise, could describe the difference between a born-again Christian and an Evangelical? Some surveys have even used the terms interchangeably, so the fact that the Barna Group’s study found such different results for these groups raises a red flag.

There’s also reason to be skeptical of the Barna Group’s conclusion owing to the math involved. The Barna Group’s discussion of this statistic focuses on the fact that only 23% of respondents had a favorable impression of Evangelicals. This number, however, includes the respondents who “don’t know” in the denominator. In other words, if you asked the question “Twenty-three percent of what?” the answer would be “Twenty-three percent of the 270 people who took the survey.” But this isn’t quite fair. It would make more sense to answer the question “How many people have a favorable impression among those who have heard of Evangelicals in the first place?” After all, if you don’t know what an Evangelical is, there’s no chance of having a good impression of them. Dropping the “don’t know” respondents from the denominator bumps the number of favorable ratings of Evangelicals to 28%, putting them in the middle of the favorable/unfavorable range re respondents’ perceptions of religious, professional and other groupings.

There is also a problem related to the sample size of only 270 survey participants. There is nothing wrong with smaller studies, per se, but the smaller size just means that we can only detect big differences between groups, and not small ones. And, there is no meaningful (i.e., statistically significant) difference between actors, lawyers, Republicans, lesbians, and Evangelicals, for they each have 23 to 25% favorable ratings. With this small sample size, the study gives no evidence that these groups are statistically different (in terms of favorability) in the general population.

Finally, even if we accept that this statistic accurately reflects public opinion (which is discussed in chapter 8 and found wanting), the picture it paints isn’t all that bad. Less than 1 in 4 (23%) of the non-Christian respondents held unfavorable opinions about Evangelicals. The rest were either positive, of no opinion, or didn’t know. This seems to be a reasonably low number, given that none of the respondents embraced the tenets of Christianity.

One’s take on these data certainly should be cautious. They may even demonstrate a positive view of Christians (just the opposite of the Barna Group’s conclusion). If a student turned this in for a class assignment, the author would tell the student that the research question is interesting, but his or her analyses and presentation need to be redone. However, the Barna Group’s findings and conclusion were catchy, so they were picked up by the media. The Atlantic magazine (July 2003) summarized this study with the title “Evangelicals and Prostitutes.” They wrote that “Non-Christians, it turns out, have a low regard for evangelical Christians, whom they view less favorably than all the above-mentioned groups except one: prostitutes.”

Christine Wicker, in her book The Fall of the Evangelical Nation: The Surprising Crisis Inside the Church (HarperOne, 2008) summarized the study as follows: “When asked to rate eleven groups in terms of respect, non-Christians rated Evangelicals tenth. Only prostitutes rated lower” (143). She did not even cite the original study, instead presenting it as an unambiguous fact reflecting high “anti-evangelical sentiment.” From Wicker’s book, a Christian organization (named Off the Map Web site) and others found the statistic and put their own spin on it. The site “A Blind Beggar” (subtitled “Devoted to the Journey of Christianity”) summarized it as “Only prostitutes rank lower than evangelicals in terms of respect in the mind of the public.” Notice that now Evangelicals are disrespected by society as a whole, not just non-Christians. Another Web site recast the statistic as “Only prostitutes rank lower than Evangelicals.” Forget respect, Evangelicals are lower in everything. The Barna Group’s statistic was not particularly well-constructed to begin with, but it got substantially less accurate and more dire with each retelling.

James C. Pakala Library Director at Covenant Theological Seminary.