REVEAL Revisited
One sociologist says Willow Creek's research may not be as revealing as we think.

The research conducted by Willow Creek and published last year in the book REVEAL: Where are you? has generated a great deal of conversation on this blog. Some have heralded the findings as conclusive evidence that Willow's popular philosophy of ministry is fatally flawed. Others have applauded Willow for the courage to be transparent about its shortcomings and seek more effective methods of making disciples. While the discussion has been stimulating, most of us lack the credentials to offer anything more than a layman's opinion about REVEAL. But not Bradley Wright. He is an associate professor of sociology at the University of Connecticut, and has written an 11 part analysis of Willow's study on his blog. Wright has summarized his take on REVEAL below.

When I go to my physician for a check-up, he starts with a series of simple tests - shining a light in my eyes, looking at my throat, listening to my breathing, and so forth. If the results of these don't seem right, he then orders more sophisticated tests, such as blood work, a biopsy, or x-rays. I would hope that he wouldn't cart me off for surgery or chemotherapy based solely on the initial, simple tests.

This illustrates how we might think about the REVEAL study conducted by Willow Creek Community Church. As described in the book REVEAL: Where are you?, this study collected data from about five thousand respondents in seven different churches. Its results have caused quite a stir. Critics point to them as evidence against the Willow Creek model of ministry. In the foreward to the book, Bill Hybels, senior pastor of Willow Creek, describes the findings as almost "unbearably painful." The findings of REVEAL, he writes, "revolutionized the way I look at the role of the local church." Coming from as successful a pastor as Bill Hybels, this is a powerful statement.

Is such a strong reaction warranted? I would say probably not, and here's why...

The type of survey used by REVEAL has its uses, but it's not well suited for evaluating the effectiveness of a complex institution like a church. It's not that REVEAL's findings are wrong, rather they are highly inconclusive. In fact, if I had to make a judgment, I would interpret the findings as generally supportive of what Willow Creek is already doing.

Technically, REVEAL used a cross-sectional survey with no comparison group and no randomization. This means they surveyed people once during a given period of time - it's like taking a snapshot of a group of people. It's the tongue depressor of survey methodology - a good place to start, but not a very powerful tool. While this type of survey does a good job in describing peoples' characteristics, it doesn't explain them. It describes "what" but doesn't explain "why." Findings from this type of survey are open to multiple interpretations, and the data themselves can't distinguish the correct one. To illustrate, let's consider some of REVEAL's findings.

January 14, 2008