Robert WUTHNOW.  Inventing American Religion: Polls, Surveys, and the Tenuous Quest for a Nation’s Faith. London, UK: Oxford University Press, 2015, pp. 247. $29.95 hardcover ISBN 9780190258900.  (Also e-book). Reviewed by Meg Wilkes KARRAKER, University of St. Thomas, St. Paul, MN 55105.

 I received excellent graduate training in research methods, especially survey research, for my Master of Science at North Carolina State University and for my Doctor of Philosophy at the University of Minnesota, thanks to Glen McCann and Roberta Simmons. Many (many!) years later, I think I do a respectable job of sensitizing my sociology students to be critical thinkers about social research, especially populations and sampling, response rates and potential compromises on validity and reliability, accurate survey construction and ‘hidden threats’ like social desirability effects. But then I turn these new scholars on to large-scale data sources with nary a thought. With that in mind, and after reading Elaine Ecklund’s back splash (“[Wuthnow] shows that numbers cannot speak for themselves and that Americans have gone too far in letting surveys and polls define our faith experiences for us.”), I was curious what Wuthnow would have to say about some of my go-to sources on sociology religion (e.g., the General Social Survey, the Pew Research Center).

            Neither of those venerable sources of survey-based data is spared Wuthnow’s critical eye.  For example, he dissects data variability between Gallup and the GSS on church attendance among Americans. From a symposium published in the American Sociological Review in 1998, and using existing data, other surveys, and a new national study he reports:

The conclusions varied. .. One report … argued that … evidence was flawed and suggested that church attendance was at least 90 percent as high as the GSS claimed. Another report that examined time-use diaries … suggested that church attendance might indeed by much lower than report in surveys. There was no disagreement that the Gallup figures were too high (p. 134). And regarding Pew’s international polls about attitudes toward the United States in Muslim countries, one Middle East specialist writing in Foreign Policy wrote: “The polls are one dimensional and filled with panic” (p. 149).

These criticisms, and others Wuthnow offers, call into question the value of surveys conducted as “must-get-the-findings-out-there-quickly!” polls. Wuthnow quotes Darren Sherkat, who goes further:

Polling is conducted by whores who violate every scientific convention that social scientists developed to make sure that polling would indeed produce high quality results. … Worse yet, indifference towards high quality data is infiltrating the social sciences (p. 149).             As I moved into the last chapter in Inventing American Religion, I agreed with Wuthnow that science negligently applied can socially construct culture and society in questionable, even damaging ways. Wuthnow’s concerns regarding surveys, especially those conducted by commercial polling organizations, come with a warning to those of us who depend on public survey research. As we ponder the latent dysfunctions of such sometimes tarnished research, Wuthnow says that the “best way to guard against [the loss of credibility in polling] is conversations across disciplines, sectors, and styles of research… (207). Maybe I should review my notes from graduate school …

            Robert Wuthnow is the Gerhard R. Andlinger ’52 Professor of Sociology and Director of the Center for the Study of Religion at Princeton University. His prolific scholarship focuses on social and cultural change, including the sociology of religion.