*Ninth Commandment: Thou Shalt Not Bear False Witness.
recently reported by Philip Brenner of the University of Michigan found that Americans exaggerate the frequency of their attendance at religious services, notably more so than do people in European countries like Austria, Germany, France, Spain, Italy, Norway, Finland, Slovenia, Great Britain, and Ireland.
This research compared two sources of information about church attendance: what people say they do based on surveys and what people actually do based on time diaries. More confidence can be had in time diaries, which are ongoing, than in surveys, which are after-the-fact. Survey data over the past few decades consistently point to church attendance rates among Americans of 35-45%, whereas time diaries suggest that the figure is closer to 25%, about the attendance rate in most European countries estimated from both sources of information. In other words, Americans over-report how often they attend church, and Americans do so to a greater extent than Europeans.
The substantive importance of these data and the discrepancy they reveal is that Americans are usually described by social scientists as highly religious, much more so than citizens of other Western nations. Plausible theories about American religiousness have followed, but they may be wrong because they have typically relied on survey data to describe a phenomenon that may not exist.
American religiousness remains exceptional, Brenner commented, but its special character lies more in the apparent importance of religiousness to one's identify than in actual behavior. Explanations are still needed for American over-reporting of religious attendance, but not for attendance per se.
Cynics may chuckle at these results, but over-reporting may or may not have anything to do with hypocrisy. It may or may not be deliberate.
What we have here is a general caution against always taking what people say at face value. I am a survey researcher, and I believe that talking to someone and taking what he or she says seriously are great first steps in social science research. But first steps are not the only ones that matter, and further information gathered in other ways always makes good scientific sense.
Survey questions posed to respondents need to be in principle answerable. Questions about the well-being of one's pancreas, for example, are probably not answerable by the typical person, although they no doubt would result in answers if asked. Church attendance is not such a mystery to a survey respondent, but if the questions are about past attendance, the inaccuracy of memory may intrude. Why it intrudes more for Americans than for Europeans is not clear.
Survey questions should not have answers that are so socially desirable or so socially reprehensible that answers cannot be taken at face value, even if the questions are readily answerable. In the United States, church attendance is obviously valued highly in most circles, and respondents may respond to questions about attendance in terms of their image of themselves as "good" people. Again, I emphasize that doing so may be inadvertent.
Positive psychologists study the good life, necessarily focusing on socially desirable topics and often relying on self-report survey data. For positive psychology to progress, the research methods must be diversified.
* As readers may know, the 9th commandment with which this entry begins is often rendered as "Thou shalt not bear false witness against thy neighbor" (Exodus 20: 16), and its original meaning may have referred specifically to lying in a court of law. Nowadays, it is often shortened as I have done and generalized to refer to any misrepresentation. In any event, please see the larger point of the blog entry ... maybe that should be the 11th commandment!