Margin Of Error: Breaking down the polls

Margin Of Error: Breaking down the polls

Marriage survey story leaves scholar dissatisfied

Posted September 15, 2015

Logo for the Margin of Error blog.

— Are self-identified Republicans happier than Democrats about their marriages? Yes, according to a recent study that was reported in The New York Times blog, The Upshot. Not only is the conclusion NOT supported by the survey data, but it also is improperly magnified by the blaring media headline.

Don’t mistake my criticisms for disliking the conclusion. I have no position about it. Instead, I am disturbed that a single poorly executed survey question could be used to make such a definitive statement about “blue versus red America” and that a respected journalist would magnify rather than dissect the dubious claim.

The General Social Survey (GSS) collected the survey data in question. Scholars like me use GSS data all the time; it’s a great survey. In this case, however, GSS asked respondents just one question to measure marriage happiness. The wording was: “Taking things all together, how would you describe your marriage? Would you say that your marriage is very happy, pretty happy or not too happy?”

Researchers prefer to have multiple measures of a concept, such as marriage happiness. Think about it – marriage is complex! Answers to that single question risk missing the variation in how people feel and what they would report if asked additional questions about their marriage. So, maybe we would get a different lay of the land with more data to work with.

The most serious problem, though, are the limited and unbalanced response options. GSS respondents could only answer by indicating they were “very happy,” “pretty happy,” or “not too happy.” Hardly anyone answered “not too happy.” In fact, no more than 4 percent of any political group said they were not too happy.

Politics and marriage satisfaction

This means that almost everyone says they are happy with their marriage. Yet, the authors of the study focus only on the 7 percent gap between Republicans and Democrats claiming to be “very happy.”

Why? Further, as the authors point out, Democrats and Republicans have different backgrounds that are significantly related to other variables known to affect marriages, such as poverty and education. Taking those factors into account, there isn’t much of a gap to explain even for that answer option (it shrinks to 3 percent).

Percent of married Americans in 'very happy' marriages

If I told you that 96 percent of men and women were at least, “pretty happy” with their marriage, I’d hope you’d think of at least two things. First, you’d agree that a headline claiming men are happier makes little sense. Second, you’d suspect a lot of people are liars. About half of all marriages end in divorce, after all. That’s a lot of happy people calling it quits! 

Being wrong about things like this can be a big deal. Misleading claims that feed partisan stereotypes tend to be sticky. Research that others and I have conducted show that once people encounter information thought to be true they continue to be influenced by it even after it is discredited. This happens when individuals fully understand the initial claim was bogus.


Please with your account to comment on this story. You also will need a Facebook account to comment.

Oldest First
View all