Thursday, March 28, 2013

Conservatives and Race Surveys: Wording Matters

How you phrase survey questions and how you report the resulting data matters a great deal within public debates. Consider the following example from a recent article.
The numbers prove it: The GOP is estranged from America
http://www.washingtonpost.com/opinions/the-numbers-prove-it-the-republican-party-is-estranged-from-america/2013/03/22/3050734c-900a-11e2-9abd-e4c5c9dc5e90_story.html
By Andrew Kohut, Published: March 22

Andrew Kohut is the founding director and former president of the Pew Research Center. He served as president of the Gallup Organization from 1979 to 1989.

Conservative Republicans are more likely (33 percent) than the public at large (22 percent) to see the growing number of Latinos in America as a change for the worse. Similarly, 46 percent of conservatives see increasing rates of interracial marriage as a positive development, compared with 66 percent of the public overall.
Let us study both points separately. First, I'll examine the perceptions of the increase in Latinos and then the issue of interracial marriage.

The wording of the first point is problematic in several ways. First, here is a rewording of the sentence in a more positive light:
A majority of Conservative Republicans (67 percent) see the growing number of Latinos in America as either neutral or a change for the better, compared to 78 percent of the public at large.
Notice that this is a mere ten percent difference. Also, the data don't examine why 22 or 33 percent of any respondents might be concerned about the rising number of Latinos. What if these people are more concerned with border security? These individuals might read questions about Latinos as proxy questions on issues of general immigration and security. It could be that conservatives are more insistent on closing borders, not that they oppose a particular minority group. Of course, the negative view is also possible — but the data are not sufficient to reach a conclusion.

That most self-identified conservatives don't have a negative view of Latino population growth seems a more important indicator of the direction of most Americans.

Then we have the second issue, that of interracial marriage. This survey question is more problematic. But, again, let us first recast the results — this time in a more negative cast:
More than a third of the public at large (34 percent) have a neutral or negative view of interracial marriage, while more than half (54 percent) of conservatives have a neutral or negative view.
The above seems horrific, but what if the meaning is being misinterpreted by pollsters and reporters?

I do not view increasing rates of interracial marriage as good, bad, or neutral. Most "conservatives" I know don't give the matter much thought, either. Among my friends, about equal numbers of "left" and "right" couples are "mixed" ethnically (a silly notion, in my view). Oddly, my gay and lesbian friends have homogenous relationships: couples share not only ethnic, but also socioeconomic backgrounds. I wonder why that is? And should it even matter? Of course we tend to meet people like ourselves.

If I don't view interracial marriage as a "positive" does that mean anything at all?

I cannot speak for Conservative Republicans, as I am not a social conservative and have little respect for the Republican Party. Most of my conservative friends care most about the freedom to marry — granted, most still limit this to heterosexual couples, though.

The social conservatives are why I reject the GOP, but we should be careful when trying to understand survey data that is incomplete and potentially biased against conservatives.

No comments:

Post a Comment