Mark Blumenthal | May 19, 2008
Topics: AAPOR , Barack Obama , Hillary Clinton
And finally the last of my my series of brief interviews conducted at last week's AAPOR Conference, this one with Brian McCabe discussing a paper he co-authored with Jennifer Heerwig that won AAPOR's award for the best student paper presented at the conference. McCabe and Heerwig are both doctoral candidates in Sociology at New York University. The paper is entitled, "Social Desirability Bias in Estimated Support for a Black Presidential Candidate."
For those intrigued by McCabe's summary, I strongly recommend the full paper, a "terrific read," as conference discussant Murray Edelman noted after Heerwig and McCabe presented their results on Saturday. The Heerwig-McCabe paper is easily the most intriguing and currently relevant I saw presented at the AAPOR conference, and worth reading alongside the cover story ("The Big Race: Obama and the psychology of the color barrier") by John Judis in the current issue of The New Republic.
Having said that, I want to pass along two cautions. First, keep in mind that the data were collected in June 2007, when Hillary Clinton, not Barack Obama was the clear front-runner for the Democratic nomination.
Second, consider the caution that Edelman voiced on Saturday (a point that may not be clear until you read the full paper): A "mode effect" may confound the estimates of social desirability bias that Heerwig and McCabe calculate by asking the same question in different ways (explained on pp. 9-12 of their paper and shown in the comparisons of "true" to "overt" support in Table 3). Edelman cited a 2006 article in Public Opinion Quarterly by Smyth, et. al. that found that respondents tend to answer more questions more completely when asked one at a time rather than when presented in a "check all that apply" list format.