The Measurement of Racial Attitudes By David C. Wilson page2

For this analysis, I use results from a 2005 Gallup Social Attitudes study, which contained a number of split-sample experiments. In one experiment, respondents were randomly assigned one of two wordings for a question related to optimism or pessimism toward the future of race relations. The primary difference between the two questions was the tone of the wordings. I classify the two questions very broadly as “positive” and “negative” for simplification purposes; however, an additional feature of the wordings is the finality (conclusive versus gradual change) of the tone. The exact wording of the items is as follows:

The sample sizes for the items were considerably different—“Negative Tone” was asked of 538 respondents, “Positive Tone” of 482—however, the respondents in each experimental sample were no different across race, sex, age, or education. The “Negative Tone” question used the traditional wording (“always” [be a problem]), while “Positive Tone” used a new wording (“as good” [as they are going to get]).

In both forms, the questions contained what are considered baseline or status quo responses, and responses that would indicate a positive change—that is, the questions didn’t provide a response asking if things would get worse. Assuming that the existence of racial problems is a fact, the baseline position would be pessimistic, and a response for change would be optimistic. Thus, an optimistic response would be either “a solution will eventually be worked out” (Negative Tone) or “will eventually get better” (Positive Tone), and a pessimistic response would be “will always be a problem” (Negative Tone) or “as good as they are going to get” (Positive Tone).

Given the nature of race relations and the firm grounding they have in the United States, the more negative tone should elicit a lower likelihood of optimism. Additionally, given the progress that has occurred in the United States relative to the past, while things are not resolved in one way or the other, they have objectively improved. Thus, the more positive tone should elicit a greater likelihood of optimism.

A comparison across the two items reveals that optimism was higher (pessimism was lower) in response to Positive Tone, and pessimism was higher (optimism was lower) with Negative Tone. In Negative Tone 45 percent of respondents were optimistic, and in Positive Tone 73 percent were optimistic. The change of twenty-eight percentage points was statistically significant at the 99 percent confidence level. The percentage difference in pessimism across the two forms was also twenty-eight percentage points. Thus, respondents were more likely to believe that race relations would “eventually get better” than that they would “eventually be worked out.” These results are shown in Figure 2.

Controlling for race, the question wording effect was the same for both blacks and whites. As shown in Figure 3, similar to the overall effect, both groups were more optimistic in response to Positive Tone than to Negative Tone. When asked about race relations in Negative Tone, 36 percent of blacks and 52 percent of whites were optimistic; however, in Positive Tone, optimism increased by twenty-nine and twenty-seven percentage points for blacks and whites, respectively. Both increases were statistically significant. Moreover, regardless of wording, blacks were less optimistic and more pessimistic than whites. This is not surprising, given the different experiences and perceptions regarding race across the two groups.