Download data and study materials from OSF
Mick P. Couper
University of Michigan
University of Illinois at Chicago
Sample size: 1120
Field period: 01/2003
This study, supplemented with funds from the University of Michigan's Survey Research Center, builds on a surprising finding in an earlier laboratory-based project (Krysan and Couper, in press) which explored race of interviewer effects in live as against "virtual" interviews. The latter is a new method for studying race of interviewer effects with a computer-assisted self-interviewing (CASI) system that uses a video of an interviewer reading the survey questions. The motivating hypothesis was that the virtual interviewer would generate social-presence-like interviewer effects, in light of recent research showing that people treat computers more like social actors than inanimate tools (Kiesler et al., 1992) as well as the less-recent evidence that the "mere presence" of a person of another race can shape responses to racial attitude questions (Summers and Hammond, 1966). Unexpectedly, the virtual interviewer condition resulted in whites giving more racially conservative responses to the virtual black interviewer than to the virtual white interviewer. That is, rather than mimicking social presence and therefore reducing negative racial attitudes, the virtual presence of the black interviewer heightened the expression of negative racial attitudes.
This finding is consistent with the psychological concept of "activation." Activation is argued to work as follows: subjects are presented with an image of a black person or some other reference to blacks-subliminal or otherwise-and negative stereotypes and sentiments about blacks are thereby triggered. Once activated, these stereotypes influence subsequent judgments by the individual (Devine 1989; Wittenbrink et al. 1997; Dovidio et al. 1997). Thus, we concluded with a revised hypothesis: when whites were interviewed by a virtual black interviewer, rather than suppressing negative racial prejudice, as social presence would predict, the image of the black interviewer increased expressions of negative racial attitudes because of activation effects.
Building on this new hypothesis, the present study was designed to see if we could clarify the activation as against social presence effects in a Web-based medium. We used a 2x2 between-subjects experimental design in which race of interviewer and social presence versus "activation" were manipulated, as described below.
The survey instrument consisted of a range of racial attitude dimensions, including social distance, racial policies, stereotypes, perceptions of discrimination, race-associated policies, and one non-racial question (asking respondents to identify their favorite actor or entertainer). Based on our earlier study, we hypothesized that questions measuring perceptions of discrimination, stereotypes, and racial policies would be most likely to show the expected effects. Our findings indicate mixed support for the central hypotheses, with the strongest support for questions about the existence of discrimination and support for racial policies. In addition, sub-group analyses suggest that the pattern of effects differs depending on respondent's political ideology.
H1: Whites who complete a web-based survey in which they are presented with a "single white interviewer" will give more racially conservative responses than those who complete a web-based survey in which they are presented with a "single African American interviewer."
H2: Whites who complete a web-based survey in which the first screen of the interview includes a montage of four small photos of African Americans will give more racially conservative responses than those who complete a web-based survey in which the first screen of the interview includes a montage of four small photos of whites.
Respondents were randomly assigned to five different conditions with equal probabilities. This represents a 2x2 between-subjects experimental design, with an additional control condition. The two experimental variables are social presence and priming or stereotype activation. We manipulated social presence by using photographs of a black or white interviewer. This image appeared at several points in the interview, and was accompanied by text intended to enhance the social presence through the use of interviewer feedback and commentary, such as: "Hi, I'm your interviewer," "Thank you for your answer," and "I appreciate your answering these questions for me." The priming or activation condition was manipulated by using a photograph of a group of either black or white persons at the start of the survey. For cost reasons, the sample was restricted to white respondents only.
The survey instrument contained questions covering several racial attitude dimensions, including social distance, racial policies, stereotypes, perceptions of discrimination, race-associated policies, and one non-racial question (asking respondents to identify their favorite actor or entertainer).
Our findings indicate mixed support for the central hypotheses, with the strongest support for questions about the existence of discrimination and support for racial policies. In these instances, we find a statistically significant interaction between the race of the person(s) in the photographs and the social presence versus activation manipulation. White respondents provide more liberal responses to the black interviewer in the single photo conditions (supporting the social presence hypothesis) and more conservative responses when presented with the black group photograph (supporting the stereotype activation hypothesis). However, the effect is modest and by no means consistent across all of the items measured. Ongoing analysis (to be presented at AAPOR in May, 2004) is exploring subgroup differences in the effect of the experimental manipulations.
We find some support for the hypothesis that when presented with a representation of a single interviewer on the Web (the social presence condition), respondents at times react as if they are interacting with a live interviewer, leading to more socially desirable responses. However, when a group photograph is shown, stereotypes tend to be activated, producing more negative responses to the target group. These effects vary across the questions being asked, suggesting that the activation of stereotypes or the engendering of social presence may vary according to the issues being asked about.
Funding for the study came both from TESS and from the University of Michigan's Survey Research Center. For cost reasons, the sample was restricted to white respondents only.
The survey was a web-based survey of persons 18 years or older, using Knowledge Networks' Web-enabled panel. A total of 1,500 panel members were contacted for the survey. Of these, 1,120 completed the survey for a completion rate of 75%. The survey was conducted in January 2003.
Devine, P.G. (1989), "Stereotypes and Prejudice: Their Automatic and Controlled Components." Journal of Personality and Social Psychology, 56: 5-18.
Dovidio, J.F., Kawakami, K., Johnson, C., Johnson, B., and Howard, A. (1997), "On the Nature of Prejudice: Automatic and Controlled Processes." Journal of Experimental Social Psychology, 33: 510-540.
Kiesler, S., Sieff, E., and Geary, C. (1992) "The Illusion of Privacy in Human-Computer Interaction." Carnegie Mellon University: Unpublished Manuscript.
Krysan, M., and Couper, M.P. (in press), "Using Virtual and Live Interviewers to Explore Race of Interviewer Effects and Features of Inter-Racial Attitudes and Interaction." Social Psychology Quarterly.
Krysan, M., and M. P. Couper. 2005. Race-of-Interviewer Effects: What Happens on the Web? International Journal of Internet Science. 1:5-6.
Summers, G.F., and Hammonds, A.D. (1966), "Effect of Racial Characteristics of Investigator on Self-Enumerated Responses to a Negro Prejudice Scale." Social Forces, 44: 515-518.
Wittenbrink, B., Judd, C.M., and Park, B. (1997), "Evidence for Racial Prejudice at the Implicit Level and Its Relationship with Questionnaire Measures." Journal of Personality and Social Psychology, 72(2): 262-274.