"I'm not a racist, but...", or why automatic stereotyping happens: Part 2

Keith Payne's work on racial stereotyping brings up an intriguing possibility. During the weapons identification task, viewers are more likely to erroneously identify a harmless object as a weapon if it was preceded by a black face compared to a white face. They are also more accurate identifying weapons after seeing black faces compared with white faces.

It's possible that both of these results are due to the same underlying mental process, but Payne's research also invites another possible assessment: that separate processes are responsible for the two different behaviors. One behavior: avoiding miscategorizing, requires viewers to control their response to an image. The other behavior: successfully categorizing an image that's consistent with a stereotype, only requires an automatic response. While the results of the experiment we discussed last week are consistent with this model (which Payne calls the "process dissociation procedure" or PDP), they're inconclusive because simpler explanations, such as simple reaction time differences, can also explain his results.

The second experiment in his study attempts to demonstrate that only the PDP can explain Payne's results.

Fifty-six non-black volunteers participated in the study. First they repeated the weapon identification task. Next, they read a short story describing a college student's day. The "student," named John, had allegedly filled out an information sheet describing himself, which was provided along with the story. On the sheet, in addition to age, major, address, and a few other details, the "African American" box was checked in a category marked "racial/ethnic background."

After reading the story, the respondents rated John on a scale of +5 to -5 for several different qualities, including likable, intelligent, self-assured, lazy, and ambitious. (In a previous study, Payne had found that in a similar task, respondents rated white people more favorably than black people after reading the same story. In this study, therefore, there was no need to include a white control story.)

As you have probably guessed, respondents who performed with a more racially stereotyped pattern on the weapons identification task also rated John more negatively (with a correlation of r=.40).

On its own, this data doesn't support the PDP model. But Payne conducted a further analysis:

This chart shows how automatic bias -- the difference in accuracy identifying weapons after seeing black versus white faces -- compares to negative ratings of the black character in the story, for two different groups. The "high control" group was less likely to misidentify harmless items as weapons after seeing black faces compared to white faces. The "low control" group was worse at this task. In both cases, the more automatic bias an individual showed, the more negative his or her ratings of the black man were. But the "low control" group's automatic bias was correlated to negative ratings much more strongly.

Payne argues these results demonstrate that executive control -- the ability to control our impulses, both in terms of stereotyping and in other realms -- is a separate process from automatic bias. Automatic biases, Payne argues, are task-specific. So stereotyping blacks as more likely to be carrying a weapon, for example, might be separate from other stereotypes. Controlling whether to act based on those stereotypes, however, is a general ability.

Payne, B.K. (2005). Conceptualizing control in social cognition: How executive functioning modulates the functioning of automatic stereotyping. Personality and Social Psychology, 89(4), 488-503.