Dissociating Automatic Associations: Comparing Two Implicit Measurements of Race Bias

Area of Science:

• Psychology
• Social Psychology
• Cognitive Science

Background:

• Implicit bias tasks often show weak correspondence, potentially due to unique automatic and controlled processing contributions.
• Understanding these distinct processes is crucial for accurately measuring and interpreting implicit biases.

Purpose of the Study:

• To examine the correspondence between automatic and controlled processing estimates from two sequential priming tasks.
• To differentiate stereotypic (Weapons Identification Task; WIT) and evaluative (Affective Priming Task; APT) associations.

Main Methods:

• Two sequential priming tasks (WIT and APT) with identical structure and timing were employed.
• Multinomial processing tree models were used to estimate automatic and controlled processing components.
• Data were collected across two studies with predominantly White samples.

Main Results:

• Stereotypic bias was stronger for Black targets, while evaluative bias was stronger for White targets.
• Overall response accuracy bias showed modest correlation across tasks.
• Controlled processing estimates corresponded more strongly than automatic processing estimates between the WIT and APT.

Conclusions:

• Findings support models of distinct learning and memory systems for different forms of race bias.
• Different forms of race bias contribute uniquely to estimates of automatic associations.
• Correspondence is stronger for controlled processing than automatic processing in implicit bias assessment.