Presentation on theme: "Miglena Sternadori, University of South Dakota GOOD NEWS IS NO NEWS? EFFECTS OF POSITIVE STORIES ABOUT AFRICAN AMERICANS ON IMPLICIT BIAS."— Presentation transcript:
Miglena Sternadori, University of South Dakota GOOD NEWS IS NO NEWS? EFFECTS OF POSITIVE STORIES ABOUT AFRICAN AMERICANS ON IMPLICIT BIAS
MAIN QUESTION Do positively valenced news stories affect the malleability of implicit attitudes? Answer sought by using a 2 (Session) x 2 (Trial Type) experimental design 2 levels of session: measurement before and after exposure to news stories 2 levels of trial type: speed of associating Black faces with either positive or negative words
WHAT WE KNOW Members of both majority and minority groups may experience spontaneous and unconscious racist reactions (Banaij & Greenwald, 1994) Humans have evolved a system that sustains stereotype-based patterns of thought (Macrae, Milne, & Bodenhausen, 1994). Implicit bias may be malleable through: Conscious attempts Exposure to positive depictions or personal experience with minority members
OPERATIONALIZATION Implicit Attitudes Test (IAT; Greenwald, McGhee, & Schwartz, 1998), measured the ease of associating Black faces with positive or negative words. Reduction of implicit bias measured by looking at the difference in average response latencies on: Compatible trials (participants respond to White faces/positive words with one key and Black faces/ negative words with another key) Incompatible trials (participants respond to White faces/negative words with one key and Black faces/positive words with another key)
HYPOTHESES: SET I Most people show greater ease of associating White faces with positive words and Black faces with negative words: H1a: The average response latency for compatible trials will be shorter than the average response latency for incompatible trials during the pre-exposure IAT. H1b: The average response latency for compatible trials will be longer or equal to the average response latency for incompatible trials during the post- exposure IAT.
HYPOTHESES: SET II After participants read the positive stories, it should become easier to associate Black faces with positive words and less easy to associate with negative words: H2a: The average response latency on incompatible trials during post-exposure IAT will be shorter than the average response latency on incompatible trials during pre- exposure IAT. H2b: The average response latency on compatible trials will be longer during post- exposure IAT than the average response latency on compatible trials during pre- exposure IAT.
HYPOTHESES: SET III Test-repetition effects expected Bias reduction would be more evident in narrowed gap between latency of associating Blacks with positive words and associating Blacks with negative words: H3: There will be an interaction between session and trial type, such that the subtraction of the average response latency on compatible trails from the average response latency on incompatible trials in pre-exposure IAT will yield a result greater than the subtraction of the average response latency on compatible trials from the average response latency on incompatible trials in post-exposure IAT.
METHOD Design: 2 (Session: pre-exposure, post- exposure) x 2 (Trial Type: compatible, incompatible) 120 students at a Midwestern university DV: average response latency in associating Black faces with either positive or negative words Procedure: Participants took the IAT, read four stories about successful African Americans, and took the IAT a second time
RESULTS Main effect for trial type, F(1,115) = 99.2, p <.0001, η 2 p =.46 H1a was supported, but H1b was not Main effect for session, F(1,115) = 50.39, p <. 0001, η 2 p =.31 H2a was supported, but H2b was not Session x Trial type interaction was significant, F(1, 115) = 6.33, p =.01, η 2 p =.05. Gap between the average response latencies in associating Black faces with positive words and the average response latencies in associating Black faces with negative words had narrowed after exposure to the positive news stories (Figure 1)
DISCUSSION Results suggest that written news have the potential to decrease implicit bias Malleability is reliable but weak Lack of support for H1b shows that exposure to stories about successful African Americans was not sufficient to eliminate bias Limitations: Within-subject design with no control group, testing effect Future research: Post-hoc tests, take into account errors in addition to response latencies; replace the IAT with the affect misattribution procedure (AMP)