Just Read: State of the Science: Implicit Bias Review 2013

This is a pair of reports from the Kirwan Institute for the Study of Race and Ethnicity at The Ohio State University (@KirwanInstitute) that is what it says – up to date reviews on the state of the science around implicit bias. Today, publishing on the 2013 update. There’s already a 2014 update that I’ll write about tomorrow, that’s even more fascinating. There’s a lot going on so I’ll bullet point my main take aways:

  • Helpful definition: The main distinction between implicit and other types of bias centers on level of awareness. Explicit biases “can be consciously detected and reported” and are “endorsed.” Implicit biases are so because they are “attitudes and stereotypes that are not consciously accessible through introspection.”
  • The two are distinct but related – someone who has a high level of explicit bias probably has a high degree of implicit bias as well. There are many studies that show that people who don’t have explicit bias, e.g. they don’t “endorse” biased attitudes, can still have significant implicit bias.
  • “Politically or socially sensitive topics” are typically hard to uncover by asking people about them because of “impression management”
  • (editorial note from me: in many parts of society, bias against people who are transgender is not sensitive, it is freely endorsed, which makes it such a fascinating topic of our time)
  • What contributes to implicit bias: Lack of attention being paid to a task, cognitive load, and ambiguity, such as a strange face vs a familiar face – e.g. a busy, rushed brain doesn’t process and compensate well.
  • Verbal versus non-verbal behavior: these are the hardest control and experience “leakage,” and are usually what people notice.
  • Non-verbal behavior: less bias equals greater speaking time, more smiling, more social comments, fewer speech errors, fewer speech hesitations. More bias correlates with blinking more and less eye contact.
  • Dehumanizing language: These can activate implicit biases, for example in a courtroom when a prosecutor uses animal imagery to refer to defendants. (editorial note from me: not unlike people who refer to human beings as “a transgender” or …much worse)
  • The Implicit Association Test: After a decade of research, we be- lieve that the IAT has demonstrated enough reliability and validity that total denial is implausible” (Kang & Lane, 2010, p. 477).
  • Concept of “stereotype threat” – fear of performing to a stereotype actually reduces performance on tests
  • “Illusion of ojbectivity” – “when people think that they are objective, rational actors, they act on their group-based biases more rather than less” – example that when jurors think a case is “racially charged” they are MORE attentive to their biases and more thoughtful. When they don’t see a case as “racially charged” they are less thoughtful and racial biases affect behavior.
  • Health care: Physicians’ bias tends to align with the communities where they reside, they are not “less biased” and show preference toward Whites over Blacks as a group. African American physicians are different and tend not to display implicit racial preferences for Whites or Blacks. Female doctors tend to hold fewer implicit biases.
  • Health care: “physicians were 23% more verbally dominant and engaged in 33% less patient-cen- tered communication with African American patients than with White patients”
  • Health care: “culturally competent” training must be performed carefully; it may activate stereotypes or over-simplify a culture, which actually activates bias.
  • Debiasing: It can be done – “Studies have shown that “instead of repressing one’s prejudices, if one openly acknowledges one’s biases, and direct- ly challenges or refutes them, one can overcome them”
  • Debiasing: Exposure to counter-stereotypic individuals is weak, eg Martin Luther King Jr., vs. Timothy McVeigh, but effective — and must be connected to relevant categories (editorial note from me: I’ll be doing this during the keynote, we’ll see how it goes)
  • Debiasing: Intergroup contact: This seems to have a large body of research around it. I posted about this previously.
  • Debiasing: Diversity training: “An entire industry around diversity education and trainings has proliferated in re- cent years, offering participants promises of reduced prejudice and greater appreci- ation of various cultures. Studies have examined whether diversity education can counter implicit biases, though the results are mixed.”
  • Debiasing: Taking the perspective of others: “authors suggested that the medical professionals imagine themselves as a minority group patient and write a story about that person’s life”
  • Debiasing: Focusing on common identity. (editorial note from me: just like when I have heard physicians who are allies say, “transgender people go off and get married and have kids just like me.”)

Ok, a lot here, and this is just a taste. A very important conclusion:

biases that we do not acknowledge but that persist, unchallenged, in the recesses of our minds, undoubtedly shape our society

Of course I see many parallels to health care. If we recognize that these biases exist, identify them in others, slow down, connect, have contact with people outside of environments where we take away their autonomy and power, we have the ability to behave fairly in the interest of their health.





Ted Eytan, MD