When examining the effect of implicit bias training, it is critical to ask: Does training reduce implicit preferences and bias, and if so, are the reductions in bias tenable? While prominent researchers continue to examine these questions, current evidence shows that programs seeking to reduce implicit bias have limited effects.

In 2012, the National Center for State Courts reviewed strategies to reduce the influence of implicit bias and convened discussions with judges and judicial educators. The resulting report identified risk factors that might increase bias (e.g., heightened emotional states) and strategies to reduce the effects of implicit bias (e.g., raising awareness, routine monitoring, and identifying triggers to remove or reduce implicit biases). The National Center for State Courts is currently updating this work.

In one particularly informative study, Dr. Calvin Lai and colleagues investigated seventeen different interventions that sought to reduce implicit racial preferences. The researchers used pre- and post-test measures of implicit and explicit bias to examine potential changes. They found that eight of the seventeen interventions generally reduced implicit preferences (for Whites, as compared to Blacks). However, the interventions only slightly reduced implicit preferences regarding Asian and Latinx individuals, two groups underrepresented in the bias literature.

The most effective intervention, Vivid Counterstereotypic Scenario, used personal reflection and a story that countered racial stereotypes. For example, participants read a story of a White man assaulting them and a Black man coming to their rescue. Across four separate studies, the researchers found that the counterstereotypic story reduced implicit preferences. The effect increased when more intense stories and photos were added to the experiment. Despite this result, none of the seventeen examined interventions consistently reduced explicit biases. In addition, whether this approach leads to long-lasting change, or how it might apply in other situations, warrants further research.

A 2012 study by Dr. Patricia Devine and colleagues examined if implicit bias is similar to a bad habit that can be broken through a combination of awareness, concern about the bias, and the utilization of strategies to take action and reduce the bias. In the study, all participants completed explicit bias survey measures (e.g., The Implicit Association Test (IAT)) across a twelve-week period. About half of the participants also watched a forty-five minute interactive training while the other half did not receive any training. This allowed researchers to examine if the training reduced measured bias. The training included discussion on the idea of prejudice as a habit, the development and activation of implicit biases, the IAT, and how implicit bias can result in discrimination.

The training also offered five strategies for participants to use to reduce bias in their own lives. The five strategies involved

1. replacing stereotypical responses with non-stereotypical responses and reflecting on why the stereotypical response occurred
2. imagining counterstereotypic individuals
3. obtaining specific individual information about members of groups to prevent stereotypic inferences
4. taking the perspective of members of stereotyped groups
5. seeking out opportunities to interact with members of stereotyped groups

Participants were encouraged to utilize the strategy/strategies that fit best into their daily lives.

As compared to those without the training, participants who received the training showed reductions in implicit preferences, in four and eight weeks after the training. In addition, these participants showed an increased awareness of real-world bias and were more concerned with discrimination and prejudice. These findings offered initial evidence that bias-reduction trainings could lead to change at least two months later. However, a 2017 replication with a larger number of participants found that the effects of the training generally declined after two weeks (rather than lasting for two months). In addition, a currently unpublished meta-analysis of implicit bias studies by a group of prominent researchers found that “changes in implicit measures are possible, but those changes do not necessarily translate into changes in explicit measures or behavior.”

While the science in this area is advancing, research currently suggests that the best bias-reduction programs will educate participants on bias, encourage counterstereotypic thinking, and provide strategies they can apply in their daily lives. Ultimately, implicit preferences are just one component within a larger social or behavioral environment that could be addressed by federal and state courts. A multi-faceted approach should consider explicit bias, implicit bias, and decision making, while also acknowledging that multiple forms of bias interact in the real world and can affect individuals of disadvantaged groups, defined broadly, differently.

For more information on bias, see

• What Is Bias and How Is It Measured
• Reducing Bias References

To share relevant research findings, contact Jason A. Cantone at fedstate@fjc.gov.