Overcoming biases

So, given the results of IATs, what aspects of the human cognitive system need to be addressed? Devine (1989) has referred to stereotypes as a form of habit derived from well-learned associations that are, in turn, caused by ‘repeated activation in memory'. This will vary from person to person resulting in low to high levels of prejudice. Stereotyping, and hence implicit bias, is, therefore, a form of automatic and largely unconscious (heuristic) processing. According to her, when one's personal beliefs overlap with a negative stereotype, then there is the potential for explicit prejudice. Where one's personal beliefs do not overlap with the stereotype, then there is a degree of conflict since cultural stereotypes are automatically triggered for both low and high prejudiced individuals. Overcoming this habit requires conscious controlled processing. A prejudice triggered by the activation of a stereotype is the result of Type 1 thinking and is enduring and difficult to counter. On the other hand, controlled processes can cause the explicit system (Type 2 thinking) to change relatively quickly - it is ‘malleable’. But, despite short-term apparent changes to expressions of explicit bias, they are likely to be short-lived.

Are there interventions that work to produce long-term changes to implicit bias; or at the very least that can mitigate the expression of bias? One way is to compare a variety of interventions to find out which ones work. Another is a scattergun approach where a range of strategies is used in a single intervention. As an example of the former strategy, Lai et al. (2014) describe an unusual research contest where a number of teams were asked to come up with various approaches that could be tested against others to see which ones, if any, reduced implicit racial preferences. The incentive for the teams of researchers who took part was to win the contest. A total of 18 interventions plus a control were used coming under the overall headings of: ‘Exposure to counter-stereotypical exemplars’, 'Intentional strategies to overcome biases', ‘Evaluative conditioning" (in this case, pairing words: Black with positive attributes and Whites with negative attributes), ‘Appeals to egalitarian values', ‘Engaging with others' perspectives’, and 'Inducing emotion'. Of these, the first three were effective at reducing implicit preferences and the last three ineffective. While the contest did not examine how long lasting the strategies were likely to be, the conclusion was that interventions that use multiple strategies are likely to be effective. In a follow up study, Lai et al. (2016) looked at the degree to which implicit racial preferences were reduced over time following the interventions that had shown promise. The time period over which they were tested was several hours to several days. Although the interventions showed that implicit preferences were malleable in the short term, none led to long-term change.

In contrast, an earlier study by Devine et al. (2012) claimed to produce a sustained reduction in bias over a three-month period. Their method was not to compare which method was most effective but rather to provide a variety of bias reduction strategies (five in all) in the hope that altogether they would induce a degree of self-regulation on the part of the participants. Their aim was to translate their ‘situational awareness into chronic awareness of biases in themselves and in society, thereby flipping the self-regulatory switch that motivates strategy use and reduces implicit bias’ (p. 1268). Based on their results, they make the case that offering a variety of strategies allows for individual differences in what is likely to motivate an individual to reduce expressions of bias by stimulating Type 2

thinking in situations that are likely to activate the (Type 1) ‘habit" of bias. "Our data provide evidence demonstrating the power of the conscious mind to intentionally deploy strategies to overcome implicit bias’ (Devine et al. 2012, p. 1277).

Studies by Bruneau, Kteily, & Urbiola (2019) provide further examples of combating racial or religious bias by tapping into collective blame hypocrisy. This occurs when people blame an out-group (Muslims in this case) in its entirety for the bad actions of individual group members, but do not blame their own group in its entirety for bad actions by in-group members. Their method taps into our desire for consistency in our thoughts and actions. Inconsistency leads to cognitive dissonance with the subsequent need to remove it. An example of their studies involved US and Spanish non-Muslim students. They were given accounts of mass violence committed by white Europeans and asked, ‘How responsible do you think [white Europeans/you] are for the actions of [Darren Osborne/Anders Breivik/Istvan Csontos]'. After that they were asked: Tn general, how responsible do you think white Europeans are for the attacks of white supremacists?’ In the second part, the procedure was repeated but with accounts of Muslim extremists and asked: ‘Fatima Wahid is a Muslim woman who owns a bakery in southern France. How responsible do you think Fatima Wahid is for the Paris attacks of 2015?" Their results showed a reduction in collective blame of Muslims both one month and one year later, particularly among those with a strong preference for consistency in their thinking.