To Blame? The Effects of Moralized Feedback on Implicit Racial Bias

Robin Scaife,Tom Stafford,Andreas Bunge,Jules Holroyd

doi:10.1525/collabra.251

Abstract

Implicit bias training (IBT) is now frequently provided by employers, in order to raise awareness of the problems related to implicit biases, and of how to safeguard against discrimination that may result. However, as Atewologun et al. (2018) have noted, there is very little systematicity in IBT, and there are many unknowns about what constitutes good IBT. One important issue concerns the tone of information provided regarding implicit bias. This paper engages this question, focusing in particular on the observation that much bias training is delivered in exculpatory tone, emphasising that individuals are not to blame for possessing implicit biases. Normative guidance around IBT exhorts practitioners to adopt this strategy (Moss-Racusin et al. 2014). However, existing evidence about the effects of moralized feedback about implicit bias is equivocal (Legault et al. 2011; Czopp et al. 2006). Through a series of studies, culminating in an experiment with a pre-registered analysis plan, we develop a paradigm for evaluating the impact of moralized feedback on participants’ implicit racial bias scores. We also conducted exploratory analyses of the impact on their moods, and behavioural intentions. Our results indicated that an exculpatory tone, rather than a blaming or neutral tone, did not make participants less resistant to changing their attitudes and behaviours. In fact, participants in the blame condition had significantly stronger explicit intentions to change future behaviour than those in the ‘no feedback’ condition (see experiment 3). These results indicate that considerations of efficacy do not support the need for implicit bias feedback to be exculpatory. We tease out the implications of these findings, and directions for future research.

Highlights

An action or judgement is implicitly biased when it is influenced by automatic mental processes which distort action or judgement
We resist giving any more detailed definition of implicit bias because, as we have argued in detail elsewhere, there are no unproblematic ways of characterising implicit biases and adopting any definition requires committing to prior theoretical assumptions that we need not take on here
The findings suggest blaming does not increase implicit bias, as measured by the Implicit Association Test (IAT), at least by any significant amount

Summary

Introduction

An action or judgement is implicitly biased when it is influenced by automatic mental processes which distort action or judgement (often without this influence being apparent to the individual). Strategies that have, with varying degrees of success, been shown to mitigate the expression of implicit bias include: attention to counter-stereotypical exemplars (Blair, 2002); the use of implementation intentions to alter patterns of response (Webb, Sheeran, & Pepper, 2010); and the inhibition of automatic associations due to negative affect, such as guilt (Amodio, Devine, & Harmon-Jones, 2007). This suggests implicit biases may be influenced

Methods

Results

Conclusion