Elsevier

Consciousness and Cognition

Volume 36, November 2015, Pages 206-218
Consciousness and Cognition

Thinking in Black and White: Conscious thought increases racially biased judgments through biased face memory

https://doi.org/10.1016/j.concog.2015.07.001Get rights and content

Highlights

  • Conscious thought about White and Black people increased biased face memory.

  • Conscious thinkers mixed up faces sharing the same racial features.

  • Conscious thinkers remembered a Black face as darker than it really was.

  • Biased face memory mediated racially biased evaluative judgments.

Abstract

It is a common research finding that conscious thought helps people to avoid racial discrimination. These three experiments, however, illustrate that conscious thought may increase biased face memory, which leads to increased judgment bias (i.e., preferring White to Black individuals). In Experiments 1 and 2, university students formed impressions of Black and White housemate candidates. They judged the candidates either immediately (immediate decision condition), thought about their judgments for a few minutes (conscious thought condition), or performed an unrelated task for a few minutes (unconscious thought condition). Conscious thinkers and immediate decision-makers showed a stronger face memory bias than unconscious thinkers, and this mediated increased judgment bias, although not all results were significant. Experiment 3 used a new, different paradigm and showed that a Black male was remembered as darker after a period of conscious thought than after a period of unconscious thought. Implications for racial prejudice are discussed.

Introduction

Despite equal rights policies, people from racial minorities still face many disadvantages. For example, studies have shown that Black employees receive lower wages than their White counterparts (Johnston & Lordan, 2014), that students of racial minorities are more likely than White students to be suspended from schools (Rich, 2014), and that Black NBA players are more often penalized for personal fouls than White players when the referee is White (Price & Wolfers, 2010). A recent field experiment on the U.S. job market found that résumés with White-sounding names such as Emily Walsh received 50% more call-backs than those with Black-sounding names such as Lakisha Washington (Bertrand & Mullainathan, 2004).

Racial discrimination is often unintended, and partially stems from automatic psychological processes. Many studies have looked at strategies to overcome or mitigate racial discrimination. A common finding in these studies is that people can successfully control discrimination when they exert conscious effort to limit the effects of biases on their judgments (Devine, 1989, Payne, 2005). For example, instructions to engage in mental perspective taking (i.e., to imagine oneself in the other person’s shoes) help judges and jurors to diminish stereotyping and in-group favoritism (Galinsky & Moskowitz, 2000). Adopting implementation intentions to control for potential bias in specific contexts (e.g., making if–then plans such as “if I encounter a Black person, then I think counter-stereotypic thoughts”) has also been found effective (Mendoza et al., 2010, Stewart and Payne, 2008). However, conscious strategies to counteract bias can also, in fact, increase race bias. For instance, instructions to suppress stereotypes (e.g., “try to be color blind”) have been shown to produce a rebound effect that increases race bias (Apfelbaum et al., 2008, Macrae et al., 1994).

In the present research we highlight another potential problem of conscious thought as a bias-reducing strategy. We propose that conscious thought increases the likelihood of biased memory representations of Black and White people, increasing, in turn, the likelihood of biased evaluations. This idea is based on the notion that conscious thought strengthens the activation of automatically activated cognitive schemata (Gawronski and Bodenhausen, 2006, Strack and Deutsch, 2004, see also Bos and Dijksterhuis, 2011, Dijksterhuis and Nordgren, 2006, Reinhard et al., 2013). Several lines of research indicate that conscious thought often supports or even intensifies the impact of automatically activated schemata on judgments, especially when information is ambiguous or complex. For example, research on the Heuristic Systematic Model showed that the valence of effortful systematic processing can be biased by heuristic cues present in ambiguous persuasive messages (Chaiken & Maheswaran, 1994; see also Petty, Wegener, & Fabrigar, 1997). Pelham and Neter (1995) found that a higher level of motivation – a condition under which people generally apply more conscious thought (e.g., Baumeister, 1984) – increased the use of erroneous heuristic cues in problem solving. Based on these findings, we hypothesize that conscious thought sustains the activation of automatically activated race categories, thereby bolstering the structuring of information in memory around race categories.

Research illustrates that the salience of race categories, in turn, increases the likelihood of stereotyping and racially biased judgments. For example, Mitchell, Nosek, and Banaji (2003) demonstrated that negative automatic evaluations toward Black people are stronger when participants focus their attention on race categories than when they pay attention to another social category (see also Macrae, Bodenhausen, Milne, Thorn, & Castelli, 1997). Payne, Lambert, and Jacoby (2002) showed that drawing attention to race categories increases stereotypical errors in a speeded weapon identification task compared to a control condition where race was not emphasized. The mechanism underlying these effects is that drawing participants’ attention to category membership activates automatic evaluations associated with these categories, which in turn influences evaluative judgments and decisions. Given that automatic evaluations of Black people tend to be more negative than those of White people, this results in judgments favoring Whites over Blacks.

There is preliminary evidence that conscious thought increases stereotyping and judgment bias. Several studies on Unconscious Thought Theory (UTT; Dijksterhuis & Nordgren, 2006, for a review see Strick et al., 2011) showed that participants stereotyped more after a period of conscious thought than after a period of distraction (or “unconscious thought”). Bos and Dijksterhuis (2011) presented participants with a person of an ethnic minority group and found that a period of conscious thought, compared to a period of unconscious thought, led to lower accessibility and poorer memory for stereotype-incongruent information, indicating more stereotyping. Messner, Wänke, and Weibel (2011) showed that a period of conscious thought, compared to a period of unconscious thought, led to objectively poorer decisions in a personnel selection context (i.e., they selected less qualified candidates in terms of the number of fulfilled job requirements) when two cues that often bias personnel selection (gender and attractiveness differences) were present. Furthermore, conscious thinkers had a stronger tendency to prefer males to females (i.e., display gender bias) than unconscious thinkers.

The present studies extend previous research in several ways. Whereas Bos and Dijksterhuis (2011) investigated how conscious thought affects memory representations, the present research went on to investigate how conscious thought affects racial prejudice, that is, evaluative judgments about people from a different race. Furthermore, in contrast to previous studies (i.e., Bos and Dijksterhuis, 2011, Messner et al., 2011), we included a control condition in which participants were given no time to think at all (either consciously or unconsciously). This way we can test whether unconscious thought has benefits above and beyond not just conscious thought, but also simply “not thinking at all”. Furthermore, we introduce two novel methods to measure racially biased memory representations.

The task used in Experiments 1 and 2 was the “Who is Who?” task. It was inspired by the “Who said what?” paradigm (Taylor, Fiske, Etcoff, & Ruderman, 1978), a task that uses memory confusion to discover the extent to which participants categorize target persons into social categories. In our version of the task, participants form impressions of four candidate housemates, two White and two Black. Each housemate is presented with a name, a face, and a number of descriptive verbal attributes. After the manipulation of immediate decision making or decision-making after conscious or unconscious thought (see below), participants are again presented with the names of the candidates and they are asked to attribute each name to the correct face, each time choosing among three White and three Black faces (the correct face and five incorrect faces). Participants’ mistakes reveal to what extent they had categorized the candidates in terms of race. Participants are more likely to misattribute a candidate’s name to another same-race face if they had remembered him as a member of that race group in the first place (Fiske, Haslam, & Fiske, 1991). Hence, biased memory is indicated by a higher proportion of “same-race errors” compared to “different-race errors”.

We used another new paradigm in Experiment 3. In this paradigm, participants saw the face of a Black male, after which their likelihood to remember him as darker (more information follows) was assessed. Thus, instead of investigating the tendency to remember a person as belonging to the White or Black race category (as in Experiments 1 and 2), Experiment 3 investigated the tendency to remember a person as more or less typical of the Black race category. This tendency is important to study in its own right because a broad literature indicates that when individuals are perceived as more typical of the Black race they are more likely to be judged according to the negative stereotypes and evaluations associated with the Black race category, an effect sometimes dubbed skin tone bias (Maddox & Gray, 2002). A research review indicated that individuals whose features are typical of the Black race generally suffer from lower social status and health compared to their less-prototypical counterparts (Maddox, 2004). As far as we know, this is the first research that tests the hypothesis that conscious thought increases the tendency to remember Black faces as more typical of their category.

The procedure of the experiments was typical for studies on UTT. In the first two experiments, participants were presented with four candidate housemates (two White, two Black), and were asked to form impressions of them. The candidates had an individual name, were described by 12 attributes each (making the decision a complex one), and a picture revealing the candidate’s race. Participants in the conscious thought condition were allowed to think about their evaluative judgments of the candidates for 3 min, while participants in the unconscious thought condition were distracted by an unrelated task for 3 min before making their judgments. Experiment 1 also included an immediate decision condition, in which participants were given no time to think at all. As racial bias is usually stronger under conditions of limited time or cognitive capacity (e.g., Macrae & Bodenhausen, 2001), we expected racial bias to be strong among immediate decision makers. In Experiment 3, one Black person was presented, after which participants thought about him consciously or unconsciously.

In the first two experiments, we expected a stronger tendency to structure information about targets around race categories among conscious thinkers (Experiments 1 and 2) and immediate decision makers (Experiment 1) than among unconscious thinkers (Experiments 1 and 2), and we expected this to mediate biased evaluative judgments about the housemate candidates (Experiments 1 and 2). In Experiment 3 we expected the tendency to remember a Black face as darker to be stronger among conscious thinkers than among unconscious thinkers.

Section snippets

Participants and design

Based on similar previous studies (Bos and Dijksterhuis, 2011, Messner et al., 2011), our aim was to recruit at least 30 participants for each thought condition. One-hundred-and-ten students from the Radboud University Nijmegen participated in return for money or course credits (28 men, 82 women, mean age = 21.63 years). Participants were randomly assigned to the thought (immediate, conscious, or unconscious) conditions. The experiment also included two priming conditions (see below) which were

Participants and design

Based on previous studies and the findings of Experiment 1, our aim was to recruit at least 30 participants for each thought condition. Seventy-two students from Utrecht University participated in return for money or course credits. The data of five participants were removed due to technical problems with the experimental software, leaving 67 participants for the analyses (36 men, 31 women, mean age = 22.22 years). Participants were randomly assigned to the conscious and unconscious thought

Participants and design

Based on the study of Ben-Zeev et al. (2014), our aim was to recruit at least 30 participants for each condition. One-hundred-and-thirty-two students from Utrecht University participated in return for money or course credits. The study was conducted in two consecutive weeks. Based on an a priori criterion, the data of eight participants (6%) were excluded from the analyses because they reported that they had seen the target face during participation in a previous study. This left 124

Face recognition

The original Black male target was correctly identified 66.1% of the time, and this percentage did not significantly differ between the thought or attribute conditions (Fs < 1.31, ps > .25). The likelihood of false alarms (i.e., Yes responses) to the lures were analyzed using a 2(thought: conscious vs. unconscious) × 2(attributes: Black vs. White) × 2(lure: lighter vs. darker) × 3(gradation: 25% vs. 37% vs. 50%) ANOVA with repeated measures on the last two factors. There was a significant main effect of

General discussion

These three experiments show that a period of conscious thought – compared to a period of distraction or “unconscious thought” – may increase racial bias in complex decisions. We found evidence that consciously thinking about Black and White individuals increases faces bias (i.e., mixing up candidates of the same race), and in turn increases the likelihood of judgment bias (i.e., preferring White individuals to Black individuals), although the face bias and judgment bias results were

Acknowledgments

The authors would like to thank Samantha Antusch for collecting the data of Experiment 3, and Avi Ben-Zeev, Tara C. Dennehy, Robin I. Goodrich, Branden S. Kolarik, and Mark W. Geisler for providing the stimulus materials for Experiment 3.

References (43)

  • C.N. Macrae et al.

    On the activation of social stereotypes: The moderating role of processing objectives

    Journal of Experimental Social Psychology

    (1997)
  • G. Moskowitz et al.

    Egalitarian goals trigger stereotype inhibition: A proactive form of stereotype control

    Journal of Experimental Social Psychology

    (2011)
  • B.K. Payne et al.

    Best laid plans: Effects of goals on accessibility bias and cognitive control in race-based misperceptions of weapons

    Journal of Experimental Social Psychology

    (2002)
  • E. Apfelbaum et al.

    Seeing race and seeming racist? Evaluating strategic colorblindness in social interaction

    Journal of Personality and Social Psychology

    (2008)
  • R.F. Baumeister

    Choking under pressure: Self-consciousness and paradoxical effects of incentives on skillful performance

    Journal of Personality and Social Psychology

    (1984)
  • A. Ben-Zeev et al.

    When an “Educated” Black man becomes lighter in the mind’s eye: Evidence for a skin tone memory bias

    SAGE Open

    (2014)
  • M. Bertrand et al.

    Are Emily and Greg more employable than Lakisha and Jamal? A Field experiment on labor market discrimination

    The American Economic Review

    (2004)
  • I.V. Blair et al.

    The automaticity of race and Afrocentric facial features in social judgments

    Journal of Personality and Social Psychology

    (2004)
  • M.W. Bos et al.

    Unconscious thought works bottom-up and conscious thought works top-down when forming an impression

    Social Cognition

    (2011)
  • J.A. Cameron et al.

    Stereotype-biased search and processing of information about group members

    Social Cognition

    (2004)
  • S. Chaiken et al.

    Heuristic processing can bias systematic processing: Effects of source credibility, argument ambiguity, and task importance on attitude judgment

    Journal of Personality and Social Psychology

    (1994)
  • T.L. Chartrand et al.

    Automatic activation of impression formation and memorization goals: Nonconscious goal priming reproduces effects of explicit task instructions

    Journal of Personality and Social Psychology

    (1996)
  • J. Correll et al.

    The police officer’s dilemma: Using ethnicity to disambiguate potentially threatening individuals

    Journal of Personality and Social Psychology

    (2002)
  • P.G. Devine

    Prejudice and stereotypes: Their automatic and controlled components

    Journal of Personality and Social Psychology

    (1989)
  • A. Dijksterhuis et al.

    A theory of unconscious thought

    Perspectives on Psychological Science

    (2006)
  • A.P. Fiske et al.

    Confusing one person with another: What errors reveal about the elementary forms of social relations

    Journal of Personality and Social Psychology

    (1991)
  • A. Galinsky et al.

    Perspective-taking: Decreasing stereotype expression, stereotype accessibility, and in-group favoritism

    Journal of Personality and Social Psychology

    (2000)
  • B. Gawronski et al.

    Associative and propositional processes in evaluation: An integrative review of implicit and explicit attitude change

    Psychological Bulletin

    (2006)
  • Johnston, D. W., & Lordan, G. (2014). When work disappears: Racial prejudice and recession labour market penalties. CEP...
  • J. Jonides et al.

    Verbal working memory load affects regional brain activation as measured by PET

    Journal of Cognitive Neuroscience

    (1997)
  • C.P. MacKinnon et al.

    Equivalence of the mediation, confounding and suppression effect

    Prevention Science

    (2000)
  • Cited by (0)

    View full text