Parents’ Role in Addressing Children’s Racial Bias: The Case of Speculation Without Evidence

Popular Suggestions for Parents

To determine the types of suggestions that are frequently offered to parents for addressing race with their children, we conducted a Google search. For the search (done on May 9, 2019), we specified that articles (a) must contain the words “race” and “children” somewhere in the title or article; and (b) could contain any of the following words and phrases in the title or article: “racism,” “child,” “parent,” “racist,” “bias,” “prejudice,” “prevent,” “eliminate,” “strategy,” “what to do,” or “reduce.” We chose search terms we thought parents might use when thinking about how to address race with their children as well as terms we thought would locate the types of articles parents would come across when reading parenting blogs or the newspaper. We intentionally omitted words that align with a particular strategy (e.g., “talk”) to avoid biasing the types of articles returned toward a particular strategy.

Our search yielded hundreds of results, and we selected and coded the top 50 popular-press articles (i.e., not peer-reviewed journal articles) that were relevant to the topic of addressing children’s racial biases. We limited our review to 50 articles for substantive and practical reasons. First, by the time we reached 50 articles, the articles started to become less specifically targeted toward parents and less relevant to how parents should address race with their children. Second, we believed parents themselves would be unlikely to engage with more than the top 50 articles. The articles were coded independently by K. E. Scott and a trained research assistant. Reliability between the coders was high (interrater agreement = 86.63%, κ = .73). K. E. Scott examined all coding discrepancies and made the final coding determination. The coding of each article and links to the articles can be found on OSF.

Before describing the advice offered to parents, we note that 66% of the articles either explicitly stated that they were for White parents or suggested that they were about addressing White children’s biases (see Table 1). None of the articles returned in the search focused solely on parents and children of color. ¹ The focus on White parents in these articles likely stems from research showing that White children display more substantial racial biases than children of color early in development (Dunham, Baron, & Banaji, 2007; Dunham, Chen, & Banaji, 2013). Furthermore, whereas parents of color engage in frequent conversations with their children about race (for a review, see D. Hughes et al., 2006), White parents are generally reluctant to talk about race with their children, which may suggest that White parents would especially need guidance about how to address issues of race and racial biases. Given the focus on White parents and children in the popular-press articles, in the present article, we center primarily on suggestions given to and research on White parents and children. From this point forward, when we discuss parents, we will be referring to White parents unless otherwise noted.

OPEN IN VIEWER

Table 1.

Suggestions for Parents to Address Children’s Racial Bias From the Popular Press

Suggestion	All articles (N = 50)	Articles that are likely or definitely for White parents (n = 33)	Articles that are definitely for White parents (n = 19)
Have conversations about race	98%	97%	95%
Read children’s books about race or featuring diverse characters	70%	61%	47%
Encourage children to embrace diversity	42%	21%	26%
Become aware of parents’ personal biases	38%	42%	32%
Model nonbiased and inclusive behavior	34%	27%	32%
Facilitate intergroup contact for children	32%	36%	37%

Note: Some articles contained more than one suggestion, so the columns sum to more than 100%.

Across the popular-press articles, one suggestion appeared in nearly every article: Parents should have conversations about race with their children to address their children’s racial biases (see Table 1). For the remainder of this article, we focus on the suggestion to discuss race with children, given its high prevalence across articles. This suggestion likely stems from the logic that if parents do not talk about race and their children are biased, then talking with children about race may be necessary to prevent or reduce biases. Another potential motivation for this suggestion derives from the literature on color blindness (i.e., the tendency to ignore or downplay the relevance of race) with adults: Many studies have established that White adults’ racial color blindness is detrimental to intergroup relations (for a review, see Plaut, Thomas, Hurd, & Romano, 2018). For example, White people’s color blindness leads White people to be seen as more biased by people of color (Apfelbaum, Sommers, & Norton, 2008), increases social distance between people of different races (Madera & Hebl, 2013), and decreases minority-group members’ psychological engagement with work (Plaut, Thomas, & Goren, 2009).

Although it would be desirable to prevent or reduce biases in children before the biases become deeply entrenched, it is not clear that parent-led conversations in this domain will lead to beneficial outcomes. Indeed, reviewing the popular-press articles suggesting that parents should talk about race brought a major concern into focus. In light of the frequency of the recommendation for parents to talk with their children about race, we are struck by the paucity of systematic research evaluating the actual impact of such conversations.

Lack of Evidentiary Base in Popular Advice

To our knowledge, there is no empirical evidence supporting the efficacy of parent–child conversations for reducing children’s racial biases. There is research showing that White parents generally adopt a color-blind approach in conversations about race with their children (e.g., Pahlke, Bigler, & Suizzo, 2012). However, reviewing Google Scholar and PsycInfo returned no empirical articles evaluating the actual impact of parent–child conversations about race when they do occur. ² Furthermore, in a recent review on the development of intergroup biases, Skinner and Meltzoff (2019) concluded that the articles they reviewed on parental socialization used exclusively correlational approaches to evaluate relations between parents’ and children’s racial biases; no articles evaluated causal influences of parental-racial socialization. In another review focused on intervention research, Aboud and colleagues (2012) found no intervention research evaluating parental engagement with their children on issues surrounding race. In sum, current searches and previous reviews highlight the complete absence of evidence and illuminate a lacuna in the literature.

Some of the conversation topics and approaches recommended in popular-press articles (see Table 2) certainly make contact with existing empirical research and therefore read as potentially advisable. The recommendation to use “fair” and “unfair” language is linked to research showing children are averse to inequality (Shaw & Olson, 2012), the recommendation to teach about the history of race relations is linked to research on effective school curricula (J. M. Hughes, Bigler, & Levy, 2007), and the recommendation to tackle race head-on (e.g., call out racism in the environment, name race) is probably linked to research on color blindness reviewed in the previous section. Critically, however, none of these recommendations are based on empirical research in which parents actually talked to their children about race.

OPEN IN VIEWER

Table 2.

Recommended Content for Conversations About Race

Recommendation	Articles making recommendation
Use “fair” and “unfair” language to discuss racial biases	43%
Call out racism in the environment (e.g., on television)	41%
Do not shush or shut down children’s questions about race	39%
Name race (e.g., label skin color when reading books)	31%
Teach about the history of race relations	27%
Ask children questions about what they think about race	22%
Tell children that everyone is equal	12%

Consider, for example, the recommendation that parents should address the history of race relations when they have conversations with their children about race. It is indeed true that one study (J. M. Hughes et al., 2007) showed that White children of elementary school age who experienced lessons presented by a trained teacher addressing the history of race relations in the United States had more positive views of African American individuals compared with those in a control group. However, it is not at all clear whether children would be similarly affected by a history of race relations implemented at home by a typical parent, who is likely to lack the type of training and expertise on the topic possessed by a teacher. We suspect that many parents would struggle to articulate the history of race relations in this country. Even parents who are knowledgeable about the history of race relations would be likely to need some guidance on which issues are developmentally appropriate, when to address current events, and how to tailor conversations for a child. More broadly, research carried out by experimenters in controlled laboratories and lessons implemented by skilled professionals may or may not be relevant to the question of whether and how parents should talk to their children about race. And this research does not clearly identify the tools parents would need to implement the recommendations in their homes.

A number of scholars who are focused on intergroup biases have discouraged making suggestions for interventions that have not been rigorously evaluated (Cox & Devine, 2019; Legault, Gutsell, & Inzlicht, 2011; Moss-Racusin et al., 2014; Paluck, 2012). Though recommendations could yield beneficial effects, it is also possible that they do nothing at all. Recommendations could also backfire and inadvertently lead to greater rather than less intergroup bias in children. So serious are these concerns that in a recent review of diversity trainings, Levy Paluck (2012) went so far as to write that “field experiments should be considered an ethical imperative, on the level of rigorous testing of medical interventions, to ensure that interventions do not create a backlash or otherwise harmful effects” (p. 185). We concur with this assessment.

Indeed, the adult prejudice-reduction literature provides a clear cautionary tale against promoting the implementation of bias-reduction or diversity-training strategies that have not been carefully evaluated. Given concerns about the problematic nature of racial biases, in recent years, businesses introduced diversity trainings across their organizations. Dobbin and Kalev (2015) evaluated the impact of diversity training in a variety of business settings and found that organizations that had such trainings were not more diverse overall than companies that did not have diversity trainings and also that minority employees left the organizations that had diversity trainings at higher rates than their majority counterparts (for additional reviews, see Bezrukova, Spell, Perry, & Jehn, 2016; Kalev, Dobbin, & Kelly, 2006; Pendry, Driscoll, & Field, 2007). Furthermore, Pendry and colleagues (2007) noted that people often respond to diversity trainings with feelings of threat, anger, and defensiveness.

In sum, to date, the primary suggestion provided to White parents for addressing race and racial biases with their children lacks the empirical backing to be broadly implemented. ³ As illustrated in the examples reviewed above, recommending an untested suggestion is ill-advised at best and irresponsible at worst. Telling parents to have conversations with their children about race to reduce bias in the absence of evidence that doing so is effective is akin to providing unevaluated diversity trainings in organizations, if not more concerning given that negative effects could be perpetuated throughout children’s life spans.

Future Directions for Scientists

Thus far in this article, we have argued that the advice offered to parents in popular-press articles focused on addressing race with their children lacks a sufficient evidentiary base. In this section, we turn to considering what would be needed to create sufficient evidence from which good recommendations could be drawn—in particular, recommendations that could plausibly lead to beneficial outcomes. Then, we consider how to effectively disseminate research findings and recommendations to parents.

A primary recommendation for creating a useful evidentiary base for parent-led bias interventions is to conduct research that includes parents and their children. The reason for this recommendation is straightforward: Studying parents and their children is the only way to understand what parents can do (e.g., what support they need, how they interpret and implement suggestions) and what effects parents’ actions can have on children. To our knowledge, no existing research has successfully engaged both parents and children in interventions to reduce children’s racial biases.

In addition to actually including parents, consideration of the existing recommendations in the popular press led us to recognize the importance of conducting research that yields specific recommendations. Specific recommendations are critical to ensure effective usage, avoid misinterpretation of recommendations, and promote parents’ self-efficacy in addressing race with their children. Recommendations should clearly detail the people for whom they are effective, the circumstances that promote their success, and the particular content that has beneficial outcomes. For example, which recommendations should be implemented at what ages? What exactly should parents say or do to realize the recommendations? Should parents enact recommendations in particular contexts or under specific circumstances? To evaluate these boundary conditions, research studies must test particular recommendations across ages, racial backgrounds, time, and situations.

Creating recommendations with precision is particularly critical given that race is a topic that fosters considerable anxiety for White adults and often leads to avoidance of issues surrounding race or interracial interactions (for a review, see Shelton, Richeson, & Vorauer, 2006). If parents are anxious about race-related conversations, they are likely to withdraw from or have difficulty with such conversations—especially in the absence of concrete guidance about what to say and do and when to have such conversations. Indeed, the literature supports the notion that even though White parents generally believe it is important to discuss race with their children, they typically avoid such discussions altogether (Pahlke et al., 2012; Vittrup, 2018). Moreover, if parents do engage in race-related conversations but exhibit considerable anxiety, their children may detect their discomfort, potentially leading to more negative outcomes. When White preschool children view interracial interactions in which a White adult is displaying nonverbal discomfort, children’s attitudes toward Black people can become less positive (Castelli, De Dea, & Nesdale, 2008). Adults’ interracial anxiety is reduced when they are given specific, structured recommendations about what to do and say (Avery, Richeson, Hebl, & Ambady, 2009), which suggests the benefit of conducting research that yields concrete, structured recommendations for engaging with children about race.

Beyond including parents in research that evaluates specific recommendations for intervention, how else might the field build a useful evidentiary base? To provide some guidance, we turn to the approach Devine and colleagues followed when building and testing an intervention for adults to reduce their own biases (prejudice-habit-breaking intervention, or PHBI; Devine, Forscher, Austin, & Cox, 2012; for a fuller characterization of the approach, see Cox & Devine, 2019). We note that our aim here is not to be prescriptive or to advocate for a particular theory or articulate a particular research program. Rather, our goal is to encourage research that will build a solid evidence base by providing a general roadmap regarding how we as a field could first think about and then act on the idea of involving parents in bias intervention work with children.

When creating an intervention to address adults’ racial biases, Devine and colleagues (2012) began by evaluating what the research literature suggested about the nature of adults’ prejudice and how prejudice might be attenuated. In doing so, they turned to a well-developed model (Devine, 1989) that conceptualized prejudice as a type of habit. They then considered the necessary components for breaking a habit—including awareness, concern, access to tools to reduce the habit, and sustained motivation. With this analysis in mind, Devine and colleagues designed an intervention that addressed each component in turn. The intervention revealed how people could be unwittingly complicit in the perpetuation of discrimination and educated people about the consequences of bias. Then the intervention taught participants strategies that had previously been shown to be effective in short-term laboratory-based intervention studies and that could be applied to facilitate their long-term goal of reducing biases. Each component of the intervention was crafted with clear intentions, and the tools provided concrete, specific behaviors or activities for adults to enact to overcome their biases. To evaluate the efficacy of their intervention, the researchers implemented multiple randomized controlled studies with different samples (e.g., college students, university professors) in both laboratory and field settings. Furthermore, they assessed a wide range of outcomes (e.g., self-report and behavioral) over time (2 weeks–2 years). The PHBI was the first and remains the only intervention that has been shown experimentally to produce long-term reductions in biases (Carnes et al., 2015; Devine et al., 2012; Devine et al., 2017; Forscher, Mitamura, Dix, Cox, & Devine, 2017).

Our recommendation is not necessarily to apply the PHBI directly to the parent-child context but rather to encourage researchers to follow a similar trajectory in mapping out their research agenda. Drawing lessons from the PHBI, researchers could similarly draw on the extensive body of theoretical and empirical work focused on the development of children’s racial biases. This literature could inform an analysis of how prejudice is developed in childhood as well as how it might be attenuated. Of particular relevance in the parent–child domain are well-articulated theories of how biases are socialized (e.g., Bigler & Liben, 2007; D. Hughes et al., 2006; Perry, Skinner, & Abaied, 2019). Furthermore, there are examples of bias-intervention strategies that have demonstrated efficacy in short-term laboratory studies with children (e.g., studies showing that perspective taking, intergroup contact, and individuation can reduce children’s racial biases; Houlette et al., 2004; Qian et al., 2017; Stathi, Cameron, Hartley, & Bradford, 2014). Overall, the results of such studies likely informed some of the recommendations listed in Table 1 but could and should be integrated into research with parents and their children.

Empirical research testing the effects of different strategies parents might implement in the service of addressing race with their children is long overdue. As in the PHBI, it will be critical that researchers focus on results from randomized controlled studies with longitudinal assessments. Such studies should be conducted in both laboratory and field settings. Laboratory studies will reveal strategies that parents can feasibly implement when they are provided with guidance and structure; field research will be necessary to evaluate whether laboratory-based findings generalize to circumstances in which everyday distractions and pressures are present. In an iterative fashion, should there be obstacles or barriers outside the laboratory, the interventions could be tweaked to enhance their efficacy in everyday settings. Furthermore, careful attention should be paid to the types of outcomes that will provide evidence that particular recommendations are actually effective: Who are the targets for whom recommendations are seeking to improve outcomes (White children? children of color?), what should be measured (social preferences on a friendship choice task in the lab? climate in classrooms?), and how do the measures researchers select operationalize relevant constructs? Creating a nuanced understanding of effective intervention techniques and conducting more inclusive research can build an evidence base from which to draw specific suggestions for parents to address children’s biases.

Dissemination of Research

Although carefully conducting and publishing research on parents and their children with specific practical suggestions is necessary, it is not likely to be sufficient to achieve the intended impact. Most parents do not glean parenting advice from psychology journals. This reality raises the question of whose responsibility it is to bring the work into the public realm and how to disseminate research in a way that accurately reflects the science and is usable by the target consumers (in this case, parents). There are a variety of productive avenues that researchers can pursue to ensure that their work will have the greatest potential for the intended beneficial effects. Here, we outline a few possible approaches that could enhance the successful and accurate dissemination of research.

One suggestion is for researchers to work with journals to provide summaries of research articles for parents that present core findings in lay terms. Such articles should clearly delineate for whom different strategies will be effective, when parents should use various strategies, and provide step-by-step guidance for the implementation of strategies, as informed by rigorous research studies. However, we acknowledge that many scientists do not have interest in writing for popular audiences or do not have skills to share their work in the popular realm. In such cases, researchers could provide assistance to journalists to write accurate and sufficiently nuanced articles. Alternatively, by working with community partners, researchers could train others to serve as the disseminators of science.

Beyond writing for the popular press, scientists in the field of bias reduction could create and moderate websites directed toward the public to review research in accessible yet specific ways. Such websites could provide parents with the background and resources to effectively address issues of race with their children. Although we raise a few potential suggestions for how to effectively disseminate research, we invite scientists to carefully and creatively consider methods for reaching the target populations. Ensuring the successful dissemination of science after the rigorous evaluation of methods will increase the impact beyond the scientific community and help bridge the all-too-frequent divide between research and practice.

Conclusion

Until research is conducted to evaluate how parents should address race with their children, the field (and parents) are in a difficult position. It is ill-advised to provide guidance absent the type of evidence we have outlined as necessary throughout this article. However, parents with young children may miss opportunities to address race with their children if they postpone engaging with race-related topics until the research exists. Given this dilemma, we propose that researchers highlight laboratory studies that provide hints about what may be effective for reducing children’s racial biases. In doing so, however, it is essential that researchers are cautious and honest about the state of the science, clearly noting when recommendations are extending beyond the existing evidence base and working to avoid the prevalent tendency to come across as authoritative absent evidence.

In sum, we hope that this article serves as a call to action for researchers in this field to conduct intervention research in the service of producing empirically supported, specific, shareable suggestions for parents who are seeking advice about how to productively address race with their children. As we move forward, we gain optimism about the potential for parental engagement to address children’s biases from two sources: First, in other domains, parents can change how children act when parents are given concrete suggestions—in particular, suggestions derived from theory, that are specific, and that have been rigorously evaluated (e.g., see Brody, Yu, Chen, Beach, & Miller, 2016, on parenting interventions to reduce risky parenting and improve children’s health; see Harackiewicz, Rozek, Hulleman, & Hyde, 2012, on parenting interventions to improve children’s educational outcomes). Second, there is a great deal of interest in and energy devoted to addressing children’s racial biases: Our Internet search generated hundreds of articles and blog posts, and these articles and posts have been read by thousands of people. Our hope is that psychological scientists can come together as a field to capitalize on that interest and provide effective, evidence-based suggestions going forward.