Understanding Disciplinary Disproportionality: Stereotypes Shape Pre-Service Teachers’ Beliefs About Black Boys’ Behavior

Abstract

The disproportionate discipline of Black male students is a pervasive problem in U.S. schools. To examine the role of stereotypes in disciplinary disproportionality, pre-service teachers were randomly assigned to read a vignette about a defiant student. Those who read a vignette about a Black student believed that the student was more likely to misbehave in the future, compared with those who read a vignette about a White student. These findings suggest that some teachers attribute the misbehavior of Black male students to more stable causes, which may lead them to alter their behavior toward these students.

Keywords

discipline policies low expectations urban education African American students Black males pre-service teachers

Disproportionate discipline is endemic within the U.S. educational system. Black students, especially males, are more likely to be suspended and expelled than students of other races. Although disciplinary disproportionality clearly exists, its causes have been difficult to ascertain. However, numerous studies have disconfirmed the hypothesis that disproportionality is solely due to true differences in the rate or types of offenses (e.g., Fabelo et al., 2011). The differential treatment of Black students is likely one cause of disciplinary disproportionality (Gregory, Skiba, & Noguera, 2010). In the judicial system, discrimination against Black males is also pervasive (Alexander, 2010; Banks, Eberhardt, & Ross, 2006). These disparities may parallel discrimination in the classroom, which has been compared with a courtroom (Weiner, 2003). Furthermore, these disciplinary patterns may be particularly impactful in urban schools, considering that (a) Black students represent a considerable proportion of students in urban contexts (Nieto, 2014) and (b) urban schools concurrently face other systemic inequities and challenges (Milner & Lomotey, 2014), which may compound their students’ vulnerability.

In this article, we first present an attribution-based approach to understanding disciplinary disproportionality. Second, we review what is known about the extent, consequences, and origins of disciplinary disproportionality. Third, we present findings from a study of pre-service teachers’ attributions about a student’s behavior in which we randomly assigned some participants to read a vignette about a Black student and others to read a vignette about a White student.¹

Theoretical Framework: An Attribution-Based Theory of Stereotypes

Throughout each school day, teachers continually decide how to respond to students’ behavior and academic performance. For example, teachers evaluate the severity of the outcome itself and the student’s perceived responsibility when they decide how to respond appropriately to a behavioral offense. These judgments of responsibility include whether the reason for a student’s behavior was internal or external to the student, whether the reason is controllable, and the likelihood that a student will behave similarly in the future (Weiner, 2006). To successfully navigate complex social situations such as these, teachers must make both quick and accurate judgments. Stereotypes allow people to make quick but biased social judgments about another individual (Wheeler & Fiske, 2005), based on some observable characteristic of that individual, such as his or her age, gender, social class, or race. High cognitive load and stress may increase a person’s tendency to rely on stereotypes (Fiske & Russell, 2010). In the classroom context, therefore, teachers may be more likely to rely on stereotypes about students whom they perceive to be members of an outgroup when they are affected by (a) the cognitive load associated with teaching and general classroom management and (b) the stress associated with managing incidents of student misbehavior.

Stereotypes, which are beliefs about the behaviors and traits that purportedly characterize a group of people, allow people to predict and explain another individual’s behavior. It has been hypothesized that all stereotypes are based on attributions that entail stable causes and characteristics (Reyna, 2000). In other words, stereotypes lead people to believe that stereotypical behaviors will recur whether or not they are controllable or uncontrollable (e.g., lack of effort or lack of ability) or due to external or internal causes (e.g., poor home life or oppositional values). This hypothesis is supported by a study of police officers and juvenile probation officers that found that participants believed adolescents were more likely to reoffend when their stereotypes about Black Americans were unconsciously primed (Graham & Lowery, 2004). In this article, we examine whether these findings can be extended to teachers’ beliefs about students.

The empirical research on disciplinary disproportionality we review below suggests that teachers’ disciplinary decisions about Black boys may be biased. Teachers may form negative stereotypes about Black students even when they are not consciously or willingly prejudiced. First, the pairing of Black faces and names with negative stimuli in the media has created and reinforced negative stereotypes that pervade American society (Massey, 2007). Second, existing disparities between White and Black Americans across various life domains also lead to and reinforce stereotypes (Jost, Banaji, & Nosek, 2004). Third, persistent discrimination in the American housing market and White Americans’ housing preferences have led to high levels of segregation in many communities in the United States (Massey, 2007); this limits contact between White and Black Americans. This lack of interaction leaves many White Americans with few positive experiences to counteract negative stereotypes.

A recent review of the literature on classroom management in urban and racially diverse schools suggests that mutual respect and positive personal relationships between teachers and students lay a foundation for successful classroom management, and that good classroom management may reduce the frequency of exclusionary discipline in schools (Milner, 2015). However, some teachers’ preexisting stereotypes may hinder the development of positive and respectful relationships with their Black students and may therefore contribute to disciplinary disproportionality. In turn, Black students who knowingly face these stereotypes may inadvertently act in ways that confirm teachers’ and administrators’ beliefs (Steele, Spencer, & Aronson, 2002). Although stereotype threat is typically conceptualized as an anxiety-related process that affects intellectual performance in high-pressure settings, this process may also affect young people who believe that their behavior will be misinterpreted by school authority figures. Anxiety or stress stemming from this concern may affect young people’s cognitive processing and consequently their behavior during tense teacher–student interactions that have the potential to result in disciplinary action.

We sought to investigate the possible role of teacher bias in disciplinary disproportionality by investigating two overarching research questions:

Research Question 1: Do pre-service teachers interpret a student’s behavior differently depending on that student’s race?

Research Question 2: Do pre-service teachers’ beliefs about how they would respond to a student’s behavior differ according to that student’s race?

Literature Review of Disciplinary Disproportionality

Extent of Disproportionality

Disciplinary disproportionality dates back to the era of school desegregation, when it was found that Black students were 2 to 3 times as likely to be suspended as White students, even in elementary school (Children’s Defense Fund, 1974). The overrepresentation of Black students in exclusionary discipline persists across the United States (e.g., Bradshaw, Mitchell, O’Brennan, & Leaf, 2010; Losen, 2011; Skiba et al., 2014; Skiba et al., 2011). Since the 1970s, the national suspension rate of Black students has increased (Losen, Hodson, Keith, Morrison, & Belway, 2015). The gap between the suspension rates of Black and White students widened for many years over this period, although it narrowed slightly between 2009 and 2012 (Losen et al., 2015). However, the narrowing of the gap was due to an increase in the suspension rate of White students rather than due to a decrease in the suspension rate of Black students.

Disproportionality exists across grades. In the United States, just 18% of preschoolers are Black children, but 48% of preschoolers who receive more than one out-of-school suspension are Black (U.S. Department of Education Office for Civil Rights, 2014). White preschoolers are conversely underrepresented in exclusionary discipline, and constitute 43% of preschoolers but just 26% of those who receive more than one out-of-school suspension. Teachers are more likely to refer Black students to the office than White students in elementary school (2.19 times) and middle school (3.78 times; Skiba et al., 2011).

Disproportionality also exists across different stages in the disciplinary process. Being categorized as Black also remains a significant predictor of receiving an out-of-school suspension after having been referred and after controlling for the severity of the behavior itself (Skiba et al., 2011). In other words, Black students also face disproportionate discipline from administrators who are responsible for determining their punishment.

Repercussions of Exclusionary Discipline

The decision to remove a student from class or school must not be taken lightly. By definition, exclusionary discipline removes the child from the classroom and instruction, and it therefore likely affects children’s learning (Gregory et al., 2010). Longitudinal research (N = 928,940) suggests that children who are suspended or expelled are markedly more likely to be retained and eventually drop out of school (Fabelo et al., 2011). Suspension is associated with lower academic achievement, grade retention, and higher drop-out rates in middle and high schools (Christle, Jolivette, & Nelson, 2005). In addition, suspension and expulsion are theorized to damage a student’s trust in authority figures, interfere with the development of positive relationships with school figures, and lead to alienation when the student does return to school (Brown, 2007; Morrison et al., 2001). These negative correlates of suspension may be particularly salient for urban youth, considering the increased use of these exclusionary actions in urban schools (e.g., Noltemeyer & Mcloughlin, 2010).

Suspension and expulsion contribute to the problem of the “school-to-prison pipeline,” which refers to the process through which some children systematically exit the school system and enter the juvenile and criminal justice systems (Nicholson-Crotty, Birchmeier, & Valentine, 2009; Wald & Losen, 2003). Individual students who are suspended or expelled are more likely to be involved with the juvenile justice system the next year, compared with students who are not subject to exclusionary discipline (Fabelo et al., 2011). This predictive relationship holds even after controlling for race, gender, economic disadvantage, disability (including emotional disturbance), poor academic performance, and poor school attendance. When compounded by disciplinary disproportionality, this risk is particularly concerning because of the discrimination Black males face at all steps of the judicial process (Alexander, 2010).

Potential Origins of Disciplinary Disproportionality

Although the existence of disciplinary disproportionality and its negative associations are undeniable, its proximal causes are less certain. Overall, however, there is little evidence to suggest that Black students are overrepresented in the school disciplinary system for purely objective reasons (for a review, see Gregory et al., 2010). The differential discipline of Black students and White students for the same offenses suggests that bias is an important factor in disproportionality (Losen, 2011). Below, we briefly review research on three alternate hypotheses. We find no evidence that suggests (a) true differences in student behavior, (b) cultural miscommunication, or (c) student socioeconomic status can fully explain differences in discipline rates.

Differences in students’ behavior

A simple explanation for disciplinary disproportionality would be that Black students commit offenses at higher rates than White students. Some have also hypothesized that Black students may engage in more severe acts of misconduct.

In fact, differences in compliance or behavior cannot account for disproportionality in discipline. Being classified as “Black” or “African American” predicts an increased risk of suspension and expulsions across various kinds of incidents (Skiba, Trachok, Chung, Baker, & Hughes, 2012). For example, in a study of adolescents, 95% of Black students implicated in a weapons-related offense were suspended, compared with 85% of White students in the same situation (Nicholson-Crotty et al., 2009). Analysis of disciplinary decisions in North Carolina schools indicates that Black students were more than twice as likely to be suspended for relatively minor first-time offenses, such as using a cell phone and violating the dress code, than White students (Losen, 2011).

Disproportionality may be especially likely when the violation is subjective or when consequences are not predetermined. A study of almost one million students found that the greatest degree of disproportionality existed for minor offenses that left the type of disciplinary action to the discretion of an administrator (Fabelo et al., 2011). Another analysis of 11,001 middle schoolers’ disciplinary records found that Black students were more likely to be disciplined for offenses that are “both less serious and more subjective in their interpretation,” such as disrespect and making excessive noise (Skiba, Michael, Nardo, & Peterson, 2002, p. 13). White students, however, were more likely to be disciplined for objective offenses such as using obscene language or smoking (Skiba et al., 2002). Finally, in a study of Louisiana students attending public schools, Forsyth, Biggar, Forsyth, and Howat (2015) found that Black students were disciplined at disproportionate rates for 28 of 32 specific offenses, including both subjective offenses such as treating an authority figure with disrespect and objective offenses such as tardiness. White students were only disproportionately disciplined for three objective offenses: possessing a pocket knife, using or possessing alcohol, and using or possessing tobacco.

Cultural miscommunication

The subjectivity of Black students’ infractions, such as disrespect, threat, and excessive noise (Skiba et al., 2002), could indicate a difference between the cultural norms and behavioral expectations of Black students and their predominantly White female teachers. It has been posited that Black students’ purported loud tone of voice, “impassioned” manner, or preferences for active learning may be interpreted as disruptive, threatening, or insubordinate (Monroe, 2005; Townsend, 2000, p. 384). However, these ostensible cultural conflicts are often assumed rather than demonstrated. The existence of group-level differences in learning style has been strongly questioned and criticized (Frisby, 1993), and this line of research does not in fact demonstrate true differences in behavior between groups of students.

A longitudinal study of 1,169 students from Grades 1 to 7 in Baltimore indicates that Black children were still more likely to be suspended or expelled from school after controlling for teacher-perceived disruptive behavior (the scale including items on breaking rules and acceptance of authority; Petras, Masyn, Buckley, Ialongo, & Kellam, 2011). In another study of 1,493 elementary students, students who were classified as Black were still significantly more likely to be disciplined (e.g., receive an out-of-school suspension) after controlling for peer evaluations of students’ overt physical aggression, relational aggression, impulsivity, pro-social behavior, and social group status (Horner, Fireman, & Wang, 2010). If cultural miscommunication played a large role in disproportionality, differences in perceived disruptive behavior would explain the disproportionality; this is not the case.

However, certain characteristics of schools themselves that are associated with disproportionate and exclusionary discipline have been identified. One characteristic that is associated with higher rates of suspension and harsher discipline in general is a larger enrollment of Black students, even after controlling for student behavior (Skiba et al., 2012; Welch & Payne, 2010). White teachers in these schools have many interactions with many different Black students; this repeated contact would seem to increase teachers’ knowledge of purported African American behavioral norms. Therefore, this finding further weakens the cultural miscommunication hypothesis.

Socioeconomic status

Disciplinary disproportionality is not limited to Black students. Students of low socioeconomic status are also disproportionately likely to be disciplined (e.g., Petras et al., 2011). This complicates the study of disciplinary disproportionality, as race and socioeconomic status are intertwined in the United States. Black Americans are much more likely than White Americans to live in poverty (DeNavas-Walt, Proctor, & Smith, 2012). In addition, individuals with ambiguous racial backgrounds are more likely to be perceived as Black when they appear to have low socioeconomic status (Freeman, Penner, Saperstein, Scheutz, & Ambady, 2011). Others’ perceptions likely affect individuals’ self-identified race (Lee & Bean, 2013), further confounding race and socioeconomic status in the United States. Consequently, it could be argued that Black students are subject to discrimination based on their household’s socioeconomic status. Experimental research has shown, for example, that teachers perceive male students with apparent low socioeconomic status to be less competent and hold lower expectations for their success (Auwarter & Aruguete, 2008).

Although socioeconomic status may be associated with biased evaluations and discrimination, statistical analyses of very large samples (Ns > 10,000) that control for differences in socioeconomic status demonstrate that Black students are still more likely to be disciplined than White students (Skiba et al., 2002; Wallace, Goodkind, Wallace, & Bachman, 2008). In a study of an urban school district that used random sampling (N = 1,169), Black elementary and middle school students had 2.02 times the hazard odds of being suspended for the first time compared with White students, even after controlling for gender, eligibility for free or reduced-price lunch, and teacher-reported levels of aggression (Petras et al., 2011). These elevated hazard odds exceeded the gap between students eligible for free or reduced-price lunch and those who were not eligible, which equaled 1.68, also controlling for other predictors.

Race-based bias

Because neither differences in behavior, cultural miscommunication, nor socioeconomic status can wholly account for the discipline gap, disproportionality may be due to bias related to perceived race. Black students are disproportionally likely to be disciplined across age groups, types of offenses, and settings, and they receive harsher consequences from administrators after having been referred for the same offense as White students. It therefore seems probable that differential treatment of students plays some role. When racial disparities exist, “it should be regarded as highly probable” that implicit racial bias is at work when one “has eliminated all conceivable non-race-related (‘racially neutral’) explanations . . . [even if] none of the relevant decision makers has reported consciously holding negative racial attitudes or stereotypes” (Greenwald & Krieger, 2006, p. 966). This hypothesis of bias, however, does not imply that most teachers and administrators are knowingly racist; much racial bias is implicit and unintentional (Amodio & Devine, 2006). To understand the role discrimination may play in disciplinary disproportionality, we sought to examine if pre-service teachers’ attributions about the behavior of student differed according to whether the student was perceived to be a Black or White boy.

Methodological Issues

Large quantitative studies have eliminated socioeconomic status and differences in rates or types of misbehavior as potential explanations for disciplinary disproportionality. Given the known inequalities in American society today and the history of state-sanctioned racial discrimination, the role of bias in disciplinary disproportionality cannot be eliminated easily. Small qualitative studies give insight into its potential role but cannot be confidently generalized. The limited amount of quantitative research on discrimination in school settings makes it difficult to estimate effect sizes and calculate sample sizes needed for adequate statistical power. In addition, studies investigating the role of discrimination are vulnerable to social desirability, which may obscure true differences.

Social desirability response set

Investigating the role discrimination or racism plays in disciplinary disproportionality is complicated by the phenomenon of social desirability. Since the Civil Rights Era and the elimination of Jim Crow laws, growing numbers of White Americans deny personal bias and the existence of discrimination even though discrimination clearly persists across various life domains (Bonilla-Silva, 2014). When people are questioned about whether they engage in discriminatory and other socially undesirable behaviors, they may consciously or unconsciously give inaccurate answers (Furnham, 1986). Therefore, explicit self-report measures cannot be used to accurately predict discriminatory behavior (Greenwald, Poehlman, Uhlmann, & Banaji, 2009; Oswald, Mitchell, Blanton, Jaccard, & Tetlock, 2013). Even subtle measures are vulnerable to social desirability (Furnham, 1986), hence it is possible that providing photographs of hypothetical children to communicate their racial backgrounds is enough to trigger this response set (e.g., Chang & Sue, 2003).

The nature of complex origins of discrimination

Self-reports of bias are also problematic because of the multifaceted, often implicit nature of bias. Discrimination can arise because of (a) people’s stereotyped negative beliefs about another group or (b) their negative affective, emotional responses to members of another group (Amodio & Devine, 2006; Greenwald & Krieger, 2006). These two types of bias predict different kinds of discriminatory behavior (Amodio & Devine, 2006). Although someone may harbor both types of biases, this is not always the case; scores on implicit measures of the two types of biases are not correlated (Amodio & Devine, 2006). Some implicit bias against Black individuals, however, is common across Americans of all education levels, ages, sexes, and political affiliations (Greenwald & Krieger, 2006).

The Present Study

In sum, researchers attempting to test for the influence of bias in disciplinary disproportionality must heed (a) the multifaceted nature of racial bias (e.g., distinguish between stereotyped beliefs and prejudiced affective reactions), (b) the multifaceted nature of attributional stereotypes (e.g., stability, controllability, and locus), (c) the implicit nature of much bias, and (d) the social norms that influence most Americans to deny or conceal racial bias.

Our study used a student’s name, within a vignette, to communicate race to participants. Research that has used stereotypical names to communicate race to employers has found evidence of racial discrimination against Black Americans (Bertrand & Mullainathan, 2004). In this field research, participants presumed that they were reading the resumes of actual job candidates and were not aware that their discriminatory responses might be noted by another individual. Therefore, any socially desirable response set was minimized. Ethical issues may preclude researchers from carrying out a similar field study in schools. However, the use of stereotypical names may be an appropriate method for the study of disciplinary disproportionality.

Based on the lack of findings in vignette-based research that drew attention to race (e.g., Chang & Sue, 2003), we attempted to minimize the effects of social desirability by randomly assigning participants to read only one of three vignettes rather than presenting contrasting ones. To further rule out the chance that participants would alter their responses because of social desirability, we included one version that gave the student a stereotypical Black name but did not state his race and another that gave the same name and an explicit statement of his race. We planned to analyze whether the randomly assigned questionnaire condition affected participants’ ratings of inappropriateness and the likelihood of the behavior recurring, their reported strategies to deal with the student, or their beliefs about the cause of the student’s behavior. Using an attribution-based model of stereotypes, we hypothesized that teachers would attribute the Black student’s behavior to a more stable cause than that of the White student and would perceive the Black student’s misbehavior as more likely to recur. According to this model, anger and blame will be elicited by internal, stable, and controllable attributions. Therefore, we also hypothesized that teachers who also attributed the behavior to a controllable cause would be more likely to endorse sending the student to the office.

Method

Participant Characteristics

The participants in this study were undergraduate education majors, or pre-service teachers, enrolled in a public university in the Midwest located in a small city near a large metropolitan area. They were enrolled in different sections of an upper-level course that was required for education majors with a focus other than early childhood education. The course focused on the use and interpretation of educational assessments. All had worked in the field full-time for a period of at least 2 weeks as student teachers. Mirroring the body of practicing teachers, the majority were women (n = 76) and identified as European American or White (n = 90). They were advanced in their undergraduate studies: 57.3% were third-year students (n = 55) and 41.7% were fourth-year students (n = 40). Participants were recruited during the Fall 2012 academic semester.

Sampling Procedures

With the approval of an Institutional Review Board, the instructor told 148 pre-service teachers enrolled in different sections of the course about the study. Students were notified that they would receive an email inviting them to participate in a research study, but they were not informed of the questionnaire’s content or purpose. The course rosters for the class were combined. The order of names was randomized, and the names were divided into three roughly equal groups. Each group received an identical email inviting them to participate in an online questionnaire for a chance to win a gift card. The email explained that the study concerned responses to behaviors in the classroom. Those who consented to participate were directed to one of the three versions of the survey, depending on which group they were randomly assigned to. We sent a follow-up email approximately 1 week after this email, and the course instructor provided a verbal reminder.

Approximately 66% (N = 98) of the pre-service teachers who were contacted responded to the survey and were included in our analyses. The pre-service teachers who chose to click on the link in the email once again consented to participate. On the next page, they read the short vignette about a seventh-grade male student, and then answered questions about the student, his behavior, and their hypothetical response to the situation.

Measure

Vignette

The questionnaire began with a short vignette describing a seventh-grade male student. We specified the child’s gender and grade because boys are more likely to be suspended than girls of the same race (Losen & Skiba, 2010), and the risk of suspension increases markedly during Grades 6 and 7 (Petras et al., 2011).

In the vignette, the student is not engaging in an assigned academic task, and he responds insubordinately to the teacher’s prompts to begin working. The situation is somewhat ambiguous because little or no information is given about the student’s background, the student–teacher relationship, or the nature of the act (e.g., tone of voice). However, the student’s behavior could easily be interpreted as disrespectful and defiant. Bullough and Richardson (2015) contended that teachers’ classroom management strategies are largely directed at minimizing students’ defiance of teacher rules and directions, and others have hypothesized that the discipline of minority students stems from fear of losing control over the classroom (Fenning & Rose, 2007). Three versions of the vignette were created. All were identical, except for the student’s name and implied or given race.

Experimental manipulation of name

To imply race, we used a name that is typically only given by White parents or Black parents, based on an analysis of birth certificate data (Fryer & Levitt, 2004). One version included a stereotypical White name (“Cody”) and an explicit statement of the student’s race (“Caucasian”), another included a stereotypical Black name (“Darius”) and an explicit statement of the student’s race (“African American”), and the last included the same stereotypical Black name but did not explicitly state the student’s race. Two versions of the questionnaire featuring the Black student were created because it was not known whether explicitly stating the student’s race would affect participants’ response patterns.

The racial connotation of each name was confirmed in a convenience sample of 35 participants for another study; participants were unanimous and confident in rating each name as having stereotypically Black or White connotations (Noltemeyer, Kunesh, Hostutler, Frato, & Sarr-Kerman, 2012). However, we note that the “stereotypical” names we used may not be typical of many people who identify as Black (e.g., recent Caribbean or African immigrants) or White (e.g., Arab Americans). In an attempt to control for socioeconomic status, we chose a stereotypical White name that was preferred by low- and middle-income parents, rather than high-income parents, at the time of the original analysis by Fryer and Levitt (2004). Socioeconomic status shapes teachers’ perceptions of student behavior (Auwarter & Aruguete, 2008), and coming from a low-income household is associated with risk for suspension (Petras et al., 2011). Other details, such as the class period and class subject, were given in an attempt to not make the student’s race or name seem out of place or conspicuous.

Dependent variables

After being told to imagine themselves as the teachers in the vignette, the participants indicated how they would respond in the situation, rated how inappropriate the behavior was, rated how likely the student was to engage in the behavior again, indicated what preventive measure they would take to prevent the behavior from recurring, and indicated the proximal cause of the behavior (e.g., having a bad day, academic frustration, etc.). The severity of the behavior and likelihood of its recurrence (i.e., stability) were rated on 5- and 6-point scales, respectively. The other questions presented the participants with a list of several options, of which they were allowed to pick one.

Data Analysis

To compare the reported inappropriateness and likelihood of recurrence across forms, we used a nonparametric bootstrap method to calculate 95% confidence intervals (CIs) for the population means (Wilcox, 2010). The bootstrap is a nonparametric technique that is used to generate a given number of samples, which are of the same size as the original sample, by randomly sampling observations with replacement (Wilcox, 2010). This technique allows researchers to estimate the sampling distribution of a statistic (such as a mean) and to then estimate CIs for that particular statistic (Haukoos & Lewis, 2005). Where true differences were apparent (Cumming, 2014), we conducted a logistic regression to further evaluate the strength and statistical significance of the associations. We analyzed nominal categorical responses using Fisher’s exact tests because our expected numbers were small.

Results

Recurrence of the Behavior

Estimation

Figure 1 depicts the distribution of responses regarding the likelihood of recurrence, a measure of attributional stability. For each version of the form, we generated 1,000 bootstrap samples and calculated each bootstrap sample’s mean (Wilcox, 2010). Bootstrapping was appropriate for our data because it “allows the researcher to make inferences from data without making strong distributional assumptions” (Haukoos & Lewis, 2005, p. 360). Although our data were skewed, the bootstrap distributions for the means were normally distributed. We used these distributions to estimate CIs for each mean. Figure 2 shows the approximate 95% CIs for the population means for the perceived stability of the behavior.

Figure 1.

Frequency of reported likelihood of recurrence by questionnaire version.

Figure 2.

Bootstrapped 95% CIs for the mean likelihood of recurrence by questionnaire version.

The bootstrap 95% CIs for the implied [3.849, 4.621] and explicitly stated [3.795, 4.387] Black conditions did not overlap with that of the White condition [3.055, 3.783]. Participants who were randomly assigned to read one of the two vignettes featuring a Black student rated that student as more likely to engage in that behavior again, compared with participants who read the vignette featuring the White student. The lack of overlap suggests a true difference between those who read either Black version of the vignette and those who read the White version, with respect to the perceived stability (Cumming, 2014).

Logistic regression

Based on visual analyses and the descriptive statistics reported in Table 1, our data violated the assumptions required for parametric tests such as ANOVA that are used for normally distributed interval data. Standard ANOVA, for example, is not robust to the existence of outliers, varying skewness across groups, and unequal variances across groups (Wilcox, 2012). We therefore conducted a logistic regression to further examine the hypothesis that the participants’ beliefs about the likely recurrence of the behavior differed according to the version of the vignette they read.

Table 1.

Psychometric Properties of the Major Study Variables.

Form	n	M	SD	Range		Skew	Kurtosis
Form	n	M	SD	Potential	Actual	Skew	Kurtosis
	Risk of recidivism
Implicit Black	34	4.235	1.156	1-6	1-6	−0.947	3.844
Explicit Black	33	4.091	0.879	1-6	2-6	0.104	3.182
Explicit White	31	3.419	1.089	1-6	2-5	0.052	1.746
	Level of inappropriateness
Implicit Black	34	2.529	0.662	1-5	1-4	0.209	2.731
Explicit Black	33	2.424	0.663	1-5	1-4	−0.045	2.746
Explicit White	31	2.387	0.495	1-5	2-3	0.464	1.215

We recoded responses from “extremely unlikely” to “somewhat unlikely” to equal 0 and responses from “somewhat likely” to “extremely likely” to equal 1. Fifteen of the 31 participants assigned to the White condition believed that the student’s behavior was likely to recur, compared with 29 of 34 respondents assigned to the implicit Black condition and 26 of 33 participants assigned to the explicit Black condition.

We conducted a logistic regression to test the significance of these differences, the results of which are provided in Table 2. Participants who were randomly assigned to read the vignette featuring a Black student whose race was implied through the stereotypical name alone had 6.187 times higher odds (p = 0.003) of reporting that recurrence was “likely,” compared with participants who were randomly assigned to read the vignette featuring a White student. Participants who were randomly assigned to read the vignette featuring a Black student whose race was explicitly stated had 3.962 times higher odds (p = 0.013) that recurrence was “likely,” compared with participants who were randomly assigned to read the vignette featuring a White student. As we included no other covariates, these results are equivalent to the simple odds ratios.

Table 2.

Odds Ratios for “Likely” Recidivism.

Form	Estimate	SE	z	p	Odds ratio	95% CI for odds ratio
Form	Estimate	SE	z	p	Odds ratio	LL	UL
Implicit Black	1.822	0.603	3.02	.003	6.187	1.897	20.173
Explicit Black	1.377	0.557	2.47	.013	3.962	1.329	11.809

Note. CI = confidence interval; UL = upper limit; LL = lower limit.

The student’s race was the only significant predictor out of those we tested in an alternate model. The reported level of inappropriateness (p = .244), attributing student’s behavior to “having a bad day” (p = .135), and participant gender (p = .668) were not related to the reported likelihood of recurrence. Our sample was too homogeneous with respect to self-identified race to examine potential associations and interactions.

Inappropriateness of the Behavior

Although the impact of the student’s race on perceived likelihood of recurrence was clear, we found no salient differences in the pre-service teachers’ beliefs about how inappropriate the behavior was in relation to the student’s race. There was significant overlap across the 95% CIs for mean perceived level of inappropriateness in the implied Black [2.309, 2.749], explicitly stated Black [2.205, 2.644], and White [2.216, 2.558] conditions. Figures 3 and 4 depict the distribution of responses regarding how inappropriate the behavior was and the bootstrap 95% CIs, respectively.

Figure 3.

Frequency of reported severity by questionnaire version.

Figure 4.

Percentile bootstrap 95% CIs for severity by questionnaire version.

Response to Student and Perceived Cause

The categorical data collected through the questions about the immediate response, future preventive strategy, and presumed cause are displayed in Table 3. These data were analyzed using Fisher’s exact test. The number of cells having expected counts less than 5 precluded the use of a Pearson chi-square test. No significant differences were found for the immediate response, p = .325, preventive strategy, p = .845, or presumed cause, p = .297.

Table 3.

Descriptive Statistics of Reactions and Perceived Cause by Questionnaire Version.

	White	Black
	Stated	Stated	Implied
Immediate response (%)
Ignore	0	0	0
Indirect request	11 (35.48%)	11 (33.33%)	11 (32.35%)
Humor	0	3 (9.09%)	0
Direct request to either comply or go to office	20 (64.52%)	19 (57.58%)	23 (67.65%)
Would send to office immediately	0	0	0
Future strategy
Nothing	0	0	0
Praise others	1 (3.23%)	1 (3.03%)	1 (2.94%)
Praise student	9 (29.03%)	16 (48.48%)	15 (44.12%)
Help immediately	3 (9.68%)	4 (12.12%)	4 (11.76%)
Wait to help	3 (9.68%)	1 (3.03%)	1 (2.94%)
Explain consequences	14 (45.16%)	11 (33.33%)	13 (38.24%)
Threaten	1 (3.23%)	0	0
Proximal cause
Skill deficit	8 (26.67%)	3 (9.09%)	9 (27.27%)
Bad day	7 (23.33%)	9 (27.27%)	6 (18.18%)
Age	1 (3.33%)	4 (12.12%)	0
Difficult personality	0	1 (3.03%)	0
Difficult teacher–student relations	1 (3.33%)	4 (12.12%)	4 (12.12%)
Home/parents	10 (33.33%)	8 (24.24%)	12 (36.36%)
Peers	3 (10%)	4 (12.12%)	2 (6.06%)

Discussion

We sought to identify if and how teachers may contribute to disciplinary disproportionality by analyzing pre-service teachers’ reported perceptions of and hypothetical responses to an ambiguously defiant seventh-grade student. In particular, we wanted to determine whether participants interpreted the ostensibly Black student’s behavior to be more stable, or more likely to recur, as predicted through an attribution-based theory of stereotypes (Reyna, 2000). Pre-service teachers who were assigned to read either questionnaire featuring the Black student did rate the behavior as being more likely to recur. Interestingly, however, these participants did not select notably more stable causes from the available options when they were asked to imagine the specific reason for the student’s behavior. These findings support other conclusions in the literature that (a) bias and discrimination may be implicit or unconscious, rather than intentional and (b) bias and discrimination are multifaceted.

Perceived Likelihood of Recurrence

Participants’ perceptions about the likelihood of recurrence, or their attributions about stability, were influenced by the race of the student. The relationship between stability and student race was present whether the hypothetical student’s race was communicated implicitly by stereotypical name alone or stated explicitly. Our results indicate that some teachers may believe that a Black male student who is ambiguously defiant is more likely to misbehave again than a White male student, even if these students have behaved in identical ways. This finding is in line with an attributional model of stereotypes that suggests stereotypes affect people’s attributions about the stability of other individuals’ behavior. Some teachers may hold negative stereotypes about Black male students that lead them to perceive and respond differently to these students’ actions. These stereotypes may be particularly problematic in urban contexts, where “ . . . teachers’ heightened concern with disruptive behavior often takes precedence over learning” (Pane, Rocco, Miller, & Salmon, 2014, p. 298).

People naturally have a lower sense of efficacy for changing an individual’s behaviors when they perceive such behaviors to be stable (Reyna & Weiner, 2001). In addition, undesirable behaviors that are perceived to be controllable (such as most acts of purported defiance and disrespect) can elicit anger from others and increase the chances that retributive, rather than utilitarian, punishment will be enacted (Reyna & Weiner, 2001). Therefore, when implicit bias and stereotypes activate such attributions, teachers may be more likely to make disciplinary referrals and send students to the office. Disproportionate office referrals then contribute to disproportionate suspensions and expulsions.

Perceived Level of Inappropriateness

Some researchers have speculated that origins of disciplinary disproportionality relate to biased interpretations of Black students’ behavior or clashing cultural norms (e.g., Horner et al., 2010). However, the pre-service teachers who participated in this study did not consider the outcome itself to be more or less inappropriate depending on student race. Heightened feelings of inappropriateness may signify prejudicial attitudes or feelings; if so, our findings suggest that stereotyped beliefs rather than prejudice per se may play a role in disciplinary disproportionality.

The Extent and Nature of Measured Bias

Our results are in accordance with theories of implicit bias. Participating teachers did not endorse a notably different pattern of immediate consequences or future strategies to deal with the student’s misbehavior across the different versions of our questionnaire. In other words, there was no evidence of outright discrimination against Black male students. Instead, biased attributions for stability may lead to discriminatory behavior during stereotype-eliciting incidents in actual classrooms (Reyna, 2000). These attributional differences likely shape teachers’ expectations and actions toward students. Subsequently, young people may inadvertently confirm their teachers’ expectations for negative behavior through self-fulfilling prophecies (e.g., Madon, Willard, Guyll, & Scherr, 2011) or stereotype threat (Steele et al., 2002).

Biased attributions may therefore lead to a vicious cycle in which teachers expect Black male students to misbehave, treat these students with suspicion, are more alert and sensitive to their misbehavior, and interpret ambiguous acts in an unfavorable light. When students sense that they are being treated unfairly, they may justifiably react negatively. In addition, being removed from class reduces instructional time and may increase the chance that students act out to escape academic work that has become too difficult. Thus, explicit or implicit judgments that Black students are more likely to repeatedly misbehave or act defiantly may contribute to disciplinary disproportionality, even if teachers do not intend to treat these students differently. This hypothesized process is depicted in Figure 5.

Figure 5.

Hypothesized student–teacher interactional process.

Implications for Future Research

We reduced social desirability by using one vignette rather than juxtaposing multiple ones or directly questioning participants about bias. However, social desirability may have affected participants’ reported immediate response to the student. It seems unlikely that none of the participating pre-service teachers would send a student to the office when confronted with the type of defiant student behavior featured in the vignette. The use of exclusionary discipline, although common in practice, is likely a behavior that is subject to social desirability. Therefore, researchers may consider incorporating more ambiguity, different behaviors, or behaviors of greater severity into vignettes. Although sending a student to the office is a binary variable—one does or does not make a referral—allowing for ordinal or continuous responses may be necessary to detect differences.

The experimental nature of our study makes our findings promising. However, the generalizability of our findings is limited by the relatively small sample size and its geographical restriction. Our sample was composed of participants who were representative of the teacher population and who were living in a bellwether state. Nevertheless, our results should be extended to practicing teachers and more diverse samples of pre-service teachers. Furthermore, our results may not generalize to all students who identify as Black. We indicated race through a stereotypically Black name (i.e., a name that is given mainly by Black Americans), but many Black parents choose names for their children that are not unique to the Black community. Given the inherent diversity among people who identify as Black or African American, further research could reveal the degree to which the stereotypes evoked by these names are generalizable.

Our results may also be limited by our methodology. Implicit measures of bias predict behavior better than self-reports (Amodio & Devine, 2006), but further research on the association between implicit bias and teacher behavior is important to validate our findings. It is also important to ascertain how pre-service teachers’ beliefs translate into behavior once they enter the field. Field experiments such as those conducted in the employment and housing sectors may be necessary to fully understand teacher decision making and disciplinary disproportionality. Bias, attributional processes, and the chain of events that leads to an office disciplinary referral are complex phenomena that interact; qualitative or mixed-methods research may provide insight into how biased disciplinary events transpire.

Relatedly, it is important to determine which teacher characteristics are correlated with more strongly biased attributions. Other research suggests that Black students referred for disrespect may be sent to the office by a small proportion of teachers who are perceived as less caring, less trustworthy, and having lower expectations by their students (Gregory & Weinstein, 2008). What leads these teachers and not others to repeatedly refer students? Instructional effectiveness and classroom management skills may be at work (Skiba et al., 2012), which might coincide with the lower qualifications of many teachers in poor, high-minority-enrollment, urban schools (Engel, Jacob, & Curran, 2014). In addition to skill level, teachers differ with respect to their beliefs about classroom management; see Bullough and Richardson (2015) for a discussion of survey instruments and potential pitfalls in the measurement of these beliefs.

A better understanding of the teacher characteristics associated with disciplinary disproportionality may help explain why some schools are subject to much higher or lower rates of disciplinary disproportionality than one would expect, given known predictors (Fabelo et al., 2011). If only a subset of teachers tend to perceive or respond to Black students in biased ways, studies of undifferentiated groups of teachers may result in an incomplete picture of disciplinary disproportionality. Researchers should identify different individual and situational characteristics that act as moderators and influence whether teacher attributions are affected by student race. Increased attention to the interactional disciplinary process itself may elucidate this pressing matter, which currently continues to negatively affect many Black children attending schools in the United States.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Claire E. Kunesh, MS, MA, is a PhD student in Human Development in the Graduate School of Education at the University of California, Berkeley. Her research interests include school discipline, the academic outcomes of immigrant and ethnic minority adolescents, and racial bias.

Amity Noltemeyer, PhD, is an associate professor and coordinator of the School Psychology Program at Miami University (Oxford, Ohio, USA). She is also co-editor of School Psychology International journal and co-manages several externally funded grants. Her research interests include resilience, school discipline, disproportionality, school climate, and multi-tiered systems of support.

References

Alexander

(2010). The new Jim Crow. New York, NY: The New Press.

Amodio

D. M.

Devine

P. G.

(2006). Stereotyping and evaluation in implicit race bias: Evidence for independent constructs and unique effects on behavior. Journal of Personality and Social Psychology, 91, 652-661. doi:10.1037/0022-3514.91.4.652

Auwarter

A. E.

Aruguete

M. S.

(2008). Effects of student gender and socioeconomic status on teacher perceptions. Journal of Educational Research, 101, 242-246. doi:10.3200/JOER.101.4.243-246

Banks

R. R.

Eberhardt

J. L.

Ross

(2006). Discrimination and implicit bias in a racially unequal society. California Law Review, 94, 1169-1190. doi:10.2307/20439061

Bertrand

Mullainathan

(2004). Are Emily and Greg more employable than Lakisha and Jamal? A field experiment on labor market discrimination. American Economic Review, 94, 991-1013. doi:10.1257/0002828042002561

Bonilla-Silva

(2014). Racism without racists (4th ed.). Lanham, MD: Rowman & Littlefield.

Bradshaw

C. P.

Mitchell

M. M.

O’Brennan

L. M.

Leaf

P. J.

(2010). Multilevel exploration of factors contributing to the overrepresentation of Black students in office disciplinary referrals. Journal of Educational Psychology, 102, 508-520. doi:10.1037/a0018450

Brown

T. M.

(2007). Lost and turned out: Academic, social, and emotional experiences of students excluded from school. Urban Education, 42, 432-455. doi:10.1177/0042085907304947

Bullough

R. V.

Jr. Richardson

(2015). Teacher perspectives on classroom management: Rules, ethics, and “crime control. In Emmer

E. T.

Sabornie

E. J.

(Eds.), Handbook of classroom management (2nd ed., pp. 283-300). New York, NY: Routledge.

10.

Chang

D. F.

Sue

(2003). The effects of race and problem type on teachers’ assessments of student behavior. Journal of Consulting and Clinical Psychology, 71, 235-242. doi:10.1037/0022-006X.71.2.235

11.

Children’s Defense Fund. (1974). Children out of school in America. Cambridge, MA: Author.

12.

Christle

C. A.

Jolivette

Nelson

C. M.

(2005). Breaking the school to prison pipeline: Identifying school risk and protective factors for youth delinquency. Exceptionality, 13, 69-88. doi:10.1207/s15327035ex1302_2

13.

Cumming

(2014). The new statistics: Why and how. Psychological Science, 25, 7-29. doi:10.1177/0956797613504966

14.

DeNavas-Walt

Proctor

B. D.

Smith

J. C.

(2012). Income, poverty, and health insurance coverage in the United States: 2011 (Current Population Reports No. P60-243). Washington, DC: Government Printing Office.

15.

Engel

Jacob

B. A.

Curran

F. C.

(2014). New evidence on teacher labor supply. American Educational Research Journal, 51, 36-72. doi:10.3102/0002831213503031

16.

Fabelo

Thompson

M. D.

Plotkin

Carmichael

Marchbanks

M. P.

III Booth

E. A.

(2011). Breaking schools’ rules: A statewide study of how school discipline relates to students’ success and juvenile justice involvement. New York, NY: Council of State Governments Justice Center.

17.

Fenning

Rose

(2007). Overrepresentation of African American students in exclusionary discipline: The role of school policy. Urban Education, 42, 536-559. doi:10.1177/0042085907305039

18.

Fiske

S. T.

Russell

A. M.

(2010). Cognitive processes. In Dovidio

J. F.

Hewstone

Glick

Esses

V. M.

(Eds.), The SAGE handbook of prejudice, stereotyping and discrimination (pp. 115-130). Thousand Oaks, CA: Sage.

19.

Forsyth

C. J.

Biggar

R. W.

Forsyth

Y. A.

Howat

(2015). The punishment gap: Racial/ethnic comparisons in school infractions by objective and subjective definitions. Deviant Behavior, 36, 276-287. doi:10.1080/01639625.2014.935623

20.

Freeman

J. B.

Penner

A. M.

Saperstein

Scheutz

Ambady

(2011). Looking the part: Social status cues shape race perception. PLoS ONE, 6(9), e25107. doi:10.1371/journal.pone.0025107

21.

Frisby

C. L.

(1993). One giant step backward: Myths of Black cultural learning styles. School Psychology Review, 22, 535-557. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&;db=a9h&AN=9402282638&site=ehost-live

22.

Fryer

R. G.

Jr. Levitt

S. J.

(2004). The causes and consequences of distinctively Black names. Quarterly Journal of Economics, 119, 767-806. doi:10.1162/0033553041502180

23.

Furnham

(1986). Response bias, social desirability and dissimulation. Personality and Individual Differences, 7, 385-400. doi:10.1016/0191-8869(86)90014-0

24.

Graham

Lowery

B. S.

(2004). Priming unconscious racial stereotypes about adolescent offenders. Law and Human Behavior, 28, 483-504. doi:10.1023/b:lahu.0000046430.65485.1f

25.

Greenwald

A. G.

Krieger

L. H.

(2006). Implicit bias: Scientific foundations. California Law Review, 94, 945-967. doi:10.2307/20439056

26.

Greenwald

A. G.

Poehlman

T. A.

Uhlmann

E. L.

Banaji

M. R.

(2009). Understanding and using the Implicit Association Test: III. Meta-analysis of predictive validity. Journal of Personality and Social Psychology, 97, 17-41. doi:10.1037/a0015575

27.

Gregory

Skiba

R. J.

Noguera

P. A.

(2010). The achievement gap and the discipline gap: Two sides of the same coin? Educational Researcher, 39, 59-68. doi:10.3102/0013189X09357621

28.

Gregory

Weinstein

R. S.

(2008). The discipline gap and African Americans: Defiance or cooperation in the high school classroom. Journal of School Psychology, 46, 455-475. doi:10.1016/j.jsp.2007.09.001

29.

Haukoos

J. S.

Lewis

R. J.

(2005). Advanced statistics: Bootstrapping confidence intervals for statistics with “difficult” distributions. Academic Emergency Medicine, 12, 360-365. doi:10.1197/j.aem.2004.11.018

30.

Horner

S. B.

Fireman

G. D.

Wang

E. W.

(2010). The relation of student behavior, peer status, race, and gender to decisions about school discipline using CHAID decision trees and regression modeling. Journal of School Psychology, 48, 135-161. doi:10.1016/j.jsp.2009.12.001

31.

Jost

J. T.

Banaji

M. R.

Nosek

B. A.

(2004). A decade of system justification theory: Accumulated evidence of conscious and unconscious bolstering of the status quo. Political Psychology, 25, 881-919. doi:10.1111/j.1467-9221.2004.00402.x

32.

Lee

Bean

F. D.

(2013). The diversity paradox: Immigration and the color line in twenty-first century America. New York, NY: Russell Sage.

33.

Losen

D. J.

(2011). Discipline policies, successful schools, and racial justice. Boulder, CO: National Education Policy Center.

34.

Losen

D. J.

Hodson

Keith

M. A.

II Morrison

Belway

(2015). Are we closing the school discipline gap? Los Angeles, CA: Center for Civil Rights Remedies. Retrieved from http://civilrightsproject.ucla.edu/resources/projects/center-for-civil-rights-remedies/school-to-prison-folder/federal-reports/are-we-closing-the-school-discipline-gap/AreWeClosingTheSchoolDisciplineGap_FINAL221.pdf

35.

Losen

D. J.

Skiba

R. J.

(2010). Suspended education: Urban middle schools in crisis. Montgomery, AL: Southern Poverty Law Center.

36.

Madon

Willard

Guyll

Scherr

K. C.

(2011). Self-fulfilling prophecies: Mechanisms, power, and links to social problems. Social & Personality Psychology Compass, 5, 578-590. doi:10.1111/j.1751-9004.2011.00375.x

37.

Massey

D. S.

(2007). Categorically unequal: The American stratification system. New York, NY: Russell Sage Foundation.

38.

Milner

H. R.

IV . (2015). Research on classroom management in urban schools. In Emmer

E. T.

Sabornie

E. J.

(Eds.), Handbook of classroom management (2nd ed., pp. 167-185). New York, NY: Routledge.

39.

Milner

H. R.

IV Lomotey

(2014). Introduction. In Milner IV

H. R.

Lomotey

(Eds.), Handbook of urban education (pp. xv-xxiii). New York, NY: Routledge.

40.

Monroe

C. R.

(2005). Understanding the discipline gap through a cultural lens: Implications for the education of African American students. Intercultural Education, 16, 317-330. doi:10.1080/14675980500303795

41.

Morrison

G. M.

Anthony

Storino

M. H.

Cheng

J. J.

Furlong

M. J.

Morrison

R. L.

(2001). School expulsion as a process and an event: Before and after effects on children at risk for school discipline. New Directions for Youth Development, 92, 45-71. doi:10.1002/yd.23320019205

42.

Nicholson-Crotty

Birchmeier

Valentine

(2009). Exploring the impact of school discipline on racial disproportion in the juvenile justice system. Social Science Quarterly, 90, 1003-1018. doi:10.1111/j.1540-6237.2009.00674.x

43.

Nieto

(2014). Afterword. In Milner IV

H. R.

Lomotey

(Eds.), Handbook of urban education (pp. 539-544). New York, NY: Routledge.

44.

Noltemeyer

Kunesh

Hostutler

Frato

Sarr-Kerman

B. J.

(2012). The effects of student and teacher characteristics on teacher impressions of—and responses to—student behaviors. International Education Studies, 5(4), 96-111. doi:10.5539/ies.v5n4p96

45.

Noltemeyer

Mcloughlin

C. S.

(2010). Patterns of exclusionary discipline by school typology, ethnicity, and their interaction. Perspectives on Urban Education, 7, 27-40.

46.

Oswald

F. L.

Mitchell

Blanton

Jaccard

Tetlock

P. E.

(2013). Predicting ethnic and racial discrimination: A meta-analysis of IAT criterion studies. Journal of Personality and Social Psychology, 105, 171-192. doi:10.1037/a0032734

47.

Pane

D. M.

Rocco

D. S.

Miller

L. D.

Salmon

A. K.

(2014). How teachers use power in the classroom to avoid or support exclusionary school discipline practices. Urban Education, 49, 297-328. doi:10.1177/0042085913478620

48.

Petras

Masyn

K. E.

Buckley

J. A.

Ialongo

N. S.

Kellam

(2011). Who is most at risk for school removal? A multilevel discrete-time survival analysis of individual- and context-level influences. Journal of Educational Psychology, 103, 223-237. doi:10.1037/a0021545

49.

Reyna

(2000). Lazy, dumb, or industrious: When stereotypes convey attribution information in the classroom. Educational Psychology Review, 12, 85-111. doi:10.1023/A:1009037101170

50.

Reyna

Weiner

(2001). Justice and utility in the classroom: An attributional analysis of the goals of teachers’ punishment and intervention strategies. Journal of Educational Psychology, 93, 309-319. doi:10.1037/0022-0663.93.2.309

51.

Skiba

R. J.

Chung

Trachok

Baker

Sheya

Hughes

(2014). Parsing disciplinary disproportionality: Contributions of infraction, student, and school characteristics on out-of-school suspension and expulsion. American Educational Research Journal, 51, 640-670. doi:10.3102/0002831214541670

52.

Skiba

R. J.

Horner

R. H.

Chung

C.-G.

Rausch

M. K.

May

S. L.

Tobin

(2011). Race is not neutral: A national investigation of African American and Latino disproportionality in school discipline. School Psychology Review, 40, 85-107.

53.

Skiba

R. J.

Michael

R. S.

Nardo

A. C.

Peterson

R. L.

(2002). The color of discipline: Sources of racial and gender disproportionality in school punishment. Urban Review, 34, 317-342. doi:10.1023/A:1021320817372

54.

Skiba

R. J.

Trachok

Chung

C.-G.

Baker

Hughes

(2012, April). Parsing disciplinary disproportionality: Contributions of behavior, student, and school characteristics to suspension and expulsion. Paper presented at the Annual Meeting of the American Educational Research Association, Vancouver, Canada.

55.

Steele

C. M.

Spencer

S. J.

Aronson

(2002). Contending with group image: The psychology of stereotype and identity threat. Advances in Experimental Social Psychology, 34, 379-440. doi:10.1016/S0065-2601(02)80009-0

56.

Townsend

B. L.

(2000). The disproportionate discipline of African American learners: Reducing school suspensions and expulsions. Exceptional Children, 66, 381-391. doi:10.1177/001440290006600308

57.

U.S. Department of Education Office for Civil Rights. (2014, March). Civil rights data collection: Data snapshot: School discipline (Issue Brief No. 1). Retrieved from http://www2.ed.gov/about/offices/list/ocr/docs/crdc-discipline-snapshot.pdf

58.

Wald

Losen

D. J.

(2003). Defining and redirecting a school-to-prison pipeline. New Directions for Youth Development, 99, 9-15. doi:10.1002/yd.51

59.

Wallace

J. M.

Jr. Goodkind

Wallace

C. M.

Bachman

J. G.

(2008). Racial, ethnic, and gender differences in school discipline among U.S. high school students: 1991-2005. Negro Educational Review, 59, 47-62. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2678799/pdf/nihms50094.pdf

60.

Weiner

(2003). The classroom as a courtroom. Social Psychology of Education, 6, 3-15. doi:10.1023/A:1021736217048

61.

Weiner

(2006). Social motivation, justice, and the moral emotions: An attributional approach. Mahwah, NJ: Lawrence Erlbaum.

62.

Welch

Payne

A. A.

(2010). Racial threat and punitive school discipline. Social Problems, 57, 25-48. doi:10.1525/sp.2010.57.1.25.SP5701

63.

Wheeler

M. E.

Fiske

S. T.

(2005). Controlling racial prejudice: Social-cognitive goals affect amygdala and stereotype activation. Psychological Science, 16, 56-63. doi:10.1111/j.0956-7976.2005.00780.x

64.

Wilcox

(2012). Introduction to robust estimation and hypothesis testing (3rd ed.). Waltham, MA: Academic Press Elsevier.

65.

Wilcox

R. R.

(2010). The bootstrap. In Wilcox

R. R.

(Ed.), Fundamentals of modern statistical methods: Substantially improving power and accuracy (pp. 87-108). Retrieved from https://link-springer-com.web.bisu.edu.cn/book/10.1007%2F978-1-4419-5525-8