Various grading systems have gained traction in recent years to make grading more useful for student learning. These systems are based on educators' professional assessments of student performance, claiming to offer more detailed, timely, and customized evaluations of student work and progress (Buckmiller et al., 2017; Knight & Cooper, 2019; Muñoz, & Guskey, 2015). These systems are often considered as competency-based grading, portfolio-based grading, and standards-based grading, (DeCastro-Ambrosetti & Cho 2005) with standards-based grading receiving the most widespread application (Buckmiller et al., 2020; Erickson, 2011; Iamarino, 2014).
Standards-based grading is when teachers consider student work in relation to a set standard, in order to determine a student's proficiency. Supporters of standards-based grading argue that it is less susceptible to teacher subjectivity compared to conventional grading approaches, where teachers might assign grades like "A" or "C" based on personal feelings or perceptions about a student (Feldman, 2019; Quinn, 2020; Townsley & Wear, 2020). This research investigates this claim by assessing whether standards-based grading systems can effectively mitigate teacher biases, particularly those related to race and non-academic factors, like attribution error. Racial bias involves unconscious associations between a student's race and academic capabilities, while attribution error occurs when teachers unconsciously factor in unrelated student attributes when assessing their work.
Standards-based Grading, Defined
Standards-based grading refers to the collaborative process of collecting and interpreting student-produced evidence to give students an accurate perspective of their proficiency, interpersonal skills, and emotional wellness (Knight & Cooper, 2019; Link & Guskey, 2022). Standards-based grading is intended to assess how well a student's proficiency measures up to a standard and is represented by a categorical label such as exceeds or meets (Guskey, 2014).
Teachers review a body of evidence against a standard and consider the relevance of any recent evidence to determine a student's mastery of standards (Brookhart et al., 2019). When teachers think about how the existing body of evidence and recent scores predict future proficiencies in student development, teachers are better able to provide a more accurate grade based on the evidence of learning students produce.
Teachers communicate their judgment of a student's growth and proficiency when reporting their performance. Standards-based reporting presents grade predictions and learning trajectories based on visible evidence of learning. These trajectories evolve with learning, making the grade book a dynamic resource for reflection, calibration, and feedback for both students and teachers.
Heuristics
Often described as cognitive shortcuts or rules of thumb, heuristics serve as mental tools people employ to simplify complex decision-making tasks or problem-solving endeavors (Kahneman et al., 2002, 2021). These cognitive strategies provide the advantage of efficiency, allowing individuals to navigate the vast volume of information encountered daily more swiftly. However, they also carry certain drawbacks, as an increased reliance on such mental shortcuts heightens the risk of errors or susceptibility to cognitive biases (Kahneman et al., 2021).
In education, one crucial area where heuristic-driven biases can have a significant impact is the process of teachers evaluating students (Kahneman et al., 2021). Teachers may inadvertently employ various heuristics in their grading processes, potentially leading to errors or biases (Kahneman et al., 2016). Further, teachers might resort to heuristics to make speedy decisions due to time constraints or the ongoing demand for classroom attention from students (Gilovich et al., 2002).
As a case in point, the "shifting standards bias" may emerge when teachers unconsciously adjust their grading criteria based on students' characteristics or backgrounds. Additionally, the "expectation bias" might come into play when teachers' grading is influenced by their preconceived expectations of a student's performance rather than the objective quality of the work itself. As such, relying on heuristics can impact how teachers grade student work (Zhan et al., 2021). Both internal (e.g. personal dispositions, values, and expectations) and external (e.g. surroundings and occurrences) elements can influence teacher decision-making. Factors like the time of classes (morning or afternoon), sleep duration, classroom environment, or preceding educational experiences can influence a teacher's judgment (Kahneman et al., 2016).
When teachers sit down with a pile of student essays, internal and external factors may influence their grading. In this regard, Kahneman et al. (2021) state, "Performance evaluations are greatly inconsistent and rely more on the evaluator than on the actual performance being evaluated" (p. 7). This inconsistency leads to unfair grades for students, with implications for their future coursework and potential careers. White individuals tend to lean towards the belief that evaluations mirror the values of the system rather than the personal judgment of an individual, although this assumption doesn't consistently hold true (Kahneman et al., 2021). As depicted in Figure 1, numerous factors impact the judgment of teachers, implying that a student's assigned score may not invariably represent their comprehension of the subject matter.
Attribution Error
We now turn to a specific type of error that influences teacher judgment: attribution error, illustrated in the upper right of Figure 1. According to Harvey et al. (2014), Fritz Heider (1958), one of the early researchers in the field of attribution theory, depicted individuals as “naive psychologists, inherently intrigued by comprehending the roots of success and failure” (p. 128). In other words, people naturally seek to comprehend why events unfold as they do. Nevertheless, the reasons they attribute to why events occur are not always correct. Attribution error, or fundamental attribution error, refers to the tendency to over-emphasize personal characteristics and ignore situational factors when judging others' behavior (Weiner, 1980, 2012).
Additionally, Wang and Hall (2018) state that, "the actors [in an event], e.g. students, tend to attribute [behaviors, outcomes] more to situational forces or constraints, whereas the observers [of the same event], e.g. teachers, are more likely to attribute [behaviors, outcomes] to the actor’s capabilities" (p. 15). Such misattributions can result in a skewed understanding of the event, engendering biases and inconsistencies (Beckman & Rodriguez, 2021; Dweck, 2018; Kelley, 1967).
Looking closer at attribution error in the classroom environment, this tendency might translate into associating a student's subpar performance solely with their supposed lack of ability or effort, overlooking other potential factors like inadequate teaching methods or limited access to resources (Cohen, 2022; Graham, 2020). Within the grading process, attribution bias may occur when the teacher focuses more on who the student is than on the quality of their work, giving a student who typically does well high grades, for example, even if the current piece of work is low quality. For example, if a teacher scores a student’s work as high quality and provides feedback that the student’s performance implies the performance was due to the student’s innate intelligence (internal, stable factors), this could be an incorrect assumption (or attribution). Or if a teacher scores a student’s work low and attributes the outcome to lack of effort (internal, unstable), this misattribution may lead the teacher to provide erroneous feedback.
Overall, attribution error can be problematic when the teacher places more emphasis on either personal internal factors or on external factors rather than the actual quality of their work. Investigation into attribution error studies how individuals articulate their perceptions regarding behavior, outcomes, and the causality of incidents (Harvey et al., 2014; Heider, 1958; Kelley, 1967; Weiner, 1974, 1980; Carson, 2019). These studies elucidate how individuals interpret success or failure, even when such explications might not be immediately apparent to the people themselves (Hollyforde & Whiddett, 2002).
Racial Bias
The second type of bias we investigate in this study is racial bias. Implicit racial biases are "associations made by individuals in the unconscious state of mind [that] cause individuals to unknowingly act in discriminatory ways" (Maryfield, 2018, p. 1), and activated racial bias often positions people of color as potentially vulnerable (Bodenhausen et al., 2010; Fiske, 1998, 2015; Rogers et al., 2020). In the presence of individuals of a specific race or ethnicity, implicit biases can automatically activate (Fazio, 2001; Herring, 2013; Koppehele-Gossel, 2020). Implicit bias research indicates that many White Americans possess biases that favor their own ethnicity and display biases against Black individuals (Greenwald & Farnham, 2000).
Despite some incremental progress, the prevailing image of a school student in the United States remains that of a White, middle-class student (Lewis & Diamond, 2015; Preston, 2007). This bias toward privileging White students is evident across various dimensions of the education system, including practices, policies, and personnel. Additionally, racial bias can manifest itself through policies that appear race-neutral on the surface, exclusionary teaching methods, and discriminatory grading practices (Lewis & Diamond, 2015; Paslay, 2021). It is worth noting that some studies (Babad, et al., 1982; Starck et al., 2020) suggest that educators exhibit racial biases similar to those of the average American.
Just like anyone else, a teacher's ability to suppress implicit biases can be compromised by various physical and psychological demands, including factors like time constraints, resource limitations, and stress. The lack of resources and time can lead teachers to create inaccurate representations of their students, which may affect their pedagogy (Spencer et al., 2016). For example, teachers may refer students from the racial majority to specific programs more often than their minority counterparts (Tenenbaum & Ruck, 2007); this is compounded since the majority of teachers are white (Howard, 2006). Jacoby et al. (2016) hypothesized "that greater implicit bias increases whites' anxiety when teaching Black students, and that the resultant distraction and depletion will diminish the quality of their instruction and, subsequently, student learning" (p. 52).
Several studies found substantial differences in the performance evaluation of students of color when the teacher subconsciously activated a general stereotype that African-American and Latino students do not score better than their White and Asian students (Ready & Wright, 2011). Jacoby et al. (2015) recruited undergraduates for a study that included cross-race and same-race lesson pairs. The researchers assigned the White participant to the instructor role and a Black or White participant to the learner role. Teachers then taught a lesson to the students. After the lesson, the participants received a five-minute discussion period. Immediately following this period, the participants were separated. The learner then had five minutes to complete an exam, while instructors were measured for explicit bias. Their findings suggest that White teachers displayed more observable anxiety-related behaviors when teaching Black learners, such as increased speech incoherence, lack of eye contact, and the teachers' choice of physical positioning in the classroom.
Warikoo et al. (2016) suggest that negative implicit associations toward low-achieving groups (stereotypes) often abound in classrooms "not only because they are automatic and difficult to control, but also because they are pervasive" (p. 2). Second, well-intentioned teachers may "sometimes act on unconscious biases towards students from stigmatized groups" (p. 3). Third, "implicit racial associations consistently correlate with problematic feelings and behaviors that emerge during interracial interactions" (p. 3). This may affect student performance and perpetuate problematic feelings between the teacher and the student.
Similarly, studies conducted by Fox (2015) and Joshi et al. (2018) explored how teacher-student racial congruence influenced a teacher's assessment of student performance. Oates (2003) identified disparities in teacher evaluations of student performance, particularly pronounced in situations where a white teacher assessed a Black student. Another study suggested that racially incongruent teacher-student relationships may indirectly contribute to poorer performance among minorities, as ethnic majority teachers appeared to have lower expectations for ethnic minority students (van Ewijk, 2011). According to Bonefeld et al. (2020), "the judgment made by teachers about students' performance did differ in terms of student characteristics that were unrelated to performance, such as immigration background and gender, in addition to differing on performance-related variables" (p. 198).
Although the literature demonstrates racial disparities in expectations and evaluation of students and academic achievements (e.g., Irizarry, 2015; Yates & Marcelo, 2014; Reardon et al., 2019), other studies exploring educator racial bias show inconclusive results. For example, Pigott and Cowen (2000) explored the extent to which perceived student race factors into teachers' judgment of student work. To illustrate, if a teacher is negatively biased toward students of color, they may see those students not as talented as White students, which may cause them to judge the student's work based on that perception (Ferman & Fontes, 2021; Jussim, 1989; Jussim et al., 2020). Indeed, "any internalized racial prejudice can activate biases and lead teachers to use discriminatory performance evaluations" (Wood & Graham, 2010, p. 177).