Minimizing Bias and Student Diversity in Assessment

Abstract

No teacher can entirely avoid the realities of student standardized assessment. But teachers must make informed choices in the classroom in regards to how students are instructed, based upon individual student needs and awareness of student diversity. There are significant questions regarding the potential biases of many standardized tests, particularly in regards to historically discriminated-against racial, ethnic, and socio-economic groups. Teachers must be aware of these questions and biases and act as advocates for their students on a schoolwide and statewide level to ensure fairness.

Ethical Standards in Assessment:
Minimizing Bias and Student Diversity in Assessment

Education is supposed to be a great social leveler. Unfortunately, many concerns have been raised regarding the ability of commonly-used educational assessment tools to provide unbiased information about all students, regardless of students’ demographic characteristics. Teachers must balance the need to prepare students for these highly pressured exam environments with the need for individual instruction and assessment. They must also be aware of the potential concerns raised about such exams in the literature and how to address them.

Needs Assessment

According to the mandatory Professional Standards for Educational Leaders (2015), educators must have a strong sense of mission and values, adhere to educational norms, and be responsive to student needs, including needs for diversified instruction and individual assessment. Yet standardized assessment is often an inevitable part of every educator’s required performance objectives. According to Standard 4: “Effective educational leaders develop and support intellectually rigorous and coherent systems of curriculum, instruction, and assessment to promote each student’s academic success and well-being” (p.1) Historically, according to the literature, many forms of standardized assessments have found to be inadequate in the manner in which they address student diversity.

As noted by Kruse (2016), cultural bias can be expressed on standardized exams based upon results, including “significantly different results for definable subgroups from apparently similar ability levels” as well as “issues with the fair and equitable interpretation and use of test results” (p.23). Cultural biases can include the use of language or references which certain socio-economic groups may be less likely to be exposed to outside of class or unrepresentative reading passages which may generate a sense of exclusion. Teachers must first act as advocates for students if they feel students are being subjected to unfair assessments. They must also strive to use as unbiased materials as possible to facilitate learning not simply for the test but to support student learning in general.

This is in keeping with the other standards articulated in Professional Standards for Educational Leaders (2015), which stress inter-professional collaboration with other instructors to more accurately understand student needs and why deficiencies may be exhibited within certain populations. Also, as noted in the Standards for Advanced Programs in Educational Leadership, (2002), families and communities should also be engaged. If there are concerns about student achievement, parents should be solicited to determine the possible cause of such issues, beyond purely looking at assessment results. Assessments, both standardized and teacher-generated, only paint a picture of student achievement during a narrow point in time. This should limit their use for determining a students’ long-term future. They should be used for informative results, not pass judgement on students, particularly given the absence of any perfect test to fully render an objective judgement about a student. Qualitative as well as quantitative assessments should be used.

Focus Areas of Need

According to Evans (2013), along with an assessments’ validity and reliability, fairness is a critical component in evaluating the ability of an assessment such as a standardized or in-class exam to produce a useful and accurate portrait of student achievement. Cultural biases act as an impediment to this. “Cultural sensitivity is more about including content, scenarios, and contexts that are relevant to people from all sorts of different backgrounds and perspectives” and rather than simply including diverse materials, it must be subject to “rigorous reviews during large-scale assessment development, often including the use of rubrics and checklists” (par. 6). In other words, it is not enough to merely offer culturally diverse stories or to change the names of students in word problems.

Bias can be reflected in terms of vocabulary, assumptions about cultural familiarity, and the likelihood of familiarity with certain types of skills that do not really reflect the students’ ability, and “occur when an item’s content matter privileges a certain kind of background knowledge or experience” (Evans, 2013, par.6). Vocabulary questions about certain kinds of foods or math problems which reflect assumptions about a student’s familiarity with certain kinds of sports might be examples of questions which exhibit socio-economic or cultural biases that the test designer may not be aware of; teachers can attempt to prepare students to encounter these questions to an extent. They should also try to avoid such intimidating questions in their own tests as well as make test designers aware of the cultural framework of their students and act as advocates in the field of education for less biased test content that more accurately reflects student ability.

Best Practices

Although all standardized tests measure students against certain norms, certain tests have found to be less reflective of biases. For example, in assessing student language ability, to reflect the needs of students raised in homes where English is not spoken as a first language, Kruse (2016) cites the use of “least biased language assessment” which re-norms student results to compare students of similar backgrounds (p.25). Rather than comparing the results of all students in mainstreamed classes uniformly, students with ESL parents might be compared against students of other ESL parents, even if they were proficient enough in English to take the same exams as their peers. There have also been calls for “dynamic assessments (i.e., sequences of instruction, testing, and retesting), the use of non-standardized measures, and test modification” to provide more nuanced results and to offer greater information about student achievement, to reflect students’ diverse backgrounds (Kruse, 2016, p.25).
Greater analysis is also needed in regards to individual results on test questions, to determine of certain questions appear to show greater bias even when assessing students of similar ability. For example, in comparing results of students of different racial backgrounds on standardized assessments in the sciences, “inference-choice items were more susceptible to issues of cultural bias than literal-choice items,” possibly due to the “fact that inference items involve applying one’s knowledge of real-life contexts, and students from different cultural groups may have different real-life contexts” (Kruse, 2016, p.25). Again, this is an issue which teachers may need to be aware of when preparing their own assessments for class and also when giving feedback in regards to how standardized assessments are used on a statewide or schoolwide level. Even tests which meet statistical standards of validity and reliability may exhibit subtle biases.

Action Plan

First and foremost, teachers must conduct a thorough audit of the types of assessments they use in their own classrooms, to screen for potential biases. They should also reevaluate the ways in which student achievement is viewed in context. This may mean speaking with parents to better understand how the education of students may be affected by the child’s home context. Socio-economic deprivation may limit student knowledge of certain assumed cultural references; students who speak English as a first language may still have a more limited English vocabulary if their parents speak English as a second language. They may also have less access to assistance with their homework.

Student achievement and progress, just like student education, should be viewed in a contextual and individual manner. Although teachers may have less influence in terms of how this is expressed in regards to standardized assessments, on a schoolwide level they have responsibilities as educators and advocates to use professional associations and connections to express their opinions about the validity of such testing. They should also be conscious of how student evaluative reports are used within the school. One of the most troubling areas in regards to cultural bias and testing is that such tests may be used to classify or place students in a manner that limits student opportunities, even if students have had to overcome significant challenges to secure they scores they have achieved. Again, context rather than scores alone should be considered.

References

Evans, S. (2013). Five characteristics of quality educational assessments. Northwest Evaluation Association. Retrieved from: https://www.nwea.org/blog/2013/five-characteristics- quality-educational-assessments-part-three/

Kruse, A. (2016). Cultural bias in testing: A review of literature and implications for music education. National Association for Music Education, 35(1) 23 –31. Retrieved from: http://journals.sagepub.com/doi/pdf/10.1177/8755123315576212

Professional standards for educational leaders (2015). Council of Chief State School Officers (CCSSO).

Standards for advanced programs in educational leadership. (2002). National Policy Board for Educational Administration.