Gathering student data during the SAT: helpful information or discriminatory practice?
I sit down, my body overcome with nerves but my head telling me to focus. Over the next four hours, I will answer a series of questions, the number of which I get correct will, I believe, be instrumental to my future. The SAT. For years, the importance of this test has steadily grown in my mind — from peers taking prep classes freshman year and the school administered PSAT to the “average accepted student’s SAT score” on college websites, the value of a strong SAT score in the college admissions process was not lost on me.
Upon receiving my bubble answer sheet, the proctor instructs us to fill out the informational sections one by one, including my name (JULIA THOMPSON), residence (SF, CA), gender (FEMALE), and ethnicity (TWO OR MORE RACES), among others. I carefully bubble in my personal responses to each without thinking twice.
Now, three years later, I am finished with my second year of college. The SAT is long behind me. Yet, I am consumed by frustration with the SAT, for many reasons, but it is particularly this beginning informational section, the one that I couldn’t care less about during the actual exam, that now draws my attention.
Stereotype threat occurs when one becomes (consciously or unconsciously) anxious or concerned about fulfilling negative stereotypes connected to a social group the individual belongs to; consequently, this anxiety adversely affects the individual’s performance, reinforcing said stereotype. Pioneered by Steele and Aronson, hundreds of studies and articles have reported how stereotype threat can adversely affect the test taking performance of groups (often minority groups such as African Americans and females) associated with stereotypes of lower intellectual ability and/or poor test-taking abilities.
The strength of stereotype threat is affected by how the task is framed — mention of a negative stereotype in direct relation to the given task increases the chance that an individual will conform to that stereotype. For instance, if prior to a SAT/ACT/AP exam, a group of students was told, “Statistically, white students perform better than African American students do on this exam,” this would heighten the chance of a greater gap between white and African American student performance.
Granted, this level of framing does not take place prior to College Board or other standardized test exams. However, the combination of what the SAT measures, along with a section mentioning ethnicity prior to the exam, has the potential to create a situation where stereotype threat could adversely affect the performance of minority students.
According to the College Board, the newly designed SAT, which launched in 2016, is “focused on the skills and knowledge at the heart of education,” measuring “what you learn in high school” and “what you need to succeed in college.” Additionally, the College Board website states “the same habits and choices that lead to success in school will help you get ready for the SAT.” Ability to perform well on the SAT is linked to the ability to perform well in school, which is traditionally correlated to intellectual ability.
Combine this mindset of what the SAT measures, which a student very likely is aware of entering the exam, with a reminder of ethnicity directly before the exam. The similarity between this situation and those that have been staged in stereotype threat research suggest that the SAT sets up the potential for stereotype threat to become triggered.
The SAT allows for a standardized comparison of students applying to college. It adds another element to the college application, one college admissions can use to compare students on a theoretically, somewhat even playing field, as compared to GPA, a number on the 4.0 scale which may represent a different level of performance according to school, teacher, class difficulty, etc.
However, while standardized tests such as the SAT may seem fair and objective on the surface, there are a myriad of factors that affect a student’s ability to perform well on the exam that are in no way correlated to intellectual ability, levels of knowledge, or academic potential. Some major factors that fall under this category include quality of education (from early childhood care up to high school), family socioeconomic status (which affects a student’s ability to access additional test prep materials, such as books, tutors, and classes), cultural upbringing (something as simple as speaking English at home can give a student an advantage over students from non-English speaking homes in reading and writing sections), and even the test itself (some students are wired to test better in the SAT format than others). Often, these factors are ones that negatively affect students from disenfranchised communities. Additionally, studies have shown that although race is in no way linked to individual intellectual ability or potential, it is becoming a growing predictor of a student’s SAT score. Many of these factors are linked to more complicated and troubling issues — including one of systematic racism in the United States and another of evaluating students in such a constraining, one-dimensional way. There is no one simple solution to fixing these large scale issues, nor an easy fix for eliminating such factors affecting student performance on the SAT. So, now, if the SAT reveals such inequity under the surface, shouldn’t we just get rid of it?
Despite these many concerning issues regarding the fairness of the SAT, the exam (nor its alternative, the ACT) is not going anywhere for the current time being. Even if the SAT or ACT were eliminated, standardized tests have become central to today’s education system, from the lowest to highest levels of education (yearly state public school proficiency exams, AP tests in high school, exams needed for professional school such the LSATs, etc.).
It is true that there are many issues with standardized testing; however, standardized tests are currently embedded in the education system in so many ways as a somewhat objective method of ensuring student grade proficiency and of comparing students for admission. Alternate methods for more effectively doing the same tasks have yet to be implemented. Given this assumption that standardized tests are here to stay (for now), how can we reduce and/or eliminate the factors that make it an imperfect method?
Stereotype threat is just one of many factors that adversely affect disenfranchised students of minority groups. And since standardized tests look to be around for the time being, I would like to offer a seemingly simple solution to mitigate this one issue of discrimination within SAT testing — move the section to the end of the test (or here’s another solution — limit the collection of students’ information to online during test registration, which the College Board already does; but for arguments’ sake, let’s just say the College Board wants to keep collecting this information on paper during the examination period).
Why isn’t just getting rid of this section altogether a viable solution? Collecting ethnicity (and gender) data on tests such as the SAT is beneficial as it contributes to studies analyzing test performance. Reports continually show that African American and Latino populations score lower on the SAT than white and Asian students, explicitly proving there is a problem within the United States education system — it does not provide students equal opportunities. Studies can be used positively to understand problems and create solutions.
The same data regarding students’ ethnicities (and genders) currently being collected prior to the test can simply be collected after the test. Since all students taking the test are required to stay for the full length of time (if you’ve taken the test, you know you can’t leave early if you finish your sections faster), the College Board could change the format of their exams so that the informational section — ethnicity section included — be filled out last by students.
The greatest potential cost? Fewer students may opt to share their ethnicity if it is conducted after the test, changing the amount of data available for study. Yet this potential cost seems minor as compared to the potential of minimizing a discriminatory element of the exam.
I have thought long and hard about this potential solution. Honestly, it seems too simple to me. If it is as simple as moving the section to the end of the test or collecting certain information exclusively online during registration, shouldn’t it have happened already? I recognize that it will not eliminate stereotype threat completely, let alone fix larger scale issues associated with standardized testing and inequity within the U.S. education system; however, shouldn’t we be doing everything we can, big and small, to mitigate the potential for unfair treatment according to background?
I invite any discussion or additional information on the topic and would love to hear thoughts and opinions!
