Rethinking standardised testing to end discrimination

In a shot heard around the United States, on 21 May 2020 the University of California’s Board of Regents suspended the requirement and use of standardised tests, including the SAT and ACT, for freshman applicants.

The University of California (UC) will be test optional for campus selection of freshman applicants in autumn 2021 and 2022 and “beginning with fall 2023 applicants and ending with fall 2024 applicants, campuses will not consider test scores for admissions selection at all, and will practise test-blind admissions selection”.

The Regents, along with some 1,200 other universities and colleges, had previously dropped the requirement for 2021 following the College Board’s and ACT’s cancelling of testing due to the COVID-19 pandemic.

The Regents have requested that the Academic Senate and the university’s administration attempt to develop a new test or adopt the existing Smarter Balanced test of high school students in time for the autumn 2025 entering class that better aligns with college readiness.

But if they fail in this endeavour, UC will eliminate the standardised testing requirement for California students.

Opponents of the widespread use of the SAT have long claimed that it promotes needless socio-economic stratification: the test favours students from upper income families and communities, in part because they can afford a growing range of expensive commercially available test preparation courses and counselling. But most significantly, grades in high school are a better predictor of academic success at UC than test scores.

The Regents’ 2020 decision echoes this view.

Yet as I chronicle in a new research paper published by my centre at UC Berkeley, UC has a long history of concern with standardised testing. In fact, UC was relatively slow in adopting the SAT as a requirement in admissions when compared to other universities with selective admissions, public or private.

Informed by this history, I offer a few observations on the Regents’ May 2020 decision and its global implications for high-stakes testing.

Political dimension not new

First, as the value of higher education increases for the individual, and for society in general, the difficulties of allocating a scarce and highly sought public good, admission, grow more intense for selective universities. Because there are generally conflicting interests in setting and influencing admissions policy at selective public universities such as the University of California, policy-making has an inherently political dimension.

Determining admissions criteria is not simply a rational choice; it is, in some form, a reflection of the internal and external politics that shape the policy behaviours of a university.

In the case of the University of California, requiring the SAT is part of a larger set of admission requirements that, over time, policy-makers adopt or modify to fit perceived institutional goals and oftentimes in reaction to the concerns of major stakeholders. UC first adopted the use of the SAT for admissions in 1979 as a counterbalance to grade inflation in California high schools.

Arbitrary decision-making?

Another axiom that is largely lost in the debates over the usage of test scores and a growing array of admissions requirements: Highly selective public universities may attempt to create relatively transparent admissions criteria, but in the end much of the decision-making is arbitrary.

For example, UC Berkeley, before the COVID-19 pandemic, received approximately 87,000 applications for the 2019-20 academic year, almost all of which were UC Eligible, a majority with 4.0 grade point averages or GPAs (inflated by Honors and Advanced Placement courses).

Yet only 14,600 or so applicants will be accepted by Berkeley. Some 60% of those admitted will enrol elsewhere, with a net enrolment target of around 6,500 students. That means rejecting some 73,000 generally highly qualified and talented students, who would statistically do very well at Berkeley.

When you have this ratio of talented and accomplished students asking for entrance to Berkeley, or to other highly selective university campuses, there are going to be arbitrary outcomes no matter how rational the admissions policies appear to be. That conclusion can be extrapolated beyond the borders of the US.

Redistributing a highly sought public good

Third, the intent of this change in policy is to provide greater access to underrepresented groups. Translated, that means an opportunity to redistribute what is essentially a zero sum: access to a selective public university that does not have the finances to grow significantly in enrolment and has a mandated limit under the California Master Plan for Higher Education to accepting students from the top 12.5% within the state’s high school graduating class.

The unanimous action of the Regents was justified largely because of the claim that standardised tests discriminate against underrepresented minority and low-income students at UC.

While the intent of the Regents’ verdict is to boost underrepresented groups, specifically Chicano/Latinos and African Americans, it will require less representation among ‘overrepresented’ groups, specifically Asian Americans who generally have benefited from the use of SAT scores in the campus selection process.

Berkeley, for example, desires to increase its Chicano/Latino population, stating an aim to become a Hispanic-Serving Institution by 2028 – a federal designation in which Chicano/Latinos represent 25% or more of the total enrolment at a college or university. In autumn 2019, Berkeley enrolled a total of 43,204 students, of which 5,855 (or 13%) were Chicano/Latino.

Whether at Berkeley or in the entire UC system, and by virtually any measure, Asian-Americans (a broad category with significant variation in socio-economic background) are significantly ‘overrepresented’. An anticipated decline in international students might provide more enrolment room for underrepresented groups.

But one might speculate that dropping standardised tests in determining UC eligibility and campus admissions will increasingly favour the state’s largest underrepresented group. Chicano/Latinos currently represent 39% of California’s population and are projected to be over 47% by 2050. Why else pursue ending the SAT at UC?

Global implications

Throughout the world, high-stake national exams determine who gets access to public universities. In places as diverse as China, South Korea, Japan, Brazil, most of Europe, Russia and elsewhere, students spend much of their secondary school careers preparing for just such a test, their fate often determined in a single sitting.

However, there is little research, outside of the US, on the actual predictive value of these tests regarding a student’s future academic performance or how high-stakes testing positively or negatively influences student learning in secondary school.

Test like China’s gaokao, a reformed senta shiken in Japan, the CSAT in South Korea, the ENEM in Brazil, the Baccalauréat in France and the UTME in Nigeria, to name only a few, use standardised testing scores as the primary way to award access to the best universities.

Yet the appropriate question is what is their predictive validity when compared to other factors and ways to evaluate the ability of a student to succeed at the university level, promote a diversity of talent and service the larger socio-economic mobility needs of society?

The debate over the SAT at UC focused on just these issues and perhaps provides a window for a more analytical approach to the use of standardised testing in other parts of the globe.

In societies with high levels of corruption, national standardised testing offered a seemingly equitable way to provide access to the best national universities.

But more and more countries, and universities, are beginning to evolve their admissions policies and practices to seek other indicators of academic promise and societal impact. In doing so, they are also recognising that talent, academic or otherwise, comes in many different forms that are not captured in testing alone.

Perhaps UC’s decision should be a shot heard around the world.

Next week in University World News: Pat Leonard responds in an article on ‘why standardised testing is necessary’.

John Aubrey Douglass is a senior research fellow and research professor in public policy and higher education at the Center for Studies in Higher Education, Goldman School of Public Policy, University of California, Berkeley, United States. He is the author of The New Flagship University: Changing the paradigm from global ranking to national relevance (Palgrave Macmillan) and Envisioning the Asian New Flagship University (Berkeley Public Policy Press). He is the founding principal investigator of the Student Experience in the Research University (SERU) Consortium based at Berkeley. See also his ‘Missing Links’ series of University World News articles: fixing the disjuncture of university missions and faculty hiring and promotion policies and the need to refocus institutional research from rankings to university needs.