1
The Use of Standardized Testing in Admissions:
Summary of Key Findings
April 2024
Introduction
In April 2020, Cornell suspended the requirement for SAT/ACT exam
scores for undergraduate applications for admission. This was largely a
response to the logistical difficulties accessing testing locations with the
onset of the COVID pandemic, but the decision also eased some concerns
that the flaws that inhere in standardized tests undermine some of
Cornell’s admissions goals.
As shown in Table 1, four of the undergraduate colleges adopted “score-
free” policies (where test scores that have been provided to Cornell by the
testing agencies are not uploaded into the admissions data reviewing
system
1
), and the remaining colleges have “test-optional” policies (where
official test scores are easily available to the reader and may inform the
admission decision).
In the spring semester of 2023—following two cycles of fall first year
admissions—the Provost charged a Task Force on Testing in Admissions
to look at the impact of removing the testing requirement by considering
the following questions:
• Does the submission of test scores alter admission chances in the test-optional
colleges?
• Do test-optional or score-free admissions policies impact the socioeconomic or
racial/ethnic composition of admitted students?
• Is there evidence that the use of test scores enhances student outcomes at
Cornell?
The Task Force delivered a confidential report in May 2023 that
summarized findings relating to the matriculating cohorts of Fall 2021 and
Fall 2022.
1
If an applicant self-reports scores on the Common Application, those scores are viewable through a PDF
of the application materials in both “score-free” and “test-optional” colleges. In the score-free colleges,
reviewers are asked not to consider those scores.
2
At the request of the Provost, these analyses were updated in the spring of
2024 by the office of Institutional Research & Planning. The updated
findings, described below, reinforce and extend the more tentative
conclusions from the earlier report.
Notable findings from both rounds of analyses include:
• There is not a clear indication that the relaxation of the testing
requirement has increased the diversity of matriculating first year
students.
• In Cornell’s “test optional” colleges (where applicants test scores
can be considered in the review process), the submission of test
scores has a substantial and statistically significant impact on the
chances of being admitted.
• Students admitted with known test scores had better academic
outcomes than students who were admitted without test scores.
Opting to provide test scores
Official test scores are provided to Cornell by the testing provider at an
applicant’s request.
2
As shown in Table 1, a minority of applicants had a
test score sent to Cornell. In Fall 2023, just 24% of Cornell’s undergraduate
applicants provided an official score.
Table 1. Percentage of Fall freshman applicants submitting an SAT or ACT score
Applicants Acceptances Enrolls
Test
policy
College
Fall
2021
Fall
2022
Fall
2023
Fall
2021
Fall
2022
Fall
2023
Fall
2021
Fall
2022
Fall
2023
Test
optional
A&S
31%
28%
24%
53%
52%
50%
67%
63%
62%
Engineering
35%
34%
30%
57%
64%
61%
71%
77%
75%
CHE
42%
33%
28%
52%
53%
45%
65%
63%
54%
ILR
32%
28%
23%
69%
43%
43%
74%
47%
46%
Brooks
3
23%
43%
53%
Score free
CALS
29%
24%
20%
38%
23%
21%
39%
24%
22%
AAP
26%
22%
18%
39%
26%
17%
40%
26%
17%
JCB-Dyson
27%
19%
12%
39%
18%
11%
40%
18%
11%
JCB-Nolan
22%
18%
18%
30%
21%
20%
30%
21%
18%
Total
University
31%
28%
24%
50%
44%
42%
59%
50%
48%
2
The request to have test scores sent to an institution may occur well before submitting a college
application, such at the time of taking the test. This may explain why some students had their scores sent
even when they [later?] applied to one of the score-free colleges.
3
The Brooks School of Public Policy became an admitting unit during the period of interest.
3
While Cornell did not receive test scores from most applicants, data from
the Fall 2022 administration of Cornell’s New Student Survey
4
indicate that
91% of matriculating first-years took either or both the SAT and/or the
ACT. Indeed, 70% had taken the SAT (or the ACT) test multiple times.
When test scores are an optional part of the admissions process, the
decision to provide a test score is a question of strategy.
5
Presumably,
many applicants guessed that providing their scores would hurt their
chances of admissions. Indeed, self-reports of test scores on the survey
indicate that those who did not submit their SAT scores to Cornell tended
to score lower than those who did (see Figure 1). Notably, it appears that
some applicants with perfect and near perfect scores did not submit them.
Figure 1. Self-reported SAT scores of Fall 2022 New Student Survey respondents, by whether scores were
submitted to Cornell by the test agencies
4
This survey of first-year students had a 79% response rate and ask students how many times, if any,
they had taken the SAT and the ACT. It also asked students to self-report their scores. The comparison of
self-reported test scores with official test scores when possible indicates that the score self-reports are
accurate.
5
In test-optional colleges, there has not been a way for the applicant to have test information suppressed
on the application if it has been provided to Cornell already. Thus, it is possible that some applicants
provided test scores to Cornell somewhat unintentionally.
4
Whether it is “strategic” to submit a score with one’s application may not
be immediately apparent to high school students: teenagers with
differential access to the resources that may help them navigate the
question to their best advantage.
Data from the 2022 New Student Survey suggest that students’ decisions to
share test scores are shaped by social background factors such as the type
of high schools they attended, their family incomes, and their access to
and use of guidance counselors. Among students who scored above 1400
on the SAT, for example, Black students were less likely than White and
Asian students to have submitted their test scores: 62% of Black students
versus 74% of White students and 79% of Asian students with these high
scores sent them to Cornell.
To the extent that students from different kinds of backgrounds are
differentially deciding to withhold scores that are strong enough to help
them gain admission to Cornell, test-optional policies may undermine
equity in admissions.
Access
Because standardized test scores are systematically correlated with
indicators of privilege, some have anticipated that removing the test
requirement could facilitate access for applicants who are low income;
Black, Hispanic, and/or Indigenous; and those who are the first in their
family to attend college.
Others have argued that test scores read in context can buttress the case
for admission among applicants who may not have the clear signals of
strong application (such as coming from a known and well-regarded high
school, informative letters of recommendation, and AP exams). Read with
an appreciation for context, an applicant with a test score that may be
below the average for Cornell students but that is well above average for
their high school may be considered a desirable admit. Test scores enable
those types of decisions.
On the whole, it does not appear that the shift in Cornell’s testing policy
has played a major role in diversifying first year students by
race/ethnicity/citizenship, first generation status, or family income (see
Figure 2). While there are some very modest shifts between the Fall 2020
5
cohort (that submitted required test scores) and the cohorts that followed
(that did not), the movement towards increasing representation of less
advantaged groups has been in process for several years. For example. the
percentage of first year students who are Black, Hispanic and/or
Indigenous (BHI) increased from 23% to 27% in the five years preceding
the pandemic. In Fall 2021, that percentage increased further to 28%, but
has declined very slightly to 25% in Fall 2023.
Perhaps the strongest evidence of a shift is with respect to first generation
status (see middle panel of Figure 2) where the percentage of the class that
is first generation increased from 16% to 19% between 2020 and 2021 and
has remained at that level in the subsequent matriculating cohorts.
However, the proportion of the class that is first generation had been
increasing prior to the policy shift as well, so too this could be considered
as a piece of a broader trend.
If the policy shift away from testing has substantially enhanced access to
Cornell, it has done so in ways that elude simple measurements.
Figure 2. Composition of first-time first-year students by matriculation semester: race/ethnicity/citizenship,
6
first
generation college student, and financial aid status
6
Students who identify as Black, Hispanic, and/or Indigenous are grouped here as “BHI.”
6
Chance of admission
As shown in Table 1 above, less than a third of applicants submitted test
scores since the test requirement was removed. However, more than 40%
of our acceptances and about half of matriculating freshmen had provided
scores.
This difference could reflect the fact that, on balance, students who submit
test scores are stronger applicants on other dimensions. That is, they
might have stronger academic records in high school, stronger letters of
recommendation and so forth.
In test-optional colleges—where the difference between applicant and
admit submission rates is much larger—it could also result from
admissions officers being more likely to admit students when the officers
could see the test scores.
Regression models that estimate the probability of admission holding
constant several other factors—including high school GPA and additional
student and high school characteristics—suggest that submitting test
scores significantly increases the likelihood of admission in the test-
optional colleges but has little or no effect in the score free colleges.
The differential impact under the two policy scenarios suggests that
admissions officers are using the test scores to inform decisions when they
are available. Given this finding, it seems prudent for those applying
under the test-optional policy to send in their test scores.
Academic outcomes
Previous studies at Cornell and elsewhere have found that SAT scores are
significant predictors of first-year GPAs.
The vast majority of students who have matriculated at Cornell—with or
without known test scores—have performed well here. However, those
who were admitted without test scores tended to have somewhat weaker
semester GPAs, were more likely to fall out of “good academic standing,”
7
7
To maintain good academic standing, a student must successfully complete at least 12 academic credits
and have a GPA of at least 2.0 each semester.
7
and were less likely to re-enroll semester after semester. These patterns
hold true holding constant students’ high school GPAs as well as other
personal and high school attributes.
There is no evidence that these differences have diminished across cohorts
of matriculants; the gap in first semester GPA has remained consistent for
all three years of new admits.
The association with GPA may be attenuating somewhat as students
accumulate more semesters of experience at the university. That is, is the
gap is smaller in the third semester than it is after the first semester. This
is encouraging for those students who have persisted, but the robust
evidence of increased rates of academic struggle and attrition remains a
concern.
The analyses of outcomes are consistent in suggesting that when
admissions officers have test scores available to them as additional
information in a holistic admissions process, they are able to use them in a
way that supports positive outcomes for Cornell students.
2023 Task Force Members
• Marin Clarkberg, Chair & Associate Vice Provost of Institutional
Research & Planning
• Jon Burdick, Vice Provost for Enrollment
• Scott Campbell, Director of Admissions in the College of Engineering
• Jason Kahabka, Associate Dean for Administration in the Graduate
School
• Alan Mathios, Professor in Public Policy and Economics
• Lisa Nishii, Vice Provost for Undergraduate Education and Professor
of Human Resource Studies
• David Shmoys, Professor of Operations Research and Information
Engineering
• Michelle Smith, Senior Associate Dean for Undergraduate Education
in the College of Arts & Sciences and Professor of Ecology &
Evolutionary Biology
