Major decisions that affect the trajectory of students鈥� lives hinge on their high school grades: in-class support they might or might not receive, college admission offers, and potentially financial aid.
But how much weight should educators and college admissions officers actually give to the course grades that show up on students鈥� transcripts? After all, researchers have found a mismatch between those grades and scores on standardized tests such as the SAT and ACT鈥攁nd that mismatch appears only to be growing.
Average composite ACT scores decreased from 22.51 to 21.90 from 2018 to 2021 while the average high school GPAs test takers reported rose in that period from 3.48 to 3.59, according to . (The maximum ACT score is 36.)
Discrepancies also exist when comparing student grades and scores from a wide variety of tests explicitly aligned to K-12 coursework, from state standardized exams to Advanced Placement tests. 69传媒鈥� grades aren鈥檛 always higher than their exam scores. Sometimes they鈥檙e lower, said Joe Feldman, CEO of Crescendo Education Group, which works with schools and districts on grading practices.
The findings reinforce a body of research that鈥檚 consistently shown grades to be inconsistent and unreliable measures of student performance. Grading methodology varies from teacher to teacher and grades themselves tend to include a mix of non-academic factors, such as student behavior, class participation, and extra credit. Plus, grade inflation has picked up since the start of the pandemic.
The ACT and the College Board advise college admissions officers to consider both grades and test scores when making admissions decisions rather than basing decisions on a single factor. Feldman, meanwhile, calls for more training for teachers on grading鈥攐ften a topic neglected during teacher-training programs鈥攕o grades more accurately reflect students鈥� understanding of course content.
鈥淲hen teachers are given the opportunity to learn how to grade more accurately, fairly, and in our words, equitably, what happens is that the grade-test consistency increases, which means the grade is more likely to match the test score across all students groups, and particularly among historically underserved groups,鈥� Feldman said.
Researchers identified discrepancies between grades and test scores
Researchers with the ACT, the nonprofit that runs the eponymous standardized test used for college admissions, reviewed the ACT composite scores and self-reported GPAs of public school students who took the test from 2017 and 2021 and subsequently enrolled in a public college in their state.
In a paper released Sept. 25, they used both to predict students鈥� first-year college GPAs. They then checked to see how those predictions fared against their actual first-year college GPAs.
鈥淲hen you examine how these trends correspond with first-year college GPA, our research demonstrates that after the onset of the pandemic, high school GPA has become a less consistent predictor, while ACT composite scores have shown relative stability in predicting first-year GPA,鈥� Edgar I. Sanchez, a lead research scientist at ACT and author of the report, said in a news release. 鈥淚n light of these findings, we must reevaluate the predictive power of high school grades and use a more holistic approach for predicting student success, so students have an accurate assessment of how they鈥檙e likely to perform in their first year of college.鈥�
The College Board, meanwhile, in late September that for the third consecutive year, the average total SAT score declined, to 1024 (out of 1600) for the class of 2024 from 1028 the previous year.
In noting that several colleges have recently reinstated standardized test requirements for admissions, the College Board said SAT scores are better predictors of college students鈥� performance than high school GPAs.
Kaitlyn Hughes, a tutor with IvyWise, a for-profit college admission consulting firm, has seen the growing discrepancy between SAT scores and grades firsthand in recent years.
鈥淚鈥檓 receiving students who have phenomenal grades, but they鈥檙e really lacking in their fundamental understanding, especially in math skills,鈥� Hughes said.
She tutored a student with all A鈥檚 who was struggling to raise her SAT math score because she completed her geometry course virtually, early in the COVID-19 pandemic.
Schooling disruptions, whether tied to pandemic closures or even long-standing challenges such as teacher turnover and vacancies, continue to affect students鈥� ability to learn the skills needed to get high scores on the SAT, which more higher education institutions are now relying on in admissions decisions. (A number of colleges .)
Feldman acknowledged that standardized tests are imperfect, reflecting disparities based on students鈥� socioeconomic status and measuring students鈥� performance based on a single-day event.
Yet colleges and universities are questioning how to weigh grades against these test scores due to serious discrepancies.
In a published this summer, Feldman and his team compared the grades students received in their second semester with the scores they received on the standardized test created to match that course鈥攚hether an AP test or required state exam. They did this with student data from multiple states and grade levels, and a mix of regular and advanced courses.
They found that nearly 60 percent of the grades did not match the student standardized-test score. Approximately two-thirds of those grades were inflated, but about a third were depressed.
鈥淲e鈥檙e making [college admissions officers鈥橾 job harder by having grades be so variable and, more often than not, inaccurate,鈥� Feldman said.
Solutions require training for teachers and college admissions decisions based on a variety of criteria
The ACT study found that using high school GPAs and ACT composite scores together improved predictions of first-year GPAs over using either measure alone. The College Board also calls for SAT scores to be a part of, not the sole factor in, admissions decisions.
But Feldman urges educators to rethink grading. This is especially necessary to address grade depression, when students understand course content, as demonstrated by their test scores, but teachers aren鈥檛 giving them proper credit through their grades, Feldman said.
The solution involves training on grading for aspiring teachers in schools of education, Feldman said. Districts need to offer such training as well. Grades shouldn鈥檛 incorporate non-academic factors from the classroom, such as attendance and behavior, Feldman said.
Families and students also need to hear about why reimagining grading for consistency鈥檚 sake is a net gain.
鈥淚f people see a greater urgency to address these deep flaws in grading that actually can be fixed and repaired with teacher training and knowledge, and if we have the will as professional educators to prioritize improving grading to be more accurate and fair, we can actually reduce the need for standardized tests because of their inherent flaws,鈥� Feldman said.
