Yale Researchers Develop New SAT Test that Could Better Predict College Success

A revised college admissions test that measures creative and practical skills as well as memory and analytical ability has succeeded in its first round of testing with college students, according to the lead investigator at Yale.

“We did better at predicting college success than the current SAT and high school grade point average, and we did better at reducing group differences,” said Robert Sternberg, professor of psychology at Yale and lead investigator of the study. “In times when affirmative action’s legal status is up in the air, the useful thing about our test is it can help accomplish the same goal as affirmative action without using affirmative action.”

Sternberg’s Center for the Psychology of Abilities, Competencies, and Expertise (PACE) at Yale, a think tank within the Department of Psychology, is developing the test in a study known as the Rainbow Project. The study is funded by the College Board, the nonprofit group that sponsors the SAT.

He said the new test, which he says augments rather than replaces the SAT, also creates much less of a discrepancy in results between various groups. The SAT has long been criticized for favoring white and Asian students over African American and Hispanic students. Sternberg said the SAT also favors traditional learners over others, such as those for whom English is their second language or whose talents lie in creative and/or practical skills.

The test is based on Sternberg’s psychological theory of successful intelligence, which maintains that there are several types of intelligence, while the SATs are not based on any theory.

“What we are trying to do is to link psychological theory to the prediction process rather than ‘let’s just see what works and use that,’” he said. “The new test helps to enhance a psychological understanding of what it takes to be successful in college.”

The new test was administered to 1,007 students at two high schools, five community colleges and eight four-year colleges across the country. The first phase of the study primarily used the data collected from the 793 first-year college students because the researchers were looking at the test’s ability to predict success in college and not in high school. Elena Grigorenko, associate professor of psychology, coordinated much of the testing process.

The exam evaluated students in three areas: analytical, practical and creative skills. The test comprised a combination of multiple-choice questions and performance-based sections. In one section students were shown cartoons and were asked to write captions for them. Students also were asked to write stories based on a list of fictional titles. Other sections asked students to consider a common difficult situation and to rate the best solutions numerically.

In Phase Two of the Rainbow Project, 5,000 to 10,000 students will take the test with the idea of making the test shorter and more effective. The purpose of the subsequent Phase 3 study, which will include an even larger group of students, will be to determine whether the test is commercially viable.

“The test has the potential to reach more students through college admissions, medical school admissions, law school admissions and graduate school admissions,” Sternberg said. “This would be better for the institution because they gain access to students who have the ability to succeed, but were not being identified that way by conventional assessment.”

Science & Technology