Facebook Twitter

What do test scores predict?

Lots of people from legislators to students like to dismiss the value of Colorado’s annual high school exams, but a new report suggests the test results may be useful as an indicator of who’s more likely to stay in college.


The report found that scores on the 10th-grade math CSAP tests were almost as good an indicator that a student will continue in college as student ratings on the admissions index that colleges use when considering applications.

The higher a student’s CSAP result, the more likely the student will complete at least 30 credit hours of college work – roughly a year of higher education.

“What I’m showing is it’s related to persistence,” said Robert Reichardt of R-Squared Research, who wrote the report for the Colorado Department of Higher Education.

“The admissions index is a slightly better predictor than the CSAP,” he said, but test scores “may be an appropriate thing to consider in admissions policies.”

Reichardt compared 10th-grade CSAP math scores and admissions index data for Colorado high school graduates who attended state colleges from 2008 through 2011. “Persistence” was defined as having completed 30 credit hours of college work. The admissions index includes student scores on ACT or SAT college entrance exams and high school academic performance as measured by grade point average or class rank.

Reichardt said he found two other conclusions of his research significant.

“I think the most significant finding of the report is the consistently lower persistence rates” for males, Hispanic and low-income students despite their CSAP scores, he said.

The lower persistence rates for low-income students were less surprising, given college costs. He’s concerned about lower persistence rates for males: “The problem of males in education in general is not one that’s being talked about a lot.”

“The second thing is … there are still huge variations between institutions in persistence,” even when student scores and indexes are comparable, Reichardt said. “I don’t know why they’re different.”

Reichardt stresses that the report has limitations, including its focus on a traditional cohort of college-going students, and that further research is needed on broader groups of students and on what other factors affect persistence.

“Our universities are dealing with a huge diversity” of students, Reichardt noted, and for that reason, more research is needed on what affects persistence for a wide variety of students.

The report was designed to address such questions as whether CSAP scores predict attendance at two-year or four-year schools, whether scores predict persistence at both levels and to compare test scores and the admissions index. The group studied totaled about 42,000 students. Their persistence rates were 60 percent at two-year schools and 81 percent on four-year campuses. Here are some key findings:

  • Hispanic and low-income students with CSAP scores equivalent to those of the entire group studied were less likely to attend four-year schools.
  • Male, Hispanic and American Indian students were less likely to persist at two-year schools, while black and white freshmen were equally likely to persist.
  • At four-year schools, persistence was lower for males and all ethnic groups.
  • Among the four-year schools, Colorado Mesa, the School of Mines and the University of Colorado-Boulder had the highest persistence rates, while Metro State, Colorado State-Pueblo and Adams State had the lowest for students with comparable CSAP scores.

Reichardt presented the report at a recent meeting of the Colorado Commission on Higher Education. The document is part of a series of studies the state Department of Higher Education has commissioned to inform development of a new master plan for Colorado’s public colleges and universities.

“CSAP” is common verbal shorthand for Colorado’s annual achievement tests. But the last CSAPs were given in the spring of 2011, and students will take the transitional TCAP tests through at least 2014. Officials then hope to roll out a permanent set of replacement tests. Learn more.

That plan, required by a 2011 law, will become the frame for performance contracts with individual institutions, contracts that may eventually be tied to funding.

The state has an aspirational goal to double the number of degrees issued by 2020. Some 44,908 degrees and certificates were granted in 2011, the most recent data available.

The 2012 legislature passed multiple bills intended to increase completion, including measures making it easier for students to combine community college and four-year credits to earn associate’s degrees, to improve remediation programs and to allow adults to earn credit for past professional experience.

It’s widely agreed that Colorado can’t double degree and certificate production just by increasing enrollment of traditional-age college students.

Some legislators and other policy makers have been critical of high school tests, arguing students don’t take them seriously because the results aren’t considered for college admissions. There have been unsuccessful legislative attempts to reduce high school testing, which is done in the 9th and 10th grades. All Colorado 11th graders have to take the ACT test.

Legislation passed earlier this year encourages school districts to give the Accuplacer tests in high school. Supporters of those tests argue that high schools can use the instant results to help students fill academic gaps before they get to college, reducing the need for remediation.