Search News


Browse Archives

News

Matching Tests to Their Purposes

April 12, 2007

Share This Story

FREE Daily News Alerts

Advertisement

College readiness and placement tests provide snapshots of what high school students learn, but states need to go beyond these indicators in order to get the full assessment picture, according to a new report from Achieve, a nonprofit organization whose focus is on K-12 assessment and accountability.

"Aligned Expectations: A Closer Look at College Admissions and Placement Tests" begins with the premise that states need to adopt tougher and more coherent systems of measuring student performance. "Rigorous standards are an important first step, but often it is the tests that dictate what gets emphasized in classrooms," the report says.

So researchers at Achieve, a group created by governors and business leaders, looked at more than 2,000 questions from college placement and admissions tests to see how well the tests measure students' preparation for college, based on benchmarks set by the American Diploma Project Network, a coalition of 29 states seeking to align high school coursework with the demands of college work.   

According to the Achieve review, the majority of college placement tests are narrowly focused on a subset of knowledge and reflect relatively low levels of rigor. "If states were to incorporate existing placement tests into their formal high school accountability systems, it might inadvertently lead to a narrowing and watering down of the curriculum," the report says.

It adds that while the ACT and SAT are credible measures of student learning, states should use caution when thinking about incorporating the tests into their assessment and accountability systems. Some states are using the placement tests as the official statewide high school exit exams, while others are debating whether students, while still in high school, should take college placement tests as a form of early feedback.

These tests don't fully measure college preparedness and in particular fall short in measuring more advanced concepts and skills, the report notes. Thus, states should create additional questions when relying heavily on these national tests to gauge how students measure up in a given state.

The report advocates the use of end-of-course tests that can be tied to specific curriculum to measure "higher level" content and skills. "They also are more sensitive to instruction because they are taken right after a student completes a course, and they allow states to monitor performance and ensure consistency of rigor across the state," according to the report.

It also calls on colleges to define what skills and knowledge high school students need to succeed at the next level, and to monitor state assessment tests in the high-level subjects such as Algebra II to ensure that they accurately measure college-level work.

Generally, the report found that admissions tests are more demanding than placement tests and better balanced in the types of questions asked. Here's the breakdown:

Reading. Admissions tests are more rigorous than placement tests here, according to the report, although the reading passages on placement tests more accurately reflect the types of reading material students will encounter in college across the disciplines.

Writing. Both kinds of tests are more rigorous than most high school tests and generally reflect the kind of writing students will be asked to do in college.

Math. Both types emphasize algebra, but the content measured tends to favor pre-algebra and basic algebra over the advanced concepts that are needed for college readiness. Although placement tests are narrowly focused on algebra, admissions tests are generally broader, measuring a broad range of topics such as data analysis, statistics and geometry, according to the report.

See all postings »
Advertisement
Advertisement

Matching Jobs

Comments on Matching Tests to Their Purposes

  • A good title and common sense
  • Posted by Duncan on April 12, 2007 at 10:00am EDT
  • It is a good title and also a common sense. However, it's easier to say than done. If that is to be achieved, I suppose we will need a lot of tests that cover different topics and levels. And when interpreting the test results, people may need to look at sub-scores.

    I, in general, prefer tests be designed by professionals. Of cause, there can have a lot of professional companies that offer options. I admired the way these professionals in balancing topics and keep score statistically consistent.

    These professionals have very difficult goals to achieve. The admission test need cover a lot of ground in order to give accurate assessment. With the amount of time given in current ACT/SAT tests, the number of question is limited and, no wonder, it may not give a good assessment for individual cases. But. On average, it may give an acceptable picture.

    When I grew up, I took admission tests that take much longer to complete - it usually take couple of days to finish all topics and which helps cover more grounds. However, we don't have the consistence that ACT and SAT have and it can really happen that the scores isn't really reflect a person's total ability. I, however, considered it is fare.

  • Assessing College Reading Demands
  • Posted by Nic Voge at UC Berkeley on April 12, 2007 at 2:42pm EDT
  • I've looked at the report by Achieve and cannot find any indication to HOW the reading demands and expectations across the curriculum in colleges and universities were determined. There is no description of a methodology for either acquiring data or the nature of the analysis that leads to their conclusions. Nor can I find any sources among the document's references that appear to include any studies that would yield data or conclusions of this kind. Being somewhat familiar with this literature, I submit that there does not exist a very comperhensive and nuanced understanding based upon an empirical analysis of college reading demands and expectations across the curriculum. (But, if someone knows of such studies, please refer me to them.) I can only assume that the authors are relying upon their impressions or personal experience. This would appear to be a rather significant limitation of this study.

    If we hope to prepare students to meet the reading and learning with text demands of post-secondary education--not to mention measure students' preparedness-- it seems essential that we actually know what they are.