David K. Scott was Chancellor of the University of Massachusetts Amherst, 1993-2001.
This is an archive of the Chancellor's Web site during his tenure.

UMass Office of the Chancellor

Put to the Test:
Teacher education faces greater accountability standards

by David K. Scott

Even though U.S. policymakers have tried for more than 15 years to improve academic performance in public elementary and secondary schools, test scores show that many high school graduates continue to lack the necessary skills for either college or work. The US public's growing concern about the quality of K-12 education is pressuring colleges and universities to strengthen their teacher?training programs.

In efforts to better prepare teachers,

  • US Secretary of Education Richard Riley has asked the higher education community to give increased priority to educating the nation's future teachers;

  • the US Congress has directed the Education Department to issue guidelines that require all colleges and universities to publicly disclose how well their graduates perform on state teacher?licensing exams; and

  • the American Council on Education has called on college and university presidents to take the lead in reforming the way campuses train education majors.

With estimates that the nation will need to recruit 2.2 million teachers in the next decade to accommodate anticipated school enrollment increases, higher education institutions can expect the focus on teacher training and qualification testing to continue to intensify.

A testy point

Whatever our view of teacher tests, they are a reality. In 1982, only 18 states mandated some form of teacher testing for certification. By 1998, only eight states did not.

While it may seem reasonable to test prospective teachers for professional competency—after all, we test medical and legal professionals—critics of current practices frequently point out that a professional organization does not administer the teacher exam, as is the case in medicine or law. Instead, commercial testing companies, which employ various sets of benchmarks, vie to have their products accepted by state governments

To ensure the development of more uniform standards, Secretary Riley has called for the National Academy of Sciences to develop national benchmarks for teacher testing.

A test case

It seems reasonable to expect that all future teachers should be proficient in basic literacy and communication skills as well as in the subject matter of their specialization. This was the focus of the tests that Massachusetts first administered in April 1998, yet 60 percent of the 1,800 undergraduate and graduate test takers, failed the cumulative three-part test.

The result led to public accusations about the test, including allegations that 10th graders could pass it and therefore Massachusetts was failing miserably at educating teachers.

Actually, experts found the test to be appropriately designed for university level students. The problems seem to have arisen in the administration of the test and in the criteria used for determining a passing grade.

Media commentary continues to focus on test takers' egregious failures in literacy and spelling. On any test there will invariably be some dismal performances, but one should differentiate between this group and the larger number of people who come close to passing. The score distribution of the Massachusetts test indicates that a large percentage of failing grades were just below the passing line, raising the question of whether the bar was set too high.

To illustrate, suppose 100 prospective teachers, based on their knowledge and skills, are exactly at the passing score for each of the three parts of the test. On any given day, because of measurement error, we can predict that only half of them will pass the first part of the test, Of the 50 who pass, only another 25 (half) will pass the second part. Of that 25, only 13 (half) will pass the third part. Because one must pass all three parts, the result is a meager 12 percent pass rate, or an 88 percent failure rate.

In Massachusetts, the policy makers chose to set the bar at an absolute level, and at a high level. Certainly this approach is fail-safe by passing only very well-prepared candidates, but it also rejects larger numbers of people who are right on the border and probably deserve to pass, and it reflects unfairly on our education systems.

When analyzing teacher test results, we need to identify those who deserve to fail, those who fail because of measurement errors, and those who fail because the bar was set too high and/or because a conjunctive passing score was required.

Post-test strategy

A more rational approach to testing would be to require that all education students pass basic literacy and communication tests before entering teacher preparation programs. This competency would not be the sole responsibility of schools of education but also of the greater institution. As such, most institutions would need to develop better integration among their colleges or departments as well as improve their general education design.

Once students pass this threshold and enter teacher education programs, say in their junior or senior year, they would then work to become competent in pedagogical and instructional techniques to ensure that they can function in the complex dynamic of the modern schoolroom.

Test questions

Following are other testing concerns worth examining.

  • Pedagogical competency. Interestingly enough, most of the current tests do not measure instructional knowledge and skill.

  • Quantitative competency. Current tests pay no attention to quantitative competency either, except for the subject-matter tests for prospective mathematics and science teachers. Many other subjects require sound quantitative and statistical judgment to distinguish sense from nonsense and fact from fiction. The failure to instill quantitative sense in all aspects of education results in poor thinking. In fact, it may account for the policymakers' failure to correctly analyze the numerical and statistical aspects of the very tests we are discussing!

  • Diversity. The dearth of education students of color is a looming problem in teacher preparation. Of 3,500 candidates in the first three administrations of the Massachusetts test, only 2 percent (65 students) declared themselves to be African American, and another 2 percent to be Hispanic American. At a time when Massachusetts is becoming more diverse (25 percent of K-12 students are of color), and when the United States is heading to 50 percent people of color, there is a desperate need for teachers of color to serve as role models,

To compound the problem, while the overall pass rate on the Massachusetts test was about 70 percent overall, only 29 percent of African-American—and 40 percent of Hispanic— American candidates passed, thereby further depleting the potential cohort for the future.

We may design all the tests we want; we may raise college entrance requirements for prospective education majors; but unless we overcome this racial and cultural disparity in our educational system, the educational outlook for Massachusetts and the nation is much more dismal than all the test results.

Campus Updates
About the Chancellor

PowerPoint Presentations

Chancellor's Report

Strategic Action
Campaign UMass

 Office of the Chancellor
 Whitmore Administration Bldg.
 University of Massachusetts
 Amherst, MA 01003

excellence within your reach
This Web site is an Official Publication of the University of Massachusetts Amherst Campus. It is maintained
by the Office of Communications and Marketing.
Copyright 2000.