Researchers Find Teacher Tests Short on Covering College Content
Teacher-licensing tests fall far short of measuring the knowledge that teachers need to bring students to high levels of achievement, a report released last week concludes.
"Not Good Enough: A Content Analysis of Teacher-Licensing Examinations" reports on a study conducted by the Education Trust, a Washington-based group that advocates higher academic achievement for all students.
The study focused on examinations of subject matter and general knowledge provided by the Educational Testing Service of Princeton, N.J., and National Evaluation Systems of Amherst, Mass., the nation's two major publishers of teacher examinations.
For More Information
|"Not Good Enough: A Content Analysis of Teacher-Licensing Examinations'' is available for $2.50 from the Education Trust, 1725 K St. N.W., Suite 200, Washington DC 20006; (202) 293-1217.|
Teams evaluated exams in English language arts, math, and science, using the highest-level tests now administered in states. Tests of pedagogical skills and knowledge were not evaluated.
The results, the study found, were generally disappointing. Most of the tests were judged to be handled easily by advanced high school students.
"We found no evidence of content at the baccalaureate level," the report says. "Not one test was up to the level of a graduating college senior."
The finding is troubling, the authors argue, because teachers need knowledge of their subject matter far beyond the K-12 level if they are to help students acquire true understanding of those disciplines.
"That we dramatically ratcheted up our standards for students without insisting on commensurate increases in standards for teachers is a chilling indictment of all of us," Kati Haycock, the director of the Education Trust, said in releasing the report.
In addition to those "low-level exams," the report says, individual states set varying passing scores. In some states, candidates can accurately answer as few as 45 percent of the questions on a test and still pass.
The study looked at the PRAXIS series produced by the ETS and state-specific examinations created by NES. Although more states use PRAXIS, a larger number of students take NES exams because they are given in populous states such as New York and Texas.
Seven states have no examination requirements for elementary-teaching candidates, while the remainder require exams that cover pedagogy and "rudimentary general knowledge and skills," the study found. In 44 states, candidates for secondary-teaching licenses must take exams, but only 29 require that they be in the subject area to be taught.
The Educational Testing Service granted a team "controlled access" to its tests, which allowed them to work through the tests just as teacher-candidates do. The team, made up of two Education Trust staff members and three outside consultants, classified each item on the tests according to the approximate grade level when the material covered would likely be taught; how challenging the question was (simple, moderate, or complex); and whether the knowledge tested was relevant to teaching.
Those findings were then validated by a national review panel.
In the case of the NES examinations, which are controlled by state agencies rather than the testing company, the study's conclusions are based on a review of complete examinations from one unnamed NES state and study guides from six others.
For elementary teachers, many states use PRAXIS I, a basic-skills test designed to screen candidates for entry into teacher-preparation programs. The ETS has a three-part series of tests for elementary teachers that includes both content and pedagogy, but most states don't use it.
Kudos for Massachusetts
One "bright spot" was the series of essay tests published by the ETS, which the report says requires candidates to demonstrate their depth of knowledge. But the essays are required by far fewer states than the lower-level multiple-choice tests.
The team was also impressed by the sample items from Massachusetts' literacy and communication-skills test, produced by National Evaluation Systems. Its questions were "of a higher degree of complexity and expectation than any of the others we looked at," the authors write. The test has generated fierce debate in the state because so many candidates have failed it.
But in general, the tests were judged to be fairly easy.
"Students who will teach high school math normally have pretty close to a math major in college, but these tests don't cover any of that material," said Lynn Arthur Steen, a mathematics professor at St. Olaf College in Northfield, Minn., who served on the national review panel for the study. "They only cover the mainstream part of the high school curriculum."
Patte Barth, a senior associate at the Education Trust and the co-author of the study with Ruth Mitchell, a principal partner at the trust, said the team was "somewhat surprised" that the tests did not cover college-level material. "We also know that there are many teachers whose qualifications far exceed what is demanded by the tests," she added. "This is not an indictment of teachers."
While agreeing that teachers must be highly knowledgeable, Mari Pearlman, the vice president for teaching and learning at the ETS, warned that the burden of ensuring teacher quality should not rest entirely on licensing exams.
Moreover, she said, no research exists to back up the Education Trust's claim that teachers would be more effective if they had "the kind of knowledge you take with you to graduate school."
Vol. 18, Issue 38, Page 5Published in Print: June 2, 1999, as Researchers Find Teacher Tests Short on Covering College Content