Performance Assessment Seen Progressing From Theory to Practice
With 20 years of experience, Clare Forseth, a 5th- and 6th-grade mathematics teacher at Marion Cross School in Norwich, Vt., could be expected to be fairly certain about how she teaches math to upper-elementary students.
But since she began two years ago to take part in her state's pioneering effort to assess student abilities on the basis of portfolios, Ms. Forseth said, her teaching has "changed 180 degrees.''
"I have to be a problem-solver myself,'' she told participants at a conference here last month. "I have to take risks and persevere,'' just as students are expected to do.
"This assessment tool,'' she added, "is the most powerful instructional tool I have ever had.''
Ms. Forseth was one of several speakers at the conference, which was organized by a coalition of education organizations to demonstrate the way performance-based alternatives to traditional multiple-choice tests are being implemented in schools.
The conference was aimed at showing that performance-based assessment has progressed beyond theory and into reality, according to Monty Neill, the associate director of the National Center for Fair and Open Testing, or FairTest.
"This is happening in real places, like states and districts, not just in journals,'' he said.
But despite such advances, conference participants cautioned, the new technology has a way to go before it displaces conventional tests.
For one thing, noted Joan Boykoff Baron, the director of assessment systems for Performance Assessment Collaboratives for Education at Harvard University's graduate school of education, researchers have yet to show definitively that the assessments meet technical standards for validity, reliability, and ability to support generalizations.
In addition, Ms. Forseth of Vermont pointed out, other testing programs, including those used to determine eligibility for federal remedial-education assistance, often conflict with the aims of performance assessments. While teachers prefer the type of instruction the new assessments promote, she said, they may be forced to teach the kind of rote skills conventional tests emphasize.
"Their heart may be in one place, but reality is in another place,'' Ms. Forseth said. "There's a tug-of-war out there.''
To solve such problems, Mr. Neill of FairTest said, policymakers and educators need to come to an understanding that would provide accountability for schools while allowing the instruction teachers demand.
The question is, he said: "How do we extract information for political purposes without undermining the serious things we want for children?''
Solving an Ant's Climb
In addition to FairTest, the sponsors of the conference included the Council for Basic Education, the International Reading Association, and the National Education Association.
The meeting comes at a time when a growing number of officials are urging a move to develop alternative forms of assessment. The National Council on Education Standards and Testing, a Congressionally authorized panel, for example, in January proposed such assessments as part of a national system of tests that would be used to gauge student performance against national standards in key subjects.
The conference participants explained that alternative assessments could improve instruction by providing much broader information about student abilities than conventional tests offer.
The Vermont program--which evaluates 4th and 8th graders in math and writing on a uniform test, a portfolio, and a "best piece'' gleaned from the portfolio--enables educators to gauge students' abilities to solve problems and communicate their solutions, not just on whether they can identify the right answer, according to Ms. Forseth.
As an example, she showed several 4th graders' responses to a problem that asked: How many days would it take an ant to climb out of a 12-foot-deep well if the ant climbs up 4 feet each day and slips back 2 feet each night?
The pupils were judged, on a four-point scale, on the quality of their understanding of the task, their approaches and procedures, their decisions along the way, and the outcomes of their activities. They also earned scores for evidence of the use of mathematical language, their mathematical representations, and their presentation.
By presenting separate scores for each aspect of a student's work, Ms. Forseth said, teachers can target their instruction in areas where students appear weak.
If a student seems unable to present information in pictures, for example, "I can show them a lot of examples of how to do that,'' she said.
Preparing Teachers, Students
Other programs discussed here also provide broader information about students' abilities, speakers said.
The Primary Learning Record, a British import used in about 40 New York City schools and in some Chapter 1 programs in California, provides detailed evaluations of primary-school children's progress in literacy skills.
But while such programs are aimed primarily at helping teachers in classrooms, the new statewide performance-based assessment in Maryland is aimed more at program improvement, said Hannah Kruglanski, the program's coordinator.
Under the Maryland system, which got under way last year, the state tests every 3rd, 5th, and 8th grader in an integrated assessment of reading, writing, language usage, math, science, and social studies. Although the state does not report individual students' scores, it does report schools' scores; in five years, schools with low performance must prepare improvement plans.
"We are not providing diagnostic information for individual students,'' Ms. Kruglanski said.
But she said the program has already reaped instructional benefits.
To help prepare students for the assessments, she said, schools purchased maps, reading materials, and calculators. And, she noted, teachers have begun to integrate their lessons, rather than teach each subject separately.
But, echoing several speakers here, Ms. Kruglanski pointed out that the new forms of assessment demand retraining of teachers to ensure that they can teach the skills the assessments measure.
And, suggested Edward D. Roeber, the director of the student-assessment compact, a consortium of states formed by the Council of Chief State School Officers, students must also be brought around to a different way of learning.
"There must be an effort to prepare teachers, parents, and
students,'' he said. "This is not just some grand experiment we hope
will succeed because 'it must.' ''
Vol. 11, Issue 28, Page 8