Two assumptions dominate the current policy debate about assessment: first, that changes in the assessment policy can be used as a lever for reforming schools, and second, that the new “authentic” forms of assessment are inherently superior to traditional standardized, multiple-choice tests. These assumptions have largely been unchallenged despite the fact that there is little empirical evidence to indicate that they are valid.
Contributors to a recent symposium in the Harvard Educational Review point out that it is doubtful that merely changing the form of assessments will improve schools or reduce educational inequities in the United States. These educators point out that there is no quick fix through assessment reform — that we can not test, examine or assess our way out of our educational problems.
Linda Darling-Hammond, Teachers College,Columbia University, writes that alternative assessment methods are not inherently equitable, nor will they necessarily improve teaching and learning unless careful attention is paid to the ways the assessments are used. If assessment reform is driven by mistrust of teachers or fails to involve teachers or is used as a lever for external control of the schools, it is not likely to be successful. Darling-Hammond argues instead for “top-down support for bottom-up reform.”
Darling-Hammond reminds educators that traditional, standardized, multiple choice tests fail to measure complex cognitive and performance abilities. In addition, they do not help teachers understand students’ strengths, needs, or approaches to learning and thereby inhibit the ability of teachers to adapt their instruction to increase student learning. Authentic, performance-based assessments can provide teachers with more useful classroom information, but Darling-Hammond states that testing policies must also be aimed at supporting the use of this information by teachers. Proposals for assessment reform differ in the extent to which they:
1. broaden the roles of educators, students and parents in assessment,
2. aim to make assessment part of the teaching/learning process and use it to serve developmental and educational purposes rather than sorting and selecting purposes,
3. support either a problem-based interdisciplinary curriculum or a coverage-oriented curriculum, and
4. regard assessment reform as part of a broader national agenda to improve and equalize educational opportunities.
These varied approaches to assessment reform predict very different consequences. Changes in the forms of assessment alone are unlikely to enhance educational equity or quality unless we also change the ways in which assessments are used.
Impact of standardized testing policies
The research of Howard Gardener, Harvard, and Robert Sternberg, Yale, shows that intelligence has many dimensions. Previously, intelligence has been assumed to be an innate, fixed and measurable commodity, and that measuring intelligence would allow people to be ranked against each other. Unfortunately, the use of standardized tests to sort students has led to ineffective as well as inequitable education.
Retaining students on the basis of achievement cutoffs on tests has proved ineffective because repeating a grade does not help a student gain ground academically and has a negative impact on social adjustment and self-esteem. Retention is costly and increases dropout rates by 40 to 50 percent. Similarly, research in minimum-skills tests for graduation indicates that even though chances of employment and welfare dependency are linked to graduation from high school, neither employability nor earnings are significantly affected by students’ scores on basic skills tests. The use of the test as the sole determinant of graduation imposes heavy personal and societal costs without any tangible benefits.
In addition, when standardized tests results are used to reward or sanction schools, this encourages schools to influence test results in a number of ways that have very negative consequences for students. Because scores on any test are sensitive to the population of students taking the test, there are incentives for schools to influence their scores by keeping out students who they fear will lower their scores. These include students who are handicapped, speak little English or come from educationally disadvantaged environments.
By labeling large numbers of these low-scoring students for special education placement, by retaining students so their relative standing will look better on grade-equivalent scores, by excluding certain students from open-enrollment schools and by encouraging such students to drop out, schools can raise their scores. In addition, capable teachers are less likely to risk losing rewards or incurring sanctions by volunteering to teach in schools with low achievement scores.
The inequality of financial support between school districts means also that schools in poor districts cannot compete with wealthier districts for good teachers or provide equal resources and materials for their students. The use of testing to reward or punish schools works against equity, and changing the test format, by itself, will not alter this.
Building and equitable system
Highly standardized, externally controlled tests have been relied upon because it was believed that teachers could not make sound decisions about what students know and are able to do. Darling-Hammond believes that the reason the United States lacks a complex, student-centered approach to teaching and assessment is because we have not invested enough in teacher education and professional development. Reforms, she recommends, must ensure that all teachers have a stronger understanding of how children learn and develop, how assessment can be used to evaluate what they know and how they learn, how a variety of curricular and instructional strategies can meet their needs, and how changes in school and classroom organization can support their growth and achievement.
Substantial teacher and student involvement in and control over assessment strategies is critical if assessment is to support a challenging curriculum for every student. Reformers believe that new assessments must be used to inform and improve teaching and learning. To do this, assessment must be integrally connected to the teaching and learning process so that students’ strengths and weaknesses are identified and used to develop appropriate instruction. Assessment should improve learning and teaching in schools. Some state assessment programs change the nature of existing standardized tests without changing the locus of control, the scoring or the uses of results.
Some new tests are still being used primarily for ranking students and schools and for maintaining control of instruction outside of individual schools. Darling-Hammond believes that if performance-based assessments are used in the same way that traditional standardized tests are used now, they are even more likely to highlight differences in students’ learning and probably won’t help either teachers to restructure their teaching or schools to improve students’ learning.
Changing assessment forms and formats without changing the ways in which they are used will not affect the outcomes of education, warns Darling-Hammond. In order for performance-based assessment to support students’ learning, it must include teachers in all stages of the development process and must be embedded in the curriculum and teaching activities. Its main aim should be to support more informed and student-centered teaching rather than to sort students and sanction schools.
Performance-based assessment must be intimately understood by teachers, students and parents. The assessment must allow for different starting points for learning and diverse ways of demonstrating competence. In order for schooling to improve, assessment must be an integral part of staff discussion and school development. Darling-Hammond believes that the goal of schooling is to educate all children well rather than to select a small, talented group to educate to a high level. Finally, as long as some students receive only a fraction of the school resources that support education in wealthier districts, testing students can not provide accountability.
“Performance-Based Assessment and Educational Equity”, Harvard Educational Review, Volume 64, Number 1, Spring 1994, pp.5-29.
Published in ERN, May/June 1994, Volume 7, Number 3.