Chapter: 1 Science & Engineering Indicators 93

Performance-Based Assessments

This method of assessment asks students to "create an answer or product that demonstrates their knowledge or skills" (OTA 1992, p. 5). They may take the form of any number of tests that evaluate student performance including conducting experiments, answering open-response questions, computing mathematics equations, presenting an oral argument, writing an essay, and creating a portfolio of work accomplished throughout the school year. According to the Office of Technology Assessment (OTA 1992, p. 18), performance-based assessments generally

allow students to create their own response rather than to choose between several already created answers;
are criterion-referenced, or provide a standard according to which a student's work is evaluated rather than in comparison with other students;
concentrate on the problem-solving process rather than on just obtaining the correct answer; and
require that trained teachers or others carefully evaluate the assessments and provide consistency across scorers.

Performance-based assessment has been gaining support as an alternative or supplement to traditional standardized tests. Proponents suggest that performance assessments more closely link assessment and instruction, more accurately measure the mathematic and scientific skills and knowledge advocated by the NCTM standards, and allow a more complete account of student academic development. By December 1992, 13 States reported implementing some sort of performance-based assessment, while 28 others reported planning or piloting stages of performance assessments (Pechman and Laguarda 1993).

One form of performance-based assessment is portfolio assessment. Students compile and submit a collection of work in a specific subject area completed during a given period of time. Supporters argue that portfolios encourage students to work to their best abilities and constantly improve their work. According to Arter and Spandel (1991), portfolios can

provide a more complete picture of student ability by incorporating measures such as motivation and persistence;
capture students' thought processes;
share with students the basis upon which they are judged, and thus align expectations and performance with assessment; and
display a chronological development of student progress.

However, while the portfolio method is useful for determining aggregate success, a recent study reports a significant lack of consistency between portfolio scorers--a lack great enough to draw the method's use into question as a reliable indicator of individual success (Koretz et al. 1992). This study suggests a need for better training of scorers. Furthermore, because many scorers are also teachers, the discrepancy in scores may point to a lack of understanding or consensus of what scorers should look for as well as of what teachers teach in the classroom.

Experiments are another useful form of performance-based assessment. This method of assessment was used in a recent IAEP study which evaluated the mathematics and science skills of over 30,000 students in four countries and five Canadian provinces.(Click here for footnote 21.) Students were tested on procedural tasks which they performed in front of an observer. In science, a majority of the questions asked students to draw on knowledge concerning the physical sciences and the nature of science; in math, students concentrated on measurement and geometry. The IAEP (NCES 1992e, p. 6) researchers discovered the following.

Scores varied widely from task to task, suggesting that the measures tap a range of skills and knowledge.
Scores on the various tasks varied significantly between countries/provinces in systematic ways, indicating real differences in performance between the various populations.
The relative performances of countries and provinces were generally different from those identified by the written tests covering related curricular areas, suggesting that this method of assessment let students demonstrate their skills in ways that were not possible with traditional written tests.(Click here for footnote 22.)

Footnote 21:

The United States decided not to participate in the project until it could evaluate the results of this assessment.

Footnote 22:

Another notable finding from the IAEP study provides insight into the different strategies students from different countries use to complete tasks. Student approaches ranged from guessing, to estimating, to calculating precise answers, depending on the strategies taught in the respective countries. For instance, Taiwanese and Scottish students tended to use precision over estimation, while those from Alberta and Saskatchewan showed a preference for estimation over precision.

ContentsSearchContinue