Jones, R. L. (1985). Second language performance testing: An overview. In Hauptman, P. C., LeBlanc, R., & Wesche, M. B. (Eds.). Second language performance testing (pp. 15-24). Ottawa: University of Ottawa Press.
Jones defines performance testing in terms of the assessment of the ability of an examinee in response to actual tasks from real language use situations. Performance tests, as opposed to other kinds of language tests, approximate the reality of a task situation. Their value resides in predictability of examinee ability in future, real-world conditions, based on observation of test-based performances (unlike other tests where examinee test ability is not necessarily a good predictor of ability with a real-world task).
Performance tests can be one of three types. Direct tests involve observation of examinees under actual conditions of language use, with minimal intrusion of the observer, and with no manipulation of the task types and language situations that might occur (this requires extensive observation to get at comprehensive knowledge of ability). Work sampling involves control over what kind of language situations will be observed, but the situations remain authentic interactions. Simulations involve contrived situations that are manipulated by the examiner in order to elicit representative types of language performance (like role-playing). These simulations are still directly related to a given task type, and examinees are able to compensate for certain proficiency weaknesses with various performance characteristics.
Validity, reliability, and practicality are still the primary criteria for good performance tests. Validity must be insured by systematic needs analyses that involve observation of task types and language situations and that involve area professionals or experts as informants for needed skills and potential task types. Reliability is dependent on scorer understanding of the test's criteria and their consistency in application of the criteria, and can be adversely affected by careless elicitation techniques. Practicality depends on program needs and educational expectations.
Jongsma, K. S. (1989). Portfolio assessment. The Reading Teacher, 43, 264-265.
Several experts offer their views on portfolio assessment in the classroom:
Portfolios come from the professional realm, and are used to demonstrate the breadth and depth of an applicant's professional capabilities. In the classroom, portfolios should be the responsibility of students, and they should represent all of the processes and products involved in the learning process. Portfolios should always be accessible to their owners, the students, and should not just be collections of student work, rather they should reflect a critical approach to organization.
Process-oriented report cards can be implemented in order to offer a more accurate and detailed picture of student learning. Literary folders can be used to reflect the "whole picture" of a student's progress over the course of a semester or year. Input from multiple sources (teacher, parent, and student) should be included.
Writing should be measured on various occasions in different modes but in natural settings. Portfolios offer an effective means of implementing these assessment goals. They "provide a more complete and accurate portrait of student writing growth" (p. 265).
