| With the recent advent of basic skills and minimum competency testing programs, the direct measurement of student writing ability has assumed increasing importance. Many large scale testing programs now require students to compose one or more pieces of writing as a part of a proficiency assessment. The design of such writing tests presents many technical, procedural, and political issues that must be resolved if the assessments are to be successful.;The issues related to the direct assessment of student writing ability that are treated in the dissertation include (1) identifying the audiences for an assessment, (2) selecting the objectives to be measured, (3) determining an appropriate theoretical base for the assessment, (4) developing a valid and reliable test, (5) designing defensible rating procedures, (6) determining appropriate test length, (7) assuring the equivalence of the assessment across testing occasions, and (8) setting performance standards.;This dissertation discusses the issues commonly faced in developing a large scale writing assessment program. These issues are delineated both through a review of the relevant research and through a case study documenting the development of one statewide writing assessment program. A discussion of the problems and successes encountered in establishing this particular assessment program provides an opportunity to suggest ways in which future practice may be improved. |
