Search

New ability estimators have been proposed by Wainer and Wright (1980) and Mislevy and Bock (1981) that are resistant against guessing and careless behaviors exhibited by some examinees. This paper presents another class of ability estimators that are resistant to departures from the underlying assumptions concerning guessing and carelessness. In addition to computing the asymptotic relative efficiency of such estimators, this paper evaluates estimators by comparing their influence curves (Huber, 1981). (Author)

The discovery of disproportionate minority enrollments in special education classes has been a subject of considerable concern to federal, state, and local education officials alike for the past decade. Because the existence of disproportions is frequently thought to result from discriminatory placement practices on the part of local school districts, many state agencies are implementing methods for detecting and taking action against offending districts. This paper discusses some of the methodological problems inherent in existing methods for determining disparities in special education placement and offers an alternative to presently-used methods. The alternative offered can be used for the simultaneous comparison of disparity across many districts and several diagnostic categories. (Author)