As validity pertains to test use rather than the test itself, using a test for unintended purposes requires a new validation program using additional evidence from relevant sources. This small-scale study contributes to the validation of the use of originally academic language tests--the International English Language Testing System and the Test…

To some extent, good testing procedure, like good language use, can be achieved through avoidance of errors. Almost any language-instruction program requires the preparation and administration of tests, and it is only to the extent that certain common testing mistakes have been avoided that such tests can be said to be worthwhile selection,…

Four essential dimensions of a performance test are detailed: directness of test method, type of criterion, standardization of conditions, and objectivity of scoring. For simplicity these factors are described as if each were dichotomous, when in actuality each is a continuum; a test method may be more or less direct, conditions more or less…

Essay tests are easily constructed, relatively valid assessments of higher cognitive processes but are harder to score reliably. Teachers using essay tests are advised to follow clearly designed objectives, construct all-inclusive, pilot-tested questions, develop a checklist of specific scoring points and a model answer for each question, and use…

This article discusses practical and theoretical problems related to the measurement of formal operations. The first section of the article discusses problems in measuring formal operations using the clinical interview method. These problems include the lack of both a standardized interview and a uniform scoring procedure. Section two discusses…

When individuals are given tests that are too hard or too easy, the resulting scores are likely to be poor estimates of their performance. To get valid and accurate test scores that provide meaningful results, one should use functional-level testing (FLT). FLT is the practice of administering to an individual a version of a test with a difficulty…

Certain testing procedures will overcome some of the problems associated with the use of essay tests. Essay tests may not validly indicate achievement because the questions included in the test may not fairly represent instructional content. Reliability may be a problem because of variations in examinee response in different situations, in test…

A summary of the writing assessment programs reviewed in this journal is presented. The problems inherent in the programs are outlined. A coordinated research program on major problems in writing assessment is proposed as being beneficial and cost-effective. (DWH)

The transcript of a six-part National Public Radio broadcast on standardized testing is presented. The first part focuses on the reasons tests are administered; these reasons are discussed by proponents and opponents of testing. Part Two contains a discussion of the possible bias of tests, and their validity. The third part discusses the…

Cross-cultural validity studies for psychological instruments may result in overestimation of structure invariance due to some items being scored on more than one scale. This problem, called the common-item effect, is investigated with some data from the literature. (JKS)

Procedures for conducting functional level testing have been available for use by practitioners for some time. However, the Title I Evaluation and Reporting System (TIERS), developed in response to the educational amendments of 1974 to the Elementary and Secondary Education Act (ESEA), has provided the impetus for widespread adoption of this…

Using criterion-referenced test item data collected in an empirical study, differences in item selection between Cox and Vargas' pretest-posttest validity index and a latent trait approach (evaluation of the item information function for the mastery score) are analyzed. (Author/GK)

Investigated certain measurement properties of the Ethical Judgment Scale. Results revealed findings of questionable validity and unacceptably low reliability for the instrument even though three different scoring methods were used in the analysis. (Author)

True-false achievement test items written by typical classroom teachers show about two-thirds of the discrimination of their multiple-choice test items. This is about what should be expected in view of the higher probability of chance success on the true-false items. However, at least half again as many true-false items as multiple-choice items…

Many testing programs include a section of optional questions in addition to mandatory parts of a test. These optional parts of a test are not often truly parallel to one another, and groups of examinees selecting each optional test section are not equivalent to one another. This paper provides a general method based on missing-data methods for…