Describes and evaluates a "controlled assessment procedure" as a holistic approach to avoiding problems of administering and evaluating traditional exams. Key characteristics include: the question is known well in advance and is broad and open-ended, students are encouraged to respond in any self-selected written format, and the rationale and…

The Ralph Nader report on the Educational Testing Service (ETS), entitled The Reign of ETS, the Corporation That Makes Up Minds, explicitly and implicitly raises serious issues concerning the testing enterprise. Major themes include the role of testing in the educational selection system, the validity of existing tests, and the corporate power of…

Using an index of answer copying developed by Houston, it was found that rearranged questions alone did not reduce answer copying, whereas rearrangement of both questions and answers effectively eliminated detectable cheating. (Author)

This article focuses on common errors associated with true-false, matching, completion, and essay questions as presented in textbook test manuals. Teachers should be able to select and/or adapt test questions which would be applicable to the content of their courses and which meet minimal standards for test construction. (JD)

This study utilized a test, re-test procedure to investigate what effects a change in the order, or sequence, of test items would have on student performance. College juniors (n=102) were given a 50-item multiple-choice and matching item test on the general subject of educational measurement. The items had no particular sequence, but had simply…

The subject of constructing criterion-referenced tests is often researched, but many technical problems remain to be satisfactorily resolved. Foremost, criterion-referenced test developers need a comprehensive set of steps for construction. In this paper, 14 logical steps for building criterion-referenced tests that refer to several different…

Effects of item presentation mode on degree of leniency bias in responses to field research questionnaires were studied. Two modes were examined: first with items measuring the same dimensions grouped together and second with such items distributed randomly. The random mode showed substantially less leniency response bias. (Author/BW)

The empirical evidence does not support the prevailing conventional wisdom that it is advisable to mix positively and negatively worded items in psychological measures to counteract acquiescence response bias. An experiment, evaluating subjects' ability to respond accurately to both positive and reversed items on a questionnaire, analyzed post-hoc…

Defends the CUNY-WAT against the charges made by Fishman (CS 731 865). Offers suggestions for those wishing to undertake research into the choice of topics for writing assessment tests. (FL)

Complex multiple choice (CMC) items are frequently used to test knowledge about repetitive information. In two independent comparisons, performance on the CMC items surpassed that of the multiple true-false clusters. Data indicate that performance on CMC items is inflated, and distractors on CMC items fail to prevent guessing. (Author/PN)

This book explains writing effective multiple-choice test items and studying responses to items to evaluate and improve them, two topics that are very important in the development of many cognitive tests. The chapters are: (1) "Providing a Context for Multiple-Choice Testing"; (2) "Constructed-Response and Multiple-Choice Item Formats"; (3)…

Discusses the development of a language aptitude test designed to help select students to study German. Four areas of assessment are memory ability, ability to deal with stress and intonation patterns, inductive reasoning ability, and ability to recognize grammatical case. (EKN)

This article invokes a literal image of test givers as measurement devices and explores the psychometric properties of these test administrator instruments. Concurrent and content validation and test-retest and parallel-forms validity are explored. (SLD)

College undergraduates (n=38) were administered identical multiple choice tests with randomly presented answer-sheets numbered either vertically or horizontally. Of the originally-scheduled four tests during the semester, tests one and three were retested with entirely different test questions, also multiple choice, resulting in scores from tests,…

After six weeks of testing college algebra students (n=84) either on recall or recognition tests, the test modes were changed without warning. Results showed that performance suffered when the test mode was changed for students anticipating a recognition test. Students anticipating a recall test did equally well in both test modes. (Author/PN)