Easy-Test has been developed to produce computer generated criterion referenced tests on any of 300 subject areas at the secondary level. The tests are drawn from over 14,000 items, can be used for diagnostic purposes, and have been shown to be both valid and reliable. Among limitations of the system are the inclusion of some items not taught in a…

For individualized or computer assisted instruction, norm referenced testing is inadequate to determine each individual's mastery on specific kinds of tasks. Hively's item forms and Ferguson's stratified item forms, both based on observable characteristics of the problems, and Scandura's algorithmic technology, positing that persons use rules to…

Identifies a number of practical decisions that must be made when designing and developing criterion referenced tests as part of a larger system of computer assisted instruction. The topics discussed include test construction, test security, item presentation, and response capturing and scoring. (19 references) (CLB)

The TITA (Totally Interactive Testing and Analysis) System algorithm for the repetitive construction of domain-referenced tests utilizes a compact data bank, is highly portable, is useful in any discipline, requires modest computer hardware, and does not present a security problem. Clusters of related keyphrases, statement phrases, and distractors…

A decision-theoretic procedure is outlined which provides a framework within which Bayesian statistical methods can be employed with criterion-referenced tests to improve the quality of decision making in objectives based instructional programs. (Author/DEP)

An item generation procedure is described which was utilized in the development of Computer Managed Review and Examination courses for the education of nurses in remote areas. The major emphases are the processes of domain definition, item writing, and item edition. Specific discussion is presented concerning methods of item construction to assess…

Describes a four-tier model for designing and developing practice and assessment in computer-based training. Topics include teaching versus assessing; levels of knowledge; taxonomy levels and intended learning outcomes; and time and cost constraints that result in criterion-referenced tests being most widely used. (Author/LRW)

As an alternative to a classical test theory basis for criterion-referenced test construction, it is proposed that a strict item-sampling model be used. The computer's role in such a model is outlined. The assumptions of the model are carefully defined and its properties reviewed. The relationship between mastery criteria and such sampling plans…

This report illustrates one way in which the technologies of testing might combine with cognitive science techniques to help steer instruction. Steering testing is brief diagnostic testing that steers, or individualizes, the course of instruction. Steering testing uses simple heuristics for reasoning about the level of a student's competence in a…

A study was made to monitor achievement of individually paced instruction. The project concentrated on designing testing procedures in group paced instructional programs to provide information to student, teachers, parents and administrators which could be used in both a formative and summative evaluation. The three objectives of the project were:…

In order to make formal instruction more responsive to individual needs, the following study implemented a procedure for transforming a conventional set of course materials into an interactive learning environment. The basic mechanism for interaction was a computer generated and computer administered quiz that sampled from concepts in the reading…

This paper highlights some special characteristics of criterion-referenced tests while comparing them with norm-referenced tests. Psychometric considerations involved in constructing a criterion-referenced test, including item analysis, reliability, and validity, are discussed, along with application of criterion-referenced testing to individual…

A replicable process for improving instruction through the consistent use of student data collected before, during, and after instruction is proposed. A rational analysis of different types of error rates (theoretical, base, posttest, instructional) and discrimination indices (base, posttest) leads to a set of rules for identifying test items and…

The Jefferson County Public School System in Colorado is conducting a study which implements a digital time-shared computer as support for a systems approach to instruction. This study currently involves one elementary school but it will support a total of thirteen schools in the future. The computer support includes computer-generated criterion…

The 75 papers included in these conference proceedings discuss testing conducted by the different brances of the armed forces. The importance of relating necessary job skills to the skills measured by the tests administered to the job applicants is emphasized. Various evaluation methods--including peer rating, aptitude testing, adaptive testing,…