Four global memory models were evaluated in 3 recognition memory experiments with 30 college students. Experiments provide receiver operating characteristic (ROC) curves. Data give a clear idea of the behavior of signal and noise distributions in recognition memory. Ways in which results support revision of current models are discussed. (SLD)

A probabilistic parallelogram model (the PARELLA model) is presented for the measurement of latent traits by proximity items. This unidimensional model assumes that the responses of persons to items result from proximity relations. The model is illustrated in an analysis of three empirical datasets from previous studies. (SLD)

A method of automatically selecting items for inclusion in a test with constraints on item content and statistical properties was applied to real data. Tests constructed manually from the same data and constraints were compared to tests constructed automatically. Results show areas in which automated assembly can improve test construction. (SLD)

Describes a goodness-of-fit statistic that considers the imprecision with which ability is estimated and involves constructing item fit tables based on each examinee's posterior distribution of ability, given the likelihood of the response pattern and an assumed marginal ability distribution. Also describes a Monte Carlo resampling procedure to…

Traditional approaches to the investigation of the objectivity of ratings for constructed-response items are based on classical test theory, which is item-dependent and sample-dependent. Item response theory overcomes this drawback by decomposing item difficulties into genuine difficulties and rater severity. In so doing, objectivity of ability…

Test items are designed to be representative of the subject areas that they measure and to reflect the importance of specific domains or item types within those subject areas. Content validity is achieved by content specification and number of items in each content domain included in the design of the test. However, largely due to the normal…

This paper examines the effect of using unidimensional item response theory (IRT) item parameter estimates of multidimensional items to create weakly parallel test forms using target information curves. To date, all computer-based algorithms that have been devised to create parallel test forms assume that the items are unidimensional. This paper…

A new test construction method based on integer linear programming is described. This method selects optimal tests in small amounts of computer time. The new method, called the Cluster-Based Method, assumes that the items in the bank have been grouped according to their item information curves so that items within a group, or cluster, are…

The methods and approaches for estimating the operating characteristics of the discrete item responses without assuming any mathematical form have been developed and expanded. It has been made possible that, even if the test information function of a given test is not constant for the interval of ability of interest, it is used as the Old Test.…

Conjunctive item response models are introduced such that: (1) sufficient statistics for latent traits are not necessarily additive in item scores; (2) items are not necessarily locally independent; and (3) existing compensatory (additive) item response models including the binomial, Rasch, logistic, and general locally independent model are…

Drawing on recently developed methodology for structural equation modeling with categorical data, this article proposes a new approach for investigating the behavior of dichotomously scored test items in relation to other relevant (observed) variables. A linear structural model relates the latent ability variable to a set of observed scores.…

Summarizes the current state of the art in test construction and contrasts it with previous conceptual models, some of which are wrong or misleading. New methodologies for item selection and review are presented, with current thinking on the specification of technical characteristics of tests. (Author/SLD)

The effect of large, within-examinee item difficulty variability on estimates of the proportion of consistent classification of examinees into mastery categories was studied over 2 test administrations for 100 simulated examinees. The proportion of consistent classifications was adequately estimated using the technique proposed by M. Subkoviak…

Two methods are proposed for the construction of weakly parallel tests based on a prespecified information function. A method is then described for selecting weakly parallel tests that are optimal with respect to the Maximin criterion. Numerical examples demonstrate the practicality of the tests. (SLD)

A Monte Carlo study, which simulated 10,000 examinees' responses to four tests, investigated the effect of item stratification on parameter estimation in multiple matrix sampling of achievement data. Practical multiple matrix sampling is based on item stratification by item discrimination and a sampling plan with moderate number of subtests. (SLD)