The consequences of formula scoring versus number right scoring are examined in relation to the assumptions commonly made about the behavior of examinees in testing situations. The choice between the two is shown to be dependent upon having reduced error variance or unbiasedness as a goal. (Author/JKS)

Investigated empirically through post mortem item-examinee sampling were the relative merits of two alternative procedures for allocating items to subtests in multiple matrix sampling and the feasibility of using the jackknife in approximating standard errors of estimate. The results indicate clearly that a partially balanced incomplete block…

The pure Rasch model was compared with four modifications of the model in a number of different simulations in order to ascertain the comparative efficiencies of the parameter estimations of these modifications. Because there is always noise in test score data, some individuals may have response patterns that do not fit the model and their…

This study investigates the amount of uncertainty added to National Assessment of Educational Progress (NAEP) estimates by equating error under both ideal and less than ideal circumstances. Data from past administrations are used to guide simulations of various equating designs and error due to equating is estimated empirically. The design…

How accurately ability is estimated when the test model does not fit the data is considered. To address this question, this study investigated the accuracy of the maximum likelihood estimator of ability for the one-, two- and three-parameter logistic (PL) models. The models were fitted into generated item characteristic curves derived from the…

It is suggested that bias--when associated with a predictor, a test, or a statistical estimator--is not always bad, in spite of the immediate negative response evoked by the word, bias. Four settings are described to illustrate situations in which a procedure should not be summarily rejected due to bias: (1) educational researchers rejected the…

Previous studies of test item bias have investigated how different groups of examinees perform differently on a given set of items. These studies imply that examinees should be treated in a certain way because they are of a particular sex or race rather than as individuals in their own right, but it is unrealistic and unfair to assume such an…

A Monte Carlo study was conducted to compare the statistical bias and standard errors of non-equivalent-groups linear test equating in small samples of examinees. One thousand samples of each size (15, 25, 50, and 100) were drawn with replacement from each of five archival data files from elementary school and secondary school teacher subject area…