This report is the sixth of a series of reports having to do with item studies of the Sequential Tests of Educational Progress (STEP) regularly administered to students in grades four, six, eight, ten, and twelve enrolled in the public school system in Hawaii. The item study summaries prepared for achievement tests administered at the various…

The final volume of The Study of Junior Colleges contains the measurements and instrumentation derived from the project for future evaluation surveys. Part One, Prediction of Student Outcomes: Multivariate Analysis of the Survey Data, provides: an introduction to the analyses; data reduction, factors, and scales; students' objectives and…

Although it has been claimed that the Rasch model leads to a higher degree of objectivity in measurement than has been previously possible, this model has had little impact on test development. Population-invariant item and ability calibrations along with the statistical equivalency of any two item subsets are supposedly possible if the item pool…

This study investigated the feasibility of concurrently and randomly sampling examinees and items in order to estimate group achievement. Seven 32-item tests reflecting a 640-item universe of simple open sentences were used such that item selection (random, systematic) and assignment (random, systematic) of items (four, eight, sixteen) to forms…

Discussed are the objectives of two introductory college biology courses for majors and how well these objectives were fulfilled. Examples of instructional objectives are given, along with sample test items. A description is given of the item analysis applied to the test items, followed by a discussion of how the test items measured whether or not…

The problem of determining the significance level which should be used in statistical tests of item validity in order to minimize type I errors is discussed. (Author/JKS)

The effects of repeaters (testees included in both administrations of two forms of a test) on the test equating process are examined. It is shown that repeaters do effect test equating and tend to lower the cutoff point for passing the test. (JKS)

Given a set of dichotomously scored items, ordering theory attempts to identify a hierarchy among those items. A graphic representation is presented for an empirically obtained hierarchy and a testing procedure to evaluate hypothesized hierarchical relationships. (Author/JKS)

A computer program for scoring written simulation tests provides individual scores and basic item analysis data. The program is written in Fortran IV and can accomodate up to thirty-five hundred options and up to ten thousand examinees. (Author/JKS)

Using 463 fifteen year old girls from five girls' schools in the metropolitan area of Sydney, this study examines four aspects of the Junior Eysenck Personality Inventory that has been subject to criticism, namely the terminology used in certain JEPI items, item response rate ceilings, factorial structure of the inventory, and the relationship…

This study compared the item difficulty, item discrimination, and test reliability of three forms of multiple-choice items: (1) one correct answer; (2) "none of the above" as a foil; and (3) "none of the above" as the correct answer. Twelve items in the three formats were administered in a college statistics examination. (BS)

This paper develops and applies three nonparametric comparisons of the shapes of two item characteristic surfaces: (1) proportional latent odds; (2) uniform relative difficulty; and (3) item sensitivity. A method is presented for comparing the relative shapes of two item characteristic curves in two examinee populations who were administered an…

The effects of statistical and cognitive difficulty ordering of test items, gender, and explicit labeling of item type on college students' test performance were assessed. A total of 137 male and 265 female undergraduates were administered two 40-item multiple choice tests. Results indicate significant effects for gender and use of labels, but…

An approximate relationship is devised between the unidimensional model used in data analysis and a multidimensional model hypothesized to be generating the item responses. Scale shrinkage is successfully predicted for several sets of simulated data. (Author/LMO)

A three subscale inventory designed by Fenigstein, Scheier, and Buss to measure self-consciousness was administered to 297 college students. Fenigstein et al.'s representation was found to fit the data in its original form. Items on the subscales differ nearly as much statistically as they do substantively. (Author/LMO)