Providing examinees with advanced knowledge of the difficulty of an item led to an increase in test performance with no loss of reliability. This finding was consistent across several test formats. ( Author/JKS)

An analysis of self-report delinquency data using four non-metric procedures for structural analysis revealed support for the existence of a general delinquency factor. However, offenses of low seriousness and victimless offenses (drinking and drug-taking items) were only weakly related to this general factor. (Author/CTM)

Factor analyses of a 75-item self-report Academic Motivations Inventory produced a set of nine interpretable dimensions which partially summarize an academic motivation domain and which can be measured with adequate reliability. The factors are discussed in relation to prior studies of the structure of academic motivation, and further research is…

Describes the rationale for and the development of an instrument that provides scales to measure five school related constructs of self concept at three school levels and cites information regarding the factorial validity of each scale based on the derived data. (Author/DEP)

This study examined whether the item response curves from a two-parameter model reflected characteristics of the mathematics items, each of which required unique cognitive tasks. A computer program performed error analysis of test performance. Cognitive subtasks appeared to influence the slopes and difficulties of item response curves. (GDC)

To determine which produced the most accurate item parameter estimates, four methods of item response theory were evaluated: (1) heuristic estimates; (2) the ANCILLES program; (3) the LOGIST program; and (4) the ASCAL program. LOGIST and ASCAL produced estimates of superior and essentially equivalent accuracy. (SLD)

Intercorrelations among multiple true-false items were examined to determine to what extent each choice can be treated as independent. Results from 157 health science and 170 medical students indicated that correlations between options from the same stem were larger than those from different stems. Methods for computing reliability estimates were…

Three administrations of two forms of a biology achievement test were conducted with high school sophomores and seniors. The 58 common items between the two tests were used for score equating purposes. (TJH)

This study examines the Average System, an alternate procedure to the Rasch Model, for equating test scores. This technique does not require computers and can be easily employed by classroom teachers. A validation experiment using 23 sets of 10-item arithmetic tests constructed from a 230-item bank is described. (BS)

Local norming data from the Brief Index of Adaptive Behavior are presented, as are construct and concurrent validity data. Participants in three studies were 430 children. Factor analytic results from Study 1 yielded an easily interpreted three-factor solution; conventional item analysis techniques yielded significant item-domain and item-total…

The purpose of this research was to recommend an item bias procedure when the number of minority examinees is too small to use preferred three-parameter item response theory (IRT) methods. The chi-square, Angoff delta-plot, and pseudo-IRT indices were compared with both real and simulated data. (Author/DWH)

From analysis of the use of a student rating form designed for multiple audiences and using a single global item, it was concluded that a multiple audience form showed distinct advantages in design. The single item most fairly and highly represented overall teaching performance for administrative decisions about retention, promotion, and merit.…

The National Board of Medical Examiners uses the Rasch model to calibrate test items, maintain item banks, equate scores, and monitor the consistency of examiner item response patterns. The model is also being used in the study of patient management problems examinations, standard-setting, and computer-based examinations. (Author/BS)

The effects of violating four item construction principles were examined to assess the validity of the principles and the importance of students' test wiseness. While flawed items were significantly less difficult than sound items, differences in item discrimination, test reliability, and concurrent validity were not observed. (Author/BW)

As a preliminary step in customizing the Stanford Achievement Test, a study was made to compare the test items at each level with the curriculum objectives of a Florida school district, classifying them as either matching, above or below grade level, or unmatching. Results are shown in table form. (TE)