Computer-Based and Paper-Based Measurement of Recognition Performance.

Autor/in	Federico, Pat-Anthony
Institution	Navy Personnel Research and Development Center, San Diego, CA.
Titel	Computer-Based and Paper-Based Measurement of Recognition Performance.
Quelle	(1989), (31 Seiten) PDF als Volltext kostenfreie Datei Verfügbarkeit
Sprache	englisch
Dokumenttyp	gedruckt; online; Monographie
Schlagwörter	Armed Forces; Comparative Analysis; Computer Assisted Testing; Discriminant Analysis; Flight Training; Military Personnel; Multivariate Analysis; Postsecondary Education; Professional Training; Recognition (Psychology); Test Format; Test Reliability; Test Validity + Suchen Sie Ihr Suchwort? Military; Militär; Diskriminanzanalyse; Multivariate Analyse; Post-secondary education; Tertiäre Bildung; Berufliche Fachbildung; Berufliche Fortbildung; Recognition; Wiedererkennen; Testentwicklung; Testreliabilität; Testvalidität
Abstract	To determine the relative reliabilities and validities of paper-based and computer-based measurement procedures, 83 male student pilots and radar intercept officers were administered computer and paper-based tests of aircraft recognition. The subject matter consisted of line drawings of front, side, and top silhouettes of aircraft. Reliabilities for both modes of testing were estimated by deriving internal consistency indices, using an odd/even item split. Prior to testing, subjects learned to recognize the aircraft silhouettes using two media: (1) paper-based form structured as a study guide; and (2) computer-based form using FLASH IVAN in the training mode. A stepwise multiple discriminant analysis was performed to determine how well the two testing modes distinguished among two groups of subjects expected to differ in their recognition of aircraft silhouettes. Computer-based and paper-based measures were not significantly different in reliability or internal consistency. The paper-based measure of average degree of confidence in recognition judgments was more reliable than the computer-based measure. The average degree of confidence measured by the two modes was more equivalent than the measures of recognition test scores. The discriminative validities of the two measures were about the same for distinguishing groups above or below the mean average curriculum grade. Using the pooled within-groups correlations between the discriminant function and computer-based or paper-based measures, the former had superior discriminative validity than the latter. Statistics associated with the canonical correlation suggested that the predictive validity of computer-based measures approximates that of paper-based measures. Four tables present study data. A 54-item list of references is included. (SLD)
Erfasst von	ERIC (Education Resources Information Center), Washington, DC