A Computer Program for Calculation of Scott's Coefficient of Observer Reliability

Bibliographic Data

ID20282129
AuthorsThomas B Gregory (0000-0002-2507-8731, Indiana University, corresponding author)
Year1970
Volume30
Issue1
Pages183-185
Publication date1970-04-01
Open AccessYes
TypeARTICLE
VenueEducational and Psychological Measurement (JOURNAL)
Journal identifiersISSN: 0013-1644 • E-ISSN: 1552-3888
PublisherSAGE Publications (PUBLISHER • US)
DOI10.1177/001316447003000128
OpenAlexW1984811529
LanguageEN
References cited3

THE past decade has witnessed the development of numerous ob-servation systems that attempt to record objectively the ongoing verbal and nonverbal events occurring in the classroom. Flander’s Interaction Analysis (1960) and the several revisions of the OScAR (e.g., Medley and Mitzel, 1958; Medley, Schluck, and Ames, 1968) are notable examples of this trend. Since such descriptive systems increasingly are being utilized as criterion measures in educational research, the problem of obtaining valid estimates of inter-observer reliability is a crucial one. Scott (1955) proposed a method of estimating the reliability of observers using any observation system that assigns events to mutu-ally exclusive categories. The method marks an improvement over conventional methods which are based on the percentage of observer agreements. The formula for Scott’s coefficient is: where Po (observed per cent of agreement) represents the percentage of judgements on which two observers agree when coding the same data independently; and Pe is the per cent of agreement to be ex

Computer program · Observer (physics) · Physics · Programming language · Reliability (semiconductor) · Reliability engineering · Statistics · Computer Science · Engineering · Hemodynamic Monitoring and Therapy · Mathematics · Psychology · Reliability and Agreement in Measurement · Statistical Methods in Epidemiology

Tools

Open DOISci-Hub