The Examination of Reliability of Vee Diagrams According to Classical Test Theory and Generalizability Theory

• Main Author: Betül POLAT DEMİR
• Format: Article
• Language: English
• Published: EPODDER 2016-12-01
• Series: Journal of Measurement and Evaluation in Education and Psychology
• Subjects: Vee diagram; Reliability; Classical Test Theory; Generalizability Theory
• Online Access: http://dergipark.ulakbim.gov.tr/epod/article/view/5000198540

The aim of this study was to investigate reliability of score obtained from vee diagrams. Therefore in this study, the Classical test theory and generalizability theory were used. The study was carried out with descriptive research model. The participants of this research were composed of 32 students at 7th and 8th grades in Niğde. The responses given to the vee diagrams were scored through the holistic rubric. In CTT, Cronbach’s alpha reliability coefficient and Kendall’s concordance coefficient for interrater reliability; in G theory, Ep2 (G) and Ф (Phi) coefficients with univariate model were computed. The Cronbach alpha reliability coefficients estimated were 0,77, 0,76, and 0,76 for each rater, respectively. For determination of interrater consistency, Kendall’s concordance coefficient was calculated as W=0,959 for the first vee diagram, W= 0,924 for the second one, W=0,963 for the third one, W=0,954 for the fourth one. For the purpose of estimating variances and percent of variances in G study, fully crossed design was applied. Generalizability coefficient for vee diagram scores was 0,918 and phi coefficient was 0,899. As a result of the G study, it was determined that the variance component estimated for the student (s) main effect accounted for 42,4% of the total variance, and it was the variance component that had the highest value. The variance component estimated for raters (p) main effect, which is the lowest variance component in all main effect, accounted for 0,1% of the total variance. According to all results, it was seen that vee diagram was reliable for determination of students’ mathematics achievement.