AN EXPECTANCY VALUE THEORY PREDICTS ACHIEVEMENT IN UNDERGRADUATE STATISTICS THROUGH ACADEMIC DELAY OF GRATIFICATION

We tested a model that integrates academic delay of gratification with Expectancy Value Theory to predict achievement in an undergraduate psychology and nursing statistics class at a metropolitan university in the southeastern United States. We analyzed measurements (n = 163: 80.4% female) of past performance, academic delay of gratification, effort, value, affect, and cognitive competence with students’ final exam score. The path model analyzed explained 14.9% of the variance in scores. Past performance inmathematics and student effort had direct effects on grades and all expectancy value theory constructs, as well as academic delay of gratification, were indirectly related to grades. We present details of our analysis and discuss theoretical and pedagogical implications of this study. First published June 2020 at Statistics Education Research Journal Archives

Verification of the Unidimensionality of Academic Delay of Gratification Scale in the Indian Context

The present study verifies the three models on the dimensionality of the construct academic delay of gratification measured with the academic delay of gratification scale prepared by Bembenutty and Karabenick (1996). Sample of the study comprises of 488 professional courses undergraduate students of Muslim minority community (277 boys and 211 girls) from law, engineering, education and pharmacy faculties of Sultan Ul Uloom Education Society, Banjara Hills, Hyderabad, Telangana, India. Exploratory factor analysis was conducted on the responses of the 10 items provided by the sample using SPSS Statistics Ver.23 to extract the factors of the construct. Confirmatory factor analysis conducted using SPSS Amos Ver.23 provided the goodness of fit measures for each of the models. The unidimensional model produced excellent fit indices. Also, one factor model satisfied Gorsuch (1983) criterion to further verify the unidimensional nature of the construct, where the percentage of variance explained by factor 1 was nearly thrice when compared by the percentage of variance explained by the next subsequent factor 2. Monte Carlo principal component analysis method also revealed single factor for this variable. Implications of the findings are discussed.

Configural, Metric and Scalar Invariance Measurement of Academic Delay of Gratification Scale

The present study measured the partial (configural, metric and scalar) invariance of Academic Delay of Gratification Scale (ADOGS) for college students prepared by Hefer Bembenutty and Stuart Karabenick (1996), with respect to gender. 488 students (277 boys and 211 girls) from engineering, law, pharmacy, and education faculties of Sultan Ul Uloom Education Society, Hyderabad, were samples of the study. Confirmatory Factor Analysis framework was used to test the three invariance, one at a time, using SPSS AMOS Ver.23. This was done by constraining the factor structure, factor loadings and intercepts for the genders boys and girls simultaneously. The SPSS AMOS CFA test provided Chi-square value and degree of freedom for the unconstrained and the constrained structures. By comparing the difference in the Chi-square test p-value for a degree of freedom of the constrained factor structure with the unconstrained structure’s Chi-Square p-value and its degree of freedom using an excel sheet calculator of Chi-square test available on the Internet, the measurement invariance of the instrument was verified. The findings of the study reveal that the scale is configural and metric invariant, as there is no significant difference between constrained and unconstrained structures. However, the scale is not scalar invariant. The educational implications are discussed.