Validation of Science Self-Efficacy Scale in the Indian Secondary School Students using Network Approach

Abstract

The Science Motivation Questionnaire – II developed by Glynn and Koballa (2006) for college students was used to validate the science self-efficacy items in it through network approach. 110 Indian secondary school students (49 boys and 61 girls) were the sample subjects of the present study. Owing to the ordinal type of the data obtained from this tool, it was validated using the fresh approach. Exploratory graph analysis (EGA) through Glasso technique revealed single cluster, confirmed further through structural consistency coefficient of 0.744 and goodness of fit estimates of the network, CFI.robust = 0.954, TLI.robust = 0.908, RMSEA.robust = 0.086 and SRMR_Bentler = 0.047, found using WLSMV based Ordinal confirmatory factor analysis. Regularized network, centrality indices, edge-weight accuracy confidence plot, correlation stability plot, edge difference plot, node difference plot and node predictability plot were also generated. The CS-coefficient of the network was obtained poor at 0.126 due to the less sample size of the study although. All these plots were generated using EGAnet, qgraph, bootnet, psychtools and mgm packages of RStudio openware. Node predictability coefficient R2 of the nodes 4 and 5 was the highest at 28.5% and 28.4 % respectively. Educational implications of the study on science education and its research are discussed.