Determinant Factors on Student Empowerment and Role of Social Media and eWOM Communication: Multivariate Analysis on LinkedIn usage

Abstract

Background/Objectives: In recent times, there is phenomenal increase in usage of Social Networking Sites like Facebook, LinkedIn etc. by college students and young professionals. This study focuses on identifying key factors that influence LinkedIn usage and the role of eWOM communication in enhancing social connectivity and engagement of students in meaningful activities to improve their social and academic standings. A theoretical model on social networking by students is proposed and the results and recommendations of this study will be brought to practical use towards student empowerment. Methods/Statistical Analysis: A preliminary survey was conducted to understand how young university students use the Social Networking Site LinkedIn and the responses were used to frame a questionnaire. A second level survey was conducted among the same set of participants by collecting their responses in five point Likert Scale. Exploratory Factor Analysis was conducted using the LinkedIn Survey responses to identify the hidden factors associated with the indicator items in the data set. Subsequently, a theoretical model was constructed using Structural Equation Modeling principles, depicting the interrelationships between the latent constructs and indicator items constituting a measurement model and a structural model. Four Hypotheses were framed such that Social Media Usage and eWOM communication have significant positive effect on Student Empowerment. Finally, Confirmatory factor analysis was done to prove the hypotheses and to analyze how well the model fits into the theory. The software IBM SPSS, and AMOS 23 were used to perform multivariate statistical analysis on the LinkedIn Survey response items. Findings: The exploratory study on LinkedIn Usage Survey responses revealed three latent factors that accounted for 69.462 percent of the total variance. The three key factors explaining the eWOM behavior of students in LinkedIn usage were Expert Opinion Seeking, Networking with Professionals and Notification of Profile Changes. The latent factors and associated relationships were used to frame a theoretical model based on SEM techniques. Based on Confirmatory factor analysis done on this model using the data set revealed that the model supported the hypotheses H1, H2, H3 and H4 and all indicators in the model significantly loaded to their respective factors and the predicting variables had a significant positive effect on the predicted variable. The factor loadings were fair to excellent ranging from .634 to .853 and the test for model fitness showed good fitness result based on value of various fitness indices which were within accepted limits. Based on CFA, the important fitness indices and their values arrived at were: CMIN/df = 2.022, NFI = 0.824, TLI = 0.887, RMSEA = 0.098 and CFI = 0.901. Improvements/Applications: The accuracy of the predicting ability of the proposed theoretical model can be improved by augmenting this research study and statistical analysis to be extended to a larger target group belonging to different institutions to achieve good model fit as well as for testing the scalability of the model. As a future work, this model can be integrated with online learning systems also with the aim of improving student engagement in the current online learning scenarios.