Development and Validation of Scale SMPS for Assessment of Procrastination Due to Social Media Among University Students

Abstract

This article explores the impact of social media on the procrastination behaviors of university students, specifically through the development of the Social Media Procrastination Scale (SMPS). The major objective is to understand the underlying factors contributing to academic procrastination due to social media use. The scale was developed with 50 items across three factors: Time management skills, engagement in academic activities, peer and social influence expert opinions led to the removal of 19 items with poor content validity (below 0.45). The remaining items were administered to university students via Google forms shared on WhatsApp groups. Exploratory factor analysis with varimax rotation validated the scale, resulting in a KMO measure of 0.771 and a significant Bartlett’s test (p < 0.001). The scale’s reliability was confirmed with a Cronbach's alpha of 0.738. The final structure consists of three factors, providing a robust tool for assessing social media procrastination. To further validate the internal structure, Confirmatory Factor Analysis (CFA) was performed using AMOS-21. The CFA results indicated ,the Chi-Square test and CMIN/DF suggest a decent fit, as the ratio is below 3.The baseline comparisons (NFI, RFI, IFI, TLI, CFI) indicate that the model fit could be improved since these values are lower than the typical threshold of 0.90.The RMSEA is below 0.08, which suggests an acceptable fit, but closer to 0.05 would be preferred. AIC, BIC, and CAIC values are provided for comparison with other models. Lower values are preferred, indicating a better fit. The ECVI value also indicates how well the model would perform on a new sample, with lower values being better. Overall, the model shows a reasonable fit with room for improvement, particularly in indices like IFI, NFI, and CFI, which are below the recommended thresholds. The RMSEA indicates an acceptable fit, and the CMIN/df ratio supports the model's overall appropriateness.