Sequential mediation of parasocial relationships for purchase intention: PLS-SEM and machine learning approach

Nora Sharkasi,Saeid Rezakhah
DOI: https://doi.org/10.48550/arXiv.2307.00005
2023-06-08
Applications
Abstract:Companies employ social media influencers SMIs due to the compelling evidence of their advertising effectiveness, however, more research is required to identify and compare factors driving their success.We investigate the effect of source influence SI on the intention to purchase I2P through the sequential mediation of parasocial relationships PSR with benign envy BE and PSR with the perceived brand influencer fit BIF.Two independent samples N=411; N=355 from Europe and Southeast Asia are used to perform: i PLSSEM analysis to obtain the model's predictive power and ii classification based machine learning ML to evaluate the model's accuracy. Moreover, within study and between studies comparative analyses are performed. We use regression analysis and split test ML technique for validation. Both samples indicate a higher role of trustworthiness and expertise in forming SI. Furthermore, comparative mediation analysis and predictive accuracy scores show that the audience related feature, BE, played a more vital role in affecting/predicting the followers' I2P than the brand-related feature, BIF. Our findings contribute to the knowledge of SMIs' credibility and comparative analysis paradigms and provide a better understanding for marketing practitioners and researchers.
What problem does this paper attempt to address?
The main issue this paper attempts to address is how the Source Influence (SI) of Social Media Influencers (SMIs) affects followers' Intention to Purchase (I2P) through the sequential mediation effects of Benign Envy (BE) and perceived Brand-Influencer Fit (BIF). Specifically, the paper aims to: 1. **Study the sequential mediation effects**: - Examine the sequential mediation effect of BIF and PSR (Parasocial Relationships) on the relationship between SI and I2P. - Examine the sequential mediation effect of BE and PSR on the relationship between SI and I2P. 2. **Compare the roles of audience characteristics and brand characteristics**: - Explore which is more important in influencing and predicting the effect of SI on I2P: human-related factors (such as BE) or product-related factors (such as BIF). 3. **Method validation**: - Use the PLS-SEM (Partial Least Squares Structural Equation Modeling) method to analyze the predictive power of the model. - Use classification machine learning algorithms to evaluate the accuracy of the model and validate it through split testing techniques. Through these studies, the paper hopes to provide marketing practitioners and researchers with a deeper understanding, particularly regarding the credibility of social media influencers and comparative analysis paradigms.