Many companies use Facebook pages (hereinafter referred to as FB pages) as a part of integrated marketing communication strategies. In previous research, despite the fact that many companies have developed FB pages, studies either took the approach of studying only a small number of FB pages, or they collected the data of a large number of FB pages, aggregated statistics. Therefore, this paper, applying consumer response data for the FB pages of 38 brands, examines the applicability of a hierarchical linear model. A hierarchical linear model is a method for accurately analyzing individual level (consumer) and a group level (FB page) hierarchical data. The results show that the group level error applies significantly better to the random effects model than to the fixed effects model. Regarding effect at the individual level, it is clear that the model assuming heterogeneity among FB pages is highly effective in explaining the dependent variable (recommended intention) by the independent variables (trust, interaction, commitment). Furthermore, the group level variables, age, gender ratio and popularity ranking, have no significant effect. Meanwhile, with respect to satisfaction and hedonic motivation, model fitness is improved not only by individual level effect but also by group level effect.
View full abstract