With the rapid development of digital technologies and the Internet, the boundaries between countries are shrinking and markets are becoming global (Oh et al., 2016). Simultaneously, cross-border online shopping is another trend that has spread across the world, and global e-commerce has now become a reality (Johnson et al., 2003; Moore, 2015). However, in prior studies related to eWOM, the national culture has received little attention among the numerous factors that could adjust the effect of eWOM (Christodoulides et al., 2012). In addition, in the real world, although individuals are frequently exposed to combined eWOM messages containing both positive and negative information about the same product, most previous studies on eWOM have focused on the one-sided eWOM valence. There are not many empirical studies on the influence of the two-sided eWOM valence on consumers’ persuasions. Thus, Study I examines the attitude effect of the two-sided eWOM valence from a cross-cultural perspective, particularly on the basis of the differences in thinking styles between the Easterners and Westerners. For this, we use a 2[Valence: positive/negative, negative/positive] x 2[Culture: East (South Korea), West (United States)] factorial design. To classify the culture, the thinking style was measured as a within-group variable. As a result, the interaction effect between valence and culture (nation) was significant. Specifically, in the East (South Korea), no significant difference existed in the changes in brand attitude depending on the two-sided eWOM valence, whereas brand attitude changes from a negative/positive presentation order in the West (United States) were significantly larger than the positive/negative presentation order.
Study II demonstrates the mediated moderation effect of perceived cognition congruency in a cross-cultural setting for explaining the underlying mechanism. Drawing on the cognitive fit theory, we present a two-sided eWOM-consumers’ perceived cognition congruency proposition: the two-sided eWOM valence that matches the information processing order consumers habitually have would facilitate the favorable comprehension (reflected by perceived cognition congruency) and assessment (reflected by changes in brand attitude) of the reviews. As a result, the two-sided eWOM valence indirectly affects brand attitude changes by mediating perceived cognition congruency. This valence directly affects the brand attitude changes in the Westerner (United States) group, which has an analytical thinking style. However, the direct and indirect effects of two-sided eWOM valence on brand attitude changes are not significant in the Easterner (South Korea) group, which has a holistic thinking style. This examination might explain why differences in the changes in brand attitude between the Easterners and Westerners were revealed through the twosided eWOM valence, thus providing in-depth insights into consumer responses for the valence.
This study expands the diversity of studies conducted on the characteristics of the two-sided eWOM. Furthermore, it is expected to provide a strategic direction and practical implications for two-sided eWOM-driven information management by organizations.

• Hyun Hee Park(Kyungpook National University, Republic of Korea)
• Jung Ok Jeon(Pukyong National University, Republic of Korea)