59 / 372
臺大管理論叢
第
26
卷第
2
期
59
shows that cognitive social capital is the most important dimension in explaining users’
knowledge sharing. With shared language and narratives, SNS users are more willing to
conduct knowledge sharing. Thus, marketers should identify and cultivate groups of people
who are more likely to build up abundant cognitive social capital through interactive
platforms such as interest-based clubs on SNSs and then motivate both group leaders and
members to engage in knowledge sharing. Marketers could also create commonly shared
language, narratives, and product storylines so that the resultant cognitive social capital leads
to users’ knowledge sharing self-efficacy and knowledge sharing behavior. Furthermore,
although relational social capital has no significant impact on either knowledge sharing self-
efficacy or knowledge sharing behavior, it still has a significant and positive influence on
psychological safety, which in turn significantly affects knowledge sharing self-efficacy.
Thus, cultivating groups of users with high levels of mutual trust, norms, obligations, and
group identification can help marketers strengthen users’ perceived psychological safety and
knowledge sharing self-efficacy.
5.2 Limitations
We collected data for this research from Facebook only, and thus the generalizability of
the findings must be taken with caution. Moreover, because the majority of respondents were
young female students, the representativeness of our sample may be limited. Our data may
differ from data collected from other types of SNSs (e.g., Twitter, Plurk) in terms of user
profiles. In addition, this article does not rule out the possibility of two-way causation
between the endogenous latent variables because it is often difficult to exclude reciprocal
relations in structural equation models using cross-sectional data. Methodological problems
of this nature might be alleviated by adopting a longitudinal research design that uses staged
data collection techniques.
5.3 Future Research Directions
One direction for further research pertains to the mediating roles of psychological safety
and knowledge sharing self-efficacy. Although we show that knowledge sharing self-efficacy
mediates the effect of cognitive social capital on knowledge sharing behavior, other
mediating variable may exist between social capital and knowledge sharing behavior. In
addition, research could refine the measurement items, which were mostly borrowed from
the organizational behavior field, and extend them to fit the SNS setting, thus improving
measurement validity. Finally, this study broadly measured knowledge sharing behavior with