49 / 414
臺大管理論叢
第
27
卷第
1
期
49
customer side of the dyadic relationship may not necessarily be reflected in the side of
service providers, and vice versa (Anderson and Narus, 1990). A possible improvement
would be to measure co-production by examining both sides of the dyadic relationship to
confirm the findings of this study, as well as to generate additional insights on the dynamic
interactions between these two parties.
Third, four items cannot capture the rich dimensionality of the co-production process
and could have undermined the reliability of measurement. Further research could develop
more extensive scales. For example, the measurement of co-production requires a more
complete conceptualization of S-D Logic. A possible improvement would be to incorporate a
broader range of dimensions, such as different elements and mechanisms embedded in
co-production.
Finally, the model of this study excluded possible antecedents and consequences of
co-production, such as customer expertise, commitment, self-efficacy, and role clarity (Auh
et al., 2007; Dong et al., 2008). In addition, this study may have missed some potential
relationships. For example, decision-making uncertainty and quality of customer interaction
may moderate the relationship between asset specificity and co-production. Furthermore,
most studies have addressed the relational benefits of co-production for the customers. A
different view on the relationship between relational benefits and co-production is that
relational benefits can facilitate customer participation in co-production. Although such
reciprocal relationships of relational benefits and co-production have not yet to be
empirically examined, the relationships between relational benefits and co-production can be
reasonably expected to be reciprocal. Therefore, this study urges researchers to evaluate an
alternative model. In summary, other variables and potential relationships may help explain
further the key relationships between customers and service providers.