Structure Correlation Between Knowledge-Oriented Culture, Knowledge Sharing, and Innovative Management at Junior High Schools - Innovative Management in Information and Production - page 319

Tabl e 8 Summary of mediation effect for mediator variable (KS) between
KOC and IM
Mediation effects of three routes Direct effect between

Variable Route A Route B Route C KOC and IM

0.69 ∗ 0.35 ∗ 0.60 ∗ 0.84 ∗

η 1 KS

* p <. 05

also reach the significant level. Finally we control A and B and see if Route C
coefficient is changed. If Route C coefficient becomes 0 or much smaller so that it
can't reach the significant level, we believe that there is a strong mediation effect. If
Route C coefficient becomes a little smaller and still reaches the significant level, we
can only claim that there is a weak mediation effect ( Barron and Kenny 1986 ). First,
Route A coefficient, 0.69, is proved to reach the significant level. Second, Route B
coefficient, 0.35, is also proved to reach the significant level. Finally, when Routes A
and B are controlled, Route C's coefficient becomes a little smaller but still reached
the significant level. Therefore it is proved that there is a weak mediation effect, and
KS is proved a weak mediator variable (Table 8 ).

7 Implication

First, according to the findings above, KOC is proved to have a significant influence
both on KS and IM at junior high schools in Taiwan. In other words, if the school's
KOC is positive, it will promote its staffs' KS and innovation. On the other hand, the
negative KOC will hinder KS and innovation at schools. Therefore, the researcher
suggests school leaders or local superintendents devote themselves to shaping a
positive KOC at schools, especially focus on “trust and autonomy” of KOC for
promoting staffs' KS and “emphasis on innovation” of KOC for promoting IM.
Besides, KS is also proved to have a significant influence on IM at junior high
schools in Taiwan. That means if the school staffs are willing to share knowledge
with each other, the IM of the school will go well. However, the research found that
staffs' willingness of KS performs better than their actual behavior. Therefore, we
suggest schools, school leaders, and local superintendents provide more incentives
to encourage their staffs to share in action.
Moreover, although SEM of KOC, KS, and IM is proved to have a good model
data-fit, KS is merely proved to have a weak intermediary effect, which implies
there are still other intermediary variables not included in the analysis. Therefore
the researcher suggests other factors such as leadership of principal, organizational
health, organizational commitment, and job satisfaction could be included for the
further researches.
Finally, this research collected data only through questionnaire survey, and some
of them are possibly not showing the true thoughts and feelings due to various
factors. Hence qualitative research methods, such as interview, field observation,
and long-term follow-up case study, are recommended in the future researches.