Following the curriculum reform trend around the world, how to achieve collaborative actions in schools had become a key element for the enactment of national curriculum (Hogan, Thompson, Sellar, & Lingard, 2018).

The promotion of professional learning community (PLC) has been seen as an effective way to provide professional development to teachers through constructing collaborative teams at school sites in Europe and worldwide(DuFour, Eaker, & DuFour, 2005; Hord, 1997; Blitz & Schulman, 2016).

However, the strategies to enhance the quantity and quality of PLC at school nationwide are under discovering. The instructional leadership by teachers, principals, and other local-level actors were mentioned as one of the most important tools(Adolfsson, & Alvunger, 2017). One theoretical approach relies heavily on the collaborative or network governance approach (Sørensen & Torfing,2005; Burns, Köster, & Fuster,2016). Traditionally, the goals of curriculum management are done through the national evaluation mechanism outside (or cooperation with) the schools(Ball, 2009). There is an emerging trend away from traditional local hierarchical governance and towards commercialized network governance ( Greany & Higham, 2018).

The network governance approach is built based on trust, joint actions, and partnerships (Considine & Lewis, 2003; Greany & Higham, 2018). there exist different dimensions when discussing the focus of governance network.

For example, Provan and Kenis (2008) provided an analytic framework from structures and relations between members in a network, and identified three typical models of governance: Shared governance networks, Lead organization networks, and Network Administrative Organizations. The shared governance network is more flat and horizontal when the Lead organization networks remain hierarchy between schools and administrative sectors.

However, how to attain a balance between top-down process and practice in the local site (i.e. school sites) remains a more and more intensive challenge for the local level sectors following the rapid change in education reform all over the world(Brown, & Duguid, 2000).

The research's argument is on if the variance between network models affecting the operation and sustaining of PLC at schools? What kind of policy practices is significantly affecting the promotion of PLC?

In Taiwan, we are processing our version of curriculum reform-oriented to "competency-based learning and teaching" (Chen, & Huang, 2017). Following the de-centralized education reform in the 1990s, the local government is authorized to the curriculum promotion in K-12. Besides, the consulting group and curriculum inspection system have had become the most crucial institution in the curriculum promotion mechanism in Taiwan since 1997. However, the impact of the consulting group on curriculum reform is questioned under the rapidly changing trends of learning revolution and curriculum development in the global competition era.

The purpose of this research lies in three dimensions as follows:1). What are the roles and strategies of the local level curriculum consulting group using in the curriculum governance process?2). What factors constrain the building of a local level curriculum development network? 3). What are the crucial elements that can help local agents to build up the governance network?

The result shows that:

1).The impact of the horizontal network was affected by the institutions setting by the hierarchical network. Both two types of the network need to be taken into consideration when discussing PLC development.

2).The horizontal network building process is heavily affected by the strategies chosen by the middle-level curriculum leaders in schools and local government.

3).The agents outside the schools showed their influence on the PLC through the horizontal network.

4).The governance network played a crucial role in the sustainable development of PLC.

5).The multiple stages regression results indicated that commitment of local government had a positive effect on the PLC under control of school environment relative variables.

Mixed-method research designs are used to answer various questions in the research. The first stage was a multiple qualitative case study (Stake, 2013). Three metropolitans in Taiwan were selected as cases to reveal the process of the network building process. The government documents are collected and analyzed in the research process, including the articles of association for the consulting groups, the plan and meeting record for the promotion of undergoing new curriculum guidelines, and the annual report of the local government of the relative programs. There are several kinds of participants were interviewed by the researcher in the research: 1)the curriculum inspectors,2)the local administrative staff, 3)the principals who were responsible for the affairs of the consulting groups, and 4)the scholars working along with the groups. There were 32 participants in 4 cities invited into the research. The interviewees are invited to participate in a 30-60 minutes face-to-face interviewed by the researcher. The records are translated into the transcript to apply for content analysis. For the data analysis, data were then entered into Maxqda software and one author worked independently to develop a list of initial codes. The grounded theory coding method is applied in the research(Corbin & Strauss, 2007). The transcript is open coding firstly and aggregated into topics, concepts, and axis. The cross-analysis of cases and axis is made and counted for the description of the network construction at the local level. The second stage was a survey. Clustered random sample strategy was used and 750 elementary and junior high schools were selected. The principal, curriculum leaders, and program coordinators were invited as participants. Of the 2993 participants, there were 1108 male teachers (37%) and 1885 female students (63%). The survey utilized a Likert-type scale to gather information about network relationships and attitudes for the governance. For the instruments, the questioner was composited with a basic data section, a network connecting experience, perceptions of curriculum collaborative governance scale(according to the results in the first stage ), teachers’ work settings(adopting from Leithwood & Jantzi, 2006), and perceptions about the Implementation of Learning Community(adopting from Wells &Feun, 2007). The network connecting experience response to 14 departments of local government was recoded to 1 or 0 to calculate the network density. The hierarchical linear regression modeling with the Maximum Likelihood method was used to evaluate the relationships among variables in cities and individuals.

The main results from the qualitative phase showed as follows: 1.The PLCs in schools and the development of local networks were both affected by the national wide institutions. Both two types of the network need to be taken into consideration when discussing PLC development. 2. The horizontal network building process is heavily affected by the strategies chosen by the middle-level curriculum leaders. However, the policy context, resources, and leadership have their effect at the same time. 3. Effective advisors from other sectors (center government, private, or third sectors) are needed to cover the limited capacities of local government. But the inconsistency of goals and difficulties on the accountability becomes other issues during the cross-hierarchy network building. 4. The loosely coupled issue network among public and private members, inspectors and teachers, scholars, and practices are necessary for the sustainability of curriculum innovation and changing schools' daily practice. 5.The regression results showed that commitment of local government, coordination with school administrators, positive teachers’ work settings had a positive effect and teachers’ work loading had a negative effect on the PLC implementation. Meanwhile the desity of network showed no significant effect on the teachers perception of PLC. Finally, for the local government, the researcher suggests the public managers put most of the resources on the demands of network building, especially on the curriculum innovation network including multiple layers and sectors. For the central government, the research reveals a model aiming for a long-term goal of sustaining curriculum innovation in an era of change. There need more collaborations with different sectors, and less abolition and merge of public sectors.

Adolfsson, C. H., & Alvunger, D. (2017). The nested systems of local school development: Understanding improved interaction and capacities in the different sub-systems of schools. Improving schools, 20(3), 195-208. Ball, S. J. (2009). Privatising education, privatising education policy, privatising educational research: Network governance and the ‘competition state’. Journal of education policy, 24(1), 83-99. Blitz, C. L., & Schulman, R. (2016). Measurement Instruments for Assessing the Performance of Professional Learning Communities. REL 2016-144. Regional Educational Laboratory Mid-Atlantic. Brown, J. S., & Duguid, P. (2000). Balancing act: How to capture knowledge without killing it. Harvard business review, 78(3), 73-73. Burns, T., Köster, F., and Fuster, M. (2016), Education Governance in Action: Learning from Case Studies, OECD Publishing, Paris Chen, H. L. S., & Huang, H. Y. (2017). Advancing 21st century competencies in Taiwan. Asia Society, Center for Global Education. Considine, M., & Lewis, J. M. (2003). Bureaucracy, network, or enterprise? Comparing models of governance in Australia, Britain, the Netherlands, and New Zealand. Public Administration Review, 63(2), 131-140. Corbin, J., & Strauss, A. (2007). Basics of qualitative research Techniques and procedures for developing grounded theory (3rd ed.). Thousand Oaks, CA Sage. DuFour, R., Eaker, R., & DuFour, R. (2005). Recurring themes of professional learning communities and the assumptions they challenge. On common ground: The power of professional learning communities, 7-29. Greany, T. and Higham, R. (2018) Hierarchy, Markets & Networks: Analysing the ‘selfimproving school-led system’ in England and the implications for schools. London: UCL IOE Press. Hogan, A., Thompson, G., Sellar, S., & Lingard, B. (2018). Teachers’ and school leaders’ perceptions of commercialisation in Australian public schools. The Australian Educational Researcher, 45(2), 141-160. Kohl, K.T. (2014). Teachers' Perceptions of Becoming a Professional Learning Community. Walden Dissertations and Doctoral Studies. Leithwood, K., & Jantzi, D. (2006). Transformational school leadership for large-scale reform: Effects on students, teachers, and their classroom practices. School effectiveness and school improvement, 17(2), 201-227. Meier, K. J., & O'Toole Jr, L. J. (2003). Public management and educational performance: The impact of managerial networking. Public administration review, 63(6), 689-699. Sørensen, E., & Torfing, J. (2005). The democratic anchorage of governance networks. Scandinavian political studies, 28(3), 195-218. Stake, R. E. (2013). Multiple case study analysis. Guilford press. Wells, C. &Feun, L. (2007). Implementation of learning community principles: A study of six high schools. NASSP Bulletin, 91(2), 141–160.