4.5 Boundary Specification 4.6 Study Design 4.7 Survey 4.7.1 Administration of Instrument 4.7.2 Data Preparation 4.9 Analysis Plan
4.9.2 Exploratory Network Analysis 4.9.1 Exponential Random Graph
Model (ERGM)
4.8 Definitions and
Operationalization of Study Variables 4.8.1 Network Structural Measure of Embeddednessss 4.8.2 Organizational Structural Measure of Embeddednessess 4.8.3 Relational Capital Variables
117
4.2RESEARCHSTRATEGY:SOCIALNETWORKANALYSIS(SNA)
This research follows the exploratory and statistical social network analysis approach found in literature studies in order to determine how the impact of firms‟ embeddedness in the centralized upstream supply network impacts upon the firms‟ relational capital outcomes. In this section, the researcher briefly discusses and justifies the adoption of the SNA methodology.
Structuring of network of relations has an important implication for actors of the various networks (Knoke and Yang, 1998). Given a collection of actors, a social network analysis can be used to study the structural variables measured on actors in the respective network. These structures involve the pattern of ties between the actors. A network analyst would seek to model these ties to depict the structure of a group. One could then investigate the impact of these structures on the functioning of the network or the influence of these structures on actors embedded within these network structures (Hanneman and Riddle, 2005).
Investigation of the implication of these structures upon the embedded firms requires a method that can analyze not only the characteristics of the actors, but also the relations between the firms that form the structures. Wasserman and Faust (1994) documented that the unit of analysis in network study is not just the actor, but consists of an entity made up from the collection of the actors and the linkages among them. An actor of the network, stressed Knoke and Kuklinski (1982) can be an individual, a team or even organisations. Consequently, the unit of analysis of this study comprises the relationships between the firms and the attributes of firms of the APMMHQ-1 upstream supply network.
In addition, Wasserman and Faust (1994) as well as Lusher (2000) argued that the typical statistical method and analysis are not adept at measuring relations. One important fact behind this argument is that standard statistical analyses disavow the existence of relationships between firms in a network through the assumption of independence of observation. However, the network approach, more specifically the Social Network Analysis (SNA), focuses on the relations between
118
firms and the pattern of the relations and the implication of the relationships (Wasserman and Faust, 1994).
The relevance of proper management and understanding of the supply network from the SNA perspective has been discussed and proposed in literature (e.g Ahuja, 2000; Corteville and Sun, 2009; Krauss, Mueller and Luke, 2004). Choi, Dooley and Rungtusanatham (2001) propose that supply network is rather a complex adaptive system consisting of both hard ties (e.g. materials) and soft ties (e.g. knowledge flow) among the organisations in a supply network. Choi and Hong (2002) map the complete supply network for the centre consoles assembly of an automobile manufacturer with three different assemblers. The authors provide several propositions regarding the operation of supply network structures, relating to the structural characteristics of formalization, centralization and complexity. Carter, Ellram and Tate (2007) stated that social network theory was a useful tool with which to study the influence in the supply chain. Kim et al. (2010) present a conceptual definition of supply network elements based on the SNA methodology.
Borgatti and Li (2009) stated that the social network analysis concepts were particularly suitable to study how the patterns of inter-organizational relationship in a supply network translate to competitive advantage. This can be achieved through management of the hard ties and soft ties in the supply network. Furthermore, according to Borgatti and Li (2009), adoption of the social network analysis to the study of the supply network will allow a better understanding of the operations of the supply networks, both at the individual level and the network level. This determines the importance of the organisations, given their position in the network and how the network structure affects individual organisations and the network performance as a whole. Consequently, this study adopted the social network analysis method strategy for data collection, analysis and reporting of results, as this is the most appropriate means for arriving at valid results and testing the hypotheses set forth in this study (Marouf, 2011). The following section will discuss the research site of this study.
119
4.3THERESEARCHSITE
The shipbuilding industry in Malaysia comprises firms that are involved in designing, building and construction of various types of ships, such as: ocean going, passenger, offshore and fishing vessels.
The shipbuilding industry has existed in Malaysia since the early 1900s. One of the preliminary shipyards was built in Kuching, Sarawak. Since then, more shipyards have emerged in the country. Shipyards in Peninsular Malaysia can be found in Lumut, Perak, Port Klang, Selangor, Kemaman, Terengganu and Pasir Gudang, Johor. Currently, there are about 70 registered shipyards in Malaysia.
The shipbuilding industry in Malaysia is largely divided into two foremost clusters comprising two principal regions, i.e.: the Eastern Cluster and the Peninsular Malaysian Cluster. The Eastern Cluster produces large vessels such as tugs, barges and river ferries. Shipyards in this cluster were considered as cost effective, viable and dynamic due to their close proximity to the main market, namely, the oil and gas sector.
In the Peninsular Malaysian cluster, the shipbuilders specialize in steel and aluminium vessel buildings for government, as well as the oil and gas sector. Figure 4.1 displays some of the various shipyards currently operating in Malaysia in the Eastern Cluster and the Peninsular Malaysian Cluster respectively.
120
FIGURE 4.2MAP OF SHIPBUILDING INDUSTRY IN MALAYSIA
Among the shipyards included in Figure 4.2 are leading shipyards such as: Malaysia Marine and Heavy Engineering Sdn Bhd (ships, vessels, FPSO - Floating, Production, Storing and Offloading and FSO – Floating, Storing and Offloading); Boustead Heavy Industries Corp. Bhd. (shipbuilding, ship repair and fabrication); Labuan Shipyard and Engineering Sdn Bhd (shipbuilding, ship repair, naval craft maintenance and oil and gas fabrication); Muhibbah Marine Engineering Sdn Bhd (shipbuilding and ship repair); Coastal Contracts Berhad (builders of barges, AHT – Anchor Handling Tugs); Kencana Petroleum (fabrication for oil and gas); Brooke Dockyard and Engineering Works Corporation (repair and building of ships and oil and gas offshore modules); Ramunia Fabricators Sdn. Bhd. (repairs and support for offshore operation); NGV Tech Sdn Bhd (shipbuilding and repair); Nam Cheong Dockyard Sdn Bhd (offshore support vessels) and Hong Leong Lurssen Shipyard Sdn Bhd (ship-building and repair). In terms of product, these shipbuilders handle Malaysia‟s normally developed simple, low-cost fibre glass boats primarily for the fishing and tourism industries. Medium-sized vessel building projects include offshore support vessels
121
(OSV), tugs, barges, patrol crafts and the like. The construction of huge vessels, on the other hand, has seen better days and is fast ceasing in business. This is mainly due to the Malaysian Marine and Heavy Engineering (MMHE) strategy of focusing only on repair and conversion, consequently leaving Boustead as the last standing large vessel builder.
Although the industry is not widely known, vessels developed by the shipyard are sold worldwide, especially to Middle Eastern oil and gas companies. The shipbuilding industry has also been acknowledged as a critical industry due to its spill-over effects. For example, the Organization for Economic Co-operation and Development (OECD) has acknowledged the industry as having strategic importance in terms of job creation, industry capacity and technological capability.
The shipbuilding industry offers great potential to the economy of Malaysia and has significant room for improvement. For instance, in 2010, the Malaysia shipbuilding industry was ranked 22nd in the world in terms of production and value, while only accounting for less than two per cent of the world vessel production.
According to the President and Chairman of the Association of Marine Industries of Malaysia (AMIM) Tan Sri Ahmad Ramli Mohd Nor, Malaysia is aiming to capture two per cent of the worldwide shipbuilding industry by 2020 from one per cent in 2010. Ahmad stated that the strategy for attaining Malaysia‟s vision to be an all encompassing player in the shipbuilding industry is to improve two important items, specifically: skilled personnel and the supply chain.
For the purposes of this study, a centralized upstream supply network of a small maritime industry seemed to be an ideal setting. A supply network in the maritime industry is a material- intensive enterprise. Much of the activity is highly dynamic and is widely dispersed throughout the network. The flow of materials and information is transferred through interactions among different firms. Because firms in a supply network operate in an environment having a high degree of complexity (Bozarth et al., 2009) and uncertainty (Wilding, 1998), these firms seek an edge through connections or interactions with members of the network. Lambert and Cooper (2000) stated that
122
the key to these issues is the on-going relationship with other partners. They stressed the importance of investigating the relationships that suppliers and customers have. Johnston et al., (2004) suggested that on-going relationships among members of the supply network increase efficiency and effectiveness of the supply network.
The focal research site of this study is located in the Peninsular Malaysian cluster. The network, labeled here as APMMHQ-1, is part of the centralized upstream supply network. APMMHQ-1 is a company in the Malaysian shipbuilding industry involved in ship repairs, maritime, engineering and related service provider matters. To date, the company has awarded contracts to local vendors and suppliers totaling RM 31 million for the development of small vessels in the region. Recently, the company invested RM 100 million to create new facilities in different locations across Malaysia to develop and service small vessels in the country. Efforts are being undertaken to determine partners for the operations. APMMHQ-1 has also crafted a vendor development program to work with the small and medium enterprises, attracting some firms to supply SBSR products and services.
APMMHQ-1‟s centralized upstream supply network was considered to be one of the best supply systems in the region through its Integrated Logistic Support (ILS) programs. Top level management was approached for possible participation in the study. After several communications about the goal of this study and the potentials' benefits for the APMMHQ-1 supply network, positive commitments were received from the top management to participate in and grant participation for this study.
In network studies, all the actors who are located within the naturally-occurring boundaries are included for analysis. Consequently, network studies do not use samples as in the conventional sense; rather, it seeks to include all the actors in some population or populations (Hanneman and Riddle, 2005).
123
centralized upstream supply network of APMMHQ-1. More specifically, this study investigates the firms operating in the upstream supply network of APMMHQ-1 relating to the supply of parts and materials for the production of Rigid Hull Inflatable Boat (RHIB) to the APMMHQ-1. In APMMHQ-1 production, the RHIB is a small, fast craft that received the highest demand from the market. Thereby, the upstream supply network for the RHIB product is one of the most active networks of firms in the APMMHQ-1 vast supply network variety.
In the following section, the researcher discusses the strategies applied in determining the network members in the APMMHQ-1 upstream supply network for the product RHIB.
4.4DETERMININGSTUDYSAMPLE
The first step of social network analysis is to determine the population of the study to be surveyed. There are two sampling units in this study, namely: the firms that occupy the APMMHQ-1 upstream supply network for the product RHIB and the ties or relationship between
them. The sampling frames for the firms and for the connections between them are nested. In
network studies, the method used to sample relations is part of the survey instrument.
As mentioned, in network studies, determining the boundaries of a network is of utmost importance in a network study (Hanneman and Riddle, 2005). To identify and define the target population within the APMMHQ-1 upstream supply network for RHIB for this study, the author combines the realist and the nominalist approaches. Nominalist and realist approaches are parts of the boundary specification strategy of this study and are discussed next.
4.4.1BOUNDARYSPECIFICATION
One of the difficulties in conducting social network research is that of determining the boundary for the network study (Wasserman and Faust, 1994). An accurate boundary specification technique will allow the network researcher to identify the target population, as well as permitting
124
may produce erroneous network measurements. The difficulties associated with setting up the proper boundary specifications in network study demand careful treatment of any particular strategy chosen by the researcher (Knoke and Kuklinski, 1982).
Laumann, Marsden and Prensky (1989) discussed the problem of specifying network boundaries in network research and studied the pattern of boundary specification techniques adopted by network researchers. In their review of various boundary specification techniques adopted and implemented by network researchers circa 1989, the authors concluded that the issues arise as network researchers have generally restricted their research to using either the realist or nominalist strategy alone. These two strategies will be discussed in the following sections.
Boundary Specification Technique: Realist Strategy
In the realist boundary specification strategy, the researcher presumes the boundary to be the limit that is experienced by all or most of the actors in the network (Knoke and Kuklinski, 1982). Such boundaries include kinship, friendship or directorships. Laumann, Marsden and Prensky (1989) described this as the vantage points of the actors in the network. Saunders (2007) relates the realist approach to define network boundaries based on actors‟ own interpretations of whether or not their organization is a part of the network in question. Thus, under the realist boundary specification technique, the inclusion and exclusion of actors inside and outside the boundary depends on whether the other actors view themselves to be part of or connected to the network members or not. A specific technique under the realist approach is the name generator or “snow ball” method (Scott, 2000). Under this technique of boundary specification, the researcher extracts names of other actors from a sample of actors and then asks these sample actors to name those actors of the particular relationships to which they are tied (Wasserman and Faust, 1994).
One application of this technique was the work of Choi and Hong, (2002). The authors applied this principle to identify the suppliers who exist in the upstream supply network of Honda in the United States for the Acura model‟s centre console assembly. In their research, the authors
125
asked the upstream suppliers of Honda's supply network to list and name the suppliers for the materials and parts for the centre console assembly of the Acura model. They excluded any organisations that were not related to the supply of the particular parts and materials.
Boundary Specification Technique: Nominalist Strategy
The nominalist boundary specification strategy is based on the researchers‟ own perceptions and constructs with regard to their theoretical interests. This involves seeking out those actors who are of interest, as well as finding out the extent of links between the actors in the network (Knoke and Kuklinski, 1982). In the nominalist approach, the researchers draw the boundary by developing a conceptual framework to serve the researchers‟ analytical purpose. In practice, under the nominalist strategy, the network analyst will determine the characteristics defining the membership of the network. Using these characteristics, the researchers will select the related actors and then proceed to study the interaction between the identified network actors. For example, legal or other formal membership requirements can represent a clear boundary construct for network research.
Limitations of Nominalist and Realist Strategy
Both the nominalist and the realist approaches each has their limitations (Diani 2002). Diani (2002) stated that the nominalist approach may sometimes cause the researcher to focus too much on certain traits or attributes of groups of actors while disregarding the concrete relationship between them, which is the essence of network research. Another concern with the nominalist approach, argued Diani (2002), is that the researchers‟ own theoretical interest could restrict or constrain the network (Knoke and Kuklinski, 1982).
The realist approach also has its limitations. Saunders (2007) argued that the realist strategy may remove from the network actors who do not define their identity in the same vein as do the rest of the network actors. Second, there is also a lack of guidance with regard to the appropriate place at which to stop the process of a snowballing procedure (Saunders, 2007).
126
Overcoming the limitations of realist and nominalist approaches requires a combinatorial strategy which is discussed as follows.