This section discusses the three-level research design and the rationale of case selection.
Research design 1
Three levels of research design are employed in the research (Table 3.2). At the first level, the research design aims at examining the phenomenon of framing effect suggested by the theoretical concepts from the behavioural decision theory. On the one hand, this level of the research design focuses on the decision makers’ perception toward the decision issues and follows the concept of detecting decision makers’
potential biases. On the other hand, it also emphasises the investigation of the choice context; specifically, the salience of the environmental cues and the sources of uncertainties. The use of the qualitative approach in this research design is to exemplify the similarities across cases. With the goal of seeking commonalities to identify general patterns, (Ragin, 1994b), the case selection strategy is to justify the representativeness of the selected cases for the validity of the plausible explanations.
Table 3.2 Research Design
Source: The author.
Following the experimental concept of theoretical sampling in the comparative approach, the cases are purposefully chosen and comparable in a defined category in which the variances of the contextual factors are minimized, yet the interested factors are saliently demonstrated (Ragin 1994, 2009). Firstly, to control the influencing factors from the different institutional and industry environments (DiMaggio, 1997;
Kaplan & Tripsas, 2008), the research setting is located in a single industry in the same country. In view of the uncertainties, the Solar Photovoltaic (PV) industry is selected as the appropriate industry setting. Taiwan is selected because the Taiwanese PV industry has experienced impressive growth for the past few years. These two conditions lead us to a specific investigation period of the years 2006-2011, to exemplify the influence from environmental change. Secondly, as the firms’ position in the industry value/supply chain is expected to influence their investment strategies (Adner & Kapoor, 2010; Teece, 1986), I focus on firms in a similar position in the industry value chain. PV cell firms are selected because they represent the vast
majority of the PV firms in the Taiwanese PV industry (PIDA, 2011; WealthPress, 2009).
Next, as the research concerns firms’ capability investments, the condition of similar networking and access to external technological resources is controlled. The participation of the government R&D innovation programs serves as the criterion.
Two programs from the Ministry of Economic Affairs (MOEA) are used, because they are benchmark government programs for R&D innovation projects in terms of both the scale and the strict review process by a group of experts from academia and the research community.
There are a total of eight PV cell firms, in four categories of PV technologies, which have been awarded MOEA programs over the past five years (2006-2011). However, not all of the eight firms could be studied, due to constraints upon resources and access. The five case firms were selected16 according to the below criteria (Table 3.3).
Firstly, three out of the four Crystalline Silicon (c-Si) firms were chosen because they had been awarded the most recent projects. The one excluded firm’s last award was in 2006. Secondly, the one selected Amorphous Silicon/Macro Crystalline Silicon Thin Film (a-Si/μc-Si TF) firm was awarded with more government innovative R&D programs in recent years than the other a-Si/μc-Si TF firm. Thirdly, the only Copper-Indium- Gallium-Diselenide Thin Film (CIGS TF) firm was selected. Lastly, the Dye Sensitized Solar Cell (DSSC) firm was omitted because the firm withdrew the DSSC business line in 2011.
16 Case firms are given as pseudonyms
Table 3.3 Case firm profile
Source: Company annual reports, MOEA projects and TSE (Taiwan Security Exchange) filing.
Compiled by the author.
Research design 2
The second level of the research design aims at examining the phenomenon of selective attention suggested by the theoretical concepts from organization decision theory. While this research design addresses the sources of variation at the organization level, it also emphasizes explaining which organizational factor or causal combination discriminates the patterns across firms. With the goal of exploring diversity in mind, the second level of research design has a stronger flavour of the comparative case approach than the first because it is “more concerned with causal complexity” in the organization.
To sharpen the comparison at the organization level, another condition is added to increase the homogeneity of the comparable cases (Ragin, 1997). With the condition of c-Si PV technology, the cases are narrowed down to the three c-Si firms. The reason for choosing c-Si technology is that this technology is relatively mature. Thus
these c-Si firms are similarly associated with a certain scale and public trading type.
In fact, these two factors obviously differentiate the three c-Si firms from the other two thin film firms (besides the technology). The exclusion of the two smaller firms not only facilitates the collection of public data but is also justified for the purpose of comparative analysis at the organization level.
Research design 3
Continuing the investigation of the phenomenon of selective attention as the source of variation, the third level of the research design examines whether individual decision makers’ attributes influence the framing effect. Specifically, the research design focuses on categorizing the relative influence of individuals’ professional backgrounds, and organizational decision contexts. With the goal of examining individual decision makers’ attributes and identifying sources of similarities and variation, the similar concept of the quasi-experimental approach is adopted in the research design, but with a modified implementation approach. The decision problems are not designed from hypothetical cases, but are drawn from the real business environment. Rather than following the verbal protocol approach in coding the specific term or semantic analysis to recognize cognitive patterns, the research design focuses on the description of the decision issue, and the analysis of the interaction with the decision context.
With the concept of the quasi-experimental approach in management cognition studies, each interviewee is purposely selected to meet the criterion of a professional background. Two functional groups – technology (R&D) decision makers and business (finance) decision makers – are categorized in the research. Initially, the
research targets ten senior managers (one for each functional group) from the five case firms. Finally, eight senior managers are available for interview. This size is considered reasonable as it matches the average group size of eight to nine in previous management cognition studies employing the quasi-experimental approach (Gregoire et al., 2009; Melone, 1994; Sarasvathy et al., 1998). Among the eight senior managers, four are in the technology function group while the other four are in the finance function group (Table 3.4). Most of the senior managers (except the deputy Chief Finance Officer) are the heads of their departments. In fact, it may be more appropriate to refer them as executives of the firms. Generally, their positions reflect their expertise in the function group. For example, of the four senior R&D managers who joined their firms since inception, three of them have PhDs and research experience in the field of PV technologies.
Table 3.4 Senior managers’ profile
Source : Company annual reports, interviews and compiled by the author.