In addition to the three reasons for choosing case study, another important aspect is to appreciate the suitability of case study to underlying components of critical realism. Six relevant points are discussed below.
7.5.1. Perspective on Phenomenon
Innovation (product, process, marketing and/or organizational) within a low-technology sector is the phenomenon under investigation. Sayer (2004) points out that the events are external and visible outcomes of the behaviour of objects/entities (marble mining and processing firms). Thus the phenomenon of innovation within the context of marble SSI is under investigation. This phenomenon manifests itself in the form of product, process, marketing and/or organizational innovations resulting from the behaviour or actions of marble firms that are influenced by the internal and external context within which they operate.
7.5.2. Perspective on Context
Critical realism provides a complex view of contexts and appreciates the interwoven nature of the epistemic transitive construction of reality. It leads us to conceptualize frameworks that focus on the interactions between context and phenomenon in question (Layder, 1993). This is particularly useful because this research applies the micro-meso-macro analytical framework (Chapter 4) to look at firm-specific LT innovations within the context of SSI while also appreciating the key role of individuals within firms. The SSI consists of elements and interactions among them (structure).
7.5.3. Perspective on Boundary
Easton (2009) suggests that a critical realist approach is more suited for
However, it is less suitable to phenomena that are highly qualitative in nature for instance human behaviours and phenomena that are highly quantitative in nature such as sales trends in industry. Suitable phenomena may include organizations or inter-organizational relationships. However, the boundaries between the phenomena and context may be flexible and subject to change in line with the nature of research or the questions it seeks to answer. This is particularly important because critical realism focuses on determining the causal mechanisms underlying the objectivist ontological world which in turn might require imposing certain limitations on the boundaries to determine causality.
Low-tech innovation cannot be characterized as a highly qualitative or quantitative phenomenon. It manifests itself in tangible forms like product or process innovation thus not having qualitative characteristics in the real sense.
On the other hand it is not really a quantitative phenomenon characterized by highly number-oriented data. However, innovation is a complex and systemic phenomenon (Chapter 3) influenced by a variety of factors. Also, this research applies certain boundaries to the phenomenon of firm-specific low-tech innovation by bringing in the ‘level of aggregation’ concept derived from SSI approach. Under this, boundaries of the phenomenon under investigation are influenced by product groups (marble of different shapes, sizes and colours excavated from mines plus different types of products in processing units), technologies and sets of activities. However, a case study is useful particularly when boundaries between the phenomena and context are not clearly drawn.
This stands true in this research because sometimes it is difficult to determine that which aspect of LT innovation has been influenced by the firm-specific factors and which one is affected by the contexts of non-firms, institutions, knowledge-base, technologies and interactions.
7.5.4. Nature of Research Questions
A central tenet of critical realism is that it attempts to identify underlying causal mechanisms that can explain phenomena (Table 7.1). This means asking questions like what caused events or how and why they occurred. The case
phenomena to explain the events that resulted from the action of entities or objects (human and non-human). The questions posed for this research study essentially ask how and why LT innovation occurs or does not occur thus helping to explain the causal mechanisms that need to be present for LT innovation to take place.
7.5.5. Flexibility in Choosing Data Collection Tools
Case study provides the flexibility of collecting data using multiple sources of evidence (quantitative, qualitative, through interviews, questionnaires and others). The use of methodological triangulation (Denzin, 2006) enhances validation of research data and research outcomes. Triangulation is a method of cross-checking data from multiple sources to search for regularities (O’Donoghue and Punch, 2003). Since critical realism focuses on establishing causal mechanisms, the choice of data collection tools and types of collected data will be influenced by the kind of causal mechanisms that need to be studied in light of research questions. Limitations of collecting data or data availability within the context of research also need to be kept in mind. This research uses literature, interviews (semi-structured in-depth and structured) and questionnaires through a multi-phased approach (discussed in Chapter 8) to help explain the existence or lack of LT innovation in the marble sector. For example, LT innovation amongst marble processing firms may be a result of certain interactions between the firms and technologies. In order to explain how these causal mechanisms exist and work to influence the event of LT innovation, interviews and questionnaires with firm owners and managers, suppliers of technologies and non-firms having a supportive role regarding technologies will be useful.
Before going into a detailed explanation of the case study design implemented in this study it is important to provide a sense of the interrelationships among key components of the research discussed so far.
Table 7.2 provides the paradigmatic assumptions leading to research methods. These methods have been explained in greater detail in Chapter 8.
Paradigmatic be explained in terms of causal mechanisms;
Iterative data collection and analysis till epistemological closure, no matter how flawed and temporary is obtained
Existing situation in terms of workings of a low-technology sector in a developing country (objects)
Existing phenomenon of innovation in the low-technology sector (events)
RQ1.1: What products, processes, organizational structure and markets do firms within the sector have or deal with?
RQ1.2: What types of innovation exist amongst firms within the sector?
Meso-firm level products, processes, organizational structure & marketing practices
Meso-firm level manifestations of innovation
Semi-structured in-depth interviews with owners/managers of mining firms within marble sector
Semi-structured in-depth interviews with owners/managers of processing firms within marble sector
Focus on understanding the SSI in term of its elements (objects/entities)
Building understanding of a low-tech SSI within a developing country (mechanisms)
RQ2.1: How are the actors or agents (firms including individuals and non-firms) setup in the sector?
RQ2.2: How do knowledge-base &
technologies exist in the sector?
RQ2.3: How do learning processes and demand exist in the sector?
RQ2.4: How are institutions placed in the sector?
Micro-individual level role of owners/managers of firms within the low-tech sector
Meso level role of firms within the low-tech sector
Macro level role of non-firms, technologies, knowledgebase, learning processes, demand and institutions within the low-tech sector
Semi-structured in-depth interviews with owners/managers of mining firms
& processing firms within marble sector Semi-structured in-depth interviews with supplier, distributor, sector expert and representative of sector support organization
Structured interviews with owners/managers of mining firms Questionnaires with owners//managers of processing firms
Explaining why or why not innovation in a low-tech sector is occurring (mechanisms)
Focus on understanding the SSI in terms of its structure (necessary and contingent relations)
Determinants of low-tech innovation and their relative influence within SSI (causal powers of objects/entities)
RQ3.1: How do firms interact amongst themselves and with non-firms?
RQ3.2: How do firms interact with (individual, firm and contextual) that influence low-technology innovation amongst firms in the sector?
RQ3.6: How much do these factors influence innovation amongst firms in the low-technology sector?
Meso-firm level interactions with non-firms Meso-firm level interactions with institutions Meso-firm level interactions with knowledgebase & technologies
Meso-firm level interactions with learning processes & demand
Determinants of innovation in a low-tech sector from SSI perspective including micro, meso and macro level determinants
Structured interviews with owners/managers of mining firms
Questionnaires with owners//managers of processing firms
Structured interviews with owners/managers and representatives of non-firms during closing phase of data collection
Table 7.2: Critical Realist Paradigm Leading to Research Methods
7.5.6. Analyzing/Interpreting Data: The Use of Retroduction
Critical realism distinguishes itself from other paradigms by using retroduction instead of relying on induction (common in qualitative/interpretivist approaches) or deduction (common in quantitative/positivist approaches).
Sayer (1992, pp. 107) describes retroduction as a ‘…mode of inference in which events are explained by postulating mechanisms which are capable of producing them…’
Lawson (1997) points out that while deduction tries to understand an event by moving from the particular to the general and vice versa for induction;
retroduction approaches events from a different perspective. The main concern here is to understand an event or phenomenon in terms of the mechanisms that caused it. Critical realism acknowledges that explanations resulting from the analysis of collected data are essentially interpretivist (especially true for this research study where all primary data comes from interviews and questionnaires involving respondents and analysis as a result of interpretations of their interpretations/responses). Woodside et al. (2005) term this double interpretation as the problem of ‘double hermeneutic’.
However, critical realism does not consider the discourse resulting from this form of data analysis to be enough in itself. Rather ‘reference to referents of the discourse need to be made’ and the researcher needs to repeat data collection (done through a multi-phased approach in this research – Chapter 8)
‘until epistemological closure, however flawed and temporary, is obtained’
(Easton, 2009, pp. 7). As mentioned earlier, retroduction is the key epistemological process that is iterative in nature (Dubois & Gadde, 2002).
Case studies are suitable in this regard because they can employ inductive as well as deductive cycles of data collection. Easton (2009) explains this by stating that;
„Deduction helps to identify the phenomenon of interest, suggests what mechanism may be at play and provide links with previous research and literature. Induction provides event data to be explained and tests the explanations… (Both) invoke causal language and the identification of mechanisms and offer the data collected as evidence.‟
This study also uses retroduction in a similar manner. It applies deductive approaches by using previous theory/concepts about LT innovation through literature review to highlight mechanisms that influence it. Further, it uses data collected from questionnaires and applies the conceptual framework (Chapter 4) to further explain mechanisms specific to the marble sector. However, it also applies inductive approaches by starting the data collection process through semi-structured in-depth interviews in a completely new context of north-west Pakistan’s marble sector with no similar study conducted previously. Outcomes from these interviews also inform the development and design of structured interview and questionnaire and help identify possible explanations of why or why not LT innovation occurs. Another set of interviews is carried out in the closing phase after analyzing first set of interview data as well as questionnaires thus ensuring the iterative nature of retroduction process. This helps offer further explanations of the causal mechanisms.