Methodology Philosophy Epistemology

Chapter 3 – Research Methodology

3.0 Introduction:

This chapter on research methodology will focus on the research question that was presented in Chapter 2, and discuss how it is embraced into the research philosophy, design, methodology and methods used in this research project.

3.1 Research Philosophy:

Research philosophy relates to the development of knowledge in a specific area and nature of that knowledge as applicable to the research project. In short, research philosophy could be assumed to embrace the personal beliefs and the way in which the researcher views the world. The importance of research philosophy within a research project has been highlighted by many researchers. Easterby smith et al. (2003), argues that it is unwise to conduct research without an awareness of the philosophical issues that lie in the background. According to Denzin and Lincoln (2000), the combination of personal belief, researcher’s experience and understanding of philosophy will underpin the formulation of strategy and selection of methods for the research project. Similarly Saunders et.al (2006) argues that research philosophy influences the researcher about the research process. Easterby-Smith et al (2003), Saunders et.al (2006) considers Ontology, Epistemology and Axiology as the three major paradigms of research philosophy applied in business and management research. Discussions will primarily focus on the ontological and epistemological perspectives of the research project. Whilst Ontology, is linked to the assumptions that a researcher makes on the nature of reality, Epistemology on the other hand is more concerned about the “general set of assumptions about the best ways of inquiring into the nature of the world’ (Easterby-Smith et al, 2003). According to Easterby-smith et al, (2003), ontology can be further classified into three perspectives: representationalism, relativism and nominalism. The table shows the details about different ontological perspectives.

Epistemology can be further classified into two perspectives namely positivism and relativism. These two perspectives unite on a common assumption that the job of the researcher is the identification of “pre existing reality” (Easterby-smith et al, 2003). Whilst the positivist’s researcher approaches the task of understanding reality through design of experiments, on the other hand the relativist achieves it through combination of “triangulation” of methods and through surveying large samples (Easterby-smith et al, 2003).

The two main approaches in management research namely deductive and inductive indicates the nature of the data to be collected and the direction of shift between data and theory during analysis (Yin, 2003). Marrying these research approaches to the research philosophies, Robson (2002), points out that representational ontology and positivist epistemology adopt inductive approach, whereas relativist ontology and epistemological approaches adopt deductive approach. To understand the philosophical approaches used in this research project, it becomes necessary to revisit and analyse the main research question presented in chapter 2:

“Is the Toyota Production System (TPS) transferable to other automotive manufacturers?”

Based on the literature presented in Chapter 2, this research begins with the basic ontological assumption that in reality TPS is transferable to other manufacturers given the fact that complete transferability is difficult to a greater extent. This is a relativist view since to establish the truth ‘consensus between different viewpoints’ must be established (Easterby-Smith et al, 2003). Furthermore, the facts are dependent on the position and viewpoint of the researcher. Therefore, as a result of the research question, a relativist ontological and epistemological approach has been adopted in this study.

3.2 Research Design

3.2.1 Selection of research design

Research design is ‘the logic that links the data to be collected and the conclusions to be drawn to the initial questions of study’ (Yin, 2003). The possible research designs available to the researchers include case research, survey research, action research, experimental design, participant observation and ethnography (Easterby-Smith et al, 2003). According to Meredith (2002) the selection of appropriate research design should be followed after careful consideration of the nature of the research and the research questions that needs to be addressed. According to Meredith (2002), the key advantages of using a case study research are relevance and exploratory depth it offers to the area of study. On the other hand, the main critics of case research are that it is time consuming, needs skilled interviewers, needs more triangulation, its limitations to draw generalizable conclusions from limited cases and lack of familiarity of its procedures and rigor (Meredith (1998); Stuart et al (2002); Voss et al (2002); Yin (2003)). According to Voss et al (2002), case research despite its critics can help not only the theory but also to the researchers by developing new ideas and insights. According to Yin (2003) case study design can be classified into two main categories: ‘single case study’ and ‘multiple case study’ based on the number of cases researched to answer the research questions. These are further classified into ‘holistic’ and ‘embedded’ case studies based on the unit of analysis within the single or multiple case studies. For instance multiple case study (embedded) research represents study from a multiple cases with a more than a single unit of analysis.

For the purpose of this research project, a single case study design was employed. To understand the interrelationship amongst the factors within the single case study, multiple component analysis ‘embedded’ (Yin, 2003) was done within this research design. Through the single case study, the production system in specific automotive firm was identified and thoroughly analysed for comparison with TPS to answer the research questions. The entire production system was subdivided into: Management, People, Production and Automation, Suppliers and Customers perspectives and each of these were considered as separate units of analysis in relation to answering the research question of transferability of TPS.

3.2.2 Possible Research Designs

Although the case study approach is selected for this study, there are other possible approaches to research design and these include: Action research & Surveys. It is quite important to understand the suitability of the individual approaches to address specific research questions before making a selection (Easterby smith et al, 2003). Table 1 explains the possible research designs for this research and justifies the selection for the most appropriate approach.

3.2.4 Potential weakness:

In selection of the case research design for the project, it was important to understand and mitigate the potential weakness within this method. The potential weaknesses of case study are: time, access and generalizability of the research findings (Meredith (1998); Stuart et al (2002); Voss et al (2002); Yin (2003)). The following discussions highlight how the potential weaknesses were addressed in this research project.

Access:

To address the issue of access to the case company, information about the key personnel was obtained and application for permission was sent out at early stage of project. The written consent was obtained from the case company for conducting the data collection through interviews, direct observations and archival documents in addition to the access of key personnel of various disciplines in the organisation.

Time:

To address the issue of time constraints in the case study research, the researcher had narrowed down the boundaries for the research. From the outset, the scope of project has been narrowed down from holistic approach of TPS towards its transferability perspective taking into consideration the available time for the project. Also time limitations were one of the main reasons for focusing on a single case organisation with embedded unit of analysis to provide an in depth analysis on topic of study.

Generalizability of the research findings:

Generalizability also known as ‘external validity’ is defined as the “extent to which it is possible to generalize from the data and context of the research study to broader populations and settings (Hedrick et al., 1993). Meredith (2002) points out that the opponents of case research argue that its “theoretic generalizability” is less because the results hold good only for particular situation, whereas the proponents of case research claim that theory developed from such studies becomes applicable to similar and non similar situations. Through the study of a single case organisation, the aim of this research project is to extend and generalize theories on transferability of TPS i.e. ‘analytic generalization’, rather than a ‘statistical generalization’ (Meredith, 2002; Yin, 2003) since the single case organisation studied here does not represent a sample or population.

From the above discussions, it is evident that single case study is realistic in terms of the nature of the study (exploratory), time limitations, gaining access and resource constraints. The arguments presented above justify the selection of case study design for this research project and supports both the research question and the research philosophy adopted in this research project.

3.3 Research Methodology – Qualitative

Case study design would be best suited to understand, the prevalence of a phenomenon in a real life context and implications of the data rather than just the measurement (Yin, 2003). Revisiting the research question mentioned in the earlier section, it is quite clear that the research objective is to map a production system of a specific firm with TPS i.e. understanding a phenomenon in real context. Furthermore, a greater degree of interpretive approach needs to be followed in order to identify the degree of closeness for implementation of TPS and sort out the transferability issues faced by manufacturing firms while emulating TPS. Thus the qualitative methodology for data collection and analysis for this research is justified.

3.4 Research Methods – Data Collection & Analysis

3.4.1 Source of Data:

Data for case studies can be collected through six sources: documents, archival records, interviews, direct observation, participant observation and physical artifacts (Yin, 2003). According to Yin (2003), these individual techniques are complementary to a greater extent and in order to develop a good case study design, researcher should aim at including as much techniques to the extent possible. Similarly Voss et al. (2002), states that multiple source of evidence (‘triangulation’) is the important principle for data collection in case research. Based on the above arguments the primary data collection for this research was carried out by semi structured interviews, direct observation, archival records and company documentation.

According to Yin (2003), the validity and reliability of the evidence collected can be maximized by three principles: Triangulation, Creating a data base and maintaining a chain of evidence. Amongst these three principles, Triangulation has the most significant impact on the validity and reliability (Stuart et al, 2002) and hence discussions presented below will focus on the elements of triangulation adopted for this research project.

3.4.1.1 Triangulation:

Yin (2003) defines ‘Triangulation’ as the use of multiple sources of evidence and points out that the any conclusion based on multiple source of evidence is likely to be more accurate, reliable and valid. According to Collis and Hussey (2003) and Yin (2003), triangulation can be classified into methodological triangulation (different methods), data triangulation (data sources) and investigator triangulation (different evaluators). In addition to the collection of qualitative data, Voss et al. (2002) argues that case research provides better opportunity to collect quantitative data with greater accuracy and reliability than survey research. The data collected through interviews were verified through direct observations to increase the validity and reliability. Hence by combined use of methods together with collection of qualitative and quantitative data ‘methodological triangulation’ was achieved in this research project. ‘Data triangulation’ refers to the collection of data from different sources to increase the reliability of results (Easterby-Smith et al, 2003; Yin, 2003). Based on these arguments, qualitative data collection for this research was primarily carried out by interviews, direct observation, archival records and company documentation. Finally ‘investigator triangulation’ refers to collection of data by different investigators so as to identify similarities and differences in the data collected (Yin, 2003). Since this research was carried out by a single researcher, investigator triangulation as suggested by Yin (2003) was not possible.

3.4.2 Data gathering:

The data gathering in case based research are usually achieved through the written and taped records of the interviews, company information and researcher’s observation (Stuart et al, 2002; Yin, 2003). It is pertinent to mention here that only hand written notes were employed for this research project since permission for audio tapes were not granted by the case organisation. The answers to the questions in the form of hand written notes were detailed immediately after each interview so as to ensure that data collection was complete. Yin (2003) points out that researcher’s bias directly affects the data collection process and findings of the research. To protect against error and researcher’s bias, the key data gathered were double checked with participants themselves and through direct observations.

3.4.3 Data Analysis:

According to Miles and Huberman (1994), the three main approaches to qualitative data analysis are interpretivism, social anthropology and collaborative social research. Social anthropology is associated with ethnographic studies whereas collaborative social research is more oriented towards action research. Since the research study was based on case research, interpretivism was the most appropriate approach to be used since it emphasizes on understanding the subject of study through social interactions (interviews and observations) and phenomenologies.

“The most serious and central difficulty in the use of qualitative data is that methods of analysis are not well formulated”

(Miles and Huberman, 1994)

The above quote acknowledges the importance of data analysis process related to research as this project mainly focuses on the collection of qualitative data. According to Collis and Hussey (2003); Miles & Huberman (1994), qualitative data for case study research design can be analysed either by ‘within case analysis’ (single case study) or ‘cross case analysis'(multiple case studies). Qualitative analysis for this research would be carried out by means of ‘within case analysis’ since this research would look for data collection from a single case study. By using multiple units of analysis (embedded) ‘within’ the same case study, the researcher was able to identify the production system in the case organisation and map it with TPS to identify and bring out the transferability of TPS.

3.5 Summary

To conclude the section on research methodology, figure 3.2 presented below summarizes the main research approach used within this research project.