• No results found

5.3 Research Methods

5.3.2 Why Case Study?

Case study design can answer key research questions via interviews, to include ‘how’ and ‘why’ (Yin, 1984). The thesis involves ‘how’ and ‘why’ questions which are best answerable through the use of a case study. According to Yin (2003), the how and why questions involve relational links that can only be enquired upon with time, not instantly. The case study is a design that seeks to explore past and current issues that affect an organisation(s), and its future is determined based on existing data. A case study design enriches the researcher with an in-depth understanding of the study, focusing on the specific problem area (Saunders et al., 2011). This confirms the argument of Sue (1997) that a case study design offers a holistic view of the research question and allows the researcher and reader to gain a comprehensive and thorough understanding of the research. Case studies are used to achieve high understanding of organisational practices (Morris and Wood, 1991), as is the case with this study. Case studies are not necessarily on individuals but also institutions, neighbourhood, innovations, decisions, services, and programmes, amongst others (De Vaus, 2001).

In line with understanding organisational practices, Scapens (2004) adds that it will be necessary to modify existing theories if they fail to provide satisfactory explanations. Thus, the explanatory case study used in the research aims at modifying some existing unsatisfactory theories in order to yield satisfactory explanatory results of the phenomenon studied: accounting and accountability practices in MFIs.

Ryan et al. (2002) discuss the differences that exist in case study research (See Table 5.1 below) and Yin (2003) explains when case study research design is preferred (see Table 5.2).

136 TABLE 5.1: Differences in Case Study Research

Type of research Positive Interpretive √

View of the world External and objective Social construction

Type of study Exploratory Explanatory

Nature of explanation Deductive Pattern

Nature of generalisation Statistical Theoretical Role of theory Hypothesis generalisation Understand

Nature of accounting Economic: decision making Objective of study Source: Ryan et al. (2002: 146)

TABLE 5.2: When Case Study Research is Preferred

Strategy/ Design Type of research

question Required control of behavioural events Focus on contemporary issues

Case study How, why No Yes

Source: Yin (2003: 5)

The case study design adopted in this study follows an interpretive approach with subjective and inductive reasoning, as those with a positivist approach fail to situate accounting practices within their historical, socio-economic and political context (Ryan et al., 2002).

The case study research steps suggested by Ryan et al. (2002) and Yin (2003) discussed below have been followed in the present study. These steps include as a first step preparation (through review of existing literature as Scapens (1990) argues, research training courses and field work/case work preparation, plotting, ethical review process, development of research

137

instruments, for example interview questions, and so on). These preparatory steps increase the research reliability by guiding the researcher through the data collection stage (Yin, 2003).

The collection of evidence is done through the gathering of data using various instruments. In the study, both primary and secondary data collection methods have been used. A triangulated data collection approach was used through semi-structured interviews, informal discussions and documentation. The approach allowed the researcher to gain an in-depth understanding and allows them to explain the changes inherent in the case study MFI’s accounting and accountability practices as well as the process and actors involved in the change. The use of in- depth semi-structured interviews allowed the researcher to ‘probe’ the respondents’ answers when there was a need to, and vice versa (Saunders and Thornhill, 2007).

Third, the assessment of the evidence collected is vital, as further discussed below. Reliability, as argued by Ahrens and Chapman (2006), is not easily separated from validity. The idea of research validity was initially developed for quantitative research with objective reasoning about the nature of reality, and this has been argued to be unsuitable for qualitative studies like the present one. This argument is supported by Chua (1986), who points out that, “social reality is emergent, subjectively created and objectified through human interaction” (p. 615). Furthermore, we cannot pretend to study the reality of how, for example, the case study MFI accounting and accountability practices function or have changed if we do not understand the roles of the actors initiating the changes or operating in the environment and independent of the theoretical interest of the researcher (Ahrens and Chapman, 2006). Hence the idea of obtaining replication in qualitative studies is inappropriate, and as Ahrens and Chapman (2006) point out, the same results should not be expected when two different researchers study the same organisation but we would rather expect their descriptions to be compatible, where the conclusion drawn by one person should not contradict that of the other. Hence, agreeing with the suggestions made by

138

Ahrens and Chapman (2006), Scapens (2004) is of the view that interpretive case studies should follow procedural reliability (where the correct and reliable research method and procedures are followed) and contextual validity. A proper research plan was developed for the study in order to pursue procedural reliability. Nearly all of the interviews that were conducted were recorded and transcribed. Field notes were also taken in addition to gathering further evidence from documents and informal discussions. The aim is to show reliability in the eventual findings, such that if the same procedure was followed by another researcher, the same/similar results will be obtained. Ryan et al. (2002) see the triangulation of research as very useful in achieving contextual validity. The aim of triangulation is to enrich the understanding of the researcher by collecting data on the same issue from alternative sources. As mentioned above, this study uses semi-structured interviews, informal discussion and documentation in order to achieve some sort of contextual validity.

Fourth is the identification and explanation of various patterns found. For more details on this step, see Section 5.5. The researcher was able to identify and explain key patterns found during the study. Miles and Huberman’s (1994) data analysis method was adopted in the study. The data analysis method consists of three complementary flows of activities: data reduction, data display and conclusion drawing and verification, as further explained in Section 5.6 below.

The fifth step is about theory development. Scapens (1990) suggests that in cases where existing theories cannot be used to justify or explain the research, a modification of existing theories can be embraced to suit the new areas under investigation. The study has therefore modified some existing institutional theories and has also done so following some suggestions/contributions from the existing literature.

139

The last step to be followed is writing up the thesis. At this point the case results and findings are discussed; making sure it is authentic, plausible and critical.

Related documents