Approaches to flexible design research

3.3.1 Case studies

This dissertation consists of a number of case studies, both in the peer-reviewed articles and as supporting case studies as presented in this dissertation. All these feed into the synthesis and further theoretical construction. Case study is an empirical inquiry that investigates a phenomenon within its real-life context. It is based on an in-depth investigation of a single individual, group, or event to explore causation and to find underlying principles. Case study is a well-established strategy where the focus is on the description and interpretation of a particular case. This particular 'case' can refer to a particular social or cultural structure or group, or be defined by other boundaries or setting. It typically involves participant observation, but also other methods can be used. The concept of population is crucial to indicate from which sample the conclusions are drawn. The initial definition of the research question is important in creating theory from the case studies. Focus is needed to avoid diluting the findings by the sheer volume of available data. Multiple-data collection methods are used, providing opportunity to triangulate the findings as well. Data analysis typically overlaps with data collection, and as such, case study is directly linked to the action research approach in this study (Järvinen, 2004, pp. 73–79).

Case studies have sometimes been criticized for having a bias toward verification, understood as a tendency to confirm the researcher’s preconceived notions. In this dissertation, this is a valid concern given the author’s long-term engagement with the cases and related action research element. Yet, as narrated by Flyvbjerg (2006), “The case study contains no greater bias toward verification of the researcher’s preconceived notions than other methods of inquiry.

On the contrary, experience indicates that the case study contains a greater bias toward falsification of preconceived notions than toward verification” (pp. 234–237).

3.3.2 Action research and interventionist approach

Action research is applied research that treats knowledge as a form of power and abolishes the line between research and social action (Mikkelsen, 2005, p. 132). Its common characteristics are that those being studied participate in the research process; research incorporates ordinary or popular knowledge; and that research seeks to raise consciousness or increase awareness, among others. The purpose of action research is to solve problems in a programme, organization, or community, rather than producing knowledge as the end result (basic research), aiming to understand the nature of issues (applied research), determining effectiveness of certain types of actions (summative evaluation), or improving the intervention (formative research) (Patton, 1990, 160–161 in Mikkelsen, 2005, pp. 132–133). Waterman et al. (2001) give the following often-cited definition that also guides this dissertation: “Action research is a period of inquiry, which describes, interprets and explains social situations while executing a change intervention aimed at improvement and involvement. It is problem focused, context-specific and future-oriented” (p. 11). The author further points out that action research is a group

activity, and has an explicit critical value basis, building on a partnership between action researchers and participants, all of whom are involved in the change process. Waterman et al.

(2001) consider the participatory process as “educative and empowering, involving a dynamic approach in which problem identification, planning, action and evaluation are interlinked.

Methodology is flexible in the sense that knowledge may be advanced through reflection and research, with both qualitative and quantitative research methods may be employed to collect data” (p. 11). The authors point out that action research can result in different types of knowledge, including practical and propositional. In this regard, also “theory may be generated and refined, and its general application explored through the cycles of the action research process” (p. 11). Robson introduces the steps (Bassey, 1998, pp. 94–95, in Robson, 2002, p.

218):

 Define the inquiry; describe the situation

 Collect evaluative data and analyse it

 Review the data, look for contradictions and tackle it by introducing change

 Monitor the change; analyse evaluative data about the change

 Review the change and decide what to do next -> going back to the inquiry

Action research, also referred to as “participatory action research”, is often defined as a cyclic process of change as shown (O’Leary, 2004): Observe (research and data collection) -> Reflect (critical reflexivity -> Plan (strategic action plan) -> Act (implementation) -> Observe … … ->

new cycle (Figure 3.2).

Figure 3.2 Cycles of action research

In this dissertation, action research was applied in articles II, III, IV, and VI, and in case studies made in Nepal and Guyana. For the Case Guyana, Knowledge, Attitudes, Beliefs and Practices (KABP) Survey was used for defining the inquiry, describing the situation, and collecting evaluative data. The outcome was then reviewed and analysed as part of action process. This approach has its roots in human health and management sciences; in the WASH sector, it has been used by such development agencies as UNICEF and OXFAM for baseline purposes. The tool can be used to collect quantitative data on the perceptions of the local communities related to WASH, the predetermined questions, or statements capturing the respondents’ knowledge, attitudes, beliefs and practices, as the name of the tool (KAPB) suggests.

The findings of the KAPB survey then fed into a qualitative step for identifying contradictions and defining the course of action for change. The outcome from KAPB surveys were triangulated by the interaction workshops at each location, mainly with the school teachers as the case study focused on school WASH as part of the Guyana Amazon Programme. KAPB survey in this case is linked to formative research, but as a baseline could be also used for summative evaluation later on (see the next chapter). The insights from years of working with action research for water sector development in Nepal have also been used in the results and discussion chapters, and hence, the action research approach has a pivotal role for the entire outcome of this dissertation.

3.3.3 Formative research and summative evaluation

Formative research aims to suggest how to improve interventions, whether it is a policy, programme, project, organization, or product. It focuses on strengths and weaknesses of the target of study to further suggest how to improve it. Formative research is about performance, and closely linked to action research (Mikkelsen, 2005, p. 133). Formative research or the term formative evaluation is also defined as “process evaluation that is used to shape and refine programme operations” (Schutt, 2004, p. 320).

Summative evaluation aims to determine effectiveness of human interventions and action (programmes, policies, personnel, products). It focuses on goals of the intervention. Its desired results include judgements and generalizations about the effective types of intervention and the conditions under which those efforts are effective. Summative evaluation is about effectiveness (Mikkelsen, 2005, p. 133). Summative evaluation characterizes Articles II, III, IV, and VI related to Nepal.

The distinction of formative and summative evaluation research is primarily one of the purposes, and the distinction is not absolute. Robson notes that, for instance, summative evaluation could have a formative effect on future developments (Robson, 2002, pp. 206–208). In this dissertation, for instance, the Case Guyana had formative research characteristics as the exercise was done to develop an appropriate WASH programme for remote schools; in other words, to improve the intervention that had been already outlined and started at the start of the survey. The same survey could serve as baseline for future summative evaluation research, and also for an impact evaluation study.

3.3.4 Interviews

Interviews as a research method have a number of applications. All interview methods used in this dissertation were in-person interviews, whether by the author herself or local enumerators/facilitators. The following definitions of three main types of interviews, all of which were used in this dissertation, were modified from Robson (2002, p. 270):

 Fully structured interviews have predetermined questions with fixed wording and coding, usually in a pre-set order. The questions are made to all respondents in a same order with the same words. Open response questions can be used. The difference to survey method is the role of the interviewer. In this dissertation, the Case Guyana and Article II on gender issues utilized fully structured survey interview.

 Semi-structured interviews have predetermined questions but these can be modified during the interview, the wording or the order of questions can be different, and the questions may be left out during the interview. In this dissertation, Articles II, III, and VI used this method.

 Unstructured interviews are interviews done over a general area of interest and concern, letting the conversation flow and develop openly.

Interviews were taken with both groups and individuals. For instance, in Article III and Case Guyana, both structured in-person interviews as well as facilitator-led group interviews were utilized. The survey interviews were used for quantitative analysis and the group interactions for qualitative analysis. All the group discussions aimed to develop specific topics further, towards actionable items.