In Table 3 below I summarise the wide range of methodologies used for the different articles.
This study used a form of ethnographic case studies, as with some authors mentioned earlier (Mosse, 2004; Crewe and Harrison, 1998; and others).
Ethnography is the study of social interactions, behaviours, and perceptions that occur within groups, teams, organisations and communities. It aims to provide insights into people’s views and actions, linked to their location of home or work and documenting their culture and views, through the collection of detailed observations and interviews. Traditionally in ethnography, the researcher was an independent and passive observer, spending long periods in the field (stereotypically, in a village) and working with a single case, or a small number of cases. In recent years researchers have become a part of the study via auto-ethnography, in which the researcher’s own thoughts and perspectives from their interactions form part of the research, incorporating reflexivity - the ‘ethnography from within’ as described by Mosse (2005, p.11).
I consider that this combination of insider and outsider role has brought insights that would otherwise not be available.
As the culture I am studying is the development aid culture, it could be said that I have been working ‘in the field’ for more than 30 years. I am looking at my own culture from the inside. The cases of the Nepal water projects can be considered to be illustrative, as they provide long-term situations with a variety of TA involvement (the dependent variable). I have been intensively working with these two projects in Nepal since 2010, with approximately three visits per year of 1-5 weeks at a time, and consider that this has given me extraordinary access and insights to both the project cases, and to my own culture as a development worker. Case studies from these projects formed the basis of many of the articles (Articles IV, V, VI and VII). The work also involved ethnography of Finns involved in development more broadly in Finland (Articles I, II and III).
Participant observation was a key element of my research with the advisers themselves, as well as the local stakeholders. Participant observation allows the researcher to observe and understand the participants in their natural world and understand their interpretations of it (Schutt, 2004). Although perhaps it is better described as being an ‘observant participant’, as suggested by Green (2011, p.38), where she is reflexively considering her own work as a development practitioner. This was an easy approach, as I am myself a practitioner. I was, at times, an active member of the TA team, and not only a passive observer.
My approach has some basis in grounded theory13, as described by Glaser (1998). I collected data from the field, and following the analysis of that data and associated literature, a hypothesis gradually emerged. The hypothesis is not pure grounded theory, as I did not begin the research with a blank slate. I had good knowledge of the field environment, and as I do not come from a background of social studies, I was not confined to fixed theoretical ideas.
Instead, I gradually found theories that appeared relevant for the world of development cooperation. The theory and topics gradually emerged, with each article.
I apply predominantly qualitative research methods. As described by Rautanen (2016), I identified the key groups and informants, aiming to gather a range of viewpoints within my overall field of technical development cooperation. I gathered information from a spectrum of informants, talking with women and men, different ages, castes, ethnicities, religions, education levels and abilities, in order to develop a broad perspective. The informants came from two main groups. The first was the cohort linked to the Nepal water projects, including people working as staff, the local NGOs contracted within the water projects in Nepal, and those they worked with, the villagers, local and national government staff, and Embassy staff. The second main group consisted of members of what I have referred to as the ‘spectrum of development cooperation’. They included Finnish development studies students, researchers, NGO and consulting company staff, current and retired consultants, MFA Finland and multilateral donor staff. My respondents were mainly limited to Finns and Nepalese. I made this choice in order to limit the field somewhat. Otherwise the huge variation of development histories and systems from many countries would have complicated the analysis.
It could be claimed that my positioning was potentially problematic and a source of bias (as discussed also earlier, with regard to the academic-practitioner role). For instance, I am the representative of the employer of some of the respondents, co-worker of some, and in an agent’s position versus principals when interviewing some MFA staff. In the case of the Finnish respondents, I do not think this was a significant issue, as there wasn’t a significant power differential (though it is fair to say that there was an element of collegial loyalty, which might make me less critical of them). With the Nepalese respondents the positioning could be more problematic. It is possible that Nepalese respondents have altered their responses, leading to potentially unreliable results. If a Nepalese had written the case study articles alone, it would have been a different story from their viewpoint. Being a foreigner clearly gives me a different viewpoint from that of a Nepalese TA or other stakeholder. Working with other authors in the case studies, this gave an
13 Grounded theory is a method for collecting data, identifying patterns through a concept-indicator model, and then constructing theoretical models based on these patterns. The aim is that the analyst should maintain an open mind, and avoid reading the literature in the area of study before writing the first draft of the emergent theory (eg. Glaser, 1998)
opportunity to triangulate viewpoints and reduce any bias. The validity of the research was also supported, by the use of multiple data sources, methodologies, types of respondents, and data treatment.
Literature review, beyond academic sources, included analysis of policies, strategies and guidelines from the MFA, project documents (available on-line and in the MFA archives; and from the projects and consulting company archives) and evaluation reports. In addition, data collection and analysis from other relevant sources was important, including OECD/DAC data and reports, financial reports of companies (available on-line) and Government of Nepal policies and strategies (available on-line or accessed through active participation in strategy development processes).
Online questionnaires were used for data collection with the Junior experts and JPOs, for Article I; with students for Article II; and with national staff in Nepal for Article V. I conducted follow-up interviews with some of the respondents. At times it can be difficult to get responses to questionnaires. For the questionnaires with project staff and with the JPOs, there was a good response rate, presumably reflecting their interest in the topic. For the students, it proved much more difficult to get responses, despite multiple visits in person to classes and re-sending the questionnaire. This probably reflects the low interest and competing priorities. However, in general as a methodological tool, I found the questionnaires gave valuable insights.
In Article I, the interviews were recorded and later transcribed, and the interviews and written questionnaires were coded (initially with nVivo, and later with Atlas.ti and Word/Excel). The approach I used for analysis was qualitative content analysis, discerning patterns and following the topics that were expressed. I found that the qualitative data analysis programs were time consuming and not so useful. I found that a more mechanical method of using coloured markers to code and sorting findings from the transcripts on Word or Excel suited me better. I used semi-structured interviews with most subsequent respondents, either taped and transcribed, or took written notes, particularly when in the field. Respondents were usually sent the general questions in advance. I also held group meetings with project staff, local government representatives and community members. In some of the work with staff, a SWOT analysis was conducted.
Table 3. Methodological framework
Article I – Spectrum of TA Article II - Learning development Article III – Unacknowledged companies Article IV – Brokers Article V – Bricoleurs Article VI – Operationalising HR & SDGs Article VII – Work with new municipalities in Nepal
literature literature review
methodol-ogydesigns case study
design
Interviews, discussions, questionnaires
informal