2.2 M ETHODOLOGICAL A PPROACH
2.2.4 Critical Reflection on the Methods
There are certainly advantages in working with national and regional research organisations as partners. First of all, it opened many doors to me when I was trying to interview scientists and policy-makers in Ethiopia. I had no difficulties in getting permission to work on the kebele level.
I was introduced to all the authorities in charge from the woreda to the village level, and I was never questioned as to what my intentions were, even during the sensitive time of the elections in Ethiopia (May 2010). However, when we (my research partners and I) started interviewing we were accompanied by DAs, kebele workers or other political agents in the villages during the first days.
The more challenging aspect was to find out if the respondents really were representative of the sample I had defined at the outset. After I had done the village mapping this was quite clear, but in the beginning, when my research partners and I were working with key informants who were assigned to us by other actors, it was difficult to assess what agenda they were
pursuing and whom they were representing. Indeed, it was challenging to get closer to those groups seen as poor and thus often marginalised and disempowered in the villages to achieve a balance in our sample. When I interviewed farmers falling into this ‘category’, they often told me that they could not participate in social institutions because of financial and labour shortages (see also 4.2.4). Thus they became silenced and invisible. In Galessa, especially, I sometimes had real difficulties in managing to talk to people who were not on the ‘official project list’, the people known as active collaborators in affairs of the GWP. Those were the people who attended trainings, carried out field trials, implemented the technologies and also benefited first of all from the GWP. When my research partner explained to the project contact farmer that I really wanted to talk to some specific people I had selected from the village mapping, he told him that one of the women was crazy and the other was too old and frail to talk to us (see also 4.2.4). Upon my insistence I finally met both of the women and found both accusations untrue. It turned out that the contact farmer had excluded them from project benefits because they could not contribute enough labour to the community nursery.
However, both of them were heads of households and therefore had difficulties in allocating free labour to communal activities.
Equally, in the area of CST2 women were presented to me as a group of ‘female-headed households’ during village mapping – otherwise women were not mentioned as potential respondents. They were even excluded from village mapping, Focus Group Discussions (FGDs) and the making of the seasonal calendar by the men who were our initial contact persons suggested by the DAs. To overcome this problem I interviewed women separately from men and I held FGDs with women only and with a female translator.
The disadvantage of working closely with national and regional research organisations is that you may lose flexibility and to some extent your independent status as an outsider, because in the eyes of the respondents you become ‘one of them’. This was particularly noticeable in Galessa, where my partner organisation had been working for more than ten years. In the beginning it was hard to move beyond long speeches praising the work of the organisation. In Ambober, there was not much awareness about the research organisation that my research partner was working for. Researchers were usually perceived as ‘someone from the government’. There the challenge was to avoid any linkage to political issues.
Working closely with partner organisations also restricted my flexibility. I was not free to negotiate my entrance to the community in my own way, but in fact there was no alternative
to the official route via the woreda and the kebele. Furthermore, I was dependent on the availability of my research partners to make a field visit, and as they were also working for their organisations, it was not always easy to find a time slot when both of us and the farmers were available. This sometimes required a lot of patience and understanding, and at times I wished I had worked with independent translators. However, it is very difficult to find translators that have both some knowledge of the topic and sufficient knowledge of English.
Especially where the farmers spoke only Oromiffa it was a huge challenge. My attempt to find an alternative translator was time consuming, and in the end I realised that the good relations I had with my research partners and their in-depth knowledge of the subject matter and the area were more important. For example, one alternative translator I tried working with struggled to translate tree names and soil conservation measures correctly from Oromiffa to English, which led to considerable confusion. From my work with the other translator I was already familiar with the terms in Oromiffa and I was confused when he provided another translation of those terms in English. Farmers would for example differentiate indigenous junipers (gatira) and foreign junipers (farendji gatira) – the latter one was actually a cedar variety, however for farmers the word for juniper was synonymous with ‘conifer’ as the range of conifers was limited to those two trees. But my translator simply translated ‘junipers’ and omitted their differentiation.
The farmers appeared to like and trust the researchers I was working with, which I concluded from the way they were interacting with them compared to other outsiders. Working with translators however has risks – as my knowledge of Amharic was limited, and non-existent for Oromiffa, I could not be sure about the actual questions and answers. And in fact it turned out that there was some bias in the translations when I double-checked some of them with another translator after my field work. I tried to address this in considering the second translations, as well as in triangulation through FGDs and in asking Ethiopian friends and other colleagues whether they considered the information correct or not.
Among the methods I found the landscape ranking exercise to be the easiest and most relaxed one to use, while the interviews were sometimes difficult because they still created some kind of formal atmosphere, and the FGDs were the most difficult to employ effectively. The FGDs were easier with the farmers than they were with the scientists, who often were reluctant to have a discussion but simply waited for new questions to be posed to them. FGDs turned out to be more of a moderated group discussion rather than a proper focus group discussion, possibly because farmers and scientists alike did not fully understand the purpose of these
semi-structured exchanges. Originally I had also planned to focus on much narrower topics, but the experiences of the interviews had shown me that this would have confused the respondents. A broader topic gives the respondents more space to explore and to slowly move towards the point they want to make. This must be understood in the political and cultural context of Ethiopia. First, free expression of opinion is still uncommon especially in a relatively public space like the FGDs due to the long history of oppression the Ethiopian people have known. Second, one heritage of this history of oppression is distrust. Third, there was a mixture of different age groups and hierarchical relationships between the different participants. This inhibited the more junior and less experienced members of the group in their participation.
At the end of my research, as I could not finish one FGD in Woglo for personal reasons, it was then completed by my research partner. I first considered omitting the transcript from my analysis, but then I realised that my role during the other FGDs in the villages was actually not more than an observer. My language skills were not sufficient to lead the discussion, and direct translation would have been time consuming and extremely disruptive. The transcript also showed me that the questions were identical with the ones I had given to my partner. I therefore included it in the study. The same applies to the landscape rankings in Wojnie and Woglo and some in Galessa. The farmers received the photographs, ranked them in terms of their preferences (as like and dislike) and then gave reasons for their choices. The role of my research partners was to explain this, hand out the photographs and record their answers. I therefore decided to include these data, even though I was not present at the time.