• No results found

Chapter 3: Research Methodology

3.11. Reflection on the Quality and Limitations of the Methodological Approach

Guba and Lincoln developed five quality criteria for qualitative research, summarized by Treharne &

Riggs (2017, 58) and referred to in Bryman (2008).

3.11.1. Application of the Five Quality Criteria

In this thesis, credibility (Bryman 2008, 377), or representativeness, is ensured by methodological triangulation; contrasting and synthesizing institutional preconditions with civil society perspectives or a secondary data analysis preparing for well-informed interview questions. Furthermore, carefully established trust and similar transformative interests between the respondents and me facilitate the research credibility. Asking how close the research comes to “reality”, one can say that on the one hand, the meta-level focus of this work leads to alienated and scattered findings but on the other hand, eventually, these findings have been synthesized comprehensively by combining different levels

28 and angles via an epistemologically and thematically coherent framework. Due to the limited number of interviews (15) and documents (23), findings are not firmly generalizable but are, in their modesty and non-conclusiveness, thus in their exploratory character, assuming, critically asking as inherent to qualitative research, internally valid.

Transferability (ibid, 378) of the findings to the context of other world regions (related to external validity) is possible when looking theoretically at the relations between commercialization trends, politicization, health equity and environmental health. Also, how policy documents problematize the two focal topics related to their root causes and approaches are partly globally applicable due to their pervasive prevalence. Concerning the major part of my findings, civil society’s role and perceptions, my empirical findings concern the European context are apart from more abstract, conceptual insights not transferrable to other regions.

Dependability, asking if findings would differ if interpreted by another researcher, is difficult to evaluate as I cannot make use of peer-triangulation or auditing (cf. ibid.). It is yet essentially enhanced by a continuous openness about methodological choices and constant reflective memo writing conducted as described – which also helped to ensure confirmability, assessing if findings reflect interview partners’ responses independently from the researcher’s biases. I have elaborate records of coded, re-coded, re-organized and commented data so that unconscious biases could be easily identified by others.

Lastly, authenticity is particularly ensured by contributing to civil society’s case of enhancing health equity and environmental health via information exchange, possibilities of reciprocal learning and enabling civil society to better understand their own broader context. Higher transformative potential is achieved by sharing policy, research and activism recommendations in this work’s concluding chapter. Even though authenticity can only be assessable after potentially publishing the results and recommendations, I consider it to be a major strength of this work.

3.11.2. Reflection on Research Limitations

Most limitations are determined by this thesis’ scope. Already the content analysis documents were only chosen if directly concerning the research topic even though health equity and environmental health are mostly determined by other, often solely economic policies or trade agreements. Yet, the detailed scrutiny of economic constraints is or will be conducted by other researchers.

Also, conclusions are limited in their extent due to missing information about the concrete implementation of instruments attached to the analysed policy targets making it difficult to assess

29 how committed actors are. This would require delving into evaluation research, conducting interviews with policy makers and investigating local authorities on policy implementations which would definitely exceed this work’s scope. Thus, an assumed gap between policy rhetoric and the “actual”

institutional space determining health equity and environmental health – is considered in the interpretation.

The scope also did not allow for the analysis of national or local policies even though most decision-making power lies with the design of national health systems and the implementation of targets at the local level. Yet, this thesis does not aim to compare health systems; it neither aspires to grasp all empirical conditions and processes. Civil society is active on all levels, just as environmental health and health equity concern all levels, and the supranational perspective is best suited as an introduction into the governance sphere.

Apart from scope limitations; the document types for the content analysis vary considerably ranging from action plans to declarations. This divergence in terms of the documents’ binding power or time frames partly renders the generalization or direct comparison of findings difficult. Yet, the classification into document types and the document-specific memos help to put findings into perspective.

One could argue that my focus on civil society excludes the voices of representatives of the European Union, national governments, health workers or for instance the pharmaceutical sector on health governance matters leading to one-sided conclusions. I anyhow disagree as this focus was chosen intentionally not claiming to research more than the civil society perspective in accordance with basic institutional settings.

In the literature, one also observes a bias towards more research on high-income Western European countries’ health systems. Trying to achieve an equal spread across the European region, I managed to speak to 4 representatives from Eastern Europe compared to 11 from Western Europe proving that I have not been entirely successful to achieve a balance – Western European CSOs seemed more accessible, for instance in terms of language for the internet search.

Moreover, during the interviews, some participants faced difficulties expressing their thoughts in English which possibly distorted their views. This was countered by giving participants the possibility to correct their transcripts afterwards and by sometimes trying to communicate in their mother tongue.

During the content analysis, I realized that extensive coding often takes data out of its context – I reconnected the data manually by working with Word documents as supplement to the analysis

30 software. While I would regret missing out on Atlas.ti’s sorting and coding options, the software yet tends to create a certain distance between the researcher and the data. Here, printing out documents, taking manual notes and drafting sketches on paper helped.