Where Gadamer explained the basis of how humans understand the world and how good interpretations are separated from bad ones, the DiaLoop model de-tails and elaborates on how the process of analysis as a concrete activity done by me as a researcher can occur, and which steps need to be taken for it to have scholarly value or indeed truth in it.
them back together again, reaching a new understanding of the phenomenon in the process. One of the ideas behind the model is to make the process of analysis transparent, in the sense of illustrating how an analysis goes all the way from raw data to making its conclusions (Alrø et al., 2016, pp. 43–46). My description of critical realism, philosophical hermeneutics, and case studies could be criticised for “black boxing” how the actual process of analysis takes place, with its focus on the underlying ideals and assumptions of the process, and the lack of expla-nation as to how this would translate into concrete analysis. Thus, my aim is to let DiaLoop be a detailed way of approaching the question of epistemology in this thesis.
The model is made up of the following eight steps (see also figure 5), which are all considered parts of the process of analysis as a whole. Although presented in a specific order here, the idea of the model is not to go through the steps one at a time from top to bottom. Rather, the steps should all be seen as parts of the process, through which the researcher will move back and forth numerous times.
This movement is the “loop” part of the title of the model (Alrø et al., 2016, pp.
51–55). Several of these terms are also used in everyday language, but it should be noted that when used in this thesis, they refer to the specific meaning they are given as described below.
1. Intuition
An impulse or sudden realisation
2. Problem
Working your way towards one or more (research) questions
3. Observation
Observed parts of the data
4. Experience
The researcher’s own experiences or reactions when analysing
5. Identification
Theoretical terms identified in the data
6. Argumentation
The researcher’s line of reasoning – describing, explaining and discussing
7. Interpretation
Overall interpretations based on synthesis of a systematic and thorough analysis
8. Patterns
Patterns found in the interpretations made
In the following, each step will be described in more detail, with examples from the process of analysis in this thesis. Four different kinds of data are involved in these steps: field notes from participant observations, video observations, inter-views, and documents.
1. Intuition
Intuition is described as an impulse, a revelation, or a sudden realisation in the DiaLoop model. It is the “aha moments” of the process of analysis, where the researcher looks at the data and gets a feeling there is something interesting going on, but without perhaps being able to put a finger on what it is, just yet. Accord-ing to Alrø, Dahl, Schumann, these impulses happen when the researcher lets go of a perhaps distanced contact with the data (observing and registering), and immerses himself in the data (2016, pp. 61–62).
These impulses cannot be explained or predicted, but are made up of ideas for what to focus on in the analysis and/or an idea about where in the data it makes sense to begin looking. This makes intuition a good line of approach to begin-ning analysis even if it later proves to be mistaken or distracts the researcher from answering the research question.
In the current study, which has very rich and multi-facetted data material, there is an inherent risk of asking a question similar to the one Steinar Kvale argues should never be asked: “How shall I find a method to analyze the 1,000 pages of interview transcripts I have collected?” (Kvale, 1996, p. 275). Intuition offers one way of making a preliminary sorting of a large and complex amount of
quali-∞
Reseacher ObservationIntuitionProblem Data ExperienceIdentifi cation Argumentation
Interpretation Patterns
Figure 5. The model depicted on the backdrop of the infinity symbol demonstrating the continu-ous movement or DIAlogue between data and the analyst, while at the same time moving between the eight steps of the model in a continuous LOOP. (Alrø et al., 2016, pp. 45–46)
tative data, thereby avoiding the 1,000-page question. While intuition can be a helpful sorting mechanism, it should also be noted that, according to Alrø, Dahl, and Schumann, intuition alone will not lead to a transparent and well-argued analysis in the end. If analysis is done primarily based on intuition, there is a high risk of making the above-mentioned spontaneous interpretations, which, they argue, need to be qualified further by other steps in the DiaLoop model (Alrø et al., 2016, pp. 62–63).
Example
During a weekend scout camp where I observed Dorthe, she spent the afternoon standing outside in the cold rain in a situation that at first glance did not seem en-joyable or worthwhile at all. As I stood out there with Dorthe, it struck me how odd or counterintuitive it was that someone would be doing this voluntarily. That initial thought made its way into my field notes, and when selecting situations for further analysis I ended up choosing it, in part based on this initial intuitive feel-ing that somethfeel-ing interestfeel-ing was gofeel-ing on there. Intuition here occurred durfeel-ing data collection, so I do not consider it to be limited to the situation of sitting in the office and looking the data.
2. Problem
According to Alrø, Dahl, and Schumann, analysis should have a certain direction or purpose. Although any research project will likely start out with a certain di-rection, in DiaLoop the direction is seen as part of the continued process, as it will likely change during the course of the analytical work. In DiaLoop, a good research question will also be based on a problem, in this case a theoretical problem.
Having a theoretical problem means there is something the researcher does not understand or is unable to explain. The authors also explain this by saying that a theoretical problem is when the actual is different from the expected. When something deviates from our expectations it poses a problem for us, and this problem con-stitutes a motivation to find answers (Alrø et al., 2016, pp. 77–78).
Example
The analysis done in this thesis started with a very open wondering about what it is that makes paid work different from volunteer work. Because I have ex-perience with both types of work myself, and found it difficult to answer the question in a satisfactory way, this posed a problem for me. I could clearly dif-ferentiate between the two, but I found it difficult to actually explain what made them different. During the process of analysis and through reading other studies and theories, this problem became more specific, and was, among other things,
developed into a focus on the influence the sectorial context seems to have, as well as the influence of the individual attitude of the person doing the work. On a smaller scale, the intuitive feeling that arose while standing outside in the rain with Dorthe has also played a role in my working my way towards a direction of the analysis and of this thesis as a whole. As described with the hermeneutic circle, every small part I have understood and interpreted in different ways has at the same time influenced my overall idea of what the direction of the thesis has been. In that sense, working with the direction or problem posed by the thesis has been an ongoing process throughout the PhD project.
3. Observation
Observation consists of everything directly observable in the data, and when referring to the actual process of analysis in DiaLoop, it seems to refer to actu-ally observed data, i.e. data belonging to the domain of Empirical as opposed to potentially observable data belong to the domain of Actual. For analysis of data to take place, it must first have been observed. Observations are characterised by being shareable with other observers; what DiaLoop considers an observation must be potentially available to everyone looking or listening. Observation is also referred to as external sensing, illustrating that it is the result of human sensing di-rected outwards towards the surrounding world. Its opposite is internal sensing, which will be described further in the next section, about the experience part of the DiaLoop model (Alrø et al., 2016, pp. 81–82).
While observation has to do with everything potentially observable, this does not mean that a) as much as possible should be observed or b) that everything observed holds equal value for finding answers to the chosen questions. Where intuition can lead to an idea about what to focus on in the analysis by insisting that the researcher immerse himself in the data, it cannot document or explain exactly what happens in the data (the situation observed), thus it is ultimately un-able to explain more precisely what is going on. The external sensing used when observing allows me to gain some distance from the subject researched and to answer questions about what concretely is happening at those places in the data where my intuition tells me something is going on. In that sense, observation can also be a stepping-stone for going deeper into the analysis. Where intuition may direct my attention to a certain situation or action, observation can move the analysis forward by starting to connect the intuition with what actually seems to be going on. Observations are examples from the data that can answer questions like, “How do you see that?”, “How does that show itself?”, or “What is that based on?” The answers to these questions are in DiaLoop partly made up of
actual observations (Alrø et al., 2016, p. 82).
Example
When outside in the rain with Dorthe and the scout children, observation has to do with registering what is happening, trying to mentally take a step back from my “intuitive idea of the counterintuitive” in the situation, and focusing more on seeing and listening as a way to gain a better understanding of what is going on.
4. Experience
Experience has to do with the above-mentioned internal sensing. Experience in the DiaLoop model refers to the researcher’s own experiences. It is the way he reacts when confronted with the externally sensed observations. In DiaLoop, experi-ence is a collective term for four different ways something can be experiexperi-enced by an individual, listed below with examples:5
Body sensing
I feel heavy, my stomach hurts, I get chills Emotional reactions
I am angry, I feel sad, I am happy, I am relieved, I am proud Inner impulse toward action
I want, I need, I have a sudden impulse to Mental reactions
I feel alert, I become interested, I am bored, I have doubts, I become curious.
These four different ways of experiencing share the same basic trait of start-ing with “I,” meanstart-ing that they all belong to the researcher himself. It is the researcher who feels heavy, has a sudden impulse, or feels bored. The question now, of course, is what these experiences have to do with conducting analysis in an academic and transparent fashion. Alrø, Dahl, and Schumann argue that our experiences as researchers can be an intriguing opening into the analysis. Our ex-periences are the result of us reacting to something in our data. We are triggered by something, which might make us curious, making us look more closely at what it was that caused us to react in this way. While both experiences and intuition can bring us into closer contact with our data, experience differs by being only about the researcher’s own reactions, whereas intuition also can be related to ideas about what is happening “out there” in the data or situation. (Alrø et al., 2016, pp. 93–95).
5 The first three are inspired by the BodyKnot model. See (Jarlnæs & Marcher, 2004, p. 211).
In addition to opening up the analysis, experiences are also an important part of the process of analysis, as they can help the researcher understand how he has reached a certain conclusion. In DiaLoop, it is argued that the researcher’s own reactions to other peoples’ actions will have a tendency to tint his interpretations of these actions. Becoming aware of your subjective influence on the interpre-tations made can help you see different possible interpreinterpre-tations of a given piece of data. If the researcher becomes annoyed or sympathises with a person who is part of situation they are analysing, this may influence (tint) their interpretation of the situation in ways the researcher is unaware of and lead to misinterpreta-tions (Alrø et al., 2016, pp. 94–105). What I consider misinterpretamisinterpreta-tions here are interpretations that cf. philosophical hermeneutics are incoherent or self-contra-dictory.
Example
If I interpret Dorthe’s contact with a user of the library as being conflictual or tense, this could have to do with the fact that I am feeling discomfort when ob-serving the conversation. But if I neither Dorthe nor the user felt any tension, and if a subsequent analysis of the conversation shows no signs of tension or conflict between the two of them based on the way they communicated, then it would most likely be my experience of discomfort that led me to this incoherent or self-contradictory interpretation. I like to think of my experiences as a criti-cal ally or teammate when doing analysis work, something I as a researcher can check in with and “say out loud” to myself. If I was feeling discomfort when the situation happened during fieldwork, or if I feel discomfort when analysing the video later, this is something I try to be aware of during the process of analysis.
Using Gadamer, I have argued that it is not possible to rule out the researcher as part of the research process, and that there is no such thing as a neutral under-standing of something. Thus the point here is not to say that if I just become aware of my own experiences I can reach neutral conclusions about the world and other people in it. The point is, that to make interpretations that are likely to hold some truth based on the Principle of History of Effect, I need to be aware of my own experiences, as they can lead me to interpretations that would likely not be deemed true by this principle. In other words, I am arguing that being aware of my own experiences helps me reach conclusions that are more coher-ent and not self-contradictory with the prescoher-ent-day historically based, somewhat common criteria for interpretation.
5. Identification
DiaLoop operates with two ways of using theory in the process of analysis:
5. Identification and 7. Interpretation, which I will argue match the previously mentioned distinction between immediate understandings and more reflected understandings. Rather than seeing them as two distinctly different terms, I un-derstand them as related to each other at each end of a continuum between directly observable and a more reflected understanding. See figure 6.
Identification refers to connecting theoretical terms to empirical data in a way that requires little explanation to be understood and agreed upon by others. The iden-tification of these theoretical terms in the data is close to being directly observa-ble or immediately understandaobserva-ble.
Example
When analysing the conversation Dorthe had with the scout children while stand-ing outside in the rain durstand-ing the camp, I began by identifystand-ing some patterns in Dorthe’s way of communicating: most of the things she said direct the children to act in certain ways. I found this pattern by identifying a theoretical term: a number of directive speech acts (Searle, 1969; Vagle, Sandvik, & Svennevig, 1993) in Dorthe’s communication. Having found many of these, I argue that there is a localised pattern in the data.
DiaLoop is based on the idea that identifying the more observable theoretical terms is a pathway to finding patterns in the data and to reaching the more re-flected understandings. The identified terms make up the systematic groundwork of well-argued broader interpretations (Alrø et al., 2016, pp. 107–123). The the-oretical terms used in the analyses will be presented in connection to the analysis where they are used.
6. Argumentation
Argumentation, or line of reasoning, is the systematic, analytical work of de-scribing, explaining, and discussing different interpretations of the data. It is the process of explicitly, systematically, and fairly answering the question of, “How do I reach the conclusions that I reach?”, and it plays the important role of showing others how the conclusion reached is a likely, durable, and relevant in-terpretation of your data (Alrø et al., 2016, pp. 126–127).
In DiaLoop, argumentation is not seen only as the text written in final version of a text, but also as the process leading up to being able to write this text. Analysis
Figure 6. Use of theory in the DiaLoop model - Identification 5. IDENTIFICATION
Directly observable 7. INTERPRETATION
Reflected understanding
as a process is described as a movement between, on the one hand, proposing a certain interpretation of something in your data, and on the other hand doubt-ing, asking questions, and requesting further explanation and justification. This movement can happen among several researchers analysing the same data, and it can happen within one researcher as an inner dialogue. In this way, argumenta-tion becomes a continuous process of inquiry and explanaargumenta-tion, and should not just be seen as presenting finished thoughts, but rather as part of building up suitable and non-self-contradictory explanations. In this sense, DiaLoop offers a way to enter into a dialogue with my own understandings of something in my data, helping me explore the data and challenge my initial understandings of it (Alrø et al., 2016, pp. 125–131).
Example
One initial interpretation of why Dorthe chose to stand outside in the rain vol-untarily is that this is an inescapable part of being a scout leader, something Dorthe has to get through. Perhaps she endures the rain because she loves sitting in the evening with the other leaders talking until late at night, but to run a camp, she knows, you also need to have activities for the children. This interpretation could be based on observing the weather, and might be influenced by my own experience of discomfort from standing outside in this weather. However, as part of the process of argumentation, I need to challenge this initial interpretation.
As I stood there with Dorthe, I could already see that she smiled, seemed quite content, acted patiently towards the children, and said nothing about being cold.
These further observations already point me away from my initial interpretation and towards starting to look at what could be enjoyable or meaningful about this activity. I thought about it while standing there and made a few mental notes, and later continued this work as I interviewed Dorthe about this situation, then fol-lowed up with a more systematic analysis of her communication in the situation, which was video recorded. Through each of these steps different details arose that either supported, challenged, or supplemented the overall interpretation I had of this situation, and the argumentation process helped me make sense of these and continually seek a coherent understanding or indeed interpretation.
7. Interpretation
Interpretation, in DiaLoop, refers to a summary or synthesis of the results of a systematic, well-documented, and well-argued analysis. It refers to involving theoretical terms or ideas, which on a more general level are used to understand, describe, and put the phenomena studied into a broader perspective than what the immediate and less reflected understanding requires. The more explanation