Abstract
The last decades have seen the completion of an increasing number of qualitative comparative research projects on teaching. Challenges and benefits which might arise from a qualitative international comparative research design have been considered. However, very little has been published on challenges and benefits which may arise from using grounded theory in international comparative research projects. This article explores some of these challenges and benefits, focusing on two methodological aspects: the emergent process of developing a grounded theory and analysing data in a foreign language. In order to illustrate the argument, an international comparative PhD project is used. The project is centred on how teachers see themselves with regards to accountability reforms in England and Germany.
Keywords
Introduction
In recent years, qualitative research in intercultural and international comparative education on teaching has increased. Research in this field covers a variety of research interests, e.g. ‘pedagogy and culture’ (Alexander, 2000), the development of teachers’ identity (Czerniawski, 2011), the influence of local and global contexts on teachers’ self-perception (Bargen, 2014b), teachers’ roles (Kelly et al., 2014), professional cultures (Rademacher, 2009) and classroom practices (Fritzsche, 2015).
At the same time, methodological challenges in comparative education are considered. A variety of publications can be found. In the Anglo-American discourse, for instance, Bray (1999), Osborn (2004) or Phillips and Schweisfurth (2007) reflect these on a more general level, and in the German speaking discourse, e.g. Allemann-Ghionda (2010), Bargen (2014a), Freitag (2014) and Parreira do Amaral (2015). Further publications deal with specific challenges of comparative qualitative research projects; for example, on the subject of contextual sensitivity by Crossley and Jarvis (2001), Czerniawski (2015), or on (inter)cultural sensitivity by Osborn (2015). Furthermore, Fritzsche and Huf (2015) discuss comparative and cross-cultural research from an ethnographic perspective.
Despite this, seemingly little has been published on using a grounded theory approach in an international comparative research project on teaching. This is surprising, considering the benefits a grounded theory approach may bring to comparative research projects. By employing a grounded theory approach, researchers may develop ‘a general explanation (a theory) of a process, action, or interaction shaped by the views of a large number of participants’ (Creswell, 2007: 63). In comparative education, a grounded theory approach may be very useful if researchers want to gain a better understanding of (new) phenomena or redefine existing theories. In particular, if you understand teaching as a complex and changing field of action, grounded theory methodology offers a suitable approach to explore and explain these changes from the inner perspective of teachers, students and parents. In the following, I will use the existing contributions in order to illustrate further benefits.
For example, Liebeskind (2012) discusses how comparisons are used in social sciences studies and in grounded theory. By doing so, she illustrates how a grounded theory approach may help to increase cultural sensitivity. She uses the term ‘Nostrifizierung’ (Liebeskind, 2012: 329) to illustrate her point. ‘Nostrifizierung’ is a term coined by Matthes (1992), suggesting that researchers may project their own culturally bound concepts to another culture, e.g. the Western concept of family life to non-European cultures. She suggests that the coding process, the theoretical sampling and the constant comparisons of the grounded theory approach help increase cultural sensitivity in comparative research. However, her work focuses on higher education and research on the employment market.
Troman and Jeffrey (2007) also use a grounded theory approach in their cross-cultural research project ‘Creative Learning and Pupil Perspectives (CLASP)’. In their paper, they focus on how data can be analysed in cross-cultural projects. They work with a grounded theory approach with a focus on ethnographic research. Of particular interest is how a ‘loosely-coupled approach to qualitative comparative research which uses a wide range of data collection methods can be effectively analysed with a qualitative synthesis’ (Troman and Jeffrey, 2007: 515). As they show, grounded theory offers here a suitable way to integrate a variety of data.
Furthermore, comparative researchers often analyse data in a foreign language. Kruse and Schmieder (2012) develop an integratives Basisverfahren (a collection of qualitative methods suitable for various research interests) in order to analyse data in a foreign language. However, their integratives Basisverfahren is not specific to grounded theory. Tarozzi (2013) discusses how far translating and a grounded theory methodology can be combined. Kruse and Schmieder (2012) and Tarozzi (2013) show that analysing data in a foreign language can be a powerful analytical tool.
However, there might also be some challenges which might arise from a combination of grounded theory and international comparative research, 1 as Waldow et al. (2013) and Falkenberg (2015a, 2015b) point out. In their workshop description, Waldow et al. (2013) see a major challenge in the combination of ‘two methodological approaches that both are comparative but follow different logics and apply different practices of comparison’ (Waldow et al., 2013). However, apart from their workshop description, no article seems to have been published so far.
In the following, I will explore some challenges as well as the potential which might arise from working with a grounded theory approach in international comparative research. I will use my own research project to illustrate some of the points I make. In my research project, I draw on a grounded theory approach to analyse how far teachers’ way of seeing themselves is related to accountability reforms in England and Germany. As principles of qualitative research are closely linked to central principles of a grounded theory, I will show that these challenges may also be relevant for other qualitative research projects.
In the subsequent sections, I will mainly focus on two interrelated methodological aspects: first, I will look at the emergent process of grounded theory which might present a challenge for an international comparative research design. Second, I will focus on a vital part of this emergent process: the analysis of data. In comparative research, it is often likely that data are gathered and interpreted in a foreign language. As I will show, this may present challenges but might also be very enriching for the analytical process.
Emergent process of grounded theory
The term grounded theory carries a variety of meanings. I draw on the grounded theory methodology according to Corbin and Strauss, in particular in the latest publication of their book Basics of Qualitative Research (Corbin and Strauss, 2015). In publications, the term grounded theory may be used to refer to the grounded theory methodology
2
or to methods which are used in a grounded theory methodology. Occasionally a grounded theory may also describe the product of a research process, i.e. a developed theory. In the following, it will be important to keep these three different meanings in mind. Corbin and Strauss (2015) define the term grounded theory as follows: Grounded theory is a qualitative methodology that aims at constructing a theory from data … What makes grounded theory unique among other qualitative methods is its approach to data collection and analysis. The researcher does not begin the research with a pre-identified list of concepts. Concepts are derived from data during analysis. Analysis begins with the collection of those very first pieces of data. Concepts derived from initial analysis guide collection of subsequent data. Each data collection is followed by analysis. This process of data collection followed by analysis continues until the researcher constructs well-integrated and dense theory.
This quotation highlights the emergent process: a theory is gradually developed by continually collecting data, analysing data, and developing codes and categories. These codes and categories are developed by making ‘constant comparisons’, which is the ‘analytic process of comparing different pieces of data against each other for similarities and differences’ (Corbin and Strauss, 2015: 85). The starting point of the research process lies in the collection and analysis of data; the endpoint is marked by a theory which is grounded in data. This emergent process implies two interrelated logics: first, that theory is gradually developed, with comparisons as a vital part of this process; second, that comparisons have a reference point in data.
In their workshop description, Waldow et al. (2013) illustrate challenges which might arise from a combination of grounded theory and comparative research, quite well: The Grounded Theory itself, which is the end product of the research process, grows so to speak from innumerable processes of these types of (micro-) comparisons. Internationally comparative studies, even case-oriented small-N-comparisons, need to use more or less abstract categories in order to compare previously specified entities. This logic of comparison corresponds more to a macro-type of comparison, which, however, may sit somewhat uneasily with the ‘constant comparative method’.
In this quotation, Waldow et al. (2013) refer to the grounded theory as a product of a research process which marks the endpoint of research. This stands in contrast to international comparative studies, which take – if you follow Waldow and his colleagues’ argument – ‘more or less abstract categories’ 3 as a starting point. These processes in research seem to be contradicting, which may present a first challenge. A second difficulty may arise through the ways comparisons are made: in a grounded theory approach, micro-comparisons seem to prevail, while in comparative education there is a tendency towards macro-comparisons. In the following, I will have a closer look at the first challenge. In the subsequent section, I will return to the second challenge.
The research process of a grounded theory can be described as emergent. This also applies to the research process of qualitative research. In his introduction to qualitative research, Creswell (2007) describes the process of qualitative research as follows: The research process for qualitative researchers is emergent. This means that the initial plan for research cannot be tightly prescribed, and that all phases of the process may change or shift after the researchers enter the field and begin to collect data. For example, the questions may change, the forms of data collection may shift, and the individuals studied and the sites visited may be modified. The key idea behind qualitative research is to learn about the problem or issue from participants and to address the research to obtain that information.
Other researchers in the field of qualitative research also embrace this emergent character. In the German speaking context, this emergent process is discussed under the term of the ‘iterativ-spiralfoermiges Vorgehen’ (a process which is characterised by its iterativeness and constant movement between data gathering, data analysis and theory development), e.g. by Flick et al. (2015), Helfferich (2013) or Kruse (2014). In a grounded theory approach, the emergent process is of vital importance. The theory gradually develops through a mutual process of data gathering, data analysis and theory development. In the German speaking context, this is discussed by, for example, Kelle (2011), Mey and Mruck (2011) or Strübing (2014). 4
If you compare the process of a qualitative research project to the process a comparative research project may take, you will find various suggestions by various authors. Common sense seems to be that there cannot be a prescribed process which suits all research projects (cf. Phillips and Schweisfurth, 2007: 101; Parreira do Amaral, 2015: 123). Still, there seems to be a tendency to suggest at least a certain level of initial planning and employing of analytical concepts before data are gathered. In the following, I have chosen a quotation by Phillips and Schweisfurth (2007) to illustrate this. They summarise their structure for a comparative research project as follows: It [the approach] suggests a systematic procedure for comparison which takes into account the essential requirements of proper conceptual analysis of a problem, the isolation of variables to be investigated after a thorough study of context, the development of hypothesis based on attempts at explanation, a revisiting of conceptual analysis based on the findings, and a consideration of any applicability of these findings. It assumes that a proper comparative study will first and foremost address a clearly defined issue in education which is to be analysed, and then, having established a strong theoretical basis for such analysis, will proceed to examine the issue in question in two or more contexts.
Parreira do Amaral (2015: 127) argues in a similar way. In order to provide transparency and a high level of reflection, a series of questions should be answered with regards to the comparison. 5 Apparently, thorough planning is necessary. This can include many decisions and assumptions which are made before the first data are gathered. There seem to be many good reasons for this. For instance, how could you make sure you compare equivalent phenomena if you do not know about them? How do you define your sample if you do not know why it would make sense to compare it? How could you plan your field access in another country if you do not know which research steps should be taken next? I do not want to neglect these issues. However, it could be that the establishment of a ‘strong theoretical basis’ (Phillips and Schweisfurth, 2007: 101) before data are gathered might present an obstacle for a qualitative research design. This issue is also discussed in the grounded theory methodology, in particular with regards to the function and use of theoretical knowledge. Strauss and Corbin speak of ‘sensitizing concepts’ which help the researcher to develop theory. However, there is a great emphasis on the emerging aspect of theory from data. It should not be forced, i.e. by using a predominantly deductive way of analysis. What I would like to suggest is to pose the questions differently: how much emergence can be beneficial for a qualitative comparative research project and how much structure and pre-defined assumptions are needed?
This contradiction between the need for planning as well as having a structure to follow and the need to keep the research process as open as possible is challenging. In my research project, I tried to solve this by employing various strategies. For instance, I planned two research stays in England and Germany. I began by conducting an interview with a teacher in Germany. The analysis of this interview helped to inform my further theoretical sampling. I also looked into research undertaken in the field and tried to analyse further relevant aspects which might also be influential for my sample. Before my first field access to England, I came up with a list of parameters I found to be influential. In England and Germany, I led six interviews with teachers, including head teachers and deputy heads. Analysing the data in greater depth, I found that some aspects I thought to be influential were less important whereas other categories started to emerge which I had not thought of beforehand. In the English sample, for instance, it seemed to be very important if schools considered themselves to be close to an Ofsted-inspection (Office for Standards in Education, Childrens’ Services and Skills). What seemed to be influential in both countries were, for instance, biographic backgrounds of the teachers. As much as my sample allowed, I tried to compare by contrasting maximally or minimally. Of course, after some time, I found that my sample did not cover all the aspects I wanted to compare. So, this was when I organised the second field access to both countries.
Compared to the idea of theoretical sampling, as it is used in grounded theory, four subsequent field accesses may not seem to be much. In their book, Corbin and Strauss (2015: 69) warn researchers to do all data collection at once as this may hinder the full development of a concept. However, they also admit that ‘this is not always possible. Sometimes researchers travel long distances to collect the data’ (Corbin and Strauss, 2015: 69). So, pragmatic reasons may interfere when following the ideal way of theoretical sampling. For instance, the funding of my research allowed two research stays abroad which I found very beneficial. Another pragmatic restriction lay in the access to schools: in both countries, I could only establish field access through gate keepers. This may also have consequences for the sample. These should be considered carefully. Even though pragmatic considerations have to be made and planning in the long term may prove to be necessary, they do not have to restrict the entire research process. In particular, when it comes to analysing the data, the emergent process can be given more emphasis. I will explore this idea further in the following section.
Analysis of data in a foreign language
In comparative research, it has been considered that analysing data in a foreign language can present a challenge. 6 If researchers do not know the foreign language, they depend on translations. ‘If there is reliance on translation and interpretation, safeguards must be in place to ensure accuracy and reliability’ (Phillips and Schweisfurth, 2007: 95). Phillips and Schweisfurth briefly explain the technique of back translation. Yet, as Osborn (2004) points out, this may be problematic.
This approach [of back translation] later came under heavy fire, however, since it was clear that such methods, although designed to produce equivalent words in the two languages, could not ensure that these literally equivalent words or phrases conveyed equivalent meanings in the two languages.
The quality of translations used in qualitative research is also addressed by Enzenhofer and Resch (2011). They suggest that translators and researchers should work closely together, if possible from the very beginning until the end of the research process (cf. Enzenhofer and Resch, 2011: 86).
Enzenhofer and Resch (2011) as well as Osborn (2004) argue in a similar way: language carries meaning. Meaning can be vague depending on the situation and the context in which the words are used. In the context of qualitative research, this issue is addressed by the terms Indexikalität and Fremdverstehen (for a comprehensive overview, cf. Kruse, 2014: 60–92). Creswell addresses this by speaking of the ‘interpretive inquiry’ (Creswell, 2007: 39). Researchers always interpret this meaning and, at this point, there is no difference whether this interpretation is made in a foreign language or in the mother tongue.
7
This idea relates to what Corbin and Strauss (2015) suggest for analysing interviews: When analysts interpret data, they are translators in the form of concepts of other persons’ words and actions. They make the voices of other persons heard and are the go-betweens for participants and the audiences they want to reach. As every language translator knows, it is not easy to ascertain meaning. Words can have different meanings from one language to another and from situation from another.
Here, Corbin and Strauss do not explicitly refer to data analysis in a foreign language. However, interestingly enough, they compare the process of data analysis with the work of a translator. In a methodological perspective, this idea is also addressed by Tarozzi (2013). He compares processes of translations and processes of a grounded theory methodology. His basic assumption is as follows: ‘The meaning of a word or proposition is not only a linguistic construction; it is also pragmatic, historical, semiotic and, in a broader sense, cultural’ (Tarozzi, 2013: paragraph 22). To translate means to interpret by using language as a medium. ‘This same process occurs in doing research: an interpretive event of social/psychological/educational … phenomena using language as a medium, after interpretive acts’ (Tarozzi, 2013: paragraph 23). Having established these parallels, Tarozzi explores in greater detail the similarities of translation processes and grounded theory processes. For instance, reading the text that should be translated is like collecting data and open coding. Both processes mean ‘an encounter with the other, the foreigner’ (Tarozzi, 2013: paragraph 25).
Having established that translating and doing grounded theory share similar processes, Tarozzi explores the benefits of coding in a foreign language. In his understanding ‘doing research in another language is a powerful analytic resource. Coding in another language requires continuous acts of interlinguistic translation so that it grows our own faculty to understanding’ (Tarozzi, 2013: paragraph 1).
Seen from this perspective, analysis of data in a foreign language makes the process of interpreting more natural: since the data is indeed in a foreign language, researchers cannot but start wondering about meaning. A natural focus will be on language, which is also an analytic research strategy suggested by Corbin and Strauss (2015: 99). Quite likely, questions will arise, e.g. on the meaning of a word, and comparisons will be made. Questioning and comparing are also two prominent strategies for analysing data to facilitate obtaining the participants’ perspective (cf. Corbin and Strauss, 2015: 88–105).
In my research project, questioning and comparing proved to be very enriching strategies when conducting the interviews and also for the subsequent analysis process. I will explore this potential in greater depth by drawing on an interview I led in England. At the beginning of the interview, I introduced myself as a researcher and teacher from Germany. I think this also influenced the way the English teacher interacted in this interview. For instance, the teacher explained what the abbreviation RME (religious and moral education, a school subject) stood for. In other interviews, I found similar explanations which were done on the side. However, there were some sequences in which the explanations were longer and led to rich sources for analysis. In the following, I will quote such a sequence. It is taken from an interview with a classroom teacher (B) who works at an English primary school. Towards the end of the interview, the interviewer and the teacher talk about students’ performance and how it is assessed. The teacher talks about his experiences. He concludes by asking the interviewer (I) a question in return (see Note 8 for interpretation of symbols used).
1 B: […] Do you have league tables of schools?
2 I: No.
3 B: No, see, yeah we’ve got that as well. (.) Yeah.
4 I: No, that is why it’s so //interesting to compare.//
5 B: //I’ll come and get a job over there.// (laughs, interviewer joins in) Sounds a lot
6 more (.) eh pleasant really in that area.
7 I: Yeah, but the system is about to change.
8 B: Is it?
9 I: Yeah.
10 B: (hard to understand, speaks very fast) God, you don’t want to copy us. (laughs)
11 I: Sorry?
12 B: (still laughing) You don’t want to do what we do.* It’s a bit too much really for/
13 (.) yeah. (..) It probably was born from a need, wasn’t it? And that’s the thing,
14 it probably/ it probably came from a system that was perceived not to work. I
15 don’t know, I haven’t been teaching that long to know but (..) yeah I think/ I think (..)
16 you can allow a certain amount of freedom with any/ with any system, but you
17 need to have some sort of standards, as well. So I think if the standards start to fall a
18 little bit, you need systems that step in, so I can see why it’s there. (.) Yeah. You know,
19 just not very nice when it’s you. When you’re at the end of it, you know, ‘cause if you
20 think you’re doing your job right and someone is telling you what to do, it’s quite
21 stressful, isn’t it? (.) Yeah.
In line (l.) 11, the interviewer has difficulties in following 9 what the teacher says. The participant has no difficulty in explaining; he readily offers a synonymous phrase (l.12). However, this is where a wider reflection process seems to start. After a short pause, the participant develops an explanation (ll. 13–18) for the existence of league tables. He finishes this explanation by expressing an understanding for the system (l.18). This sequence (ll.13–18) offers a rather distanced, observing perspective. To me, it appears as if the teacher was taking an outsider’s perspective. Almost opposite to this is, however, is the following sequence (ll. 18–21) in which he seems to take the insider’s perspective. He draws back on his own experience of being ‘at the end of it’ (l.19). There seems to be a conflict which might centre on his professional autonomy and external control (ll.19–21).
Two contradictory perspectives are taken on one and the same phenomenon (league tables). Interestingly enough, the first perspective which I would describe as the outsider’s perspective admits a justification for the way the system is. Here, the teacher opens up the macro-perspective as he speaks of ‘systems’ (l.18). This is contrasted with the process of being controlled at a micro-level. At this point I would like to reconsider what seemed to be a contrasting logic earlier on. In this sequence, the macro-level seems to be connected and relevant at the micro-level and hence should not be seen as separate or contrasting (cf. quotation by Waldow et al. (2013) above). However, I would rather suggest searching for interrelations. For instance, this sequence inspired me to look at official documents and compare these with the contexts participants touched upon in the interviews. In this case, for example, I looked at an official document issued by the government on accountability. What I found surprising was the frequent recurrence of the words evidence and performance. This led me to the question of how far corresponding documents in Germany would have similar or different recurrences. Analysing these, I found the word Entwicklung (development) and Qualitaet (quality) as well as combinations including these words such as Schulentwicklung (school development), Qualitaetsentwicklung (quality development), Entwicklungsziel (the aim of development) or Qualitaetssicherung (quality assurance) to name but a few. These differing recurrences could be interpreted in the light of different accountability regimes and corresponding test systems. So in this case, more abstract categories and concepts provided a further, very fruitful source for analysing and interpreting my data.
The above quoted sequence could be also interpreted as a dialogue in which the researcher and participant actively constructed meaning. This corresponds very well to the form of interview I have chosen, which draws on problem-centred interview. Witzel and Reiter (2012) describe the problem-centred interview as follows: In short, the PCI [problem-centred interview] invites respondents to co-construct and reconstruct problems together with the interviewers in an interactive and interpretive process … In a discursive dialogue the researchers’ prior knowledge meets the respondents’ practical knowledge. This exchange provides the chance to develop the researchers’ social scientific constructs (of the second degree) in a dialogue with the respondents’ common-sense constructs (of the first degree).
To my knowledge, very little reflection has been done when it comes to choosing a suitable interview form when a foreign language is used. Czerniawski (2015) reflects on what it might mean when participants use a foreign language. Reflecting on his experiences, Czerniawski concludes: I cannot possibly account for how these interviews, in content, would have varied had I carried them out in Norwegian. In other words I am sure that there were interpretations made of questions I posed that highlighted linguistic differences; however it is impossible for me to say what these might have been. All I can do is draw attention to the fact that I was aware of their potential existence.
I would even go further and suggest that ‘linguistic differences’ (Czerniawski, 2015: 81) may also play a role in interviews which are led in the mother tongue. However, they may be harder to discover. In qualitative research, this is taken into account systematically: qualitative research wants to find out how social reality is constructed. The researcher is a part of the construction process and cannot be set as ‘neutral’ or ‘eliminated’ from the analysis. A variety of strategies have been developed in (comparative) qualitative research to reflect and analyse this construction process systematically. What are the impacts of participants communicating in a foreign language?
Czerniawski (2015) briefly explains that participants tend to need more time to answer questions when they use a foreign language. At the same time, these answers seemed to be more ‘focused’ (Czerniawski, 2015: 81). I am not sure if focused implies here that answers were formulated on a more abstract level, e.g. by giving summaries, descriptions and well-constructed arguments. If so, this may also have implications on the choice of the interview form. For instance, if you choose a narrative interview as it has been developed by Schütze (1983), 10 researchers rely on the participant’s ability to produce Stegreiferzaehlung (their ability and readiness to tell stories about their lives off the cuff). Ideally, the interviewer asks one question at the beginning. The participant tells about his/her life and is not – in the best scenario – interrupted by the researcher. Here, it may be difficult for the analysis if participants tend to use rather more abstract forms of telling about their lives, i.e. descriptions and argumentations. Additionally, it may be necessary to ask more questions to follow what is being said. So, careful considerations should be made when either researcher and/or participant communicate in a foreign language as they may be influential on the conduct as well as the interpretation of the interview.
Conclusion
This article explored the potentials and challenges which might arise from using a grounded theory methodology in an international comparative research project. It focused on the emergent character such a study might bring as well as the use of a foreign language in interviews. I would agree with Tarozzi that ‘doing research in another language is a powerful analytic resource, when the researcher is using an inductive and comparative method aimed at generating theory’ (Tarozzi, 2013: 27). However, researchers are quite likely to meet contradictions which seem to result from different logics. These may not be easily resolved; however, moving in between these logics and actively searching for possibilities of combinations may be very beneficial for the research project. To me, linking international comparative research with a grounded theory methodology is indeed an emerging process which should be given more attention in future research.
Footnotes
Declaration of conflicting interest
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
