positivism and evidence based practice

first posted on Janshs blog Feb 2009

One of my fellow students posed the question: what is Evidence Based Practice and is it the same as positivism?

I have been wrestling with this thought for a couple of days now.

Positivists take the view that knowledge is developed by applying methods from the natural sciences – that is, that it is based on behaviour that is observable rather than on values or reasons. This means that it is based on experimental evidence or (at least) on quantitative analysis.

E891 Educational Enquiry: Study Guide prepared by Martyn Hammersley, pages 18-21:-
EBP movement began in medicine (EBM), using, for example randomized controlled trials (RCTs) which are intended to be easily replicated in other circumstances. However, even in medicine, the EBM approach may well be tempered with experiential evidence. This might also involve observation and interviewing.

So it might be appropriate to say that EBM/EBP may use evidence from a wide variety of quantitative and qualitative data but it must be valid and reliable. For EBP to affect practice, it is likely that generalisability will also be sought, although some (Schofield, 2007, Dellinger and Leech, 2007, Ely et al., 1999) might view this as being less of a necessity, provided that thick descriptions (Geertz, 1973) are used so that the reader can determine the degree of fit with her/his own context.

So I do not think that postivism and EBP equate totally, since there is an opportunity for EBP to draw on qualitative methods.

Dellinger, A. B. and leech, N. L. (2007) ‘Toward a Unified Validation Framework in Mixed Methods Research’ in Journal of Mixed Methods Research, Vol. 1, No. 4, 309-332
Ely, M., Vinz, R., Downing, M. and Anzul, M. (1999) ‘On Writing Qualitative Research: Living by Words’. London: Falmer Press
Geertz, C. (1973) The Interpretation of Cultures. New York: Basicflooks
Hammersley, M. (2007) E891 Study Guide for Open University
Schofield, J. (2007) ‘Increasing the generalizability of qualitative research’ in Hammersley, M. (ed.) Educational research and evidence-based practice. London: Sage

Communities of Practice

Just a few musings of me, ... I'll try to add to this later but I'm late for work!

In a community of practice (I think original idea Lave and Wenger), I would suggest that members share tacit understanding and tacit knowledge (worth looking up Polanyi) which may cross other cultural boundaries. So, for example, teachers in many countries may share similar beliefs about the nature of how people learn, or want to learn. I guess that this could be because of personality traits shared by people who join certain communities but we do also sometimes actually set out to enculturate newcomers - this might be through fairly formal means such as asking them to demonstrate evidence against a set of competences in order to 'qualify' and it can also be less formal through discussion and observation of peers.

I think this may be why #phdchat on twitter has developed as so much more than a professional or social network. There is, perhaps, a community of researchers.

"The concept of the researcher as “bricoleur” or “carpenter of her (sic) own personal theoretical constructs”, is put forward by Ely et al. (1999: 1, 4) who state that it is important to “breathe into our words the life we have experienced”. This is not to say, as Glaser emphatically states (2002: para. 9), that the researcher composes the story on behalf of the interviewees. What it does mean is that writer is attempting to express the experiences of the respondents in a way which is based on tacit, shared understandings that exist between them." (Moreland, 2007: 90)

other related blog posts
stigmergic collaboration
learning theories - is apprenticeship still valid?
The memory of social networks

Further reading

Athannasiou, N. and Nigh, D. (2000) ‘Internationalization, Tacit Knowledge and the Top Management Teams of MNCs’. Journal Of International Business Studies. 31: 47-487

Lave, J. and Wenger, E. (1991) ‘Learning and Pedagogy in Communities of Practice’. In J. Leach and B. Moon (eds.) Learners and Pedagogy. London: Paul Chapman Publishing.

phronetic research

Just some stuff I wrote in my thesis (let me know if you want refs)

Human nature is such that it is highly likely that the researcher will bring some personal theories to the start of a grounded theory study, even if these are weakly defined at this stage. Thomas and James (2006: 7) criticise the concept of GT, partly because of this issue of personal experience. Mjøset (2005: 5) is also challenging GT as a methodology. However, he pinpoints an important issue: “Grounded Theory [….] does not aim to generalize to a broader population.” The importance of the GT approach is that the situations being explored are too complex to explain using only a priori theories or hypotheses that might have been formulated before the research was undertaken and based upon generalisations from other situations. The only way that a satisfactory fit can be made between theory and actuality is by generating articulated hypotheses in context as they emerge.


Wright Mills (1959: 6) propounds the idea of a ‘sociological imagination’ to describe the link between the socio-economic background of the phenomenon being investigated and the life history of the participants in the research (the respondents and the researcher). The underlying aim of Wright Mills’ concept of the sociological imagination was to contribute to the discussion about the validity of social sciences at a time when the natural sciences were still much more highly thought of.

Anfara and Mertz (2006: xiv) cite Silver (1983) as defining theory as “a unique way of perceiving reality” leading not only to a deeper insight but also to a “fresh and different perception of an aspect of the world”.
The tacit understanding that I have as an insider researcher, coupled with the richness of the data, and alongside my approach to GT/situational analysis gives me just this kind of profound insight and leads to emergent grounded theory. Although Anfara and Mertz present a linear relationship from concepts/events to constructs (“clusters of thoughts”) to propositions (or “expressions of relationships among several constructs”) to theory ( “a ‘set’ of propositions”), the process is probably nested or looped, since the researcher may move backwards and forwards between concepts, constructs and propositions before any kind of substantive theory emerges. Even then, the emerging theory may drive the analysis back to the very beginning and start off a new loop. For Anfara and Mertz (ibid: xvii) a theory is useful if it “tells an enlightening story […] that gives new insights and broadens your understanding”.

Anfara and Mertz (ibid: xxii-xxiii) cite Cresswell (1994: 86), stating that, for him, ethnography or phenomenology involves starting with a framework whereas in grounded theory the data collection and analysis takes place before “using theory”. This is something of an oversimplification of the situation and indicates a clear cut dichotomy on theoretical stances.
Figure 3.1 (overleaf) lists Cresswell’s five ‘research traditions’ of qualitative research, along with his paradigmatic assumptions and ideological perspectives, as referred to by Anfara and Mertz (2006: xxii – xxiii).

For example, I could take 1, 2, 3, and 4 from figure 3.1, map them on to a, b, and c, and ‘underpin’ them with all of the ideological perspectives listed. This would lead to my use of narrative transcripts with my own ontology/epistemology/methodology clearly determining the constructs and propositions that I develop. My ideological perspective is such a mixture of those stated above that it would be difficult to divorce them from my whole conceptual framework.

figure 3.1 – Cresswell’s guide to a research study design

Anfara and Mertz do go on (ibid: xxiii, xxiv) to consider the work of Merriam (1998), who indicates the ways in which our framework as researchers is linked to a particular literature base and disciplinary orientation – making the whole process ‘theory laden’. They also (ibid: xxv) refer to the work of Mills (1993: 103, 144) who defined theory as an ‘analytical and interpretive framework’ which is for helping the researcher to make sense of a situation. Mills, as reported by Anfara and Mertz, perceives that there are both implicit and explicit theories “underlying” any investigation – even when a theory is “purported to be emergent”, the study will have been framed by beliefs and initial propositions.
As Anfara and Mertz write (ibid: xvii), it is the “diversity and richness” of these initial analytical frameworks that encourage researchers to “see in new and different ways what seems to be ordinary and familiar”.
Flyvbjerg (2001) takes a different view of the role of the social sciences in being able to generate theory. He asserts that there is little point in trying to generate empirical theories, such as in the natural sciences, since the very matter being subjected to investigation cannot be suitable for that approach. He bases his work on the Aristotelian division of knowledge into episteme (scientific, know why, irrefutable), techne (craft, know how) and phronesis (ethics, values, variable, context dependent). If we are investigating what is basically a value-laden human phenomenon, Flyvbjerg suggests that we should take a phronetic approach to our methodology. In such methodology, he suggests we ask the following questions (2001: 60):

1. Where are we going?

2. Is this desirable?

3. What should be done?

4. Who gains and who loses; by which mechanisms of power?

As he states, this means that there is an implicit investigation of values but he denies that this lays the approach open to criticisms of relativism since (ibid: 132) a “focus on relations of values and power” leads to interest from others than the researcher. These others will evaluate the research themselves so the phronetic researcher must deliberately assimilate the context and learn from it – without ‘going native’.

Henstrand (2006: 16) actually embraces the idea of ‘going native’, stating that “proponents of ethnomethodology” actively encourage the researcher to become involved with the group being studied in order to “understand the contextual meanings and avoid distorting the vision of the world” – here she cites Adler and Adler (1987: 32); “going native is the solution rather than the problem”.

On the topic of the richness of data and how this can lead, if not to a simple theory, then certainly to a clearer understanding of the situation, Flyvbjerg (ibid: 133) says that phronetic researchers must begin by asking “little questions” and, with patience and attention to detail, focus on Geertz’s (1973) “thick description”. Flyvbjerg would probably agree with Wright Mills: the problem with Grand Theories is that they are too general to ever be empirically observable.

Grounded theory in contemporary studies makes use of a rich array of data, building in an assimilation of contextual partiality. This potent combination allows the researcher to use the joint processes of data collection, literature search and content analysis to probe deeply into relationships and power interactions. The emerging themes and substantive hypotheses which are formed can then be used to continually develop the investigative area.

What is the truth that is your tacit knowledge?

I well remember the first residential weekend when I started my doctorate. My research proposal had been accepted – albeit with some fairly lengthy feedback about how it would need to be refined; I had written a progress report based on said feedback; and here I was, bright-eyed and ready to make my mark on the world of research. Or maybe not quite as ready as I thought: dear reader, indeed I was most unready.


The first group discussion started; there were, perhaps, six students and six supervisers. “So, let’s all begin by describing our personal theoretical frameworks – Jan perhaps you’d like to start us off?” Holes in floors sprang to mind most readily – where was the framework for my escape? I had to admit that I had no idea what the question meant. Well, they may all have thought I was stupid but at least I hadn’t proved it.

The months and years passed by. Still, I was searching for that elusive framework to my research. Just how did my methodology underpin my methods? I think I ‘got’ it just in time. Since then, I have supervised others and also had informal discussions with friends undertaking the same sort of study. It seems to me that this is a crucial factor when jumping through the academic hoops to achieve the status of doctor and a licence to do unsupervised research.

Like many things in life, it comes to some people easily and for others it is more hard won. For some it arrives early on. For still others, the ideas do not seem to crsytallise until after the event. Theoretical frameworks sometimes have recognised ‘names’ – but not always.

Are these the key questions that novice researchers must ask themselves:
• What is the ‘truth’ that for you is tacit knowledge?
• How will the methods that you employ to make the tacit become explicit relate to your theoretical framework?

Bear in mind that your ‘theories’ or ‘framework’ do not have to have ‘names’. For example, if you employ an observation pro forma for the videos of the lessons, will that have an empiricist stance? On the other hand, if you carry out interviews, will this have more of a Grounded Theory approach because you are unsure what themes might emerge?

I am following up with three bits of writing I did ages ago. As always, comments most welcome please.

The following is something I wrote some time ago:

“As a novice researcher in the postmodern era, I turned to some writing on the topic to see if I could find a fit with my approach to Grounded Theory (GT). Clarke (2005: xxiv) describes postmodernism in an easily accessible manner when she cites Fontana (2002: 162) explaining that postmodernism abandons “overarching paradigms and theoretical methodological metasystems”. For Clarke, postmodern research involves embracing complexity, partial hypotheses, and contradictions. She continues (ibid) by highlighting the fact that, from the postmodern perspective, all forms of knowledge are socially and culturally produced – and this includes both natural and social sciences, as well as ‘lay knowledge’.

In other words, all knowledge is situated. This has very real implications for my approach to GT since it implies that my tacit knowledge of my field of inquiry is not only valid but also enables me to construct analytical methods which pertain directly to the phenomena being measured.

Clarke proposes three ‘maps’ which are intended to be used as analytic exercise and she asserts (ibid: xxii) that through mapping the data against these three axes, the “analyst constructs the situation empirically” – so that the situation becomes the ‘unit of analysis’.

1. situational (considering relationships)
2. social world/arenas (meso-level interpretations of the context)
3. positional maps (positions taken along axes of difference, concern and controversy in the inquiry)

By using this approach, Clarke states (ibid: xxiii) that the researcher can “simultaneously address” narrative, texts, and – most importantly for me – the influence of change and of power relationships.

She writes that the outcome of these mapping should be ‘thick analyses’ (citing Fosket 2002: 40), like Geertz’s (1973) ‘thick descriptions’ – looking at all the interrelations within the situation.

Pole and Lampard (2002: 132-133) discuss ways in which the interview is a means of exploring a situation “via their (the participants’) personal recollections”. They state that it is too simple to view the researcher’s epistemological stance as the main determining factor in choosing a methodology for a particular study. If one’s epistemological stance is related to a personal view of knowledge, then I agree with them. For me, the purpose of research is to extend one’s own view of knowledge – its nature, sources and limits. On the other hand, one’s ontological position may well have a great bearing on the chosen methodology, since it is related to a particular domain of study which has been chosen by the researcher, rather than by the other participants.”

The following is something that evolved out of the prep I did for my viva:

Sunday, May 24, 2009

what is a theoretical framework?
What is a theoretical framework, where do you get one, and how does it affect the research and the researcher[1]?

• The researcher starts with some preconceptions – which could be termed bias but Glaser (2002) asserts that this is not so because these notions are acknowledged and described.[2]
• These preconceptions determine the researcher’s personal ontology (what s/he perceives as truth).
• This ontology, in turn, determines the researcher’s proposed defined epistemology (what s/he perceives as knowledge) of a certain situation or context which is under examination.[3]
• Thus, the researcher will develop a set of research questions aimed at interrogating the proposed epistemology.
• The ontological perspective and the epistemological stance will determine the methodology of the investigation (including the methods of data collection and analysis)[4].
• The methodology is validated by these perspectival linkages – the sources of knowledge in the domain chosen by the researcher.
• The whole process described above defines the personal theoretical framework for the particular researcher in the particular context. The exact moment that the framework is defined varies according to the research approach being used (Grounded theory, ethnography, phenomenology, etc.). Framework ≠ methodology. Frameworks in qualitative or mixed methods research can be based upon a mixture of small, medium or grand theories (personal, interpretive or paradigmatic).
• Theoretical frameworks in qualitative or mixed methods research do not necessarily have ‘names’.
• The theoretical framework defined by the researcher is unique, a perception of reality expressed through the data (including literature search) accumulated, analysed and (most importantly) interpreted with clear and thorough descriptions of the context.[5]

References
Anfara, V. A., Jr., and Mertz, N. T. (2006) (eds.) Theoretical Frameworks in Qualitative Research. California, USA: Sage
Glaser, B. G. (2002, September). ‘Constructivist Grounded Theory?’ Forum Qualitative Sozialforschung / Forum: Qualitative Social Research. 3, 3 http://www.qualitative-research.net/fqs-texte/3-02/3-02glaser-e.htm (accessed 20.02.2004 and 05.03.2004) 24 paragraphs
Pole, C. and Lampard, R. (2002) Practical Social Investigation: Qualitative and Quantitative Methods in Social Research. (2002) Essex, England: Pearson

[1] Based upon Denzin and Lincoln (2003:30) cited in Anfara and Mertz (2006:xx1): see references for full citation.
[2] It is important to remember Glaser’s points (2002) about the partiality of the researcher. The perspectives of the researcher (in conceptualising the grounded theory) do not affect the way in which the reader acquires it; therefore no bias can reasonably be claimed. We must show a commitment to realism and we cannot be completely objective so partiality is not a failing.
[3] Although Anfara and Mertz (2006: xiv) present a linear relationship from concepts/events to constructs (“clusters of thoughts”) to propositions (or “expressions of relationships among several constructs”) to theory ( “a ‘set’ of propositions”), the process is probably nested or looped, since the researcher may move backwards and forwards between concepts, constructs and propositions before any kind of substantive theory emerges. Even then, the emerging theory may drive the analysis back to the very beginning and start off a new loop.
[4] Pole and Lampard (2002: 132-133) state that it is too simple to view the researcher’s epistemological stance as the main determining factor in choosing a methodology for a particular study. If one’s epistemological stance is related to a personal view of knowledge, then I agree with them. For me, the purpose of research is to extend one’s own view of knowledge – its nature, sources and limits. On the other hand, one’s ontological position may well have a great bearing on the chosen methodology, since it is related to a particular domain of study which has been chosen by the researcher, rather than by the other participants.
Footnotes 2,3 and 4 from my thesis
[5] Silver (1983), cited by Anfara and Mertz (2006: xiv), sees theory as “a unique way perceiving reality, an expression of someone’s profound insight into an aspect of nature”.

I have also written the following about Mixed Methods research:

What is mixed methods research?
According to Burke Johnson et al. (2007: 121), one definition of mixed methods research, is that it entails “within research paradigm mixing”. In their discussion of the search for a definition of mixed methods research, they remind us that: “The classical pragmatic philosophers (i.e., Pierce, James, Dewey) had it right when they pointed out that the present is always a new starting point.”

Is mixed methods research an opportunity to develop a shared ontology? This would fit well with the description provided by Burke Johnson et al. (2007: 113) of mixed methods research as an approach which attempts to “consider multiple viewpoints, perspectives, positions, and standpoints”.

Perhaps mixed methods research is based upon an epistemological rather than an ontological stance. In his article outlining the paradigmatic history related to mixed methods research, Morgan suggests a pragmatic approach rather than a worldview definition. He considers that methodology should be at the centre of the research design, placing it between methods and epistemology (2007: 68/69). Reflecting upon this has given my original evolutionary and situational approach to grounded theory a slightly altered perspective. Morgan’s exposition (ibid: 71) of abductive reasoning that “moves back and forth between induction and deduction – first converting observations into theories and then assessing those theories through action” is a good description of how a study can be progressed.

As Burke Johnson et al. (2007: 113) describe it, mixed methods research takes account of multiple perspectives. For me, this is the essence of postmodernism. Clarke (2005: xxiv) describes postmodernism in an easily accessible manner when she cites Fontana (2002) explaining that postmodernism abandons “overarching paradigms and theoretical methodological metasystems”. For Clarke, postmodern research involves embracing complexity, partial hypotheses, and contradictions. She continues (ibid) by highlighting the fact that, from the postmodern perspective, all forms of knowledge are socially and culturally produced – and this includes both natural and social sciences, as well as ‘lay knowledge’.

In other words, all knowledge is situated. This implies that tacit knowledge of the field of inquiry is not only valid but also enables the researcher to construct analytical methods which pertain directly to the phenomena being measured.

Burke Johnson (2007: 116) cite Collins et al. (2006) and their “four rationales for conducting mixed methods research: participant enrichment …., instrument fidelity ……., treatment integrity …., and significance enhancement…..” The first of these (participant enhancement), involves ensuring that respondents are all suitable for the aim of the investigation. Instrument fidelity requires that all measurement instruments are also appropriate. Treatment integrity involves maintaining the reliability of any interventions that are made by the researcher. Finally, significance enhancement involves employing thick descriptions (Geertz, Fosket) and thus validating the findings through a process of ‘crystallisation’ (Ely et al.) and providing enough information for the reader to determine a degree of fit (Schofield).

References (I really will get round to these refs)

• Burke Johnson, R., Onwuegbuzie, A. J., and Turner, L. A. (2007) Towards a Definition of Mixed Methods Research, in Journal of Mixed Methods Research, Vol 1, number 2, pp. 112 - 133, SAGE, Thousand Oaks, California
• Clarke, A. E. (2005) Situational Analysis: Grounded Theory After the Postmodern Turn. California, USA: Sage
• Ely
• Geertz
• Fosket
• Morgan, D. (2007) Paradigms lost and Pragmatism regained: Methodological Implications of Combining Qualitative and Quantitative Methods, in Journal of Mixed Methods Research, Vol 1, number 1, pp. 48 – 76, SAGE, Thousand Oaks, California
• Schofield

Reflexivity

Something I wrote ages ago:

The importance of a reflexive approach is, perhaps, defined by what might happen if researchers are not reflexive. If the subjectivity of the researcher is denied or ignored, then the analysis takes no account of the reactions of the subjects to the interviewer and thus a large section of the social interaction that has taken place is not considered when constructing frameworks or theories in action.


During the course of my study on both the Open University Masters’ and Doctor of Education programmes, one particular facet of my own development which has been of interest is the development of a more reflective approach to my reading, to my research, and to my practice as a teacher. [Assignments] ... formed part of my audit trail – documenting the changes in my outlook to research in general, to Grounded Theory, and to the research questions that I... [was] investigating.

Reflexivity is closely bound up with, though separate from, reflective practice. The question of reflexivity on the part of the researcher is bound up with the historical development of qualitative action research. The issue of insider research has caused many researchers (including me) to have concerns that the very act of being part of the context being explored will change that context and affect the respondents to the detriment of enabling a theory to emerge. Hellawell (2006: 1) writes about reflexivity, referring to Shacklock and Smyth (1998), who give us the notion of “the conscious revelation of the role of the beliefs and values held by researchers in the selection of research methodology for the generation of knowledge”. As Hellawell points out this indicates a “deliberate self-scrutiny in relation to the research process”.

In other words, to be reflexive, the writer/researcher needs to reflect on her/his own potential impact on the phenomena being investigated. In fact, we should use this reflexivity to monitor the research process, and then, as Finlay (2002: 210) writes, the process is made clear and obvious and “personal experience is transformed into public, accountable knowledge”.

It is important, as Finlay points out, to maintain transparency by clear documentation of all interactions (in my case, interviews), and how these led to the next interaction (for me, how the analysis of each interview, along with my literature search, affected the schedule for the next set of interviews).
There remains a concern that the researcher has put her/his own inference on the analysis of the context. In fact, the constant process of re-analysis could even be regarded as muddying the waters rather than clearing them. Perhaps these concerns could be said actually to be the defining nature of late twentieth and early twenty-first century sociological writing. As Finlay writes, we have become very concerned with being self-analytical.

Finlay proposes certain ‘maps’ to negotiate what she terms the ‘swamp’ of reflexive research. The following gives a very brief account of each of these maps.

• Introspection – thinking about how I feel as the researcher and documenting this for the benefit of the reader, so that s/he may determine a fit with other contexts.

• Intersubjectivity – considering how the respondents might feel and discussing this as a means of further analysis of the way that the research has developed.

• Mutual collaboration – actively discussing with the respondents the relationships that have formed as part of the research

• Social critique – here, we might be seeking to describe a particular social phenomenon (for example, how the leadership style of the Head teacher and her/his professional relationships affect the process of change management within a school) –if using this map, it will be important to consider any power imbalance between the researcher and the respondents.

• Discursive deconstruction – here, Finlay warns of the need to be careful not to over analyse. In deconstructing, we may be in danger of not putting back together again. Like Pole’s flower, do the individual parts of the flower come back together to give us the idea of the beauty of a flower?

Finlay, L. (2002) ‘Negotiating the swamp: the opportunity and challenge of reflexivity in research practice’. Qualitative Research. 2, 2: 209-230 Thousand Oaks, CA: SAGE

Hellawell, D. (2006) ‘Inside-out: analysis of the insider-outsider concept as a heuristic device to develop reflexivity in students doing qualitative research’. Teaching in Higher Education. 11, 4: 483 - 494

Just my view on Grounded Theory - please feel free to disagree

What is Grounded Theory?


The principle of generating theory through textual or contextual data is usually associated with the methodological approach first used by Glaser and Strauss (1965) in their work with the terminally ill. They coined the phrase ‘grounded theory’, and this approach is widely used in all disciplines of psychology, as well as other human sciences. ‘Grounding’ is not simply a research technique but is based upon the developments about learning and understanding as well as being related to the debates about the social generation of scientific knowledge.

The development of grounded theory, or GT, came at a time when qualitative research rested mainly on the use of quantitative methods such as questionnaires. There was a preoccupation with testing abstract theories rather than developing theories that related directly and empirically to the context being studied. Glaser and Strauss chose the term ‘grounded theory’ to describe an iterative process of sampling and analysis of qualitative data collected from concrete settings such as interviews, participant observation, and archival research.

Wilhelm Dilthey (1977) argued that it would be a mistake to pursue causal explanation at the expense of establishing understanding (Verstehen or ‘meaning’). The work of Schutz (1962) also uses the idea of Verstehen in terms of interpreting everyday understandings. Blumer (1969) employed an essentially symbolic interactionist perspective, which sees human interaction as a complex system of gesture, symbols, and language that is embedded in the social context. From this perspective, the only way to unravel or make sense of this complexity is to engage in interpretive work within the everyday context in order to construct a theory that attempts to explain the social reality at that time and in that place.

The method of grounded theory research is associated with developing a coding system that is constantly analysed and refined throughout the process with the generation of theory as its goal. This coding system is a key feature but it is important to recognise that the grounded theory approach is not just content analysis or thematic analysis, although both of these are certainly means of enhancing reliability and validity. Grounded theory goes further than being a form of content analysis because it has the goal of generating theory, and to this end it has two other key features: constant comparison and theoretical sampling. Constant comparison relies on continually sifting the data as new categories emerge and theoretical propositions suggest themselves. Theoretical sampling involves the active sampling of new cases as the analysis proceeds. This is very important in terms of defending arguments based on the analysis of a body of data, because it means that new cases and categories can be actively sought to test theories in action. There is thus a dynamic relationship between the collection and the analysis of data, which other methodological stances do not necessarily have. Analysis begins as soon as data is available, rather than waiting for a predefined amount of data to be collected. Essentially, the literature review which precedes and runs in tandem with gathering contextual data can be seen as part of this grounded theory approach since the reader will always be redefining her/his working hypotheses as a result of the work of others and will also be using these emerging categories to shape the new research.

Human nature is such that it is highly likely that the researcher will bring some personal theories to the start of a grounded theory study, even if these are weakly defined at this stage. Thomas and James (2006: 7) criticise the concept of GT, partly because of this issue of personal experience. Mjøset (2005: 5) is also challenging GT as a methodology. However, he pinpoints an important issue: “Grounded Theory [….] does not aim to generalize to a broader population.” The importance of the GT approach is that the situations being explored are too complex to explain using only a priori theories or hypotheses that might have been formulated before the research was undertaken and based upon generalisations from other situations. The only way that a satisfactory fit can be made between theory and actuality is by generating articulated hypotheses in context as they emerge.

Wright Mills (1959: 6) propounds the idea of a ‘sociological imagination’ to describe the link between the socio-economic background of the phenomenon being investigated and the life history of the participants in the research (the respondents and the researcher). The underlying aim of Wright Mills’ concept of the sociological imagination was to contribute to the discussion about the validity of social sciences at a time when the natural sciences were still much more highly thought of.

Anfara and Mertz (2006: xiv) cite Silver (1983) as defining theory as “a unique way of perceiving reality” leading not only to a deeper insight but also to a “fresh and different perception of an aspect of the world”.

The tacit understanding that I have as an insider researcher, coupled with the richness of the data, and alongside my approach to GT/situational analysis gives me just this kind of profound insight and leads to emergent grounded theory. Although Anfara and Mertz present a linear relationship from concepts/events to constructs (“clusters of thoughts”) to propositions (or “expressions of relationships among several constructs”) to theory ( “a ‘set’ of propositions”), the process is probably nested or looped, since the researcher may move backwards and forwards between concepts, constructs and propositions before any kind of substantive theory emerges. Even then, the emerging theory may drive the analysis back to the very beginning and start off a new loop. For Anfara and Mertz (ibid: xvii) a theory is useful if it “tells an enlightening story […] that gives new insights and broadens your understanding”.

Anfara and Mertz (ibid: xxii-xxiii) cite Cresswell (1994: 86), stating that, for him, ethnography or phenomenology involves starting with a framework whereas in grounded theory the data collection and analysis takes place before “using theory”. This is something of an oversimplification of the situation and indicates a clear cut dichotomy on theoretical stances.

Figure 3.1 lists Cresswell’s five ‘research traditions’ of qualitative research, along with his paradigmatic assumptions and ideological perspectives, as referred to by Anfara and Mertz (2006: xxii – xxiii).

click to enlarge

For example, I could take 1, 2, 3, and 4 from figure 3.1, map them on to a, b, and c, and ‘underpin’ them with all of the ideological perspectives listed. This would lead to my use of narrative transcripts with my own ontology/epistemology/methodology clearly determining the constructs and propositions that I develop. My ideological perspective is such a mixture of those stated above that it would be difficult to divorce them from my whole conceptual framework.

Anfara and Mertz do go on (ibid: xxiii, xxiv) to consider the work of Merriam (1998), who indicates the ways in which our framework as researchers is linked to a particular literature base and disciplinary orientation – making the whole process ‘theory laden’. They also (ibid: xxv) refer to the work of Mills (1993: 103, 144) who defined theory as an ‘analytical and interpretive framework’ which is for helping the researcher to make sense of a situation. Mills, as reported by Anfara and Mertz, perceives that there are both implicit and explicit theories “underlying” any investigation – even when a theory is “purported to be emergent”, the study will have been framed by beliefs and initial propositions.

As Anfara and Mertz write (ibid: xvii), it is the “diversity and richness” of these initial analytical frameworks that encourage researchers to “see in new and different ways what seems to be ordinary and familiar”.

Flyvbjerg (2001) takes a different view of the role of the social sciences in being able to generate theory. He asserts that there is little point in trying to generate empirical theories, such as in the natural sciences, since the very matter being subjected to investigation cannot be suitable for that approach. He bases his work on the Aristotelian division of knowledge into episteme (scientific, know why, irrefutable), techne (craft, know how) and phronesis (ethics, values, variable, context dependent). If we are investigating what is basically a value-laden human phenomenon, Flyvbjerg suggests that we should take a phronetic approach to our methodology. In such methodology, he suggests we ask the following questions (2001: 60):

1. Where are we going?

2. Is this desirable?

3. What should be done?

4. Who gains and who loses; by which mechanisms of power?

As he states, this means that there is an implicit investigation of values but he denies that this lays the approach open to criticisms of relativism since (ibid: 132) a “focus on relations of values and power” leads to interest from others than the researcher. These others will evaluate the research themselves so the phronetic researcher must deliberately assimilate the context and learn from it – without ‘going native’.

Henstrand (2006: 16) actually embraces the idea of ‘going native’, stating that “proponents of ethnomethodology” actively encourage the researcher to become involved with the group being studied in order to “understand the contextual meanings and avoid distorting the vision of the world” – here she cites Adler and Adler (1987: 32); “going native is the solution rather than the problem”.

On the topic of the richness of data and how this can lead, if not to a simple theory, then certainly to a clearer understanding of the situation, Flyvbjerg (ibid: 133) says that phronetic researchers must begin by asking “little questions” and, with patience and attention to detail, focus on Geertz’s (1973) “thick description”. Flyvbjerg would probably agree with Wright Mills: the problem with Grand Theories is that they are too general to ever be empirically observable.

Grounded theory in contemporary studies makes use of a rich array of data, building in an assimilation of contextual partiality. This potent combination allows the researcher to use the joint processes of data collection, literature search and content analysis to probe deeply into relationships and power interactions. The emerging themes and substantive hypotheses which are formed can then be used to continually develop the investigative area.


References (sorry cardinal sin, a couple missing Mills, and Adler and Adler)
Anfara, V. A., Jr., and Mertz, N. T. (2006) (eds.) Theoretical Frameworks in Qualitative Research. California, USA: Sage




Blumer, H. (1969) ‘The methodological position of Symbolic Interactionism’. In H. Blumer, Symbolic Interactionism. Englewood Cliffs, New Jersey: Prentice-Hall



Creswell, J. W. (1994) Research design, Thousand Oaks, CA: Sage



Dilthey, W. (1977) Descriptive Psychology and Historical Understanding (R. M. Zaner and K. L. Heiges, Trans.). The Hague: Martinus Nijhoff (Original work published 1894)



Flyvbjerg, B. (2001) Making Social Science Matter. Cambridge, UK: Cambridge University Press,



Geertz, C. (1973) The Interpretation of Cultures. New York: Basicflooks



Glaser, B. G. and Strauss, A. L. (1965) Awareness of Dying. Chicago, Illinois: Aldine



Henstrand, J. L. (2006) ‘Seeking an Understanding of School Culture: Using Theory as a Framework for Observation and Analysis’. In Anfara, V. A., Jr., and Mertz, N. T. (2006) (eds.) Theoretical Frameworks in Qualitative Research. California, USA: Sage



Mjøset, L. (2005) ‘Challenges to Grounded Theory’. Paper for the 37th World congress of the International Institute of Sociology, Stockholm, July 2005

Thomas, G. and James, D. (2006) ‘Reinventing grounded theory: some questions about theory, ground and discovery.’ British Educational Research Journal, 32, 6: 767 – 795, Routledge



Wright Mills, C (1959) The Sociological Imagination. New York: Oxford University Press

literature as data

Long time since I posted here but #phdchat on @twitter has prompted me even though I was not around to join in on this particular occasion the summary http://is.gd/LnU8m8 done by @SavvyOD has really got me thinking.

As I could basically be said to be in the Grounded Theory 'camp' I do regard the whole process of research as potentially yielding data. Thus, one's own bias (properly described and acknowledged) becomes data ( Glaser, B. G. (2002, September). ‘Constructivist Grounded Theory?’  Forum Qualitative Sozialforschung / Forum: Qualitative Social Research. 3, 3 http://www.qualitative-research.net/fqs-texte/3-02/3-02glaser-e.htm   for a much better way of explaining than I can). And I think any researcher (GT or otherwise) will wish to link their literature review to their own empirical data. But how can this be done?

There are lots of different types of software and apps around now that can certainly aid the process ( as @lizith amongst others points out on #phdchat ) but does that help with developing a framework for actually using and linking your highlighted clips and references with other data? I am not sure that the following is really getting me to where I want to be. Comments welcome!

click to enlarge