In this article we look at Grounded theory and you may find it useful to reflect on its relationship to the NLP Modelling Process as you read. Grounded theory is found within a ‘progressive’ qualitative research framework which suggests there is a relationship between the researcher, the participants and the context allowing the researcher to be part of the research setting.  This acknowledges that we have some effect on what is being studied, especially with regard to encouraging participants to share their vulnerabilities. It is different to the more traditional ‘naturalist’ approach which sees reality as fairly straightforward, observable and verifiable by fact, similar to a positivist approach. Like NLP Modelling it suggests that people are individual products of social interaction, and that personality, beliefs, values, preconceptions, drivers etc may be the results of this social interaction. A completely positivist approach would assume they exist in the same way within all people which is at odds with the NLP pre-suppositions.

There are several considerations in the design, and in social science research a lack of quantifiability is not now considered to be a problem although there is an opposing belief,that knowledge can be gained by observations made as though the investigator is outside the study with no preconceived notion of what is being observed.  I dispute this, as humans cannot be objective and a researcher is likely to hold some expectation, belief or values when conducting research and the participant may pick this up. The result is that people can and do construct their own maps of the world as NLPers are aware.

Grounded theory does not aim to provide any final solution (Charmaz,2003)  although, it is expected that as a result of the study, recommendations for developing practice can be discovered.  The aim is not to test any specific hypothesis although some hypotheses could emerge during coding and could lead to further testing and verification.

Grounded theory can be traced from the 1940s to Glaser and Strauss (1967) who came up with the systemic analysis of large amounts of qualitative data. Use of terminology such as ‘coding’ and ‘comparison groups’ has left their work open to criticism of neo-positivism (Ashworth,1996) and this could be justified if strict rules are followed with no flexibility.

The key process in using grounded theory is the coding; breaking data down into component parts which are given names. Data are coded as they emerge, being a main difference to quantitative work which requires the data to fit into preconceived codes. We are looking for continuous and descriptive evidence from which to draw patterns and understand what the aggregated experiences are telling us, rather than asking the individual to interpret their own experience.  This is helped by one of the strengths of grounded theory in that it gives the freedom to consider other influences which affect behaviour e.g. metaprogramme styles.  It is important to be able to incorporate these influences, of which participants may be unaware, while not falling into the ‘researcher as expert’ role.

This is a methodology which is participatory and collaborative, recognises ‘difference’, ‘perceived truth’ and ‘reality’, includes evolution rather than static accounts and provides a partial account of the current state of the context.  It allows us to compare data as results emerge, in order to formulate emerging concepts, whilst providing rigour through procedures, and gives a sense of comfort in the findings.  It can be difficult to know when to stop because the context is moving forward through a period of exceptionally rapid change.

How do we do it? The core of Grounded Theory is the Coding. Strauss and Corbin (1990) devised a system of coding practice to reduce transcripts into meaningful chunks or categories and bring together patterns. There are three stages, each progressively interrogating the data for more specificity as we would do with the Meta Model;

  1. Open codes; highlighting key passages from the transcripts, assigning passages to concepts, working through all the transcripts collecting quotes to saturate the concepts and then chunking down into categories by refining the concepts and deleting and amalgamating some.
  2. Axial codes; making connections between the categories and linking them to each other, to sources and to consequences. These are the ‘real world’ phenomena.
  3. Selective coding; which selects the central or core categories from the most common and revealing codes, relates and compares them to other categories and provides the backbone of the storyline from which the results are derived or in modelling terms ‘the difference that makes the difference’.

Quotes taken from interview transcriptions provide evidence within individual open codes. The quotes can vary in length from a word to a paragraph. Some quotes could be used across a range of codes where multiple possible concepts (Strauss & Corbin1998) are found to exist. Considering these multiple quotes is important, as the quotes could have various characteristics which impact on the open codes given so we need to go back and review the context in the transcript and put the quote against the code it most reflected. We are continually checking whether we are discovering anything new or different when making categorisation decisions from the labels given by participants and by ourselves, checking whether these labels produced any differences to the meaning.

Eventually codes and content became repetitive, adding no new information and at that point we reach ‘saturation’ and stop. The outputs are best presented diagrammatically, however the diagrams produced by coding software such as NVIVO can be too unwieldy and cumbersome for inclusion in our reports.

Although criticised for reducing the data to a level of variables very like the quantitative methods it is supposedly a reaction against, the process of repeatedly returning to the data ensures that the results are firmly grounded in the participants’ experiences.  This does help keep faith with participants’ stories while checking whether their comments are individual, repeated across the sample group or are only related to one situation which is not seen to be significant.

The relationship of the interviewer and interviewee is a distinguishing feature of grounded theory. The grounded theory approach is a useful, general and flexible strategy for analysing data in complex areas. Bulmer (1979) cited in Bryman(2004) has questioned whether it is possible for the researcher to suspend their awareness of relevant theories and concepts until a quite late stage in the process of analysis. I would suggest that the sensitive researcher will be aware of emergent themes early in the data analysis and this is influenced by personal filters from beliefs, values, curiosity and previous experience. However, the professionalism of the researcher will enable him/her to suspend judgement and approach the study with an open mind, although the sensitivity can provide a focus and enable a building on the work of others.

In conclusion, the grounded theory approach means we are not aware of some of the concepts until quite late in the data analysis and in a study of this type, we need to be conscious of playing into the hands of critics of NLP and the dangers of reliance on unproven anecdotal evidence. This is done by using multiple codes and comparisons to reduce the effects of random error and to ensure that both the accuracy and the precision of the data are addressed. The high intra and inter-participant data convergence suggested that the validity (Table;1.1) would be high (Easterby-Smith, Thorpe and Lowe,2004).

Table 1.1; Introduction to Perspectives on validity, reliability and generalisability, adapted from Easterby-Smith, Thorpe and Lowe (2004) by Vanson (2009).

 

Positivist Phenomenologist
Validity Do the measures correspond closely to reality? Does the study clearly gain access to the experiences of those in the research setting?
Reliability Will the measures yield the same results on other occasions? Is there transparency in how sense is made from the raw data?
Generalisability To what extent does the study confirm or contradict existing findings in the same field? Do the concepts and constructs derived from this study have any relevance to other settings?

 

To sum up, there are competing views regarding grounded theory (Charmaz,2000) which can be argued to be objectivist in that it aims to uncover a reality that is external to social factors  (the world exists despite the human being), or constructionist, where social reality is not independent of human action.  Whilst Glaser, Strauss and Corbin may neglect the role of the researcher and focus on the generation of data, this does not imply that they are indifferent to the constructivist approach which can be invaluable when collating the data and taking the NLP modelling concept to the next level.

Dr Sally Vanson; DBA, MSc, ChCIPD, PCC , Certified Master Trainer of NLP (sally@theperformancesolution.com)

Main References;

Berger, P. and Luckmann, T. (2002). The Social Construction of Reality. Contemporary sociological theory: 42.

Bryman, A. (2004). Social Research Methods. New York, Oxford University Press.

Charmaz, K.(1995). Grounded theory. In J.Smith, R.Harre & L.Langenhove (Eds.) Rethinking methods in psychology (pp.27-65). London:Sage.

Charmaz, K. and Smith, J. (2003). Grounded theory. In N. K. Denzin & Y. S. Lincoln (Eds.), Strategies of Qualitative Inquiry. (2) pp. 249-291).

Corbin,J. (1998). Alternative interpretations:valid or not? Theory & Psychology, 8(1), 121-128.

Easterby-Smith, Thorpe et al. (1991). Management Research, an introduction: London:Sage

Elliot, J. (2007). Using Narrative in Social Research; Qualitative and Quantative Approaches. London, Sage Publications.

Glaser, B. & Strauss., A. (1967). The discovery of grounded theory; strategies for qualitative research. New York:Aldine.

Glaser,B. (1978). Theoretical sensitivity, advances in the methodology of grounded theory, Mill Valley: Sociology Press.

Glaser,B (1992). Basics of grounded theory analysis; emergence vs. forcing, Mill Valley: Sociology Press.

Glaser, B. (1994). More grounded theory methodology: A reader, Mill Valley: Sociology Press.

Holliday, A. (2002). Doing and writing qualitative research. London: Sage Publications.

Strauss, A. C., J. (1990). Basics of Qualitative Research. California: Sage.

(Visited 1,284 times, 1 visits today)

Leave a Comment