Theory in qualitative educational research

When I started to learn "grounded theory" methodology. I asked myself "What is a theory?" as "theory" for me was a big thing and I felt it's extremely tough to create a theory... I digged the answer... It's always good to make it clear what your understanding about a term or a concept.

Theory in qualitative educational research provided by the British Educational Research Association (bera) raises some issues relating to theory as a term, a concept and an agent in educational research.

Question about thematic analysis and grounded theory

If you have a similar question, have a look the ideas about grounded theory and thematic determination on the Methodspace.

Situational analysis and grounded theory

This website is about "Situational Analysis", where the author, Prof. Adele E. Clarke, has answered some critical questions about grounded theory. 

Statistics in grounded theory research

People asked questions like "how many interviews should I have?", "how do I know it's saturated?" There are some tips I got that may help you to check your research. They are not rigid rules, but for your references only.

• The literature review demonstrated that saturation normally occurs between 10 and 30interviews. Although saturation might occur after the tenth interview, it is good practiceto test the level of saturation by conducting a few more interviews. ... Thus, it would be wise to anticipate 30 interviews in order to facilitate pattern, category, and dimension growth and saturation. It is only through the quality of the data that meaningful and valid results are developed, so it is essential that the researcher ensure that saturation has occurred (Thomson, 2011, p.50).
• Saturation is achieved when one has conducted between 25 and 30 interviews (Suddaby, 2006).
• 10-15 minutes to 40 minutes interview will reach theoretical saturation (Wuest, 2007).
• Do not put red and green slices next to each other, because 5% of population cannot distinguish red from green. (Fink, 2003, p.9).
• A 40 minutes interview may take you 2 days to transcribe. One hour interview will be transcribed into about 15 pages text.

Bibliography:

• Coyne,  I.T. (1997). "Sampling in qualitative research. Purposeful and theoretical sampling; merging or clear boundaries?". Journal of Advanced Nursing. 26 (3), 623-630.
• Fink, A. (2003). The survey kit, vol. 10: How to report on surveys, 2nd ed. London: Sage.
• Morse, J. M. (1994). Designing funded qualitative research. In: Denzin, N.K. & Lincoln, Y.S. (eds.), Handbook of qualitative research, 2nd ed., pp.220-35. Thousand Oaks, CA: Sage.
• Mason, J. M. (2010). "Sample Size and Saturation in PhD Studies Using Qualitative Interviews". Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 11 (3), Art. 8.
• Suddaby, R. (2006). "From the editors: What Grounded Theory is Not". Academy of Management Journal, 49 (4), 633–642.
• Thomson, S. B. (2011). "Sample Size and Grounded Theory". JOAAG, 5 (1). 45-52.
• Wuest, J. (2007). "Grounded Theory: the method". In: Munhall, P.L. (ed.), Nursing research: a qualitative perspective, pp. 239-272. Sudbury, Mass. ; London: Jones and Bartlett.
• A discussion about "Grounded theory sample size" on the Methodspace - home of the Research Methods community.

Grounded theory design - Self-Assessment

This website provides additional information for users of Educational Research: Planning, Conducting, and Evaluating Quantitative and Qualitative Research, Second Edition, by John W. Creswell. The Chapter 14 is about Grounded Theory Design. It may be limited to access, but the Self-Assessment section is accessible and may be useful for researchers.

A "good" code

A code may be changed later over the process of your research. However, it doesn't mean you can name it without pondering it. With same data, different people may identify different labels, but some are "good" and some are "bad". Please bear in mind what you need to consider when you are coding.

Glaser (1998) and Locke (2001: 69) suggested to ask these questions when you do coding:

• What is happening?
• What is basic problem faced by the actors here?
• What category or what aspect of a category does this incident suggest?
• What does this incident suggest this is a theory of?

Again, 5 elements of a "good" code suggested by Bryne (2001):

• a label
• a definition of what the theme concerns
• a description of how to know when the theme occurs
• a description of nay qualifications or exclusions to the identification of the theme
• examples, both positive and negative

Bryne, M. (2001). "Data analysis strategies for qualitative research-Research Corner", AORN Journal. Dec, 2001.
Locke, K.D. (2001). Grounded theory in management research. London: Sage.

Doing Qualitative Research Using Your Computer: A Practical Guide - Chapter 9

This is the companion website to <<Doing Qualitative Research Using Your Computer: A Practical Guide>> by Chris Hahn (Sage Publications).

The book uses straight-forward, everyday language to walk readers through the research process, drawing on a wide range of examples to demonstrate how easy it is to utilize software that you might already own. The author used a diagram to describe the qualitative research coding process. However, Chapter 9 - Level 3 and Level 4 (Theoretical Concept) Coding, which a grounded theory researcher more concerned, is unavailable on the website.

Questions about coding in GT (2)

Q:
In Grounded theory, I am confused of which coding process shall I follow.

A:
I suggest you to have a clear idea of what different researchers advocated. You need to know what to code firstly and then know how to find the relationships between codes and develop your core category and build your theory. 

The following comparison may help as a start.

A coding process by Glaser and Strauss is different:

Glaser & Strauss (1967)	Glaser (1978)		Strauss & Corbin (1998)
Open coding	Open coding	Substantive coding	Open coding
	Selective coding
Theoretical coding	Theoretical coding		Axial coding
			Selective coding

Selective criteria for core category:

Glaser (11 Criteria) (1978)	Strauss & Corbin (6 Criteria) (1998)
central reoccur frequently more time to saturate connections not be-forced clear & grabbing implication for formal theory carry through completely variable is also a dimension of the problem prevent to other sources of establishing a core see the core category in all relations it can be a kind of theoretical code	central appear frequently no forcing of data sufficiently abstract grows in depth and explanatory power is able to explain variation

Glaser (1978) indicated foundational 6Cs as coding family.

• Cause 
• Consequence
• Condition
• Context
• Covariance
• Contingency

If Glaser's approach is difficult to understand, have a look Spradley's (1979: 111) 9 semantic relationships. It may offer you a start to code your data, and combining with Grounded theory coding process, you will then build a map of your code relationships. Furthermore, will have categories and core category.

• X is a kind of Y (strict inclusion)
• X is a place in Y (spatial)
• X is a result of Y (cause-effect)
• X is a reason for doing Y (rationale)
• X is a place for doing Y (location for action)
• X is used for Y (function)
• X is a way to do Y (means-end)
• X is a step in Y (sequence)
• X is a characteristic of Y (attribution)

(Spradley, J. A. (1979). The ethnographic interview. New York: Holt, Rinehart & Winston.) 

Question about grounded theory, Ethnography and Phenomenology

Q:
I just started my research. I'm confused the use of grounded theory, Ethnography and Phenomenology...

A:
You may have a clue by reading my thesis, chapter 3.2.

It may help if you read this book and think what your research stands on? 
Creswell (2003). Research Design: Qualitative, Quantitative, and Mixed Methods Approaches, 2nd ed. London: Sage.

For example, ask yourself what your research process probably is? (read P.6 Alternative knowledge positions)
ask yourself what strategy of inquiry you (read P.13 Alternative Strategies of inquiry).

Other resources:

• Franchuk, J. E. (2004). Phenomenology vs. grounded theory: The appropriate Approach to use to study, the perspectives of instructional librarians on their roleIn information literacy education for undergraduates.
• Goulding, C. (2005). "Grounded theory, ethnography and phenomenology: a comparative analysis of three qualitative strategies for marketing research". European Journal of Marketing, 39 (3/4), 294-308. 
• Starks, H. &Trinidad, S.B. (2007). "Choose Your Method: A Comparison of Phenomenology, Discourse Analysis, and Grounded Theory". Qualitative Health Research. 17(10). 1372-1380.
• A teaching martial by Brian J. Brown. "Qualitative methodologies: ethnography, phenomenology, grounded theory and more". 

Question about symbolic interactionism and grounded theory

Q:
What is symbolic interactionism in grounded theory? 


A:
Grounded Theory is informed by the pragmatism and has the ideas of the symbolic interactionist. I quote Parker & Roffey (1997, pp.216-217) and Denzin (1989, p.5) statements here to let you see the connections between Symbolic interactionism and Grounded theory.


Similarities:

• Interacting individuals produce and define their own definitions of a situation;
• People can engage in self-reflexive behaviour (i.e. assessing the contextual meaning of their own actions and reactions), and
• humans interact with each other in negotiating a poisition in relation to each other.

Differences:

• Interactionists regard (observation of) human interaction as their basic source of date
• Grounded thoery includes data sources such as interviews, written reports and documents that relate to the research phenomenon.

Questions about coding in GT (1)

Q:
What is really code? I mean only events or objects or both. When I read it seem to code only codes. After coding, you generate a category and take some code as property of category. Can I say a sub category is a property of the main category?

A:

Yes, a code is just a label. However, a good code defines of what the theme concerns, describes of the theme occures and examples both positive and negative. As codes are developed, it is useful to write memos known as code notes that discuss the codes. When you try to explain the code and why you named it, you will have the meaning of it rather than just a piece of code!

Which approach you follow: Glaser's or Strauss&Corbin's or others? With the coding method developed by different authors' approach, you may follow different coding process as each of them has its own ceriteria for coding and categerising. 

In some cases, a sub category may be developed to be a property of a core category. But it's not all the case. A code can belong to more than one concept. Similarly, at the next analysis level, a concept may contribute to and belong to several categories.  

I try to recall how I understood the codes, categories, properties, process, dimensions... Many notes in my handbooks in different dates. This may be an example of how I started (a memo on 09/May/2007). 

After read an article by Carlson and McCaslin (2003), I jotted down my understanding about their research and drew a diagram for myself.

Carlson, N.M. & McCaslin, M. (2003). "Meta-Inquiry: an approach to interview success". The Qualitative Report, 8 (4), 549-569.

Questions about my thesis

Q:

On Appendix 15, you identify "online identity" as category and "traceability of identity", "anonymity" and "balance between online identity and real identity" as properties. 

When I go on page 287, you put "perception" as a property of "online identity" and the dimension - "true self". I was expecting the previous property to be repeated here.

For example for How - you mention "anonymity" as category.

A:

On p.287, there are some examples of open coding at the Anchoring/Centring stages (which is a way of explaining how I conducted Grounded Theory as an approach). They are stage results. On pp.279-285, it gives the full list of codes that I had at the end and how I put them into categories/dimensions/properties that Grounded Theory requires (it also meams the categories were considered saturated). They are final codes.

If you read Chapter 3 carefully (particularly 3.3.3), you will see that a code related to a piece of data may be changed many times before it's saturated. With the process of data collection and analysis, I needed to add analysis into my interpretation of a code that I defined at the begining and I needed to add analysis explanation when I did each change (these are memos). For example, I found "Online identity" is emerged in my data and I defined it as a cateogry. If it is a category, it may have properties/dimensions/subcategories. Then I compared my data and codes and see which can be justified as properties/dimensions/subcategories or if they cannot, I categoried them as a single code for later data comparasion. "Anonymity" was categorised as a "How" subcategory of "Online identity". But later I realised that "How"/"What"/"When" suggested by Strauss and Corbin is used to help clarify the relationship between codes in detail. I then put "Anonymity" as a conceptual level code which became a property on p.279. Similarly, some other codes on p.287. I might either drop/rename, combine them into another code at a conceptual level, or left them as an open code at a textual level.

Questions about using GT in mixed methods research

A:
What is your opinion about mix methods, such as following case study method procedure for research process and use GT for data analyis because to me data analysis in GT is more systematic compare to other methods?

Q:
If applying GT coding method (I suppose it's Strauss & Corbin's approach) helps you clarify or understand quantitative data, it's fine to use mixed methods in your research study and state that you use GT coding method to analysise data in your thesis. However, be careful that mixed methods research is much more than simply combining a qualitative method and a quantitative method. To some extent, it is not a real GT study because a GT study aims to generate a theory and mostly it's qualitative. Meanwhile, you need to reckon if it's good to title your research study as a GT study.

Questions that relates to my article

In your article, you stated that GT is applicable mainly in the following cases:

• The researcher seeks to create a theory about issues of importance in people's lives and specifically focuses on human interaction or aims to explore new territory (Denscombe, 2003).
• The study of new socio-technical phenomena (Ferna'ndez, 2004).
• The area of interest is a new developing one and does not have a long, firm and empirically based literature yet (Goulding, 1999).

Q:
With all points you have mentioned above, I doubted that how can we make sure that the area that we are going to study is new.

A:

It's not surprising. For me, this kind of questions went along with me till I submitted thesis. It is a part of the research. You will know it when you've done thoroughly literature review, discussed with your supervisors, colleagues, read literature widely about your area. Someone who has been told to undertake a research project by their supervisors/funders may know what is new based on their supervisors'/funders'/research groups knowledge; someone who wants to do a further research study about their interested topic may know what's new because they keep their knowledge updated.

Q:

In my case, I happen to see a lot of research in this topic. Could you please give more explanation about the points that given above with some example. I have read those articles but still could not understand.

A:

In my article, the above points are examples. It does not mean that a GT research study needs to have all the requirements listed there. Also, I didn't suggest more new researcher to apply the GT. It has to fit your research purpose. If there are already many studies in your area, you need to think about why you want to use the GT; ask yourself: Are there alternative methodology and why other approaches are not suitable for your study?

For example, Denscombe's idea: it emphasises the purpose of a research study. The research about dying by Glaser and Strauss (1967) is a typical example on this point. The researcher want to explore some phenomenon (often in social science) that has not been fully explained by existent theories in that area. They intend to interpret the phenomenon by an emerged theory based on what they found in their research. It's definitely a qualitative research study as data is difficult to be measured and quantificat.

Fernandez's idea is clear, for instance, recently there was a piece of news said that women are more likely to date with a man who has an iphone (iphone was released on June 2007). Researchers who explore the iphone related phenomenon is doing a study of new socio-techical phenomena. However, it does not mean they need to use the GT approach.

Goulding's idea can been seen in my research as I was doing research about blogging by higher education students. The phenomenon of using blogging in education was new in 2005 and there was a gap in the literature that my research can contribute to.