Dear Professor Gibbs, Thank you for this series of videos, i personally found them extremely helpful in explaining such a complex technique in an understandable way.
I am just starting work on my master's degree, and I have been trying to grasp grounded theory. These videos are very helpful. I will be recommending them to my colleagues. Thank you.
Ideally you should find some new themes that are not simply reflections of your initial 50 questions. You may find some codes or themes that apply only to one ethnic group but you will probably also find some that apply to all the interviewees. Your original codes based on your questions and the additional codes/themes you have created can then all be developed through axial coding to selective coding, by combining and redefining them and trying to pull out the core categories cont.
These core categories will probably apply across all cases. In general you do not want a separate coding scheme for each interview. Most codes should apply across all cases/interviews. However, some may only apply to some subgroups. There may be differences between ethnic groups, but there may also be other kinds of difference that cuts across ethnic groups (e.g. age, gender or educational differences). Good luck with your work.
I think your question is actually about theoretical sampling, although this can arise as a result of the constant comparison of codes with codes and codes with data. The development of your theory in this way might suggest you need to interview new types of respondent. You can't predict this ahead of your analysis. On the other hand you may have a hypothesis at the start that Egyptian students in Belgium and in your home country may have different experiences and relationships to events... cont.
in Egypt. And you'd need to have good reason for thinking that. Then you could plan your sample to reflect it. But I can't think of any reason for that unless, perhaps, your home country is Tunisia, where such a political change has already been experienced and where that might have affected the Egyptian students studying there in ways the students in Belgium were not.
Dear Prof, I am struggling with the coding process. Do I need to complete all the coding steps/ phases every time data is collected or 'open coding' only until all the data collection is done? I mean, say I have 2 interview transcripts at my hand now. With this, do I need to do open coding only and then collect some more data; and then do open coding again for new interview (data)? Or, with the two transcripts (for example) I have to do open, axial, and selective coding before I move onto further data collection? Then repeat the process? Thanks so much for your video presentation.
I have a question on constant comparison and theoretical saturation: I am doing my masterthesis on social media use during the Egyptian revolution trying to capture people's views on the alleged influence social media like Facebook have had on it erupting right there and then. Since I am not in a position to go over there to interview, the only people I interview are Egyptians living in my home country and f.e. Erasmus students. ( read further in my next post)
Does constant comparison mean I should start looking for a subcategory of Egyptians for whom the theory won't hold or slightly alters? I was thinking of making a distinction between erasmus students that have experienced it live or that have been using it over there AND Egyptians that were in my home country ( not Egypt) at the time and may therfore hold other views on the subject. Thank you for helping me out
Dear Professor Gibbs, I'm wondering there are many types of qualitative data coding such as content analysis, constant comparative method, narrative analysis, etc How do we know which one of these is the right one for our research. What considerations should be taken into account? And, Are there any books that explain this in detail ? Thank you
Orana, To some extent, the different analytic approaches used reflect different research questions and even different epistemological positions. A good recent collection of essays that examines a range of analytic strategies and data types is Flick, Uwe (ed) (2013) The SAGE Handbook of Qualitative Data Analysis. Los Angeles: Sage.
Gibbs, Graham R. (2012) Grounded theory, coding and computerassisted analysis. In: Understanding Research for Social Policy and Social Work: Themes, Methods and Approaches. Understanding Welfare: Social Issues, Policy and Practice series . Policy Press, Bristol, UK, pp. 337-343. ISBN 9781847428158
Dear Prof Gibbs, I am a masters level student planning my dissertation and would like you to kindly clarify a few things. I aim to perform thematic analysis using an inductive approach but at masters level I am not intending to generate any new theories. Are the terms "open coding" "closed coding" "saturation" etc used exclusively for Grounded theory? Surely it is not expected that everyone who uses an inductive approach has to perform grounded theory research. What approach would you advise students in this category use to describe their research? And what is the best way to demonstrate methodological rigor if study design does not meet such prescriptive theoretical approaches? Thank you.
It sounds like you have undertaken semi-structured interviews based on your 50 questions. Although grounded theory suggests an inductive approach to coding, it makes more sense for you to start with codes that capture the issues in your 50 questions. This means you start with a coding scheme that cuts across all your 74 cases (I assume each interview is a separate case). Next you should develop the coding scheme inductively, based on a close reading of the transcripts. Cont.
Dear Sir, I have a question. I did 74 interviews under the same topic (each interview had 50 questions) . 45 for an ethnic group, and the rest for another. I want to analyze it with GT. I'm wondering if I should do it 45 times( and then 29 times) separately, one by one, from open coding --> axial coding--> selective coding, or should I gather the answers from the same questions, and then code them? if I did it separately, would I found different core categories within interviews
Dr. Gibbs, you're much more than amazing. This video is surely not for only us but for generations
Dear Professor Gibbs,
Thank you for this series of videos, i personally found them extremely helpful in explaining such a complex technique in an understandable way.
I am just starting work on my master's degree, and I have been trying to grasp grounded theory. These videos are very helpful. I will be recommending them to my colleagues. Thank you.
These videos are awesome, the best resource of qualitative analysis I've found. Thank you, Dr. Gibbs!
Ideally you should find some new themes that are not simply reflections of your initial 50 questions. You may find some codes or themes that apply only to one ethnic group but you will probably also find some that apply to all the interviewees. Your original codes based on your questions and the additional codes/themes you have created can then all be developed through axial coding to selective coding, by combining and redefining them and trying to pull out the core categories cont.
Thank you for putting this here. This has helped me complete my paper tonight, April 30,2021. 11:35pm,CST
Dear Sir,
Thank you SO MUCH for these videos and the wonderful lecture! It is indeed very informative and helpful!
These core categories will probably apply across all cases. In general you do not want a separate coding scheme for each interview. Most codes should apply across all cases/interviews. However, some may only apply to some subgroups. There may be differences between ethnic groups, but there may also be other kinds of difference that cuts across ethnic groups (e.g. age, gender or educational differences). Good luck with your work.
Thank you so much, Professor Gibbs. Your replies help me a lot. Thank you!
I think your question is actually about theoretical sampling, although this can arise as a result of the constant comparison of codes with codes and codes with data. The development of your theory in this way might suggest you need to interview new types of respondent. You can't predict this ahead of your analysis. On the other hand you may have a hypothesis at the start that Egyptian students in Belgium and in your home country may have different experiences and relationships to events... cont.
in Egypt. And you'd need to have good reason for thinking that. Then you could plan your sample to reflect it. But I can't think of any reason for that unless, perhaps, your home country is Tunisia, where such a political change has already been experienced and where that might have affected the Egyptian students studying there in ways the students in Belgium were not.
Very helpful. Thank you.
Dear Prof,
I am struggling with the coding process. Do I need to complete all the coding steps/ phases every time data is collected or 'open coding' only until all the data collection is done? I mean, say I have 2 interview transcripts at my hand now. With this, do I need to do open coding only and then collect some more data; and then do open coding again for new interview (data)? Or, with the two transcripts (for example) I have to do open, axial, and selective coding before I move onto further data collection? Then repeat the process?
Thanks so much for your video presentation.
I have a question on constant comparison and theoretical saturation: I am doing my masterthesis on social media use during the Egyptian revolution trying to capture people's views on the alleged influence social media like Facebook have had on it erupting right there and then. Since I am not in a position to go over there to interview, the only people I interview are Egyptians living in my home country and f.e. Erasmus students. ( read further in my next post)
Does constant comparison mean I should start looking for a subcategory of Egyptians for whom the theory won't hold or slightly alters? I was thinking of making a distinction between erasmus students that have experienced it live or that have been using it over there AND Egyptians that were in my home country ( not Egypt) at the time and may therfore hold other views on the subject. Thank you for helping me out
Dear Professor Gibbs, I'm wondering there are many types of qualitative data coding such as content analysis, constant comparative method, narrative analysis, etc How do we know which one of these is the right one for our research. What considerations should be taken into account? And, Are there any books that explain this in detail ? Thank you
Orana,
To some extent, the different analytic approaches used reflect different research questions and even different epistemological positions. A good recent collection of essays that examines a range of analytic strategies and data types is Flick, Uwe (ed) (2013) The SAGE Handbook of Qualitative Data Analysis. Los Angeles: Sage.
Thank you so much for taking your time answering my questions :)
how would you like us to cite you?
Gibbs, Graham R. (2012) Grounded theory, coding and computerassisted analysis. In:
Understanding Research for Social Policy and Social Work: Themes, Methods and Approaches. Understanding Welfare: Social Issues, Policy and Practice series . Policy Press, Bristol, UK, pp. 337-343. ISBN 9781847428158
Dear Prof Gibbs,
I am a masters level student planning my dissertation and would like you to kindly clarify a few things. I aim to perform thematic analysis using an inductive approach but at masters level I am not intending to generate any new theories. Are the terms "open coding" "closed coding" "saturation" etc used exclusively for Grounded theory? Surely it is not expected that everyone who uses an inductive approach has to perform grounded theory research. What approach would you advise students in this category use to describe their research? And what is the best way to demonstrate methodological rigor if study design does not meet such prescriptive theoretical approaches? Thank you.
Thank you soooo much! great resource
within interviews from an ethnic group?
Thank you!
Your lectures are really informative!
It sounds like you have undertaken semi-structured interviews based on your 50 questions. Although grounded theory suggests an inductive approach to coding, it makes more sense for you to start with codes that capture the issues in your 50 questions. This means you start with a coding scheme that cuts across all your 74 cases (I assume each interview is a separate case). Next you should develop the coding scheme inductively, based on a close reading of the transcripts. Cont.
Dear Sir,
I have a question.
I did 74 interviews under the same topic (each interview had 50 questions) . 45 for an ethnic group, and the rest for another.
I want to analyze it with GT. I'm wondering if I should do it 45 times( and then 29 times) separately, one by one, from open coding --> axial coding--> selective coding, or should I gather the answers from the same questions, and then code them?
if I did it separately, would I found different core categories within interviews
This is good. But it is also hard!