On the Evaluation of Grounded Theory Studies
When considering the evaluation of a grounded theory study, I believe that the first matter that should be taken into consideration is what has come out of the study. Rather than "Rich Thick Description," which is the goal of most contemporary ethnographic approaches (in the Geertzian tradition), a grounded theory study yields substantive theory. A theory may be thought of as an integrated set of hypotheses that has explanatory and predictive power. As such, the resultant theory should be both conceptual (in that it may be abstracted from the context(s)/condition(s) it was discovered in and subjected to different parameters and retain its conceptual integrity) and it should be dense (in that it subsumes multiple indicators of a pattern that has been named conceptually).
This is what a grounded theory studies "comes up with." Considering this, in a grounded theory study it is not acceptable to bring as a lens to the research an existing theory, theoretical framework, coding scheme, pet concept, or "great man" ideology. Conceptual categories must earn their way in as they are discovered in the data. Good concepts have "grab" in that the implications are immediately apparent to colleagues in the field and readily prompts them to "flood out" with their own experienced indicators of the discovered concept.
Now it is important to turn to the methods of the methodology. The researcher of a GT study should have employed the tools: open coding, theoretical coding, theoretical sampling, memoing, constant comparison, theoretical sorting, and theoretical writing. Also, a GT study should yield a core category.
Like other methodologies, a GT study should be systematic, rigorous, appropriate, and fair.
Unlike hypothesis testing approaches, a lengthy literature review should not precede the writing of the theory however, connections to existing theory, research, a particular field and so forth may be constructed in the writing once the theory has been written.
I feel compelled to point out a few warning signs of poor GT studies or write ups or that they that may in fact not be GT studies:
1. There is no theory that comes out of the study
2. There is no mention of core category or core variable
3. GT is cited only in terms of coding
4. There is no mention of memoing (memoing is where the action is in a GT study and is commonly ignored in reifications of GT).
5. There is an existing theoretical framework, theory, or ideology that is brought to the research
6. The study yields vague categories that do not carry a notion of fit, relevance, and grab with general implications. Another way to think of this would be to say that the categories do not "hang together" in a coherent and obvious way. This may be a sign of a lack of conceptualization.
7. A very large number of categories or codes are reported and then collapsed into others and then into others. While this MAY be appropriate in some circumstances, I see this frequently as leading to large vague categories that don't integrate well into theory. A lack of memoing usually accompanies this as does the use of NVivo.
8. There is a large amount of quoting from the data that is used to paint descriptive accounts. This is common in qualitative studies but can also be said of quantitative studies as well.
9. The study is an evaluation of an intervention, innovation, product, approach, or pet concept.
10. The study makes use of logic, philosophical neologisms, or super-think rather than data to prove concepts or make assertions.
When considering the evaluation of a grounded theory study, I believe that the first matter that should be taken into consideration is what has come out of the study. Rather than "Rich Thick Description," which is the goal of most contemporary ethnographic approaches (in the Geertzian tradition), a grounded theory study yields substantive theory. A theory may be thought of as an integrated set of hypotheses that has explanatory and predictive power. As such, the resultant theory should be both conceptual (in that it may be abstracted from the context(s)/condition(s) it was discovered in and subjected to different parameters and retain its conceptual integrity) and it should be dense (in that it subsumes multiple indicators of a pattern that has been named conceptually).
This is what a grounded theory studies "comes up with." Considering this, in a grounded theory study it is not acceptable to bring as a lens to the research an existing theory, theoretical framework, coding scheme, pet concept, or "great man" ideology. Conceptual categories must earn their way in as they are discovered in the data. Good concepts have "grab" in that the implications are immediately apparent to colleagues in the field and readily prompts them to "flood out" with their own experienced indicators of the discovered concept.
Now it is important to turn to the methods of the methodology. The researcher of a GT study should have employed the tools: open coding, theoretical coding, theoretical sampling, memoing, constant comparison, theoretical sorting, and theoretical writing. Also, a GT study should yield a core category.
Like other methodologies, a GT study should be systematic, rigorous, appropriate, and fair.
Unlike hypothesis testing approaches, a lengthy literature review should not precede the writing of the theory however, connections to existing theory, research, a particular field and so forth may be constructed in the writing once the theory has been written.
I feel compelled to point out a few warning signs of poor GT studies or write ups or that they that may in fact not be GT studies:
1. There is no theory that comes out of the study
2. There is no mention of core category or core variable
3. GT is cited only in terms of coding
4. There is no mention of memoing (memoing is where the action is in a GT study and is commonly ignored in reifications of GT).
5. There is an existing theoretical framework, theory, or ideology that is brought to the research
6. The study yields vague categories that do not carry a notion of fit, relevance, and grab with general implications. Another way to think of this would be to say that the categories do not "hang together" in a coherent and obvious way. This may be a sign of a lack of conceptualization.
7. A very large number of categories or codes are reported and then collapsed into others and then into others. While this MAY be appropriate in some circumstances, I see this frequently as leading to large vague categories that don't integrate well into theory. A lack of memoing usually accompanies this as does the use of NVivo.
8. There is a large amount of quoting from the data that is used to paint descriptive accounts. This is common in qualitative studies but can also be said of quantitative studies as well.
9. The study is an evaluation of an intervention, innovation, product, approach, or pet concept.
10. The study makes use of logic, philosophical neologisms, or super-think rather than data to prove concepts or make assertions.
No comments:
Post a Comment