I'm interested to know the most important points that a researcher should take into consideration to make the interpretation and description of coded qualitative data in the most effective manner.
I think, after you coded, u hv to highlight the main themes of the result, as well as the sub-themes that refers to the main one.
Reflexivity. You must be able to be reflexive as much as you can on your own role in the coding process as researchers, and how your world view shape the coding of data into sub-themes and themes.
After finishing coding you have to further refine your results. By using contrasting and comparative approach and identifying negative cases you look for variations within the themes. Then you have to look between the themes to see how different themes relate to one another and in what kind of relationships they might be.
You also need to be able to document issues of trustworthiness and develop essentially an audit trail. If you have more than one coder, keep track of how you came to the decisions and consensus on your codes and higher order concepts.
Your role as a researcher is very important. Yes being reflexive is the key. Do you agree with the themes generated? You might have other words or phrases in mind as the themes emerge. How you perceive and interpret these themes as the researcher is influenced by various factors.
Before data analysis you must address these topics:
1. Was there instrumental growth or decay? Did the observer coded in a manner which was constant all over the observation sessions? If there were several observers/coders did they behave as if they were a single observer? (i.e., reliability, validity, coding stationarity ...). Was the questionnaire the same and always applied in the same manner? Were the selected questions to ask the same?
2. Were the categories really independent from one another or did they overlapped? You had to address this when you initially defined and selected your units; but while carrying out the observations you might have found that some units overlapped or were interchangeable. This has to be corrected. You might have to split or lump categories, or drop others altogether.
3. Decide of the beginning and end of some units. Some units have a tendency to be more like states, more or less continuous. What criteria are you going to apply to separate a given occurrence from another occurrence of the same unit just following? Is it a continuation of the same unit or rather a transition to another instance of the same unit?
4. Do you have sufficiently observed? Were enough data gathered and for long enough periods? You have to study the power of your gathered data (in the light of your hypotheses). You have to examine the frequencies of your units. This has to be checked while you accumulate data.
5. Often raw observations have to be transformed into data. Data are used to test your research hypotheses; to be relevant to a given hypothesis a datum must correspond to a concept that was incorporated into your hypothesis; the datum is usually of a higher level than the observational units.
For ex. , one hypothesis might state that one group would be more active than another. How do you define active using your observations? How do define this higher level category using more primitive ones in your observations?
I surely forget other important aspects to be considered, but enough for the moment.
- Badges
- Science method
- Similar topics
- Methodology
- Research Methodology
- Qualitative Research