Analyzing Qualitative Research Data
Interviewers have the choice of taking notes during their interviews or tape-recording them for later analysis. The latter is preferable so that the interviewer can concentrate on listening and responding to the subject during the interview and not be distracted. Note taking also may result in increased interviewer bias as only those responses perceived as being most relevant or interesting at the time may be recorded. Those being interviewed may also wonder why certain points are being recorded and not others (supporting the incorrect notion that there is a way I should respond). Tape recording provides complete data for analysis. Some subjects however are reluctant to be tape-recorded. The ideal tape recorder would be small and unobtrusive. It should provide high quality recording and turn the tape over when needed for longer interviews. A counter function would be very useful when analyzing data.
Transcribing is the procedure for producing a written version of the interview. The researcher then has a full “script” of all that was said during the interview. Transcribing is a very time consuming process taking as much as 4-5 times longer than the interview itself. The researcher may also take notes from the recording or just transcribe portions of the interview. Good transcription also indicates how feelings and meanings that were nonverbally communicated (through tone, hesitation, etc.) should be communicated on paper. Take the phrase,” I hope it works.” It might be said in a variety of ways with different meanings.
“I hope IT works.” (not sure that it will - may need something else.)
“I hope it WORKS.” (hasn’t always – hope it will this time.)
“I hope it works.” (like we all do…)
“I hope it works?” (maybe I do, maybe I don’t)
Listening and noting how things are said will help the researcher detect positive and negative feelings, certainty or uncertainty about a topic, or enthusiasm or reluctance to discuss a topic. These points can be critical in processing the meaning.
Another procedure used extensively in grounded theory research designs is constant comparative analysis. This is where the researcher conducts the first interview, analyses that data and uses findings in structuring the second interview and so on.
Analysis of qualitative data
Analysis of data involves summarizing and communicating it in a manner that summarizes the data collected and presents the results in a way that communicates the most important ideas. Quantitatively, this is accomplished through standardized statistical procedures that compare data to one another and give the reader a picture of the outcome. In qualitative data analysis there is also a desire to communicate the “big picture” and use data to describe a phenomenon and provide understanding regarding what it means. The basic process in quantitative and qualitative starts the same way – giving every item a label or code so that we can recognize similarities and differences between each one.
Content analysis is one common qualitative analysis method. The qualitative researcher most often does not have a system for precoding data like quantitative researchers do. The researcher reads the text over and over and may listen to the tape until the relevant or interesting points emerge. These may be written in the margins or on another sheet. Then the researcher may look at their notes comparing them looking for ideas that occur several times for categories or themes. All possible categories are included at first as they may be reduced or combined later. Major and minor categories may be identified. Links among categories may be revealed. As the “big picture” emerges, data may seem to fit into another category or into two. The researcher works through all transcripts in this way. The notes and categories are then reviewed and major themes may emerge. The researcher must be careful to constantly check their ideas back with their original notes and data. The researcher continues with this process until they are sure the findings accurately represent the data.
Computerized data analysis
Software packages exist that make the analysis of qualitative data easier. A variety of packages with an increasing range of features have been developed.
All of the packages assist with the process of categorization and relating of categories. They help the researcher with organization and save time, however the identification, assignment and rechecking of categories still requires a lot of skill and input from the researcher. In other words the actual analysis and interpretation of data are still done by the researcher and not the software program. The most popular of these programs is called NUD*IST.
Now complete assignment 4 in your exercises!