Learning Outcomes
Understand
- The value of research memos and diaries
- Validating qualitative analyses
Remember
- mixed methods research
Apply
- planning a qualitative data collection
Mixed methods research
Mixed methods research combines clear quantitative components, which focus on the tabulation and numerical analysis of data, with clear qualitative components, which focus on meanings, metaphors, and concepts. For example, if Lazar et al. had also analysed the usage patterns of the devices their participants bought, this would have constituted quantitative data.
For true mixed methods research, the quantitative component needs to be substantial. If you collect basic demographic information (age, gender, education) about your interviewees, then this is a type of quantitative data, but if you only use it to contextualise your interview results, then you are not doing quantitative data analysis.
Likewise, if you collect free text in a survey, and then count the number of times people use certain words, you are doing quantitative data analysis – you are counting and measuring words.
Research Diaries and Memos
There are many aspects of data collection and interviewing that are potentially relevant to how you interacted with your interviewee, but that might not be reflected in your notes from the interview, or in your recording. Therefore, it is important to write everything you remember down after you’ve conducted the interview. Think about aspects such as
- what was the environment? How did it feel?
- how did the interviewee appear to you?
- was there anything that the interviewee did that made you feel bad, strange, happy, sad?
- what was your overall impression of the interview?
- how did you feel before starting the interview? nervous, happy, well prepared?
- were there points during the interview where you didn’t know what to say or do?
All of these things will affect your interpretation of what your interviewee said. There is no right or wrong here. If you write about your own feelings and perceptions, this can help you figure out which interpretations of your data might be due to your own biases, mood, or feelings that day, and how your own behaviour might have affected the data that you get from the interviewee. Writing all of this down also helps other researchers who code the same data.
(By the way, quantitative data can also be affected by context. See for example the phenomenon of white coat hypertension.)
Validating Qualitative Analysis
Part of the answer to this question depends on how you define truth and reality. Within qualitative analysis, researchers often emphasise that social reality is shaped by people’s ideas, values, concepts, and expectations. These are highly subjective. When trying to determine what other people think and value, researchers will inevitably be affected by their own background. Qualitative analysis recognises this and over time, analysts have incorporated several safeguards, such as:
- codes and themes are linked to the data that motivate them, which makes it possible to review and revise analyses
- constant reflection by the researcher about their own potential biases
- constant comparative method: as new codes and concepts emerge, they are compared to existing data, and new data that fits existing codes and concepts is compared to the old data under these codes to see whether the emerging picture is still coherent.
- coding by more than one person. This is not a panacea, because if both people have similar biases, and are not aware of them, then they can arrive at the same misinterpretations. However, if both researchers have been trained in qualitative methodology, this can reduce the risk of bias.
- Using quotes to illustrate key themes. This allows other people to check and independently validate the analyses of the researchers, in particular when there may be cultural misinterpretations. It is also in keeping with the general tenet of documenting your reasoning, so that others can check your findings, which holds across both qualitative and quantitative research
Additional Resources
Open coding (take 1) (take 2) versus axial coding
Overview of mixed methods research with further links.