This article aims to discuss the qualitative data analysis process and the importance of having human judgement and using software to aid the process, which does not replace the researcher. This article provides a brief overview of the differences between qualitative and quantitative data analysis, including sample selection and sample size, as well as qualitative data analysis, specifically the thematic analysis method, and the use of software in data analysis.
The Difference
Qualitative data analysis differs from quantitative data analysis; it is a time-consuming and laborious task compared to quantitative data analysis. It is also important to emphasise that in the qualitative approach, we do not intend to generalise the findings, that is, the quantitative approach. Often, students do not understand the amount of data they have to grapple with till they complete the data collection stage. This understanding is vital for the timely completion of the project. In terms of sample selection, it is not a random sample; instead, we often select a purposefully chosen sample using the snowball method.
Size Matters
For anyone new to qualitative research methods, I would like to provide an understanding of the scale and size you are dealing with when conducting a one-hour semi-structured data analysis, which can produce approximately 12 pages of transcription. If you conduct 30 interviews, the amount of data is quite substantial. Remember, if you have to transcribe these manually, that is also a time-consuming task. Sometimes, you can use apps for this, but anonymity and the confidentiality of the data should also be top priorities when selecting apps and outsourcing options. When choosing the sample size, it is a matter of selecting a diverse sample as much as possible, even if it is a purposively selected sample. In other words, ensure the sample includes individuals from diverse age groups, sexes, races, and geographies.
Where To Begin & How to Find Themes.
Often, students say that they do not know where to begin and feel confused, at times overwhelmed. It is essential to understand the structure and how these elements are interconnected.
We intend to achieve the overarching aim of the research by accomplishing the research objectives. Then, this common thread continues into the literature review, which examines the existing body of research in relation to the research objectives. The culmination of the literature review is the development of the research framework, which explains the relationship between the various themes and components discussed throughout the literature review.
Therefore, it is a good starting point for identifying major themes by following one of the approaches: your data analysis journey begins.
- Questionnaire
You can identify the main themes by going through the questionnaire, as each question represents a theme, and each question intends to explore a specific area.
- Themes from the literature review
When we explore the existing literature to create a literature review for the given study, the existing knowledge is structured under sections. So those sections are the main themes.
- Research Framework
The research framework represents different components and identifies main themes based on these components.
- Free method
Create themes as they emerge, and no pre-identified themes.
Once you have identified primary themes through a literature review, questionnaire, or research framework, the next step is to let the data in the transcriptions speak freely; this will enable the researcher to identify new main themes and sub-themes that emerge from the data. That freedom should always be allowed. This process is called coding. Also, you need to revisit after every few interview transcripts analysed, review the structure of the themes, whether they need rearranging, and create journal entries for any insights or patterns you perceive. These journal entries are essential in the later stage of theory building and asking questions from the data.
If you are using software like NVivo, the terminology we use is “main themes,” which are referred to as parent nodes, and “sub-themes,” which are referred to as child nodes. Nodes are where themes are stored or arranged.
How to Use Software for Qualitative Data Analysis
Using software for qualitative data analysis has been a point of academic debate for several years; fortunately, the situation has now changed, and software for qualitative data analysis is widely used. Qualitative researchers are quite protective of their research landscape, and their main argument is that the qualitative approach is more inclined to study human emotions, ideas, and behaviours in detail; hence, introducing machine-led software approaches does not align with the qualitative approach. In addition to that, the following reasons also push researchers to opt for a manual approach
- A deeper level of familiarity with data
Researchers who argue for a manual approach argue that it allows researchers to familiarise themselves with the data on a deeper level, gaining a better understanding of the subtle contextual meaning. The automated coding can not give that familiarity as the researcher does not need to go through the raw data over and over again, resulting in the mechanisation of the process.
- Affordability and scale
If the study is small-scale and the researcher does not have access to the qualitative data analysis software, then manual approach is appropriate.
- Learning the approach
Understanding how the manual approach works is crucial for early-career researchers. It is essential to know the process, which involves using sticky notes, coloured highlighter pens, cutting and transcribing notes into strips, and categorising them under different themes.
However, the main disadvantage of this manual approach is that it is time-consuming to identify and manually arrange pieces of paper into identified categories. It also poses challenges in scalability and maintaining consistency with larger studies. This method involves human bias, as human judgment is always present in any situation.
Embrace Technology with Understanding
Understanding the qualitative researchers’ deep-seated reasons for being reluctant to embrace software-led technology is important when, as a researcher or student, one navigates one’s own approach to the project. The involvement of technology and various software at different stages of the research has been on an increasing trend, which I perceive as technology as an enabler, a deeper engagement with participants and data. The supporting reasons are
- Wider accessibility
Reaching a diverse set of participants has become easy with technology. You can reach participants in different regions geographically at a fraction of the cost using online platforms to conduct interviews. Video interviews are increasingly being used, which enables the grasping of body language cues and establishes credibility and trustworthiness.
- Deeper understanding, quality and trustworthiness of data.
Unlike decades ago, recording interviews has become the norm (of course, you need to obtain consent). You can utilise sophisticated platforms that offer recording capabilities, some of which also provide transcription facilities. This new advancement provides greater accessibility to a much deeper level of diverse data. Researchers need to possess the necessary technology know-how and understand when to intervene in the process to enhance the quality, trustworthiness, and accuracy of the information. Start automatically generated transcription notes that the researcher can check over for accuracy and completeness. This will help the researcher to familiarise with the data; also, save time and money.
- Being Familiar with the data
In the data analysis stage, my approach is to read the entire interview transcripts from start to finish to gain a holistic understanding of the participants’ viewpoints and to become familiar with the data before identifying themes. (If project is large in scale, conducting interviews and data analysis are most probably done by different individuals; familiarity with the data is paramount).
- Identifying themes, revisiting and revisiting
Identifying the major themes using a questionnaire, literature review, or elements of a research framework constructed in relation to your study comes next, and then let the data talk and emerge the sub-themes. Follow the same process with interview transcription of several participants. Revisit the coded data once again to see whether you can identify further missing themes. You can follow this process until you code all the interview transcriptions. Meanwhile, if you have noticed any patterns or associations of ideas, be sure to record them in your journal entries. This is particularly helpful when conducting data interrogation and theory building.
- Context to the code
The most unique feature that modern data analysis software does is that researchers can view the particular theme appear in which context, in other words, the software gives context to the code.
- No auto-coding,
As a researcher, it is difficult to support when researchers entirely depend on auto-coding. But you can select a sample of transcribed interviews and perform the auto-coding as a pilot study. Then you can compare whether you have missed any themes. This way, you are trying to strengthen your approach.
As a final note, as a qualitative researcher operating in an interpretive paradigm, you bring your own worldview to the table. These perspectives and conceptual frameworks determine what and how to code; what questions to ask when interrogating raw data. You cannot hand over making those decisions entirely to software; however, it can be used as a tool to aid the process.