Principally and procedurally, qualitative data analysis techniques are different from quantitative data analysis techniques. The process of collecting qualitative data, which generally focuses on interviews and participatory observation, analyzes the data in the form of textual analysis of the transcripts or unstructured field notes.
Read also: Qualitative Research Methods Unlike quantitative research where the raw data seems ready to be analyzed, qualitative data analysis techniques tend to use data that is not ready to be processed. Therefore, a systematic process in data processing needs to be applied.
The qualitative data analysis techniques that will be presented here use general techniques that have been simplified from the method written by social research expert Alan Bryman. I will briefly review step by step along with examples to make it easier to understand.
However, before we get into the technical area, I need to first state some of the main principles in qualitative data analysis, such as the approach and its basic operations. Understanding, although at a glance, the main principles are important so that the implementation of the technical steps taken later does not deviate from scientific principles. Also read: Scientific Method: Definition & Examples
Approaches in qualitative data analysis
In general, there are at least two approaches in qualitative data analysis, namely induction and grounded analysis. The grounded model is often confused with deductive but it is not always the same. I follow Alan Bryman who distinguishes qualitative analysis into the two approaches because it is quite digestible. Here is a brief explanation of the two:
Induction analysis approach
The analysis with this approach starts from the hypothesis that has been held by the researcher before going to the field. Similar to the actual quantitative approach, but more open to field findings.
After going into the field, the researcher checks whether the data obtained confirms or disproves the hypothesis. If the field data finds cases that refute the hypothesis, the researcher moves to two options: redefine the hypothesis to exclude cases that are deviant (refute the hypothesis) or reformulate the hypothesis.
Redefining the hypothesis and excluding any deviant findings will lead to the completion of the analysis process. Meanwhile, reformulating the hypothesis will lead to examination or even collection of field data again.
For example, the research that will be conducted is on ‘Consumption of Islamic literature for young people in the capital city who decide to migrate’. Researchers compiled research questions to determine what types of Islamic literature were consumed by youth who chose to migrate. The hypothesis developed is that young people in the capital who migrate tend to consume Islamic literature written or produced by groups related to Indonesian Islamic organizations such as NU and Muhammadiyah, for example.
The results of interviews and observations show that it turns out that those who migrate tend to consume translated Islamic books written by Middle Eastern scholars. Thus, the data do not support the hypothesis. The process of induction analysis provides an opportunity for the researcher to redefine the hypothesis or reformulate the hypothesis. Qualitative research is always open to field findings, therefore, the data analysis process almost never stops at one point.
Next, I will review qualitative data analysis techniques with a grounded theory approach. This approach has a longer process and is more commonly applied in many qualitative research.
A grounded theory analysis approach
Qualitative analysis using grounded theory has been used extensively among social researchers. The definitions of what is grounded theory are quite diverse. Here I don’t need to narrate because it takes up too much space. For a glimpse of the understanding that can be used to follow this post, in my opinion, it is enough to see grounded theory as the null hypothesis in the initial research process. For the record, it does not mean that the hypothesis is impossible to develop.
In short, the researcher makes research questions without thinking about the hypothesis in his head first. However, researchers still conduct literature reviews or literature reviews to find out which theories have been applied in research with related topics. The theoretical knowledge from existing studies is used to be developed by applying theoretical sampling.
Read also Grounded Theory: Definition, Examples & Research Methods What is theoretical sampling? According to Glaser fan Strauss (1967) theoretical sampling is
“The process of collecting data to produce a theory in which the researcher collects, codes, and analyzes the data and decides what kind of data he wants to collect and where to get it in order to develop the theory he is developing. This data collection process is controlled by a theory that is being developed either substantively or formally. “
From this definition we understand that theoretical sampling is an ongoing process, not a final result.
After theoretical sampling, the grounded approach takes the researcher to collect field data. The next process after the data is collected is coding. Coding is one of the key steps in applying qualitative data analysis techniques.
The two approaches I have reviewed above are likely to show a difference in qualitative research analysis. As explained earlier, this post will explain in simple terms how qualitative data analysis techniques are applied. I will explain step by step with examples. I will try to simplify the procedure made by Bryman for easy understanding.
Read also Research Approach: Examples & Explanations
Qualitative data analysis techniques
We start with coding. Coding is a basic step in the operationalization of qualitative data analysis. At the coding stage, it means that the (initial) field data has been collected. Even so, it is possible for researchers to go to the field again to collect data again. Thus qualitative research tends to shape the circulation process in its analysis.
Coding
Coding can be done manually or using qualitative data analysis software such as Atlas.ti or Nvivo. The coding process is also called indexing. The researcher coded the text or narrative (data) that was found. For example, an imaginary interview transcript shows the following narrative:
◊ Interviewer: Is there any Islamic book that has changed your way of thinking so that you decided to emigrate?
♦ Informant: I like reading the writings of contemporary Islamic scholars from the West such as John L. Esposito and Tariq Ramadan. But what makes me love reading books is when I was assigned to make a review about the Muslim Brotherhood. I came to know who Hasan Al Banna was and how he was able to establish such a large organization. The only book I saw in two volumes reading it was the book of Hasan Al Banna, an Egyptian intellectual. From that book I came to like Islam, whereas what I did before was full of sins.
Researchers can code the text into “Middle Eastern literature”, for example. Or it could be written with a “statement why you like reading”. Various methods of coding can be used to facilitate systematic analysis efforts.
Also read: Research Interview Guidelines
The research we are doing is still the same as the example above, which is about ‘Consumption of Islamic literature’. The examples used here are only imaginary examples for online learning purposes only.
From the theoretical sampling we have developed, it is found that Islamic literature produced by entities related to NU and Muhammadiah is widely circulated in bookstores and becomes public consumption. However, young people who decide to migrate tend not to consume this literature.
Create a draft
The results of coding the data show that translated Islamic books from foreign authors have inspired many young people who consume them to migrate. Researchers need to check field data again or even collect data again to ascertain whether new data are needed. If the data is saturated, the coding results can be used as a concept. Middle Eastern literature is a concept. Researchers can interpret what is meant by Middle Eastern literature, what is meant by NU literature, what is meant by Muhammadiyah literature, and other literature, according to field data.
Create categories
After the concepts used are clear, the researcher can arrange categories. For example, make a list of which informant’s statements are included in ‘Middle Eastern literature’, ‘local literature’, and so on. Making categories or categorization is not rigid. Researchers can do this by coming up with other terms that have been mentioned by the informants in the interview. For example, ‘Islamic books ngepop’, ‘radical Islamic books’, and so on. The next step, the researcher explores the relationship between each category and ensures that the data (usually in the form of text or narrative) fall into the appropriate categories.
Make a hypothesis
From the categories that have been compiled, the researcher can make a hypothesis. For example, ‘young people in the capital who have migrated prefer to read Middle Eastern literature rather than Islamic literature from NU or Muhammadiyah’. This hypothesis shows that books by local authors do not encourage people to migrate or are only read by young people who do not need to emigrate. Of course, the hypothesis must be tested first.
Obtain analysis results
As previously stated, qualitative data analysis is almost always not linear. Researchers go to the field, return to work, go to the field again and so on to get quality results. Qualitative research has an emphasis on the quality of research results, not quantity. After the hypothesis is tested, the researcher can confirm existing theories, develop theories or create new theories. The results of the analysis are the results of studies that are ready to be tested and presented to the public.
