Dr. Philip Adu is a Methodology Expert at the National Center for Academic & Dissertation Excellence (NCADE) in The Chicago School of Professional Psychology (TCSPP). In this post, he takes us through the stages of qualitative coding and, along the way, outlines the features he finds most useful.
Think about analyzing interview transcripts: capturing significant information from the data and putting them into ‘containers’, called codes (
What about labeling each code created, providing a brief description about each of the codes, and documenting your reflections?
You then sort the codes into ‘big containers’ called parent codes based on their content similarities and unique relationships.
Lastly, you use the parent codes (i.e. themes) to address the research question(s). Here's an illustration of the high-level process:
Imagine doing all these using a user friendly qualitative analysis software such as NVivo. It contains features which help you to work on multiple data, run queries, code significant parts of data, add descriptions and memos (reflections) to the codes generated, create illustrations to better display your findings, and brainstorm ideas using the ‘Mind Maps’ function in NVivo.
The coder’s findings (such as themes, models, and/or theories) don’t only represent the data but reflect his/her subjective intent and thought process, background, and experiences.
To ensure credibility, the coder needs to be transparent in the coding process. There are three main qualitative data analysis stages that NVivo could be effectively used for to maintain transparency, attain consistency of the labels or nodes created, and reach meaningful findings with visual representations.
1. Pre-coding stage (getting to know your data)
I always start the coding process with a review of all the data, to familiarize myself.
The ‘Query’ command (in NVivo) for example, is a great tool to use so as to know the kind of words participants use and how often they are used.
The ‘Word Frequency’ result could be displayed as a ‘Word Cloud’ with varied word fonts depicting the number of times the words are utilized:
In addition, I sometimes want to know how a specific word or phrase is used in a participants’ responses. So I use the ‘Text Search’ command resulting in the creation of a ‘Word Tree’.
The ‘Word Tree’ shows words or phrases used before and after the searched word. This helps in knowing the context in which it was used:
2. Coding stage (assigning labels to the nodes)
When coding, the consistency of codes or nodes generated is very crucial. It helps you to easily see the relationships between nodes and sometimes figure out underlying ideas and meanings among them.
So how do you maintain consistency of nodes?
One strategy is to select specific coding method(s) consistent with the research question to create labels for the nodes (Saldana, 2013).
3. Post-Coding stage (presenting your findings)
The main purpose of conducting a qualitative analysis is to generate themes to address the research questions (Adu, 2013).
After coming up with themes, the next stage is to present the findings.
How the results are presented to your audience could impact the credibility of the results. To ensure credibility, a researcher should present each theme with its respective meaning and evidence from the data. Adding a visual representation of the themes, their relationships, and related ideas helps the audience to better understand the findings (Saldana, 2013).
The NVivo ‘Explore’ function is an excellent tool to create project maps, concept maps, charts, and cluster trees just to mention a few.
In conclusion, data analysis is a very intensive process, with qualitative researchers carefully and systematically reducing the data to themes to address the research question. To attain credible findings, the coder should extensively document the process and clearly explain how they arrived at the findings.
NVivo is a great qualitative analysis software which could be used to code data, document the data analysis process, and present a visual presentation of the results so as to increase credibility.
Adu, P. (2013, November 22).
Qualitative analysis coding and categorizing.
Saldana, J. (2013).
The coding manual for qualitative researchers. London: Sage
ABOUT THE AUTHOR
Dr. Philip Adu is a Methodology Expert at The Chicago School of Professional Psychology (TCSPP). His role is to provide support to dissertating students in TCSPP addressing their methodology related concerns. You could access some of his webinars at the ‘Methodology Related Presentations – TCSPP’ YouTube Channel. He completed his Doctoral degree in Education with a concentration in Learning, Instructional Design and Technology from West Virginia University (WVU). Dr. Adu recently authored a book titled, “A Step-by-Step Guide to Qualitative Data Coding” (available on routledge.com or amazon.com). You could reach Dr. Adu at firstname.lastname@example.org and @drphilipadu on Twitter.