Data Analysis for Qualitative Research: 6 Step Guide

man doing qualitative research" width="900" height="600" />

Data analysis for qualitative research is not intuitive. This is because qualitative data stands in opposition to traditional data analysis methodologies: while data analysis is concerned with quantities, qualitative data is by definition unquantified. But there is an easy, methodical approach that anyone can take use to get reliable results when performing data analysis for qualitative research. The process consists of 6 steps that I’ll break down in this article:

1. Perform interviews(if necessary)
2. Gather all documents and transcribe any non-paper records
3. Decide whether to either code analytical data, analyze word frequencies, or both
4. Decide what interpretive angle you want to take: content analysis, narrative analysis, discourse analysis, framework analysis, and/or grounded theory
5. Compile your data in a spreadsheet using document saving techniques (windows and mac)
6. Identify trends in words, themes, metaphors, natural patterns, and more

To complete these steps, you will need:

1. Microsoft word
2. Microsoft excel
3. Internet access

You can get the free Intro to Data Analysis eBook to cover the fundamentals and ensure strong progression in all your data endeavors.

What is qualitative research?

Qualitative research is not the same as quantitative research. In short, qualitative research is the interpretation of non-numeric data. It usually aims at drawing conclusions that explain why a phenomenon occurs, rather than that one does occur. Here’s a great quote from a nursing magazine about quantitative vs qualitative research:

“A traditional quantitative study… uses a predetermined (and auditable) set of steps to confirm or refute [a] hypothesis.

“In contrast, qualitative research often takes the position that an interpretive understanding is only possible by way of uncovering or deconstructing the meanings of a phenomenon.

Thus, a distinction between explaining how something operates (explanation) and why it operates in the manner that it does (interpretation) may be [an] effective way to distinguish quantitative from qualitative analytic processes involved in any particular study.” (bold added)

((EBN))

Learn to Interpret Your Qualitative Data

This article explain what data analysis is and how to do it. To learn how to interpret the results, visualize, and write an insightful report, sign up for our handbook below.

Step 1a: Data collection methods and techniques in qualitative research: interviews and focus groups

Step 1 is collecting the data that you will need for the analysis. If you are not performing any interviews or focus groups to gather data, then you can skip this step. It’s for people who need to go into the field and collect raw information as part of their qualitative analysis.

Interviews

Since the whole point of an interview and of qualitative analysis in general is to understand a research question better, you should start by making sure you have a specific, refined research question. Whether you’re a researcher by trade or a data analyst working on one-time project, you must know specifically what you want to understand in order to get results.

Good research questions are specific enough to guide action but open enough to leave room for insight and growth. Examples of good research questions include:

• Good : To what degree does living in a city impact the quality of a person’s life? (open-ended, complex)
• Bad : Does living in a city impact the quality of a person’s life? (closed, simple)

Once you understand the research question, you need to develop a list of interview questions. These questions should likewise be open-ended and provide liberty of expression to the responder. They should support the research question in an active way without prejudicing the response. Examples of good interview questions include:

• Good : Tell me what it’s like to live in a city versus in the country. (open, not leading)
• Bad : Don’t you prefer the city to the country because there are more people? (closed, leading)

Some additional helpful tips include:

• Begin each interview with a neutral question to get the person relaxed
• Limit each question to a single idea
• If you don’t understand, ask for clarity
• Do not pass any judgements
• Do not spend more than 15m on an interview, lest the quality of responses drop

Focus groups

The alternative to interviews is focus groups. Focus groups are a great way for you to get an idea for how people communicate their opinions in a group setting, rather than a one-on-one setting as in interviews.

In short, focus groups are gatherings of small groups of people from representative backgrounds who receive instruction, or “facilitation,” from a focus group leader. Typically, the leader will ask questions to stimulate conversation, reformulate questions to bring the discussion back to focus, and prevent the discussion from turning sour or giving way to bad faith.

Focus group questions should be open-ended like their interview neighbors, and they should stimulate some degree of disagreement. Disagreement often leads to valuable information about differing opinions, as people tend to say what they mean if contradicted.

However, focus group leaders must be careful not to let disagreements escalate, as anger can make people lie to be hurtful or simply to win an argument. And lies are not helpful in data analysis for qualitative research.

Step 1b: Tools for qualitative data collection

When it comes to data analysis for qualitative analysis, the tools you use to collect data should align to some degree with the tools you will use to analyze the data.

As mentioned in the intro, you will be focusing on analysis techniques that only require the traditional Microsoft suite programs: Microsoft Excel and Microsoft Word. At the same time, you can source supplementary tools from various websites, like Text Analyzer and WordCounter.

In short, the tools for qualitative data collection that you need are Excel and Word, as well as web-based free tools like Text Analyzer and WordCounter. These online tools are helpful in the quantitative part of your qualitative research.

Step 2: Gather all documents & transcribe non-written docs

Once you have your interviews and/or focus group transcripts, it’s time to decide if you need other documentation. If you do, you’ll need to gather it all into one place first, then develop a strategy for how to transcribe any non-written documents.

When do you need documentation other than interviews and focus groups? Two situations usually call for documentation. First, if you have little funding, then you can’t afford to run expensive interviews and focus groups.

Second, social science researchers typically focus on documents since their research questions are less concerned with subject-oriented data, while hard science and business researchers typically focus on interviews and focus groups because they want to know what people think, and they want to know today.

Non-written records

Other factors at play include the type of research, the field, and specific research goal. For those who need documentation and to describe non-written records, there are some steps to follow:

1. Put all hard copy source documents into a sealed binder (I use plastic paper holders with elastic seals).
2. If you are sourcing directly from printed books or journals, then you will need to digitalize them by scanning them and making them text readable by the computer. To do so, turn all PDFs into Word documents using online tools such as PDF to Word Converter. This process is never full-proof, and it may be a source of error in the data collection, but it’s part of the process.
3. If you are sourcing online documents, try as often as possible to get computer-readable PDF documents that you can easily copy/paste or convert. Locked PDFs are essentially a lost cause.
4. Transcribe any audio files into written documents. There are free online tools available to help with this, such as 360converter. If you run a test through the system, you’ll see that the output is not 100%. The best way to use this tool is as a first draft generator. You can then correct and complete it with old fashioned, direct transcription.

Step 3: Decide on the type of qualitative research

Before step 3 you should have collected your data, transcribed it all into written-word documents, and compiled it in one place. Now comes the interesting part. You need to decide what you want to get out of your research by choosing an analytic angle, or type of qualitative research.

The available types of qualitative research are as follows. Each of them takes a unique angle that you must choose to get what information you want from the analysis. In addition, each of them has a different impact on the data analysis for qualitative research (coding vs word frequency) that we use.

1. Content analysis
2. Narrative analysis
3. Discourse analysis
4. Framework analysis, and/or
5. Grounded theory

From a high level, content, narrative, and discourse analysis are actionable independent tactics, whereas framework analysis and grounded theory are ways of honing and applying the first three.

Content analysis

• Definition: Content analysis is identify and labelling themes of any kind within a text.
• Focus: Identifying any kind of pattern in written text, transcribed audio, or transcribed video. This could be thematic, word repetition, idea repetition. Most often, the patterns we find are idea that make up an argument.
• Goal: To simplify, standardize, and quickly reference ideas from any given text. Content analysis is a way to pull the main ideas from huge documents for comparison. In this way, it’s more a means to an end.
• Pros: The huge advantage of doing content analysis is that you can quickly process huge amounts of texts using simple coding and word frequency techniques we will look at below. To use a metaphore, it is to qualitative analysis documents what Spark notes are to books.
• Cons: The downside to content analysis is that it’s quite general. If you have a very specific, narrative research question, then tracing “any and all ideas” will not be very helpful to you.

Narrative analysis

• Definition: Narrative analysis is the reformulation and simplification of interview answers or documentation into small narrative components to identify story-like patterns.
• Focus: Understanding the text based on its narrative components as opposed to themes or other qualities.
• Goal: To reference the text from an angle closer to the nature of texts in order to obtain further insights.
• Pros: Narrative analysis is very useful for getting perspective on a topic in which you’re extremely limited. It can be easy to get tunnel vision when you’re digging for themes and ideas from a reason-centric perspective. Turning to a narrative approach will help you stay grounded. More importantly, it helps reveal different kinds of trends.
• Cons: Narrative analysis adds another layer of subjectivity to the instinctive nature of qualitative research. Many see it as too dependent on the researcher to hold any critical value.

Discourse analysis

• Definition: Discourse analysis is the textual analysis of naturally occurring speech. Any oral expression must be transcribed before undergoing legitimate discourse analysis.
• Focus: Understanding ideas and themes through language communicated orally rather than pre-processed on paper.
• Goal: To obtain insights from an angle outside the traditional content analysis on text.
• Pros: Provides a considerable advantage in some areas of study in order to understand how people communicate an idea, versus the idea itself. For example, discourse analysis is important in political campaigning. People rarely vote for the candidate who most closely corresponds to his/her beliefs, but rather for the person they like the most.
• Cons: As with narrative analysis, discourse analysis is more subjective in nature than content analysis, which focuses on ideas and patterns. Some do not consider it rigorous enough to be considered a legitimate subset of qualitative analysis, but these people are few.

Framework analysis

• Definition: Framework analysis is a kind of qualitative analysis that includes 5 ordered steps: coding, indexing, charting, mapping, and interpreting. In most ways, framework analysis is a synonym for qualitative analysis — the same thing. The significant difference is the importance it places on the perspective used in the analysis.
• Focus: Understanding patterns in themes and ideas.
• Goal: Creating one specific framework for looking at a text.
• Pros: Framework analysis is helpful when the researcher clearly understands what he/she wants from the project, as it’s a limitation approach. Since each of its step has defined parameters, framework analysis is very useful for teamwork.
• Cons: It can lead to tunnel vision.

Grounded theory

• Definition: The use of content, narrative, and discourse analysis to examine a single case, in the hopes that discoveries from that case will lead to a foundational theory used to examine other like cases.
• Focus: A vast approach using multiple techniques in order to establish patterns.
• Goal: To develop a foundational theory.
• Pros: When successful, grounded theories can revolutionize entire fields of study.
• Cons: It’s very difficult to establish ground theories, and there’s an enormous amount of risk involved.

Step 4: Coding, word frequency, or both

Coding in data analysis for qualitative research is the process of writing 2-5 word codes that summarize at least 1 paragraphs of text (not writing computer code). This allows researchers to keep track of and analyze those codes. On the other hand, word frequency is the process of counting the presence and orientation of words within a text, which makes it the quantitative element in qualitative data analysis.

Video example of coding for data analysis in qualitative research

In short, coding in the context of data analysis for qualitative research follows 2 steps (video below):

1. Reading through the text one time
2. Adding 2-5 word summaries each time a significant theme or idea appears

Let’s look at a brief example of how to code for qualitative research in this video:

Step 5: Compile your data in a spreadsheet

After you have some coded data in the word document, you need to get it into excel for analysis. This process requires saving the word doc as an .htm extension, which makes it a website. Once you have the website, it’s as simple as opening that page, scrolling to the bottom, and copying/pasting the comments, or codes, into an excel document.

You will need to wrangle the data slightly in order to make it readable in excel. I’ve made a video to explain this process and places it below.

Step 6: Identify trends & analyze!

There are literally thousands of different ways to analyze qualitative data, and in most situations, the best technique depends on the information you want to get out of the research.

Nevertheless, there are a few go-to techniques. The most important of this is occurrences. In this short video, we finish the example from above by counting the number of times our codes appear. In this way, it’s very similar to word frequency (discussed above).

A few other options include:

1. Ranking each code on a set of relevant criteria and clustering
2. Pure cluster analysis
3. Causal analysis

We cover different types of analysis like this on the website, so be sure to check out other articles on the home page.

How to analyze qualitative data from an interview

To analyze qualitative data from an interview, follow the same 6 steps for quantitative data analysis:

1. Perform the interviews
2. Transcribe the interviews onto paper
3. Decide whether to either code analytical data (open, axial, selective), analyze word frequencies, or both
4. Decide what interpretive angle you want to take: content analysis, narrative analysis, discourse analysis, framework analysis, and/or grounded theory
5. Compile your data in a spreadsheet using document saving techniques (for windows and mac)
6. Identify trends in words, themes, metaphors, natural patterns, and more

1. Source text [↩]

About the Author

Noah

Noah is the founder & Editor-in-Chief at AnalystAnswers. He is a transatlantic professional and entrepreneur with 5+ years of corporate finance and data analytics experience, as well as 3+ years in consumer financial products and business software. He started AnalystAnswers to provide aspiring professionals with accessible explanations of otherwise dense finance and data concepts. Noah believes everyone can benefit from an analytical mindset in growing digital world. When he's not busy at work, Noah likes to explore new European cities, exercise, and spend time with friends and family.