Written by Erika Yi, Ph.D.
After my last post, several people asked me about the tools I used for the coding process. This article will be focusing on the qualitative analysis tools you can use to code your data. The methods I’ve used differ in pricing and functionalities, so you can decide which method is the most useful to you. Note that this review is by no means comprehensive, I can only speak to the tools and functions I actually used.
Before you start coding, make sure you get very familiar with your data. Regardless if your data is transcripts or notes, you should know the rough content of each part of the scripts and the approximate location of the important information you’re looking for. Delving into coding without knowing your data thoroughly is never an option if you want your hard work to render data-based insights.
Method 1, the old school
Tools needed: a pair of scissors and some folders
This is the very first coding method I learned in graduate school. Assuming you’ve familiarized yourself with the data set, now all you need is to print the transcripts and cut them up. Your cut snippets reflect your codes, each piece of paper is a code unit. You can use sticky notes to attach the code names to each piece of paper so if you ever decide to change the code name you can simply remove the sticky note and attach a new one to it.
After you’ve cut up and organized all the paper clips, you will need a large empty space — the floor of a large room or a large empty table. You can organize the coded paper clips into groups, move them around to see which structure communicates cohesive themes the best, and then use the folders to hold the paper clips together and name the folders by your themes.
In the end, you will have several folders with pieces of paper in it. From that, you will be able to analyze the themes more in-depth and engage in storytelling based on your codes.
Pros: low tech requirement, free to use (as long as you have scissors and folders)
Cons: difficult to digitize the codes, pieces of paper could go missing after a while, not environmentally friendly
Method 2, the digital adaptation
Tools needed: any word processor and spreadsheet
For those who don’t want to print off paper copies of your scripts, this method can come in handy. You will need a word processor with highlighting and commenting functions and a digital spreadsheet.
Use different highlights to color code your data, and use comments to give your code units proper code names. The first page of your spreadsheet should be a summary of all the codes, you can move the codes around to categorize them. In addition, each code should have a separate sheet page wherein you will copy and paste the text snippets from the word processor.
In the end, you will have coded text documents and one spreadsheet, the codes of the spreadsheet should reflect all the codes appeared in the text document.
Pros: low cost, codes are digitized
Cons: only suitable for smaller data set, larger amount of data will be very difficult to sort and order using a spreadsheet
Method 3, professional qualitative analysis tools
Tool #1 NVivo
NVivo is hands down the most powerful qualitative analysis tool there is on the market. This software supports a wide range of data import from pure texts to images and videos and even data from social media sites such as Twitter and Facebook. It also has very powerful functions that go beyond just coding. For example, you can establish connections and create clusters between data sets, such as mapping out the relationship among a group of participants.
The data visualization function of NVivo can help a researcher generate easy-to-read charts and diagrams that quantify the results. However, the automatic coding function doesn’t work as well as expected. This is understandable, after all, the critical insights of a researcher are part of the qualitative analysis process and it will be almost impossible to replace that human touch with a machine.
Pros: probably the most powerful and comprehensive qualitative analysis tool, supports team collaboration
Cons: steep learning curve, high price, software runs heavy due to its rich features
Tool #2 Dedoose
Dedoose is a web-based qualitative analysis service that charges a monthly fee to the users, but only during the month that you log in. Dedoose doesn’t require installation, you simply upload the data to its server and use the website to code. Dedoose supports text documents, images, and videos. Similar to NVivo, Dedoose also has data visualization feature that helps us to generate informative analysis and help researchers detect patterns. For mixed method researchers, Dedoose supports both qualitative and quantitative analysis on the data, so that both qualitative and quantitative data can be analyzed and organized in the same place.
Pros: no need to install, affordable monthly fee, state-of-art data encryption, supports mixed method, supports team collaboration
Cons: requires an internet connection, can be too feature-heavy for a web browser
Tool #3 F4analyse
F4analyse can be installed on Mac, Windows, and Linux systems. This lightweight coding tool is excellent for transcribing and coding interviews. You can import the audio file and transcribe the texts directly in F4analyse as you listen to the audio recording. The transcripts will have automatic timestamps for each line, making it very easy to locate a sentence in the original recording. The coding of the transcripts are easy to use, you can color-code your codes, drag and drop them to form structures, and filter text snippets based on your search criteria.
F4analyse supports multiple languages, an important factor for any multilingual researcher. Although it is not as feature-rich as NVivo and Dedoose, F4analyse does generate a number of tables and charts that help you visualize your data.
Pros: affordable pricing, lightweight, easy to use
Cons: only supports text coding, can only import and export RTF format
Tool #4 Delve
Delve is a lightweight, web-based qualitative coding tool that’s still under development. After chatting with Delve team members LaiYee Ho and Alex Limpaecher, I tested Delve to code some small projects. The interface of Delve is very clean and straightforward, and the coding process is easy to learn and use. Similar to previously mentioned tools, you can move around your codes to form structure and categories. You can assign different descriptors to a script for better search and filter — very useful for generating personas. You can manage several research projects at the same time, each project is stored separately. Overall, Delve is a very convenient tool for coding text data as an individual or as a team, and I’m looking forward to using more of the Delve features.
Pros: no need to install, lightweight, easy to use, beautiful minimalist interface, supports team collaboration
Cons: requires an internet connection, some features are still under development
Bonus, use digital notebooks to document your codes
If you’re a OneNote or Evernote user, you can use the tagging and page organizing functions to document your scripts that reflect your code structure. A digital notebook is a suitable bonus for all the above mentioned coding methods. Digital notebooks are better at text processing and editing than the coding tools and a spreadsheet. The notebook page structure helps to visualize the code structure better than a word processor as well. In addition, they are accessible through multiple platforms online and offline, so that you won’t be limited by internet access when analyzing your data.
The most important tool
As a researcher who has used all the above tools in my 8 years of research, I can see the benefits in all of them. However, a tool is only as good as the one who uses it. Relying too heavily on any of them and expecting it to do the work for you is unrealistic. Researcher themselves are in fact instruments to measure and analyze problems, these tools are just there to make the process easier for you. After all, the most important research tool, my dear researchers, is you!