Data Visualization of Qualitative User Research

While it is extremely important to gather qualitative data such as user interviews, it can be difficult to make sense of such ‘unstructured information’.   Often the biggest challenge is knowing just where to start.  For example consider trying to make sense of the transcriptions from several dozen user interviews - you might have a mix of positive and negative feedback, anecdotes, opinions and narratives.  You could painstakingly sift through all of the conversations, highlighting meaningful terms or search for keywords.  

But data visualization tools, typically thought of in terms of quantitative data, can also help to understand qualitative information.  Data visualization has been promoted as an effective means to present data, but its enormous value in analyzing data has been largely overlooked. 

One such example is IBM’s Many Eyes.  While this web application set is known to many for its beautiful graphing capabilities, I find the text visualization tools most valuable for analyzing qualitative data.  In particular, the Word Tree in Many Eyes “lets you pick a word or phrase and shows you all the different contexts in which it appears. The contexts are arranged in a tree-like branching structure to reveal recurrent themes and phrases.”

When applied to qualitative data (e.g. interview transcripts, free-text survey comments), the Word Tree allows a researcher to quickly scan through text-based content by searching via keyword or phrase.  For example to see what a group of users said about a particular product feature, the researcher can create a word tree around the feature (e.g. “portable” or “installation”) or around particular terms that are likely to indicate problems (e.g. “difficult to”, “but”).  

Visualizing the interviews around critical terms provides a starting point for reviewing and understanding qualitative data in an efficient manner.  It is not a substitute for thoughtful analysis, but a head-start. 

Note: One significant caveat in the case of Many Eyes is that all submitted data is publicly viewable, so it’s not always suitable for proprietary data analysis – but it is free.