What is textual analysis?
Textual analysis is a user research analysis technique for parsing data from textual user responses. These responses could come from user research surveys, interviews, support tickets, forums and comments.
Why should I use textual analysis?
Textual analysis can help to help explain the reason for a finding in a user research studies. Especially in quantitative studies, you often find out how much something is happening without getting a picture as to why. The textual analysis can help support other research methods by providing deeper explanations.
Textual analysis is also very useful for copy writing and prioritizing. For example, if you are deciding between the words “share”, “follow”, and “subscribe”, a textual analysis can help you to determine which of the words or phrases your users are more likely to use.
How do I conduct a textual analysis?
You first need a data set of user responses. Often, a three to five question qualitative survey that focuses on open-ended questions and essay field responses will provide a good source of information.
Simple forms of textual analysis involve splitting text by punctuation and counting the frequency of terms. Some tools, such as SurveyMonkey and ConceptCodify, provide textual analysis features within them. More complex techniques use a set of algorithms under the subject of natural language processing. The Python programming language includes such a toolset called NLTK. Several user interface driven tools are available both by open source and proprietary license.
What are the limits of textual analysis?
Textual analysis requires a large set of existing user responses. Timeliness is of the essence with these sort of studies, as user expectations and feedback can change quickly. Also, textual analysis isn’t necessarily a primary research technique; it is more often used to support findings from other types of research, such as analytics data, multivariate testing, or usability testing.
Questions, comments, suggestions? Let us know!