Page 311 -

300 CHAPTER 11 Analyzing qualitative data

statistical methods used to evaluate the validity and reliability of the coding com-
pleted by human coders. The general strategy discussed in this chapter is just one
of the many approaches available for analyzing text and other qualitative informa-
tion. Substantially different strategies may be used for different disciplines, such
as literature or art.

11.2 GOALS AND STAGES OF QUALITATIVE ANALYSIS
The goal of qualitative analysis is to turn the unstructured data found in texts and
other artifacts into a detailed description about the important aspects of the situation
or problem under consideration. This description can take many forms, including
textual narratives, graphical diagrams, and summary tables. These items can often be
combined to provide the range of perspectives needed for understanding the underly-
ing complexity.
According to Corbin and Strauss (Corbin and Strauss, 2014), qualitative data
analysis consists of three stages. We start with a data set containing information
about our problem of interest. For example, the problem can be related to challenges
faced by or unique needs of a specific group of users (i.e., people with visual dis-
abilities, senior citizens, children, etc.). It can focus on a specific technology such
as a gesture-based input method, a photo sharing application, or 3D printing. The
problem can also examine the interaction behavior in a specific context, such as text
entry on a small screen while the user is constantly walking. Via analysis, we hope
to identify major themes and ideas that describe the context, activities, and other
perspectives that define the problem. In the second stage, we drill down into each
component to find relevant descriptive properties and dimensions. In many cases, we
need to understand not only the nature of each component, but also how they relate
to each other. In the third stage, we use the knowledge we gained from studying each
individual component to better understand the original substance and make infer-
ences about that substance.
For example, we might analyze chat logs to study the online behavior of Internet
users. Reading these logs, we might notice that three factors, namely personality,
education, and computer-related experience, are repeatedly found to influence us-
ers’ online behavior. We continue to study each of those three factors and how they
relate to each other. We study the literature in psychology and sociology to under-
stand the types of personality, how an individual forms and develops a specific
personality, and how a specific type of personality affects an individual’s social
behavior. Once we have a thorough understanding of the three factors, we can tie
the knowledge back to the original texts and examine how each of the components
affects a user’s online behavior. Specifically, we might use our literature review to
identify specific personality or educational behaviors that might influence online
behaviors, This application of experience and contextual knowledge is critical for
the appropriate interpretation of qualitative data and the entire knowledge discov-
ery process.