Page 310 -

CHAPTER
11
Analyzing qualitative data

11.1 INTRODUCTION
In Chapter 4, we discussed how to use significance tests to study quantitative data
and measures such as speed, error rate, distance, adoption rate, and rankings. The
well-defined nature of quantitative measures makes them appealing options for many
studies. When we can clearly specify our measures of interest and how they are to be
measured, research methods and analytic procedures can be clearly defined, making
study design reasonably straightforward. Of course, complications may arise, but we
don’t need to worry about the definition of the underlying units used to measure task
completion times.
Our discussions of case studies, interviews, and ethnography introduce markedly
different kinds of data associated with research questions and analysis methods that
are not quite so clear-cut. Rather than searching for numerical measurements, these
qualitative studies attempt to study texts, observations, video, and artifacts to under-
stand complex situations. Analysis of these data often raises challenges that rarely
raise with quantitative data, as we struggle to interpret ambiguous comments and
understand complex situations. To make matters worse, we don’t even know what the
“truth” is—as multiple researchers might (and often do) have different perspectives
on the same situation.
Acknowledging these challenges, social science researchers have developed re-
search methods designed to increase rigor and validity in analyzing qualitative data.
Qualitative methods do not aim to eliminate subjectivity—instead, they accept that
subjectivity is inherent to process of interpreting qualitative data, and they strive to
show that interpretations are developed methodically to be consistent with all avail-
able data, and representative of multiple perspectives.
In this chapter, we present an introduction to qualitative research, discuss-
ing techniques for ensuring high-quality analysis of qualitative data that is both
reliable and valid. We introduce the process of coding, which assigns labels to
observations from text or other qualitative data forms. We specifically focus on
grounded theory (Glaser and Strauss, 1967), the starting point for many qualitative
analyses. The use of content analysis to extract categories from diverse “texts” is
described, along with a discussion of the analysis of two very important forms
of qualitative data: text and multimedia. In order to control the impact of subjec-
tive interpretation, a commonly accepted coding procedure should be adopted and

Research Methods in Human-Computer Interaction. http://dx.doi.org/10.1016/B978-0-12-805390-4.00011-X 299
© 2017 Elsevier Inc. All rights reserved.