Qualitative HCI Research. Ann Blandford
2014; Rajkomar et al., 2015) and interacting with information (e.g., Blandford and Attfield, 2010; Makri et al., 2008a; Makri and Warwick, 2010). The first of these brings challenges, particularly in engaging with patients and dealing with sensitive issues within complex healthcare processes. The second brings a different kind of challenge: that interacting with information is often not the primary focus of someone’s activity; it is a secondary activity that they barely notice, so gathering useful and reliable information about users’ interactions can be difficult. Using these and other experiences, we review challenges and provide advice for designing and conducting qualitative HCI research.
1.1 AN OVERVIEW OF QUALITATIVE APPROACHES AND METHODS IN HCI
There are many, many approaches and methods for qualitative research. Some of them have names, such as Ethnography, Contextual Inquiry, Focus Groups, Grounded Theory, Interpretive Phenomenological Analysis, Discourse Analysis or Thematic Analysis; others do not. Some—such as Contextual Inquiry and Grounded Theory—are widely used in HCI, while others—such as Interpretive Phenomenological Analysis and Discourse Analysis—are not; we focus on the more widely used methods and approaches in this book. Some of these names have precise meanings; others are often used as generic descriptors of qualitative research. For example, Grounded Theory (GT) has been described as a “bumper sticker” (Bryman and Burgess, 1994) to cover a broad range of qualitative approaches, even though there are strong principles underpinning GT proper. This makes it particularly important for HCI researchers to be open and transparent when explaining and justifying the qualitative approaches they have adopted. When writing up an approach, it is essential for researchers to explain in detail what they did and why, giving reasons for adopting, adapting or combining particular established approaches.
Denzin and Lincoln (2011) discuss a research process in terms of five phases, or levels of activity. The first phase is the researcher—you!—who comes to the study with their individual history, experiences, values and understanding; the researcher shapes the research, and should be aware of the role they are playing in the research.
The second phase is the research paradigm. In Chapter 6, we discuss research paradigms in terms of quantitative and qualitative approaches that are widely used in HCI. In brief: quantitative research is most commonly applied to test pre-determined hypotheses, whereas qualitative approaches aim to describe and explain phenomena in a rich, often exploratory, way. Denzin and Lincoln (2011) identify four major paradigms for qualitative research: positivist and post-positivist; constructivist-interpretive; critical (Marxist, emancipatory); and feminist-poststructural. Given the aims of HCI studies, focusing on the design and use of interactive technologies, qualitative HCI research generally fits within the first two of these paradigms, and this book focuses on the constructivist-interpretive paradigm. This paradigm assumes a subjective reality that is shaped by the interpretations of researchers and study participants. This can feel uncomfortable at first, particularly to those who have been brought up in a classic scientific paradigm where it is assumed that there is an objective reality “out there” and that the role of research is to establish what it is. This book is intended to provide tools and techniques to conduct high quality interpretive qualitative HCI research.
Denzin and Lincoln (2011) label their third phase “research strategies.” This is the phase that focuses on the strategy for addressing the research question or purpose of the study. Their list of strategies includes several that are commonly used in HCI, including ethnography, participant observation, ethnomethodology and GT. We discuss these approaches in Chapter 6, after discussing the particular methods that make up a study (Chapters 4 and 5).
Their fourth phase is “methods of collection and analysis.” They include interviewing, observation, autoethnography and focus groups as data collection methods; to this list, we add think aloud as a technique that is particular to HCI (Chapter 4). We separate out analysis (Chapter 5), focusing particularly on Thematic Analysis (Braun and Clarke, 2006) as a technique for data analysis that is widely used in HCI.
The final phase according to Denzin and Lincoln (2011) is that of interpretation and evaluation. The questions in this phase are broadly: what can be learned from this study, how confident can we be in the findings and how might they be reported? In this book, we present this in terms of reporting the study (Chapter 7) and delivering the highest possible quality research (Chapter 8).
1.2 THE SPACE OF INTERPRETIVE QUALITATIVE STUDIES IN HCI
There are several important dimensions on which qualitative studies in HCI vary:
• The focus of the study: Qualitative studies in HCI, by definition, focus on current or future technology design and use. But there are still many possible questions that the study might address—e.g., “how does our new design compare with our competitor’s design?” or “what are the privacy implications of introducing this new technology?”
• Who provides the data: Most studies involve the current or intended future users of the system of interest. Occasionally it is necessary to work with surrogate users—e.g., when real users are too busy to take part, or too expensive to recruit. Many studies also involve stakeholders in the system, such as domain experts who understand at least some of the users’ technology needs. In healthcare, for example, medical practitioners might provide input on system functionality they believe their patients need.
• Where the data is gathered: For many studies, particularly observational ones, it is important to do the data gathering in the “field”—i.e., in the real world, where the technology will be used in practice. But interviews may be conducted away from the situation of use, and think-aloud studies that focus on the individual’s interaction with a specific system often take place in controlled (“laboratory”) settings.
• How the data is gathered: Most studies gather data through observation, interviews, focus groups, or diaries. Some studies use existing data such as incident reports, product reviews or system documentation.
• How the study is structured: Some studies have a clear stepwise structure, from devising research questions, to gathering data, to counting responses and producing results; others are more exploratory and iterative, which can include interleaving data gathering and analysis, as they refocus questions and find more meaning as they engage with the data. This is discussed in more detail in Chapter 6.
• The relationship between the analyst and the data: Some studies presume an objective reality to be “out there,” and so two independent people can analyse the same data and discover the same conclusion; other studies recognise the role of the researcher in shaping the analysis and creating a narrative. We discuss positivist and interpretivist traditions that underlie these positions in more detail in Chapter 6. Throughout this book, we focus primarily on the interpretivist tradition.
This makes it sound as if there is enormous scope for variability. In practice, every study needs to be coherent: the approaches to gathering and analysing data need to be well suited to the research question, as discussed in more detail below.
In the following chapters, we expand on these themes to help you plan, conduct and report on your qualitative HCI study. Planning a study can seem overwhelming at first, but a focus on the purpose of the study, together with the courage to commit to early writing and to just getting on