Page 11 -

Foreword

Many disciplines are hyphenated combinations, such as bio-informatics or physical-
chemistry, but human-computer interaction (HCI) spans a broader range of topics
than most. As a result, HCI researchers often draw on multiple diverse research
methods, even in a single paper. It is just possible that HCI's remarkable successes in
academic publishing and in widely used technologies stem from its diverse research
methods.
While the traditional scientific method was a solid foundation for HCI, controlled
laboratory studies with tests for statistically significant differences were never quite
enough to deal with the ambitions of HCI researchers. We also embraced interviews,
surveys, and focus groups, sometimes in fresh ways, to capture impressions of dis-
tinctive users and elicit suggestions, reactions, frustrations, and fears. Ethnographic
observation and anthropological methods were also applied to study computer users
“in the wild,” which meant going to the place where people worked, lived, or played
to see what actual use was like. As researchers shifted from studying the immediate
out-of-the-box experience to understanding the evolution of user experiences over
weeks and months, long-term case studies and time diaries became more common.
A larger step for HCI researchers was to incorporate iterative engineering pro-
cesses and design thinking. They had to overcome resistance from traditional re-
searchers who believed that controlled experiments were the best way forward. Over
the years still newer methods tuned to the needs of businesses were developed, such
as usability testing and expert reviews, to accelerate the development process, rather
than refine theories. A major step forward was the development of A/B testing which
contrasted two slightly different user interfaces in actual use over a period of days or
weeks with thousands of actual users. Web designers were able to make rapid prog-
ress in determining which features led to greater commercial success.
Another novel approach has been to crowdsource research, by putting up online
experiments available to many users or to use services like Amazon Turk to hire
hundreds of participants for experimental studies. In recent years still newer meth-
ods based on big data analyses of millions of tweets or social media posts changed
the game dramatically. The online availability of so much data about human perfor-
mance led theoreticians and practitioners to study whole communities at scale in
realistic settings.
I am pleased that the authors have used the distinction between micro-HCI and
macro-HCI to organize thinking about when to apply one research method or an-
other. Short-term perceptual, motor, or cognitive tasks can be studied by micro-HCI
methods such as controlled experiments, but long-term trust, community develop-
ment, or satisfaction are better studied by macro-HCI methods. I am also pleased that
the authors encourage readers to reach out to other research communities to learn of
their methods, to partner with them in policy initiatives, and to convey the opportuni-
ties that HCI presents for bold new directions and powerful impact.

xxi