Page 151 -

138 CHAPTER 6 Diaries

6.2 WHY DO WE USE DIARIES IN HCI RESEARCH?

Diaries fill the gaps in HCI research methods between observation in naturalistic set-
tings, observation in a fixed lab, and surveys (Hyldegard, 2006). Many say that con-
trolled studies in controlled settings (such as usability labs) are ideal and others say
that observing users in their natural settings (such as homes or workplaces) is ideal.
However, in many cases, it is not feasible to either bring users into a fixed setting or
visit the users in their natural setting. In addition, having observers present in either
setting can sometimes change the actions of the users (Carter and Mankoff, 2005).
Diaries are especially good at studying usage patterns that cross multiple technolo-
gies, multiple locations, and multiple environments (Hayashi and Hong, 2011). For
instance, to study how people utilize passwords across multiple systems, devices,
networks, and locations, at both home and work, a diary study was used, with 20
participants, who recorded 1500 password events over a 2-week period (Hayashi and
Hong, 2011). Of the various research methods described in this book, very few of the
methods could have accurately collected data from so many different technical and
physical environments.
All research methods have strengths and weaknesses, and by using two or three
different research methods, you can often get a much better understanding of phe-
nomena than you would with only one research method. For instance, Kientz et al.
developed a technical solution (called FETCH) to help blind people track everyday
items, such as keys, iPod, remote controls, and sunglasses. A small Bluetooth tag
was added to these items to help in tracking. While a laboratory study was con-
ducted first, a controlled laboratory study clearly would not be sufficient to deter-
mine how this approach to finding items could be used in someone's daily life. After
the laboratory study discovered some needed improvements in the interface, a diary
study was used in which participants would track when they lost items that they
needed and how long it took them to find the item. In the first 2-week phase of the
diary study, the participants did not use FETCH. During the second 2-week phase,
the participants used the FETCH system and recorded when they lost items and how
long it took to find them (Kientz et al., 2006). The diary study was then followed up
with interviews with the participants. The use of the diary in conjunction with other
methods strengthened the findings of this research project. However, while ideal, it
is sometimes not possible to use two or three different research methods, due to time,
cost, or participant availability.
It might seem that surveys are an appropriate solution compared to diaries, as
they allow users to record data in their own settings and time, and surveys reach a
geographically distributed set of users. However, surveys can lead to biased data
in behavioral research in some situations, and diaries offer some advantages over
surveys in certain research situations. In many cases, diaries are used in conjunction
with other methods; when this is possible, it is ideal, as adding one research method
often ameliorates the shortcomings of another method. Diaries allow for collect-
ing of more detailed research than surveys, which often use predefined questions
and allow little flexibility for respondents. Alaszewski said it best, “While survey