Page 159 -

146 CHAPTER 6 Diaries

entry would be appropriate, even if the participant is unable to make an entry at that
time, and the participants can then be reminded to make the diary entry as soon as
possible (Chong et al., 2014).
The crucial factor in choosing the media should be the type of media that will be
most natural for the diarists in their everyday life. For instance, if participants will be
performing the tasks of interest while sitting at their computer, it might make sense to
use word processing, spreadsheets, or web-based forms (see the Diary Study of Task
Switching sidebar). However, if participants will be recording diary entries about the
use of mobile devices, you would expect these entries to occur while the participants
are on-the-go (see the Recording Entries on the Go sidebar).

DIARY STUDY OF TASK SWITCHING

Czerwinski et al. (2004) did a diary study of task switching between
different projects (and related interruptions) during a week. They were
trying to examine how interruptions impact on task switching, with the
end goal of improving how user interfaces support users recovering
from interruptions. They used a diary study, which they felt was most
appropriate since there were no existing empirical studies of tools for
dealing with task switching and recovery. In addition, diary studies,
because they take place in the users' natural settings and tasks, have high
ecological validity.
Eleven users took part; all were professionals who multitasked
among at least three major projects or tasks. Before the diaries started,
users filled out a baseline survey with demographic information and
perceptions about computers. The diaries were recorded using an
Excel spreadsheet, where the researchers had labeled columns for each
parameter that they wanted to track. For instance, the researchers were
interested in learning how users defined tasks, at what level of granularity.
The diaries also tracked the difficulty of switching tasks and the amount
of time spent on the tasks.
Due to the qualitative nature of the data, two researchers tested and
validated the rich coding scheme. Using the coding scheme, first frequency
counts and descriptive statistics were carried out on the diary data, followed
by regression analysis. Among the significant findings were that users
reported an average of 50 task shifts over the week, and that long-term
projects, which involved multiple documents and involved more revisits,
were very hard to return to, once interrupted.
A time diary form from Czerwinski et al. (2004) appears on the next page.