Page 156 -

6.4 What type of diary? 143

participation in the research should remain anonymous (see Chapter 15 for more
information on human subjects protection).

6.4 WHAT TYPE OF DIARY?

There are a number of different methodological decisions to make when using dia-
ries in HCI research: What type of diary? How will the diary be recorded (paper or
electronic)? For what period will the users be asked to keep the diary? Diaries are
typically kept for a period of 1 or 2 weeks (Rieman, 1993). Any longer than that and
participation tends to drop off.
At a high level, diaries can be split into two types of purpose: feedback and
elicitation (Carter and Mankoff, 2005). A feedback diary is one in which the data
from the diary itself provides the feedback to the researchers. The feedback diary
is the data collection method; the diary is not meant as a springboard to anything
else. In an elicitation diary, the data recorded in the diary is used for prompting,
when interviews take place at a later point, and the users are encouraged to ex-
pand upon each data point (see Chapter 8 for more information on interviews).
Feedback diaries usually focus on the events that interest the researcher, whereas
elicitation diaries usually focus on events that interest the user. Feedback diaries
tend to have instructions for users that they should make a diary entry when a
certain event or threshold occurs. Elicitation diaries tend to encourage users to
make diary entries based more on events that have meaning to the user. Feedback
diaries can be more accurate (since users record events on a regular basis as they
occur) and more objective but elicitation diaries can provide a view that is more
representative of what the user is feeling (Carter and Mankoff, 2005). In a similar
fashion to a survey, with an elicitation diary users must recall in a later interview
what has occurred and this can lead to bias. However, the data points recorded by
the user in the elicitation diary can provide some level of validation, which does
not exist in a survey.

6.4.1 FEEDBACK DIARY
Feedback diaries come in many different formats, but probably the most important
research question in a feedback diary is how often a diary entry is made. For in-
stance, what event, time, or threshold triggers the need for the user to make a diary
entry? Users could be asked to make a diary recording when an event occurs, such
as when they feel frustrated with an interface, or when they complete a certain task.
Users could be asked to make a diary recording at a set time every day (say, 9 p.m.),
or during a specified time period (say, from noon to 6 p.m.). Users could be inter-
rupted throughout the day at random times, to get a random sample of the user's
daily life (Carter and Mankoff, 2005). Historically, an individual in this type of
time diary study wears a beeper and must record what they are doing whenever the