OBSERVATIONS AND LESSONS LEARNED The discussions of the

By admin on October 31st, 2007

OBSERVATIONS AND LESSONS LEARNED

The discussions of the first phase were useful to identify the
requirements that could improve the application and the user
interface. We noticed that the playback of the video in playback
mode is more entertaining when two people recorded a recipe
together rather than a single one. When a single user is
recording a recipe, it can happen that she forgets to mention a
step. When two people cook together, the spontaneous
conversation among the two of them, such as Can you please
pass me the cheese? or Can you turn down the fire? helps to
keep track of the process. It also creates a more fun experience
for the users that are recording as they often end up in jokes or
small talks about the recipe. The feeling of a natural social
setting makes the whole cooking session more interesting for
the user who plays the video back and the personalization of
content is much stronger. For this reason, after the first test, we
agreed on the protocol to cook with two people in all further
tests.
Given that cooks are already managing a set of tools, we
observed that the manipulation of an additional one, i.e., the
pen, soon became inconvenient, especially when it was
misplaced on the kitchen counter as it happens with cooking
tools. After this observation in the first tests, the pen was
hanged with a piece of string next to the display, so as to
support more casual use.
We also noticed that more than one camera would better
capture the whole event. A closer look at commercial TV
cooking shows revealed that they use up to seven distinct
camera positions. Indeed it is important to have a close view at
the location where the food is actually prepared (e.g. the fire
and the counter), as well as on the face and movements of the
person who is cooking, in order to keep the desired sense of
presence.
The cognitive walkthrough tests with users from outside of the
project team gave us some useful insight both on the usability
of the interface as well as on the whole experience. To this
respect it became clear the difficulty to evaluate experiences in
a hybrid context, in an environment which is supposed to
recreate an everyday life one, but is actually still far from the
intimacy and acquaintance people have in their own real
domestic environments. Even though some testers were
employees of the lab and know the environment of this specific
kitchen, the activities they normally perform there are different
from proper cooking: e.g. coffee preparation, food warm-up,
food storage. When both the employees and the external testers
were asked to use the application and to cook there, we noticed
that a certain stress was generated by the pure fact of cooking
in an unfamiliar kitchen, using the interface, talking aloud or
observing a video, and, in the cognitive walkthrough, being
observed by a researcher. Some users, in particular, were most
concerned about the result of their cooking, being aware that
the team would taste it afterwards. Thus, the feedback they
provided was mostly focused on this aspect rather than the
whole experience of using the application and how this could
affect their mundane activities.
In these tests we realized that for the assessment of the whole
experience and of its social meaning it would be necessary to
let users try the application in their own kitchens they feel
familiar with. Furthermore, the expected increase in motivation
due to social relationships needs to be validated in a more
intimate social setting. One of the testers explicitly asked to
have the video of his cooking session to show it to his
girlfriend.
Therefore we plan to connect different
kitchens/cooks that have a real familiar relationship.