Information Technology Reference
In-Depth Information
streaming back data from apps, we were better
placed to focus interview questions on observed
behaviour. For example, one interviewee was ob-
served to have scored negative points on his first
few days of use and on questioning him on this
fact it was revealed that he had misunderstood the
game rules at first. This mechanism also allowed
us to select interviewees based on interesting
patterns of behaviour. For example, one player's
location data appeared to suggest that she played
the game while travelling, and on speaking to her
she confirmed that she regularly played it on her
daily commute to work.
Given the flexibility of the tools for interacting
with users and studying log data, we were able to
use the evaluation tools to ease the task of redesign.
This reflects one of our research goals, which is
developing means to quickly and appropriately
adapt software to suit the changing contexts and
interests of users. For example, as an answer to
the task “What could be improved about Yoshi?”,
one user (anonymised here as Helen) commented
that plantations were often too full. Helen was
invited for interview, and the interviewer then
raised this point to obtain further detail. Helen
explained that, as plantations auto-generate fruit
at a rate of one per hour, they would often be full,
which she felt was to the detriment of the game. In
particular, Helen described a situation where she
would empty a plantation before leaving for work
in the morning, and wanted to collect a seed from
work to plant when she got home. However, by
this time the previously empty plantation would
have around 10 pieces of fruit in it again, which
would have to be picked first and fed to unwilling
yoshis, leading to a points penalty.
Following this interview, the game designers
agreed that this was a valid criticism that should
be addressed if it reflected a common concern
or problem among users. We again used the task
mechanism to consult our user-base at large. A
question was added as a task in the game, in the
form of a vote as to whether to introduce this
feature, and exactly what form it should take.
We presented three options: (A) leaving the game
unchanged, (B) players could burn empty planta-
tions to stop them re-growing (as suggested by our
interviewee) and (C) even full plantations could
be burned, which would also destroy all the fruit
that had grown. 17% voted in favour of leaving
the game as it was, while 29% were keen to see
option B and 54% selected option C. The chosen
feature was therefore implemented and distributed
in a new Yoshi version, thus beginning another
iteration in our design process.
On detecting that Helen had installed the new
version, we contacted her again to gauge her
reaction towards the new feature and she replied
positively, agreeing that the version implemented,
although not the design she had suggested, was
the better of the new options. Around the same
time, we included another vote on the new fea-
ture, consulting the opinion of the user-base at
large after they had had a chance to use it. Users
responded with approval, with 94% agreeing that
they liked the new feature. This demonstrated to
us a significant benefit in this iterative approach
of conducting design by engaging with users at
both a micro and macro-scale, and letting the
results of one feed into the other.
DISSCUSSION
As systems like Ego and subsequently Hungry
Yoshi push further into the wild, away from
controlled and constrained settings, the spatial
distribution and temporal duration of users' ex-
periences grows. Space and time play key parts
in understanding the increased levels of uncer-
tainty introduced by the evaluation of Ubicomp
systems used in this way. Responding to this
changing perspective directed us to exploit more
agile evaluative techniques to tackle the rising
uncertainty—i.e., the seemingly uncontrollable
aspects of an experience—that naturally go hand-
in-hand with embedding interactive systems into
users' lives. It is only through increased awareness
Search WWH ::




Custom Search