How People Really Use the Mobile Web - Human-Computer Interaction and Innovation in Handheld, Mobile and Wearable Technologies - page 203

in the second field study, the motivation/behavior
framework was at least 96% accurate.
In examining the data from Part Three, we
found similar results for motivations and behav-
iors. There were only 4 instances of “other” moti-
vations (3%) and 6 instances of “other” behaviors
(4%). As in Part Two, the “other” instances that
could not be classified were primarily due to lack
of sufficient detail reported by the participants
to determine the specific context. Thus, for the
survey respondents, the motivation/behavior
framework was at least 96% accurate. Thus we
found the framework sufficiently well defined to
cover the vast majority of mobile Web motivations
and behaviors currently exhibited by active U.S.
mobile users.

contextual data from the interviews. Of those, 7
reported a new location of Gym and 6 reported
a new location of School (leaving 7 unclassified
or 5%). Examples of locations that did not fit
into the original classification scheme or occur
often enough to warrant a new category included
“library” and “church.” In addition, several in-
stances did not include sufficient detail to identify
a physical location for the setting including “at a
meeting” and “at a job interview.”
Of the 167 responses in Part Three, there were
22 reported instances of “other” locations (13%).
We found 17 of those “others” commented that
they accessed the mobile Web “everywhere.” We
omitted those instances because they were too
general to be meaningful, leaving 5 “others” (3%)
that all self-identified their location as “school.”
Therefore, based on the findings from Parts
Two and Three, we added the two new location
categories (see Table 3) of “School” and “Gym”
for physical settings, defined as follows:
Context 11. At School: While in a classroom
setting or on a school campus.
Context 12. Gym: While at a physical fitness
facility.
These additions make the contextual frame-
work at least 95% accurate.

Modifications to Context Schema

Based on the data analysis conducted for Parts
Two and Three, we identified the need to modify
the contextual classification to improve the overall
accuracy of the framework.
In Part Two, participants reported a total of 134
locations in which they accessed the mobile Web.
Note that the survey instrument that participants
used to track their instances of use allowed for
reporting more than one activity per location,
so the number of instances of motivations and
behaviors (177) was higher than the number of
reported locations (134). In addition, an error oc-
curred in some survey pathways where users were
not able to select the applicable location from a
pre-set list and had to choose “other” and specify
a location. This resulted in a higher instance of re-
ported “other” locations. However, the additional
descriptions participants provided, combined with
the additional detail gathered through in-person
interviews, allowed us to accurately identify
instances where “other” locations could be clas-
sified into existing location categories. Out of
134 instances of reported locations, there were
33 reported as “other,” of which 20 could not be
reclassified into an existing category based on

iPhone Motivations, Behaviors,
and Contexts of Use

We applied the framework to understanding
whether a specific device experience made a dif-
ference in how people used the mobile Web. In

Table 3. Re-classifications of physical settings

1. On Transit 6. Recreation Site

2. Walking 8. Other's Home

3. At Work 9. Car Driver

4. Service Facility 10. Car Passenger

5. Store 11. At School

7. Home 12. Gym