Information Technology Reference
In-Depth Information
Method
We selected three computer science courses of different years of study: a second
year introductory lecture on sorting algorithms, a third year lecture on computational
engineering and a lecture for graduate students on ubiquitous computing. A total of
29 students (5 female, 24 male) were recruited among the attendees of these lectures.
Their semester of study averaged out at 5.6 (
SD
1). Participation was voluntary
and no compensation was given. To avoid bias, none of the students was personally
known to us nor did they attend one of our courses.
Before the lecture, each participant received an Anoto pen and a digital paper
printout containing the slides of the current lecture. Each A4 sheet contained two
slides on the left-hand side and empty areas for taking longer notes to the right.
The user was trained for three minutes on how to make annotations and to classify
them using the digital pen. The task during the lecture was to make annotations on
the printout the same way the participants usually do. In addition, we asked them
to semantically classify and/or to share annotations using the paper buttons when
it seemed appropriate to them. All these activities were digitally captured by the
CoScribe system for subsequent analysis.
After the lecture, the participants filled in a standardized questionnaire for quan-
titative feedback on the paper-based user interface. The questionnaire contained 25
closed and open questions related to the printed interface, to the digital pen, to the
lecture and to personal information. Finally, in each of the three sessions, we con-
ducted a semi-structured group interview with three to six participants. The goal was
to gather additional qualitative insights into benefits and shortcomings of the current
design and to brainstorm about which further functionality would be helpful. These
interviews were videotaped and varied in length from 35 to 70 minutes.
In the statistical analysis of the questionnaire data, we investigated correlations
between Likert-scale items which were five-point scaled and performed χ
2
-tests and
t-tests to identify significant group differences. All these tests were based on a level
of significance of 95 %. Moreover, we analyzed the annotations that were made
during the lecture. This comprised a statistical analysis of frequencies, positions
and types of annotations.
=
3
.
Results and Discussion
Document Annotation
Participants considered digital pen and paper for annotat-
ing lecture slides to be easy to use. All users reported that annotating printed lecture
slides with the digital pen worked reliably and as they had expected. In the question-
naire, the participants judged the use of CoScribe about as distracting as traditional
pen and paper (
M
=
2
.
7 on a scale ranging from 1=more distracting to 5=less dis-
tracting,
SD
=
.
7,
N
=
29), but significantly less distracting than using a laptop
(
M
001) (see Fig. 5.9). This
indicates that the novel techniques can be well integrated into the existing ecology.
=
4
.
5,
SD
=
1
.
0,
N
=
29) (
T
=
9
.
0,
df
=
26,
p
<.
