Biomedical Engineering Reference
In-Depth Information
signal that indicates his/her current position in the interface, as it happens in visual navi-
gation. Spatial cues are uttered by a stereo-audio overview, which simulates the position
of selected nodes within a Cartesian coordinate plane; the information identification and
memorization is strengthened by a verbal feedback voiced by an integrated synthesizer.
Three different sonification models have been tested for WhatsOnWeb. The sound's vol-
ume, pitch, tone, blinking, and grid reference are used to transmit visual features in a
univocal way. In the first model, called the VolumeSonification model, the Euclidean dis-
tance coding for a node compared to a significant reference is rendered through the sound
volume level, whereas the panning is used to strengthen the node detection on the abscis-
sas axis as absolute information. The second model, called the BlinkAndPitchSonification
model, conveys spatial relations through an independent mapping of the two axes of the
Cartesian plane (
x
,
y
), respectively, with the frequency of the sound blinking together with
panning and with the note pitch. Finally, to optimize the graphic representation in terms
of sound, the PanAndPitchSonification model has been created solely considering the pan-
ning for the
x
-axis and the pitch for the
y
-axis.
16.2.3.3 A Usability Evaluation
In this section, we describe an experimental analysis of the re-engineered and sonificated
WhatsOnWeb (Mele et al. 2010). This analysis evaluated the usability of the different visual
layouts: TreeMap, Layered Radial, and Spiral TreeMap.
16.2.3.3.1 Experimental Procedures
Phase 1:
The first phase investigated the usability of the sonificated WhatsOnWeb by an
expert evaluation. Three experts with more than 5 years of experience in the usability evalu-
ation assessed the software by the Nielsen's heuristic list (Nielsen 1994). A user scenario
was performed to test each of the implemented layouts. In particular, the experts' tasks
were to test the usability and the layout differences among the three models of sonification:
PanAndPitch, PitchAndVolume, and BlinkAndPitch. The heuristic evaluation identified a
small set of usability issues with a medium and high level of severity, suggesting that it was
necessary to redesign the layout. Finally, all of the evaluators suggested unification of two of
the sonification models—PanAndPitch and BlinkAndPitch—proposing a new model called
PanAndPitchBlinking. The PanAndPitchBlinking model conveys spatiality through the two
axes of the Cartesian plane (
x
,
y
) by using the panning technique (
x
-axis) and the note pitch
(
y
-axis) and it uses the blink effect to represent the rank order of each vertex.
Phase 2:
After the usability issues reported by the experts were fixed, a usability test was
performed with two groups of participants: four totally blind users and four sighted users
(mean age 28 years, equally distributed by sex). This phase of evaluation aimed at investigat-
ing both the quality of users' interaction with the visual and sonificated WhatsOnWeb and
the users' satisfaction. To achieve these evaluation goals, we used the Partial Concurrent
Thinking Aloud (PCTA) (Mele et al. 2009) and the System Usability Scale (SUS) (Brooke
1996) questionnaires. Each user tested WhatsOnWeb after a clear and essential description
of the task and a preliminary exploration (lasting 3 min) of the layout. The experimen-
tal task, provided by a scenario, consisted of an exhaustive search of the meaning of the
word “Armstrong” by using the WhatsOnWeb search engine. The keyboard navigation
was performed by using either three typologies of layout—Radial, Layered, and Spiral
TreeMap—or the PanAndPitch Blinking sonification. At the end of the evaluation session,
all of the subjects were interviewed about their layout preferences and finally they were
asked to complete the SUS survey.
