Information Technology Reference
In-Depth Information
processes for learning or memorizing a task, which requires specific experimental
protocols enabling us to measure the influence of the auditory modality on these
same processes.
An example illustrating this approach is the methodology developed around the
device, called Ballancer, designed by Rath and Rochesso [RAT 05]. It consists of a
tangible interface made up of a wooden bar on which a virtual marble rolls, the
movement and position of which the user controls by tilting the bar to one side or
another; the sound of the ball rolling is managed by a computer sound synthesis that
enables us to modify, among other things, the size, weight and shape of the ball as
well as the texture of the bar (i.e. the properties of the ball/bar contact, see
[RAT 05], for more details about the device). The objective of the study is to
measure the influence of auditory feedback on control of the position of the ball by
the user.
The experimental method implemented to this effect is built around a scenario
consisting of immobilizing the virtual ball in a given region of the bar. The
experimental protocol measures improvement in the performance of the “player”
over time by varying the parameters of the game (position and dimensions of the
targeted zone) conjointly with the auditory parameters (properties of the ball and
bar). Globally, the study shows that the auditory feedback improves the use and
especially favors learning of the device (see [RAT 06], [RAT 07], [RAT 08] for
more details about the methodology and results).
8.5.4. Review of evaluation methods of the state-of-the-art studies
This section clarifies the methods used in the works presented section 8.3 in
order to measure the efficiency of the proposed sonification principles (i.e. earcons,
spearcons and scrollbar). The results of the evaluation are also reported and
discussed.
8.5.4.1. Measure of the recall rate
The recall rate here corresponds to the ability of sonified menu users to
remember the item corresponding to a given earcon. Initially each item of the menu
is presented with the associated earcon, then only the earcons are presented and the
user must find which item on the menu they correspond to.
In an initial experiment, Brewster [BRE 98a] tests the aptitude of listeners to
learn earcons associated with a hierarchy with four levels and 27 nodes (see section
8.3.1). He proposes an evaluation procedure in three phases:
- learning phase 1: the participants become aware of the hierarchy and
associated sounds. Each sound is played only once;
Search WWH ::




Custom Search