THE USABILITY OF MUSIC THEORY SOFTWARE: THE
ANALYSIS OF TWELVE-TONE MUSIC AS A CASE STUDY
Tuukka Ilomdki
Sibelius Academy
tuukka.ilomaki@iki.fi
ABSTRACT
Computer applications are an everyday tool for music
analysts, composers, and music theory students. While
these applications are a welcome tool to be used in the
classrooms and research labs, their effectiveness could
be improved by focusing on their usability.
The usability of a user interface can be evaluated and
even measured with respect to the goals of its users. In
order to demonstrate the evaluation of a user interface, I
present an experiment in which the efficiency of user
interfaces is assessed in the context of three scenarios or
''use cases.1"
Based on the experiment, I discuss some basic
principles of usability theory, such as affordances,
minimization of navigation, error handling, immediate
feedback, and data visibility. The evaluation of these
principles suggests some new types of music theory
applications.
1. INTRODUCTION
The popularity of personal computers has given impetus
to the development of music theory software. It is now
commonplace to employ computer applications for
learning music theory fundamentals, generating row
matrices, analyzing pitch-class sets and K-nets, and so
on. While many of these applications undoubtedly serve
their purpose, it is worthwhile to investigate how their
usability could be improved. Instead of merely
examining what these applications can do, I will here
explore how they do it.
Computer applications can be evaluated from a
number of perspectives. In research contexts it is usually
only required that the application produces the correct
result in a reasonable time. If a broader audience is
targeted, however, the user's experience depends not
only on whether the application returns the right result,
but also on the ease with which the user can interact
with it.
Virtually everybody agrees on the maxim (or truism)
that a user-friendly user interface makes software more
usable. There is, however, significantly less agreement
on what constitutes a user-friendly user interface. For
example, we should not be satisfied with a popular but
trivial generalization that a graphical user interface is
automatically better than a command line interface.
Unfortunately, a user-friendly user interface seems more
often to be an issue for the marketing department than
for the research department.
In this paper I argue, based on usability theory,
against a deeply rooted myth that the quality of a user
interface is a matter of opinion. In order to demonstrate
the evaluation of a user interface, I will offer a hands-on
experiment in which I present two user interfaces and
measure the time that it takes to finish a given task. The
reading of a stopwatch is certainly not a matter of
opinion. According to Nielsen [4], "Clarifying the
measurable aspects of usability is much better than
aiming at a warm, fuzzy feeling of user friendliness."
Furthermore, the evaluation of some of the principles of
a good user interface suggests some appealing directions
for music theory applications.
2. USABILITY THEORY
Usability theory (or computer-human interaction) is a
relatively young branch of computer science that has its
roots in cognitive psychology. Even if it does not yet
belong to mainstream computer science, the awareness
of its importance seems to be growing. While some of
the principles of usability theory are gradually making
their way to the mainstream software applications, the
underlying research is less known.
In order to critically discuss usability we need to
define what we mean by it. According to The
International Organization for Standardization usability
is "the extent to which a product can be used by
specified users to achieve specified goals with
effectiveness, efficiency and satisfaction in a specified
context of use" (ISO 9241-11). The crux of this
definition is that it does not define usability per se; it
only defines usability with respect to the goals of the
users. Hence, in order to design a good user interface
and evaluate its usability we must first discover the
goals of the potential users. With respect to usability,
the "features" of a computer application are
insignificant. The only thing that matters is how
efficiently the users can achieve their goals. (This does
not mean, however, that some user interface designs
would not be terrible independently of the user's goals.)
Nielsen [4] prefers the term "usability" to the vague
umbrella "user friendliness." He divides usability into
five categories: learnability, efficiency, memorability,
error handling, and user satisfaction. The relative
importance of these depends on the type of application -
we can accept a steeper learning curve on software
controlling a nuclear power plant, than on an application
that prints a row matrix. Each of these categories can be
evaluated. In the following I will focus mostly on
learnability, efficiency, and error handling. Learnability
denotes the ease with which a first time user can interact
with the user interface, and efficiency denotes the
93
0