|
|
While much research has been conducted in
multimedia learning in the discipline of instructional technology, a
similar line of inquiry exists in the discipline
of computer science known as algorithm visualization. Algorithm
visualization is concerned with finding novel ways to illustrate
complex algorithms visually. Algorithm visualization for
pedagogical purpose is generally implemented as a graphical animation
and are sometimes supplemented with narration to facilitate the
learning process. There are number studies that have been
conducted using the underlying theories of cognitive constructivism,
individual differences, dual-coding, and epistemic fidelity.
Hundhausen, Douglas, and Stasko (2002) conducted a
systematic meta analysis of existing research studies on algorithm
visualization learning effectiveness.
The corpus of studies included a number of independent variables
including form (static versus dynamic), content, persistence, scope,
abstraction, interaction and many more pertinent manipulatives (Hundhausen et al., 2002).
A summary of the results of 24 independent experiments shows that
46% of the experiments showed statistical significance in favor of
algorithm visualization as opposed to a non-algorithm visualization
configuration. Forty-two percent were not significant, and the
remaining were classified as uncertain or inconclusive.
One of the primary conclusions of the meta analysis conducted by Hundhausen,
Douglas, and Stasko is that the experimental research studies indicate
that algorithm visualization is most effective based on how a student
uses it, rather than what is seen when the technology is being used.
This implication means that well-designed multimedia
visualizations may not be fruitful if the instruction is not carefully
designed to support their use. In the next section, you will see
a few applications of algorithm visualization.
|