| |
This research project
supported by the Engineering and Physical Research Council is investigating
the impact of distributing information over Multiple External Representations
(MERs) on learning and problem Solving, and follows on from previous work on
MERs (for more information see the
Multi-Representational Learning Environments homepage).
Research on the benefits of MERs for learning and problem solving has provided evidence that demonstrates that they can confer learning benefits if people manage the complex tasks associated with their use. However, much of this research has concentrated upon presenting representations that express the same information but do so in different formats (informational equivalence, computational non-equivalence). Yet a common application of MERs is to distribute the information over a number of separate representations, each one displaying a subset of the total information.
The four key objectives
of this research are:
1) To explore how people learn to perform a complex task when
instructions are provided either in a single complex representation or many
simpler representations. 2) To investigate if these instructions are better presented
in multiple representations of the same format or multiple representations of
different formats. 3) To examine how changing the redundancy between representations
influences learning by providing multiple redundant representations or multiple
distinct representations. 4) To consider if transfer across related tasks is facilitated
when instructions are presented either in one complete representations or many
individual representations. To test the effects
of varying the distribution of information on learning and problem solving the
Alchemist's Factory task was designed and implemented in Macromedia Director.
The task requires the learner to create gold. They must buy or create the chemicals
they need, which they must then manipulate in specific ratios in the appropriate
apparatus, which must be set to the correct settings. The software allows the
instructions for operating the factory to be provided in a variety of ways.
The two factors important for this research are the amount of information per
representation and the format of the representation. The Alchemist's Factory
allows us to manipulate these factors when representing the instructions to
the learner. Experiment one explored
if the way the instructions for operating a complex device are represented influences
problem-solving and learning. For this experiment the instructions describing
how to make gold were presented either in one complex representation or 4 simpler
representations of the same form, which each only presented a subset of the
necessary information. The forms of representation used in this experiment were
tables, diagrams and text. The hypotheses were: 1) Does the form of representation
influence efficiency and effectiveness of problem solving? 2) Does the complexity
of representation influence efficiency and effectivness of problem-solving?
3) Does the form and complexity of representation interact to influence efficiency
and effectiveness of problem-solving? 4) Does the form and complexity of representations
interact to influence learning?The Alchemist's Factory
Experiment 1
Click on the links below to see example screen shots of the representations used in experiment 1
The results from this study have shown that learners found the optimal solution more often when given the instructions as text (single or multiple) or as a single complex representation. However it was found that the single text representation was associated with significantly slower performance in the task. The learners recalled more about the task with text representations, irrespective of whether they were given the text as a single complex representation, or in four simple representations. The experiment has confirmed that representations that display instructions in such a way as to increase the cost of operating with them can paradoxically lead to better performance.
We are currently carrying out further experiments where we are investigating if instructions are better presented in multiple representations of the same format or multiple representations of different formats, and if transfer across related tasks is facilitated when instructions are presented either in one complete representations or many individual representations.
Ainsworth, S.E. & Peevers, G.J. (2003) The Interaction between Informational and Computational Properties of External Representations on Problem-Solving and Learning. 25th Annual Conference of the Cognitive Science Society.
