Diagrams and Descriptions in Acquiring Complex Systems Julie Heiser and Barbara Tversky (jheiser,{bt}@psych.stanford.edu) Department of Psychology Jordan Hall, Building 420 Stanford, CA 94305 USA Abstract Complex systems such as a car brake, circulatory system, or legislative system can be conveyed by language or diagrams. Such systems can be presented from structural or functional perspectives. In three experiments, we examine communicating structure and function of mechanical systems (bike pump, car brake, pulley system) by text and diagrams in relation to mechanical ability. By adding arrows, structural diagrams can be enriched to convey functional information. Inferring structure from function was easier than inferring function from structure. Participants high in mechanical ability outperformed low participants except when text perspective matched question perspective. Those with low mechanical ability are at a disadvantage, especially for inferring function from diagrams. Comprehension of complex systems depends in sensible ways on perspective, medium, and ability. Conveying Complex Systems When we learn about a new digital camera, attempt to troubleshoot a broken-down car, or try to understand a new finding in neuroscience, we need to understand a complex system. Despite the ubiquity of contact with complex systems, understanding them or interacting with them can be frustrating. The frustrations are due not only to the complexity of the systems but also to the inadequacy of instructions and explanations. Effective explanations of complex systems have a complexity of their own. Effectiveness depends on the perspective of the information to be conveyed, on the medium of conveying the information, and on the ability and expertise of the learner. Some of these complex interactions have been examined in previous work, though finding generality in the conclusions has been elusive (e.g., Hegarty, et al., 1990; Hegarty, et al., 1993; Mayer & Gallini, 1990; Morrison and Tversky, 2000). More clarity may be achieved by an analysis of the information to be conveyed relative to characteristics of the media and to qualities of individual differences. Information about complex systems is of two types: structural information, the configuration of parts or topology of the system, and functional information, the sequence of operations and outcomes. The configuration of parts has a spatial or metaphorically spatial structure, and the sequence of operations has a temporal, causal structure. The primary media for conveying complex systems are language and diagrams. With an increasing emphasis on visual displays of information, we found it important to investigate the success that diagrams have in comparison to text in communicating this information. Structural information should be effectively conveyed in diagrams because diagrams use elements and relations in space to convey actual topology. Furthermore, arrows indicating the sequence of operations can be added to a diagram to convey functional information. There are conflicting results on the relations between medium and ability. Some studies show that people with low ability benefit from diagrams and others show that people with low ability have difficulties extracting information from diagrams (Hegarty 1992; Larkin & Simon, 1987; Mayer, 1989) An analysis of information perspective may reconcile these conflicting findings. In particular, low ability participants or novices may be able to extract structural but not functional information from diagrams. Functional information must be inferred from diagrams, in contrast to structural information, which is explicit. Three experiments examine the interactions of medium, content, and ability in the comprehension of complex systems. We use three systems that have been used with success in previous literature, a pulley system (adapted from Hegarty & Just, 1993), car brake and bicycle pump (both adapted from Mayer & Gallini, Experiment 1: Descriptions from Diagrams Diagrams of complex systems are excellent for conveying structural information as they use space and the elements in it to convey real or conceptual elements and the relations among them. Adding arrows may facilitate conveying functional information as arrows indicate the temporal sequence of operations. Participants were asked to describe what is depicted in a diagram of a complex system, without and with arrows.
[1]
Joan K. Gallini,et al.
When Is an Illustration Worth Ten Thousand Words
,
1990
.
[2]
Barbara Tversky,et al.
The (in)effectiveness of animation in instruction
,
2001,
CHI Extended Abstracts.
[3]
M. Just,et al.
Constructing mental models of machines from text and diagrams.
,
1993
.
[4]
Herbert A. Simon,et al.
Why a Diagram is (Sometimes) Worth Ten Thousand Words
,
1987,
Cogn. Sci..
[5]
R. Mayer.
Systematic Thinking Fostered by Illustrations in Scientific Text
,
1989
.
[6]
P. David Pearson.
Handbook of reading research.
,
1990
.
[7]
Masaki Suwa,et al.
What do architects and students perceive in their design sketches? A protocol analysis
,
1997
.
[8]
Herbert A. Simon,et al.
Why a Diagram is (Sometimes) Worth Ten Thousand Words
,
1987
.
[9]
Julie Bauer Morrison,et al.
Does animation facilitate learning? An evaluation of the congruence and equivalence hypotheses
,
2000
.