Posted by: soniahs | November 4, 2010

Exam readings: Using visualizations in science education

These two readings are from the literature on science education, about the importance of visualizations for science. These two authors focus on different topics in the broad area of visualizations and science education.

Barbara Tversky. “Prolegomenon to Scientific Visualizations.” in John K. Gilbert (ed.) Visualization in Science Education, pp 29-42. Dordrecht: Springer, 2005.

Summary: Tversky uses the analogy of scientific visualizations (and viz in general) as maps to aid understanding. Effective maps select important information and even distort it for emphasis (schematize it); abstract relationships are often thought of in spatial terms (e.g., good=up), and this mapping seems to hold meaning/be non-arbitrary. While maps are composed of elements (icons + morphograms [simple schematic shapes that are vocabulary-like: lines, arrows, etc.]) and the spatial relations between them, trees and graphs are composed of elements in an order or subset relationship (metaphorically, not directly spatial.) She gives a few examples of how we interpret maps (e.g., bar graphs suggest containers & make comparisons; line graphs suggest links & convey trends.) Tversky outlines two cognitive design principles: congruence (structure/content of viz should correspond to desired mental structure/content) and apprehension (structure/content should be readily & accurately perceived and comprehended.) She discusses two types of narrative in science viz: structure and process (the latter being more complex to depict.) For her, visual narratives should use analogy as well as present facts. While clarity and brevity are good in many situations, complexity sparks discovery and insight, so there are places for multiple types of diagrams.

Comments: Tversky’s general goal is to make use of schematic cognitive structures in the mind for design. She suggests several strategies for conveying concepts about process, including animations, arrows, and series of diagrams (as well as verbal descriptions.) She feels that animations are poorer in analogy, etc. than comic book format is (b/c animation mainly allows temporal links.) Perhaps interactivity would help address some of this concern about making different types of links.

Links to: Tufte 1, 2 (ideas about simplicity); Zhang & Norman (discussion of distributed cognition)

John K. Gilbert. “Visualization: An Emergent Field of Practice and Enquiry” in Science Education.” In John K. Gilbert, Miriam Reiner, and Mary Nakhleh (eds.) Visualization: Theory and Practice in Science Education, pp. 3-24. Dordrecht: Springer, 2008.

Summary: Gilbert discusses three levels of representation for scientific models: macroscopic, sub-microscopic (e.g., atoms, cells,) and symbolic (qualitative abstractions). External visualizations are used to create internal mental models; a key skill for full understanding is metavisualization, the ability to acquire, monitor, integrate, and extend from visualizations. He suggests two ways of classifying models: purpose (e.g., viz can be larger, smaller, show only processes, etc. of the subject) and dimensionality (e.g., 3-D ball & stick chem. models, 2-D diagrams, 1-D equations.) For metavisualization, people need to be able to understand the representation conventions for different dimensions, be able to translate between modes, construct their own representations, and solve problems using analogy by visualizations. He discusses challenges for mastery of conventions at different levels: macro representations are often taught in labs (they correspond with visible world); sub-micro level creates particular challenges for 3-D structures, but there’s a range of strategies for 2-D structures (e.g., diagrams, animations); and at the symbolic level one issue is differentiating between multiple systems (e.g., for chemical equations.) A key problem is being able to translate between levels (macro-micro-symbolic) or dimensions.

Comments: Traditional approach to mental models (internal vs. external), rather than distributed cognition. A lot of summary of classification systems and lists of skills needed to be visually literate. Goes into some detail about teaching strategies for developing metavisualization skills, which is not my main area of focus (except that multimedia may be good for this purpose.)

Links to: Zhang & Norman (distributed cognition view)



  1. […] & Hassett (discuss the social context of design, rather than design principles); Nersessian, Gilbert (cognition & […]

  2. […] to: Kostelnick & Hassett (rhetorical design); Tversky (different emphasis on maps as data […]


%d bloggers like this: