Asia-Pacific Forum on Science Learning and Teaching, Volume 11, Issue 1, Foreword (Jun., 2010) John K. GILBERT The role of visual representations in the learning and teaching of science: An introduction Previous Contents Next

Metavisual capability

Importance

All the ultimate explanatory entities in science are too small to be seen with the naked eye. Consequently, a full understanding of a scientific phenomenon  -the possession of ‘expert scientist’ status in it – requires an individual to be able to mentally construct, to move between, the three types of representation: macro, submicro, symbolic. This capability has been described as metavisualization and as:

---- involving the ability to acquire, monitor, integrate, and extend from, representations’ (Gilbert 2005)

A key issue for science education is how to support students in getting to this level of performance. The first issue is what, in detail, does ‘expert performance’ involve?

Criteria for the display of metavisualization

Metavisualization is shown by a number of capabilities i.e. that of being able to:

• Demonstrate understanding of all the codes of representation for all the modes of representation and their constituent forms. As has been shown above, these codes are complex and many, perhaps most, have not been coherently expressed in the literature. For example, even those of the form collectively called ‘diagrams’ are diverse.

• ‘Translate’ between the various modes for a given model. For example, the school-level model of the ‘ideal gas’ can be expressed in concrete/material, in diagrammatic, and in mathematical equation, modes and hence forms. A full understanding of the model requires a student to be able to readily access and to appreciate the explanatory scope of each of them.

• Construct a representation of a model for a given purpose. For example, if students wish to show the function of the arterial/venous system, then a ‘circuit diagram’ is the most appropriate form.

• Use a visualization to make a prediction. The only way that the scope and limitations of a given model of a phenomenon can be established is by making and testing predictions about its behavior. Such predictions are made by imaging possible properties on the basis of a representation.

• Use an existing visualization as the source of an analogy with which to represent an apparently very different phenomenon. For example, the Bohr model of the atom seems to be an analogy based on the heliocentric model of the solar system of planets.

Cognitive psychologists debate whether a full display of  visualization - what I have termed  metavisual capability (Gilbert 2005) - is innate, or the result of suitable experience, or an interaction between the two. This is a manifestation of the ‘nature or nurture’ debate about human capabilities (Newcombe 2005). It does seem that males are better at visualizing then females and that, although the differences can steeply decline under the impact of suitable training, they never entirely disappear (Halpern 2005). Whatever the causal mechanism behind the development of visualization at any time, there seems little doubt that students do show a range of quality of performance which does impact directly on their performance as scientists.

Students’ problems: their nature and origin

Students’ metavisual capabilities have been mainly investigated in the field of chemistry(Wu 2004), perhaps because it is there that visualization has the most readily obvious saliency as a determinant of attainment. Several general problems have been identified. First, the conventions of representation for the submicro and symbolic types are often not understood (Kosma 1997). This seems hardly surprising, given that all the codes are not systematically taught. Second, students find the interpretation of a phenomenon presented in the symbolic type difficult to interpret into the corresponding submicro type (Krajcik 1991). Perhaps the complexity of navigating through the intricacies of the symbolic type provides too great a cognitive load and requires too much working memory for this ‘translation’ to be readily achieved. Even where a particular mode can be focused on, moving between the sub-forms of representation are found problematic (Keig 1993). Lastly, when all these problems are absent, whilst representations of the macro type can be both understood and produced, linking a given macro representation to the corresponding submicro and symbolic types is found difficult (Ben-Zvi 1988).

The problems suggest that specific teaching interventions are needed if ‘expert chemist’ status is even to be approximately approached.