Fostering Strategy #7: Learners translate between representations

(most recent update 24jan2018) (return to workshop front page)

Contributors: Alexey Leontvey and Tim Shipley

Description:

Learners are given one or more visualization of a phenomenon and are asked to:
  • Transfer information from one visualization onto a different visualization of the same phenomenon.
  • Given one visualization, create a different type of visualization of the same phenomenon, and/or
  • Reflect on and articulate the rules that govern the translation from one form to another form, and the advantages/disadvantages of each.

Examples:

Types of visualizations:

  • Multiple representation of molecular structures (Video on converting Newman to Fischer projections)
  • Map versus block diagram versus cross-section in geology
  • Historic forms of representations versus modern forms
  • Plan view versus elevation view in architecture or engineering
  • Statistical graphs, i.e. pie chart versus bar graph

Pedagogical approaches

  • Show one representation; ask learner to draw a different form of the representation.
  • Show one representation; ask learner to pick the other form of representation showing the same phenomenon from an array of choices.
  • Show multiple representations; ask learner to pick out the ones that all represent the same phenomenon.
  • Card sort; provide multiple representations on cards, and ask learner to sort them into groups that represent the same structure or phenomenon.
  • As a wrap up to the translating activity, ask students to consider several specific situations in which a person wants to communicate a specific aspect of the science to a specific audience. Which form of representation would be best and why?

Affordances of this strategy/what it is good for:

  • This strategy place learners into situations where they have to exercise (for a learning activity) or demonstrate (for an assessment) BOTH their knowledge of the phenomenon under study and their mastery of the representational techniques.
  • Visualizations are models and all models are incomplete. Different visualization techniques leave out different aspects of the referent phenomenon. The process of translating between types of visualizations can help the learner build a complete and coherent mental model of the phenomenon in which higher order attributes are foregrounded by the way all of the different visualizations fit together appropriately.
  • Learners may come to appreciate that there are inevitably limitations to any given representational choice.
  • Learners may come to appreciate that different forms of representation can afford different perspectives or insights about the same phenomenon.
  • Translating fluently between different forms of representation is a necessary professional skill for learners who go on in STEM careers.

Potential pitfalls & challenges:

  • The cognitive load required to simultaneously work with two representations is high, pedagogical strategies such as directing attention and using gestures that support working memory and reduce cognitive load would likely be beneficial.
  • Learners may combine representations incorrectly, merging some pieces from one type of visualization and some pieces from another, to end up with an incoherent union.
  • Learners may learn tricks or shortcuts as to how to make commonly-tested translations from one commonly used form of representation to another, without actually building a functional mental model of the phenomenon. Test-prep oriented instruction may favor these shortcuts. Such shortcuts may help an expert work efficiently, but could be counter-productive for a learner if they by-pass an opportunity to build or improve the learner's mental model.

Emergent insights:

  • One reason that this approach may be so powerful is because structural information about the referent from the visualization has to pass through the mind and get transformed there. When this strategy is working correctly, the learner extracts structures from one visualization or visualization type, builds or draws on a mental model of these structures, and then externalizes their mental structures onto the other type of visualization. The newly-constructed or newly-reinforced mental model remains behind when the exercise is over.
  • The mental model left behind at the end of the exercise may be richer than either of the external visualizations used in the exercise. For example, the mental model may be a three dimensional model, even though the visualizations were 2D. Or the mental model may be dynamic, even though the visualizations were static.
  • One way to think about this strategy is in terms of simultaneous satisfaction of constraints; the mind seeks to build an internally consistent mental model that takes into account the constraints from each separate representation.

Researchable questions:

  • Do the strategies that support learners coordinating complex single object structures in mathematics and chemistry also support all types of representations (e.g., large-scale-multiple-object and dynamic)?
  • Does translating between two representations yield a new form of mental model that has emergent properties (i.e. is greater than the sum of the two representations)?
  • Is translating from one representation into another a general skill, such that being good at it in one domain predicts success in another?
  • Can a progression be developed to support development of expertise at transferring between representations? What practices in fostering proficiency at translating between representations in one domain best support learning transfer to other domains?
  • What, if any, is the relationship between translating between different forms of representations and general cognitive spatial transformation skills such as visualizing transformations of an object and transformations of a scene?
  • Does this activity promote awareness of the limitations of a given representation, and conversely does making students aware of such limitations support learning to translate?

References & Credits:

  • Shah, P., & Hoeffner, J. (2002). Review of graph comprehension research. Educational Psychology Review, 14(1), 47-69.
  • Ainsworth, S. E. (2006). DeFT: A conceptual framework for learning with multiple representations. Learning and Instruction, 16(3), 183–198.
  • Ainsworth, S. E. (2008). The educational value of multiple-representations when learning complex scientific concepts. InVisualization: Theory and Practice in Science Education (pp. 191–208). Springer Netherlands. http://doi.org/10.1007/978-1-4020-5267-5_9
  • Hinze, S., Rapp, D., & Williamson, V. (2013). Beyond ball-and-stick: Students' processing of novel STEM visualizations. Learning and Instruction, 26, 12–21. http://doi.org/10.1016/j.learninstruc.2012.12.002
  • Richland, L. E., & McDonough, I. M. (2010). Learning by analogy: Discriminating between potential analogs. Contemporary Educational Psychology, 35, 28-43.
  • Stieff, M., Bateman, R. C., Jr., & Uttal, D. H. (2007). Teaching and learning with three-dimensional representations. In J. K. Gilbert (Ed.), Visualizations in Science Education (pp. 93-118): Springer.