Representation Translation with Concrete and Virtual Models in Chemistry

Principal Investigator: 
Project Overview
Background & Purpose: 

This project examines the uses of both concrete and virtual models in promoting representation translation (translation between different diagram formats) in organic chemistry. We will examine whether model use improves students’ ability to translate between molecular representations, what aspects of models are most effective for representation translation, whether instruction can alter the frequency and quality of model use and whether the usefulness of models is different for students with high and low spatial ability.


The research will be carried out at the University of California, Santa Barbara, the University of Maryland at College Park, and the University of Illinois at Chicago Circle.

Research Design: 

The project has a comparative research design and will generate evidence that is descriptive [design research and observational], associative/correlational [quasi-experimental], and causal [experimental and quasi-experimental]. Original data are being collected from students in college-level organic chemistry classes using assessments of learning and observation [videography].

The main task in our experiments will be to translate between alternative diagrammatic representations, an essential skill that all students of organic chemistry must master. We have developed an 18-item instrument to test this ability. Each item of this instrument gives the student one of three diagrammatic representations of a molecule (Newman, Fischer or Dash-Wedge) and asks the student to draw one of the other molecules. We examine the drawings for overall accuracy and specific errors. We also videotape the students while they are performing the diagram translation task, with and without models and code how they use the models and the gestures they make during the representation translation task.

We will analyze the accuracy of students’ drawings of diagrams, other aspects of their drawings, such as the perspective from which they drew the diagram of the molecule, and we will categorize common errors in the diagrams drawn. We will also code and analyze how they interact with the diagrams, for example, whether they pick them up and align them with either the given diagram or the diagram to be drawn and we will code their gestures during the diagram translation.

We will use analysis of variance to compare the accuracy of students’ drawings as a function of experimental manipulations, such as the presence or absence of physical and or virtual molecular models. We will also use regression analyses to examine how use of the models and spatial ability predicts accuracy of students’ drawings.


In pilot studies we have found large variability in how much and in how students use molecular models in translating between alternative diagrams of molecules. Students who were provided with models performed better than control students, who received no models.

Within the model groups there was large variability in how much students used the models. Students who used the models more often had better performance on the diagram translation task. However, several students did not use models when they were provided. Students who are given models but who do not use them perform no better than controls.

Publications & Presentations: 

Stull, A. T., Hegarty, M., Stieff, M. & Dixon, B. (2010). Does manipulating molecular models promote representation translation in chemistry? In A. Goel, M. Jamnik & N.H. Narayanan (Eds.) Diagrammatic Representation and Inference (Proceedings of Diagrams 2010). Berlin: Springer-Verlag.

Other Products: 

The project will develop new instruments for measuring outcomes, as well as new materials for teaching representation translation in organic chemistry.