Observed Human Errors in Interpreting 3D visualizations: implications for Teaching Students how to Comprehend Geological Block Diagrams

Block diagrams are commonly used to communicate three dimensional geological structures and other phenomena relevant to geological science (e.g., water bodies in the ocean). However, several recent studies have suggested that these 3D visualizations create difficulties for individuals with low to moderate spatial abilities. We have therefore initiated a series of studies to understand what it is about the 3D structures that make them so difficult for some people and also to determine if we can improve people’s understanding of these structures through web-based training not related to geology or other underlying information. Our first study examined what mistakes subjects made in a set of 3D block diagrams designed to represent progressively more difficult internal structures. Each block was shown bisected by a plane either perpendicular or at an angle to the block sides. Five low to medium spatial subjects were asked to draw the features that would appear on the bisecting plane. They were asked to talk aloud as they solved the problem. Each session was videotaped. Using the time it took subjects to solve the problems, the subject verbalizations of their problem solving and the drawings that were found to be in error, we have been able to find common patterns in the difficulties the subjects had with the diagrams. We have used these patterns to generate a set of strategies the subjects used in solving the problems. From these strategies, we are developing methods of teaching. A problem found in earlier work on geology structures was not observed in our study, that is, one of subjects failing to recognize the 2D representation of the block as 3D and drawing the cross-section as a combined version of the visible faces of the object. We attribute this to our experiment introduction, suggesting that even this simple training needs to be carried out with students encountering 3D block diagrams. Other problems subjects had included difficulties in perceptually recognizing variations in layer thicknesses, difficulties in recognizing an internal structure from the visible cues on the block walls, difficulties in mentally constructing objects and intersections that were not perpendicular, and difficulties in keeping track of the number of folds of a layer, and thus, the number of intersections of the layer with the bisecting plane. All of these problems suggest that web-based games giving mass practice with these variations in block diagram representations are likely to give any person appropriate skills in their interpretation. The time to complete the drawings and the errors in the drawings were also correlated with quantifiable properties of the diagrams, e.g., number of layers, number of folds in the layers, angle of bisection of the plane, etc. These will be used in further research to organize the training from easy to hard problems following what is known already about mass practice and developing abstracted skill sets. The plan is to also make the training adaptive, that is, to provide practice in those areas where an individual user is having the most problems.