Isometric Projection as a Threat to Validity in the PSVT:R

This work describes a study of graphics interpretation in a common test of spatial ability, the Purdue Spatial Visualization Test: Rotations (PSVT:R). Tests of spatial abilities have long been considered to assess a cognitive ability of manipulating shapes in the mind's eye, and are often used for student assessment in engineering educational contexts. However, researchers are increasingly finding that some widely-used assessment instruments are confounded by other factors that do not necessarily fit within the confines of the traditional understanding of spatial ability. The purpose of this study is to investigate the images used as stimuli in the revised PSVT:R to determine whether they are naturally perceived as three-dimensional forms. The PSVT:R utilizes black and white isometric line drawings of shapes of various complexity, which test takers are supposed to “mentally rotate” into different orientations and select the correct view of the shape from an answer bank. Some researchers have suggested that the isometric shapes on the PSVT:R can look confusing and can be mistaken for flat patterns instead of 2D images. Hoffman's rules related to the principles of generic views explain why many of the shapes would be difficult to interpret as 3D forms. For example, lines that would not be coinciding in 3D coincide in the isometric drawings in the PSVT:R. In our study, we showed a subset of the shapes on the revised PSVT:R to a group of 111 engineering students enrolled in an introductory engineering graphics course, and had them perform a sorting task to determine which shapes they viewed as real 3D shapes and which they did not. Our results showed that at least 19% of the answer bank shapes in the PSVT:R were not viewed as real 3D shapes by most participants, and at least 38% were not viewed as real 3D shapes by some participants. We conclude that the use of isometric views in the PSVT:R presents a threat to instrument validity if the test is to be considered to assess mental rotation ability. Performance on the test is likely impacted by the test-taker's acquired knowledge of the conventions of isometric drawing in the Cartesian coordinate plane and the use of deductive reasoning to relate ambiguous generic views to 3D forms.