The missing piece in block design tasks: Resolving performance differences elicited by designs with identical stimulus parameters.

Abstract

Block design stimulus variables (perceptual cohesiveness, response uncertainty, partial components) have been previously shown to impact test performance. However, relevant studies suggest designs that "look" different should elicit similar performance, where these stimulus parameters are identical. The current study sought to explore other, previously unidentified parameters. To control for known parameters, designs with no perceptual cohesiveness and maximum uncertainty were presented-rotated in increments of 90° and with block edges either cued or uncued-to 65 university undergraduates. Even with cohesiveness and uncertainty constant, performance differences between designs that "look" different were elicited. Post hoc, we hypothesized two possible contributory parameters and coded the designs accordingly. Regression analysis suggested redundancy of adjacent blocks (i.e., coherent regions within the design) appeared to reduce the actual uncertainty of some designs. On the other hand, local and global symmetry, as defined by Palmer (1991, The Perception of Structure, pp. 23-39, The American Psychological Association), did not seem to predict block design performance above and beyond the effects of coherent regions. The advantages of a more complete understanding of stimulus contributions to test performance are discussed.