Testing the principle of orthogonality in language design

  • Authors:
  • Edward M. Bowden;Sarah A. Douglas;Cathryn A. Stanford

  • Affiliations:
  • Department of Psychology, University of Michigan-Flint, Flint, MI;Department of Computer and Information Science, University of Oregon, Eugene, OR;Hewlett Packard, Everett, WA

  • Venue:
  • Human-Computer Interaction
  • Year:
  • 1989

Quantified Score

Hi-index 0.00

Visualization

Abstract

Research has shown that organization plays an important role in memory. This study applies these findings to the design of a command language. The concept of orthogonality was used to maximize the internal organization of a text-editing command language. In Experiments 1 and 2, this orthogonal language was compared to an organized, but nonorthogonal, and an antiorganized language on measures of predictability, recall, and performance. Subjects in the orthogonal language condition performed better than subjects in the other conditions on all measures. In a third experiment, steps were taken to eliminate possible confounding effects of mnemonics. The orthogonal language was compared to the organized language on measures of recall and performance. Even without the aid of mnemonics, subjects in the orthogonal language condition performed better on the recall test than subjects in the organized language condition. In addition, analysis of keystroke data revealed that subjects using the orthogonal language required less time to think of appropriate commands to accomplish their tasks. General steps necessary to design an orthogonal language are discussed.