Picture Recognition: A Human Factors Study of Four Illustration Types

Abstract

Documentation and job aids for the installation and assembly of mechanical apparatus have traditionally made extensive use of figures and illustrations. Evidence suggests that using illustrations to supplement text instructions will be effective only if the illustrations they contain can be quickly and accurately understood. Surprisingly, there has been little research on what makes a good technical illustration. The present study was designed to see what effect illustration type has on performance in a task which primarily involves recognition. Twenty individuals with education or work experience in mechanical drawing and twenty individuals without such experience took part in the study. Each group consisted of equal numbers of male and female participants. Stimulus materials for the experiment were 16 geometric objects of uniformly high colorfulness assembled from two or more Playskool® Bristle Blocks in various combinations and four 35-mm slides depicting four illustration types for each of the 16 stimulus objects. The slides showed the four types of illustration of each stimulus object: a color photograph, a black and white photograph, an isometric line drawing, and a perspective line drawing. Objects were evenly divided among four complexity levels according to a pilot scaling study. Sixteen blocks of slides were presented tachistoscopically on a back projection screen. They appeared in the same orientation and subtended approximately the same visual angle as the objects. In each block, half of the slides showed one of the four illustration types, a target item, and half depicted a non-target item. Before each block, the participant studied one object for 10 seconds. After each slide, the participant was asked to judge from memory whether or not the slide depicted an object just seen. Response time to each slide as well as ‘yes’ and ‘no’ responses were recorded. Results showed that all the interesting effects occur for target items. Of greatest interest is the significant effect of illustration type on recognition time. Photographs resulted in faster recognition than line drawings and color photographs were faster than black and white photographs. We also found that increasing object complexity increased the time to recognize illustrations of that object.