Effects of perceptual training of sequenced line movements

Abstract

The effects of consistent training on sequential stimuli were examined in a series of six experiments. A stimulus consisted of three line segments which occurred one after another such that each line segment's orientation and position on the visual channel could be different from the other lines of the stimulus. Subjects practiced these sequences in search and detection tasks for 17 h in both varied mapping (VM) and consistent mapping (CM) conditions. Subsequent experiments compared performance on the two practice conditions with results of previous experiments which used letters and digits as stimuli. The results indicated that the CM stimuli required less attention than the VM stimuli. The differences seen, however, were smaller than those found in previous experiments with letter and digit stimuli. Large differences between the CM and VM groups were found when the individual line segments of the sequential stimuli were manipulated. Shortening and lengthening the line durations had little effect on CM stimuli and large effects on VM stimuli. When the line segments were overlapped with each other, subjects indicated, in a subjective ratings task, that the CM and VM stimuli were perceived differently. Finally, presenting rotated versions of the trained stimuli destroyed any performance advantages the CM stimuli had over the VM stimuli.In the LaBerge (1973) theory of information processing, memory is conceptualized as a network of nodes. When sensory stimulation occurs, the information is analyzed by feature detectors. Output from selected groups of detectors then feed into a higher level pattern code. LaBerge (1976) found that, with extended practice, the processing of novel characters became less sensitive to attentional manipulations. Shiffrin and Schneider (Schneider & Shiffrin, 1977; examined attentionalload in detection and divided attention tasks. The theoretical perspective taken by Shiffrin and Schneider (1977) was similar to that of the LaBerge theory. Briefly, they conceptualized memory as a large collection of nodes that could be interrelated and associated to each other through learning and practice (see Shiffrin & Schneider, 1977, pp. 155-164). When a stimulus is perceived, one of two processes, or a combination of the two processes, is utilized. One of the processes, an automatic process, utilizes a relatively permanent set of associations in long-term storage that is built up through hours of practice. This automatic process occurs when the sequence of nodes becomes active in response to a particular input configuration. On the other hand, a conThis project was supported in part by funds from Office of Naval Research Contract NOOOI4-86-K-0107, w. Schneider, Principal Investigator. Correspondence should be directed to R. Eberts, School of Industrial Engineering, Purdue University, West Lafayette, IN 47907.

Automatic Information Processing and High Performance Skills

“…Thus, the demands on short-term memory to retain pattern elements in these systems can be somewhat less than the load apparently imposed in the Eberts and Schneider (1986) work.…”

“…Previous work with automatic processing of spatial pattern information (e.g., Eberts and Schneider, 1986;Schneider, et al, 1982) produced somewhat mixed results regarding the capability of subjects to fully automatize the processing of such information. Eberts and Schneider (1986), for example, reported several studies concerning the effect of extensive practice on the detection of line segment patterns.…”

“…Eberts and Schneider (1986), for example, reported several studies concerning the effect of extensive practice on the detection of line segment patterns.…”

“…That advantage was maintained when the number of channels to be monitored was increased. However, CM performance was affected by the number of channels to be processed; this led Eberts and Schneider (1986) to suggest that only a partial form of automatic processing had been achieved in the CM condition.…”

“…Eberts and Schneider (1986) conducted a series of experiments with dynamic spatial patterns that required the integration of pattern elements over time. The results of these experiments had produced evidence of partial automaticity.…”