Equipment of many different types is operated by multiple control levers located in close proximity to each other. Selecting the wrong control, or moving the correct control in the wrong direction, has the potential to cause serious injury (e.g. Burgess-Limerick & Steiner, 2006). The objectives of this research were to use a virtual reality simulation of a generic equipment control task to examine: (1) the consequences of shape coding on selection error rate and reaction time; (2) the effect of mirror versus place control layout on shape coding and reaction time when the side to which the controls levers were placed changed; and (3) the consequences of different directional control-response relationships in the horizontal and vertical plane on direction errors and reaction time. Three experiments were completed by a total of 168 participants. Participants were presented with two identical virtual octagonal rods that could elevate/lower, slew left/right, extend/retract and change colour red/blue. One of the images was computer controlled and the other was controlled by four control levers, mounted on a stand, which could be altered in orientation and effects. The task for the participants involved replicating the movement of the computer controlled image by manipulating the four levers positioned on their right or left side. Each participant completed ten blocks of 16 trials, with each trial requiring sequential movements of each of the four control levers. Twenty-four participants completed Experiment One, using horizontal controls and a control-response relationship in which moving each of the horizontal control upwards caused the stimulus to elevate, slew right; extend; or change its colour to red, referred to as control-response relationship 1 (CRR1). Each participant was randomly assigned to start with shape or non-shape coded controls, changing to the reversed coding condition after five blocks. Fewer selection errors occurred in shape coded conditions in the first five blocks, although the difference was not statistically significant. Reaction times decreased as a function of block, and were higher in shape coded trials. There were very few elevation directional errors. Slew and extension directional errors were less likely when there was a spatial correspondence between the movement of the control and the response. Horizontal controls, and the same set of control-response relationships was used in Experiment Two. Each of the 48 participants was randomly assigned to start the experiment with the controls located on their left or right side, whether to complete the experiment with shape or non-shape coded controls, and finally whether to transfer sides to a mirror or place arrangement after five blocks. An effect of coding on selection errors occurred only in the block immediately following a change of sides. However, the effect of coding on selection errors was only seen among participants transferring to a place arrangement, and there were no differences in the number of selection errors between transfer conditions for either coding conditions, nor were there any differences in reaction times between transfer conditions. Reaction times decreased as a function of block, and longer reaction times were measured in the shape coded trials. An advantage of a mirror arrangement when changing sides was suggested when controls were not shape coded, however no advantage was evident when the controls were shape coded. Ninety-six participants completed Experiment Three. Controls were horizontal or vertical, and data were collected for two directional control-response relationship conditions in each orientation: horizontal CRR1; horizontal CRR2 "up" = lower, slew left, retract and blue; vertical CRR1 "away from operator" = elevate, slew right, extend and red; and vertical CRR2 "away from operator" = lower, slew left, retract and blue. Each participant was randomly assigned to complete the experiment with the controls horizontal or vertical, using shape or non-shape coded controls, in the CRR1 or CRR2 condition. The effect of shape coding on reducing the rate of control selection errors in the first block was not significant, however reaction times for correct movements were consistently longer for shape coded conditions. Elevation directional errors in the horizontal CRR1 condition was minimised in all cases, regardless of whether the control movement and the response were spatially corresponding or not. Similar results were seen for the extension control in the horizontal and vertical CRR1 conditions. In other situations, directional errors were less likely when there was spatial correspondence between the movement of the control lever and the response. While only limited evidence was found to indicate that shape coding reduced the probability of selection errors being made in this situation, the observation that the time taken to make correct responses was consistently longer in the shape coded conditions could be interpreted from an information processing, stage model, perspective as indicating that the additional information available in the shape coding was being processed during the decision making stage. It may be that the effect on selection error rate would have been larger if more controls were involved. The results indicate that directional compatibility effects are not as straightforward as the current standards and guidelines suggest. In most, but not all, situations it is important to ensure that control-response relationships correspond spatially.
| Identifer | oai:union.ndltd.org:ADTP/253951 |
| Creators | Mrs Veronica Krupenia |
| Source Sets | Australiasian Digital Theses Program |
| Detected Language | English |
Page generated in 0.0022 seconds