Percepiual and Motor Skills, 1975,40, 167-170. @ Perceptual and Motor Skills 1975

EFFECT OF SPEED ADAPTATION O N PERFORMANCE IN A SELF-PACED TRACKING TASK STEPHEN MILLER AND BILHA WEIDER The City Uniuwsity, London Summary.-12 Ss were required to perform a self-paced tracking task designed to simulate 'hazardous' road conditions. Those 6 Ss who had previously been exposed to a simple, high-speed version of the task travelled faster and made more errors on the hazardous task than 6 control Ss who performed [he hazardous task throughout the experiment. The findings supplement earlier studies of the effect of speed adaptation on the judgment of speed and suggest that, in addition to perceptual distortions, an acquired motivation for speed may be an important factor in the behaviour of drivers leaving high-speed roads.

After travelling at high speed, drivers tend to underestimate their subsequent speeds when required to decelerate. This effect has been calibrated by Denton ( 1966, 1967), and Schmidt and Tiffin ( 1969) have shown that the underestimation of speed increases as a function of the time spent travelling at the higher, adapcation speed. Since drivers make limited use of their speedomecers when changing speed (Denton, 1965), ic has been suggested that distortions in the subjective judgement of speed may be an important factor in accidents, particularly at those locations where speed-adapted drivers are required to decelerate, e.g., a motorway exic (Denton, 1966; Schmidt & Tiffin, 1969). This type of argument presumes that speed estimation, or factors closely related to this, determine how fast a driver chooses to travel. It may be, however, that more complex factors influence the control of speed, such as the driver's perception of momentum and 'safety' or motivat-ional states associated with prior exposure to high speed. If such factors are operative, then our knowledge of distortions in speed estimation following adapcation cannot be used to predict the driver's actual behaviour. Rather, it becomes important to examine the direct effects of speed adaptation on the driver's choice of speed and level of risk taking in the post-adaptation period. The aim of this experiment was to demonstrate such effects in the laboratory using a self-paced tracking cask bearing some resemblance to driving. An attempt was made to simulate the conditions under which a driver emerges from a high speed road or motorway and continues to travel on a more hazardous route. It was predicted chat speed adaptation would influence both S's speed of performance and error race in the post-adaptation tracking cask.

METHOD Materials The track or 'roadway' consisted of two parallel lines 1/2 cm apart drawn in red ink on paper tape. The tape was drawn downwards past a window 2 %

168

S. MILLER

&

B. WEIDER

cm wide and of variable height. The speed of the tape could be varied by a foot-pedal control within the range 0 to 45 cm per second. A pen recorder was used to mark off the tape at 1-sec, intervals, so that the speed could be subsequently calculated for various sections of the track. The curvature of the track was variable, consisting of alternating right and left bends. In the 'simple' condition, designed to produce speed adaptation, the bends were very slight {extending for 20 to 25 cm with a lateral displacement of up to 1 cm from the centre of the tape) and preview of the oncoming track was set at 20 cm. In the 'hazardous' condition there were sharper bends ( 4 to 10 cm in extent with up to 1 cm lateral displacement), and preview of the track was reduced to 6 cm. Subjects Ss were 12 male engineering students at the City University. range was 20 to 22 yr. and all Ss had driving experience.

The age

Desiglt and Procedure The 12 Ss were divided randomly into two equal groups. The experimental group performed the simple (speed adaptation) task for 20 metres of track (Stage 1 ) followed by the hazardous task for a further 5 metres (Stage 2 ) . The control group performed the hazardous task throughout; at Stage 1 they did so for 3 min. (an approximate match to the duration of Stage 1 in the experimental group), and at Stage 2 the controls continued with the hazardous task for a further 5 metres. Each S, tested individually, placed a pencil at the foot of the window and was instructed to follow the curvature of the track by making only lateral movements of the pencil. He was told that there was no time limit and chat the object was to control the speed of the tape so as to ensure that the path of the pencil remained inside the track. The tracking began at a pre-set speed of 2 cm per second, but thereafter was under S's control. For all Ss the tape was stopped for 10 sec. between Stage 1 and Stage 2 to allow for the reduction in preview of the track in the experimental group. The experimental Ss were alerted to this change and the associated increase in track curvature in their initial instructions.

RESULTS Stage 1 Table 1 shows the mean speed of experimental and control Ss at three successive phases of Stage 1 (first, middle and final third). A two-way analysis of variance with repeated measures on one factor was conducted. As expected the experimental Ss were able to maintain much higher speeds on the simple (speed adaptation) task than could the control Ss on the hazardous task (F = 63.4, df = 1/10, p < ,001). The effect of time on task was insignificant ( P = 2.49, df = 2/20, p > .1) as was the interaction between this factor and the

SPEED ADAPTATION IN SELF-PACED TRACKLNG TABLE 1 MEAN SPEEDOF EXPERIMENTAL AND CONTROL SUBJECTS DURING FIRST, MIDDLE,AND FINAL THIRDOF TASK,STAGE1 (ns = 6)

Time on Task

Experimental Group

Conuol Group

First third Middle third Final third

experimental/control effect ( F = 2.31, d f = 2/20, p > . I ) . The experimental group also produced a substantially lower mean error rate than the control group (0.3 and 4.6 errors per metre, respectively). In this context an error was defined as a movement outside the track boundaries. Stage 2

Table 2 shows the mean speed and error rates of experimental and control subjects at Stage 2, viz., when both groups were performing the hazardous task. The duration of this task was too short to permit reliable analysis of possible trend effects through time. The differences between experimental and control group means on both measures were significant when subjected to the MannWhitney U test ( p = ,001). TABLE 2 MEAN SPEED AND

ERROR RATESOF

Measure

I

EXPERIMENTAL AND

CONTROLS U B J B ~ SSTAGE , 2

Experimental Group

Conuol Group

9.9

5.1 4.1

Speed (M per second) Errors per metre

7.8 DISCUSSION

How is a driver's choice of speed influenced by prior exposure to continuous high-speed travel? This question has received surprisingly little direct attention, although indirect evidence is available from studies which demonstrate the underestimation of speed following adaptation (e.g., Schmidt & Tiffin, 1969). In this experiment we made an initial attempt, in the laboratory, to examine the direct effects of speed adaptation on subsequent performance in a self-paced tracking task. The results showed that the experimental (speedadapted) Ss travelled almost twice as fast as the control Ss on rhe same task, and produced almost twice as many errors. Furthermore, the experimental group made virtually no reduction in speed when transferred to the hazardous, postadaptation task despite initial instructions which emphasized accuracy rather than speed. Further research is needed to determine the extent to which such effects occur under real driving conditions where errors are associated with significant

170

S. MILLER & B. WEIDER

costs for the driver. Nevertheless, our results suggest that adaptation effects cannot be explained purely in terms of distortions in speed estimation; such an explanation would predict misjudgements in the extent of deceleration following speed adaptation (Denton, 1967), but can hardly account for the zero deceleration observed in the experimental group. Rather, it would seem that prolonged exposure to high speed induces a tendency to continue at that speed despite increases in the risk of errors. This tendency might best be described as an acquired motivation for speed and may be related to what Welford ( 1968) describes as 'the predominance of initial experience.' Further study of the time course of this effect, its role in real driving conditions and methods of counteracting the motivational consequences of speed adaptation is needed to supplement what is already known about the perceptual component of speed adaptation. REFERENCES DENTON, G. G. The use made of the speedometer as an aid to driving. Road Research Laboratory, Laboratory Note No. LN/861/GGD, Harmondsworth, 1965. DENTON, G. G. A subjective scale of speed when driving a motor vehicle. Ergonomicr, 1966.. 9.. 203-210. DENTON, G. G. The effect of speed and speed change on driver's speed judgement. Road Research Laboratory Report LR97, Harmondsworth, 1967. J. Distortion of driver's estimates of automobile speed as a SCHMIDT,F., & TIFFIN, function of speed adaptation. Journal o f Applied Psychology, 1969, 53, 536-539. WELFORD, A. T. Fundament& of skill. London: Methuen, 1968. Accepted November 1,1974.

Effect of speed adaptation on performance in a self-paced tracking task.

12 Ss were required to perform a self-paced tracking task designed to simulate 'hazardous' road conditions. Those 6 Ss who had previously been exposed...
142KB Sizes 0 Downloads 0 Views