Civil Engineering Reference
In-Depth Information
• Subjective evaluation can also be used. One can ask the driver to use a scale of, say, 1
to 5 to continuously rate the difficulty or stress of the traffic conditions as they keep
driving. This measure usually provides a reasonably good assessment of stress.
• Excretion of stress hormones is more useful for assessing long term stress than for
temporary stress in traffic. For example, this measurement can show the cumulative
effects of stress from a day of hard work.
Several other potential measures may be considered, but typically heart rate variability,
galvanic skin response, and subjective evaluation are the most appropriate in the traffic
scenario.
As mentioned above, the independent variables in the traffic environment are design-
related, such as road illumination, the width of the road, and traffic density. These
variables matter most to traffic engineers in designing roads. For example, to increase
traffic capacity, a traffic engineer must decide between options, such as making the road
wider or building more traffic lanes. The latter option is more expensive, so there are
several trade-off decisions to be made.
EXAMPLE
Helander (1978) measured galvanic skin response (GSR) during different types of traffic
events. Fifty test persons drove an experimental vehicle on a rural narrow road which was
24 km long. An experimenter was sitting in the passenger seat next to the driver, marking
every traffic event on a keyboard as it occurred. The traffic events were recorded on a
digital recorder together with galvanic skin response and variables describing vehicle
behavior, such as brake pressure and steering wheel angle.
The average brake pressure and the average GSR were calculated for 15 traffic events
(see Table 3.1). The traffic events were then ranked in order, so that the event with the
greatest average brake pressure obtained rank 1 for brake, the traffic event with the
greatest average GSR obtained rank 1 for GSR, and so forth. From Table 3.1 it is clear
that in most cases the rank orders of brake pressure and GSR follow each other perfectly.
Spearman rank order correlations were calculated between brake and GSR (Siegel and
Castellan, 1988). Before performing the statistical analysis, we noted that for events 15
and 13 the driver was passing or being passed, and there was no reason why he should
brake. If these events were excluded, the rank correlation coefficient between brake
pressure and GSR is R
s
=0.95, which is statistically significant with p<0.0001. The
significant/? value means that the correlation coefficient could have been obtained by
chance only in one case out of 10,000. The finding is therefore conclusive.
We can conclude from this study that traffic events which require drivers to brake will
also be perceived as stressful. To design a less stressful traffic environment one could try
to construct traffic environments where the traffic flow is smooth and reduces the need
for braking. The traffic flow on freeways and highways with several lanes and good
illumination is usually much smoother than on rural country roads and requires less
braking. Freeways are also about four times as safe as rural country roads
