The purpose of this arrangement was to improve the validity of results. One of the problems in driver behaviour monitoring is that the awareness of being the object of monitoring changes behaviour, and consequently produces biased results. An experimental design was used so that the baseline measurements comprised all drivers driving with studded tyres. During the first and the second winter, studless drivers were using the test car with studless tyres and those having studded tyres on their own cars were driving the test car equipped with studded tyres respectively. Totally, some 16 OOO kms of driving with instrumented cars on the fixed, preselected routes in variable winter conditions were recorded. Moreover, the studless and studded drivers were driving pairwise the same route in 10 minutes intervals without being aware of each other.
On the other hand, there were clear behavioural changes recorded by the instrumented cars.
In general, unstudded tyre drivers were driving somewhat more carefully on slippery road surfaces and less carefully on bare road surfaces than the control drivers with studded tyres.
On motorways under good driving conditions, speed levels clearly increased during the first winter. Unstudded tyre drivers were driving 5-6 km/h faster than the control drivers using studded winter tyres. On slippery road surfaces, when negotiating steep curves and on approaches to splippery intersections, drivers using unstudded winter tyres were driving 2-4 km/h slower than the control drivers. Those using unstudded tyres were also keeping longer headways with vehicles in tront. While the mean headway for the unstudded drivers was 45 meters, the respective figure for the studded drivers was 34 meters in the platoon travelling at the speed of about 70 km/h. Moreover, unstudded drivers were braking more gently than the control drivers under very slippery conditions, but the difterence in the frequency of using the brakes was marginal between the two groups.
Drivers in both groups also recorded incidents they classified as problematic or risky when driving their own cars. There was not a statistically significant difference between the groups.
It seems, to some extent, that drivers take into account the type of tyres they are using. Also the compensation is not sufficient under all conditions, especially at intersections. Moreover, the increased speed levels of unstudded drivers under good driving conditions may potentially create more accidents when the results are projected to the whole driver population.
TAPANI M�KINEN
Technical Research Centre of Finland