Factors related to road users
Riders or rider groups who ride more kilometres are more exposed to the dangers of road traffic and will usually have more accidents. Therefore, when studying other factors that may contribute to accidents, the accident figures have to be corrected for kilometres travelled. Such a study can be based on national statistics. However, reliable and detailed data on kilometres are scarce. Other designs to study rider factors that may contribute to accidents are case control and questionnaires.
There is a problem when the results of all these studies are compared: the more information on rider characteristics in the study, the more accidents with no or less severe injury.
In all studies the age of the rider has been found to be important. Young riders have much higher accident rates, even if corrected for (lack of) experience. This has to be explained as a result of age related psychological factors. Unfortunately there is no information on accident rates for the very young ages at which mopeds or motorcycles are allowed in some countries. Trends in risk by age have been different for different modes over the last 20 years.
Experience as a rider is also found to be related to accident rates of motorcycle riders, although not in all studies. There are different types of experience: years of riding, recent or frequent riding, familiarity with a specific motorcycle and familiarity with specific conditions. All types of experience contribute to a lower accident rate to some extent. Riding experience may not be sufficient to overcome the extra problems created by adverse weather conditions during winter.
Several studies emphasise the high accident rate of riders of a sports motorcycle. Some studies indicated a higher accident rate for recreational riding which may have to do with a combination of different conditions and a different riding style. Other groups of riders with a relatively high accident rate were only found in single studies: winter riders in Great Britain, urban riders in New Zealand, inexperienced riders of custom bikes in the Netherlands.
Other factors influencing PTW safety are psychological factors, influencing rider motivation and riding style, and the increasing numbers of older motorcyclists.
There are several violations of the law which are common to PTW riders and may contribute to accidents such as speeding, drinking, riding without a valid licence, tampering of the engine. The proportions of riders with such violations vary from country to country.
Rider safety is also influenced by the perceptions of other road users. Other road users appear often to fail to perceive PTW riders and to some extent this lack of perception is made worse by the behaviour of the PTW riders.
Age, comparing transport modes
As an example of using national statistics figure 4 shows accident rates per billion kilometres per vehicle type and per age group for two 10 year periods. The accident figures are actually vehicle occupants fatalities (i.e. riders/drivers as well as passengers) in the Netherlands. Kilometre data are from an ongoing national travel survey. There are several interesting observations to be made.
Figure 4 Fatalities per 10*9 kilometres for different modes of transport by age of driver or rider for two periods in the Netherlands (Source:SWOV)
For the most recent 10 year period, car occupants of all ages have a much lower fatality rate than moped or motorcycle occupants. E.g. between 40 and 50 years of age (which is the age group with the lowest rates for all three vehicle types) the rate is 2 for car occupants and 37 and 52 respectively for moped and motorcycles. There is a strong relation between these fatality rates and age.
The moped fatality rates show a U-shaped curve, with equally high rates for young (15-17) and old (60-65) occupants and the lowest rate between 25 and 50 years of age. The fatality rates for car occupants show a U-shape too, with the highest rate for young car occupants (18-19) and the very old (75+) and the lowest level between 40 and 60 years of age.
The trend with age for motorcyclists is different with fatality rates falling from age 20-25 and a lower rate for 18-19 years than for the next youngest group.
The higher fatality rates for young riders/drivers as compared to middle aged ones has to do with both inexperience as a rider/driver and a difference in age related (psychological) factors. The higher rates for old riders/drivers are influenced by their poorer physical condition, resulting in more serious injuries from accidents.
Disregarding the age differences gives accident rates of 4 for car occupants, 68 for moped and 73 for motorcycle. However, the strong relation of these fatality rates with age means that a comparison of vehicle types over all age groups is less meaningful since for instance moped fatalities are mostly young and old, motorcycle fatalities mostly in the middle ages and car fatalities mostly middle age and older.
Trend in risk by age over the last 20 years
A comparison between the two 10 year periods shows that for moped and car occupants the fatality rates are lower for the most recent years, but only for older age groups. For moped occupants the lower rates start from age 40 and car occupants already have lower rates at 30 years. A partial explanation for the difference in trend for young and old rider/drivers could be that older riders/drivers in the recent 10 year period are on average more experienced than riders/drivers of the same age in the earlier period; younger riders/drivers are relatively inexperienced in both periods.
There is a striking difference in fatality rates between the two periods for motorcycle occupants. For the early 10 year period there was a continuously falling rate with age, starting with the youngest age group. For this age group only the fatality rate has decreased. But from age 20 there was a substantial increase in fatality rates from the first period to the most recent 10 years. There are two possible explanations. Firstly, many riders of the older age groups in the recent period are starting riders and thus are less experienced than riders of the same age in the earlier 10 year period. In addition, the motorcycles they ride tend to be heavier and more powerful. Secondly, the Netherlands have introduced a graduated licensing system in 1996. From then on riders of 18-20 years could only ride motorcycles with restricted engine power. A more definitive explanation would require more detailed information which is not available.
Severity of accidents
The rates in figure 4 are for fatalities only. Including accidents with less serious injuries would change some of the observations. Based on fatalities plus hospital admissions, the rates for moped occupants for all age groups are much higher than for motorcycle occupants. This is because moped occupants are more likely to become involved in an accident (corrected for kilometres travelled) but with less serious injuries than motorcycle occupants. This finding illustrates the importance of the method of sampling accidents in a study on the safety of PTW’s. A sample with less severe injuries can lead to different conclusions than a sample with fatalities only.
Age and experience
Although the accident statistics show a clear effect of age, the question whether this effect is a result of the actual age of the rider, rather than his level of experience remains unanswered. To tease out the effects of the two factors, extra information is needed about actual driving experience, or by comparing accident involved rivers with accident-free riders (case-control studies).
Additional accident information
In figure 4 the effects of age and experience are combined in the fatality rates per age group. To separate these effects information is needed on experience of riders involved in accidents and of riders not involved. This requires a special study design to obtain this information. Noordzij & Vis  obtained extra information on 926 motorcyclists involved in accidents resulting in hospital admission in the Netherlands in 1993 as well as from a special national survey of 3000 motorcyclists. Motorcyclists were divided in young (under 25 years of age) and inexperienced (less than five years riding a motorcycle), old and inexperienced and old with (five or more years) experience. The extra information also included type of motorcycle as touring, sports or custom. The rates of accidents per million kilometres show:
- Accident rate for young motorcyclists is one and a half times higher than for older riders with equally limited experience
- For older motorcyclists the accident rate is lower with more experience
- The rates for sports motorcycles is approximately double that of touring motorcycles and the rate for custom motorcycles is also higher than for touring motorcycles but only for inexperienced riders.
Case control studies
A slightly different study design to study the effects of contributory factors is the case-control study. With this design accident facts are recorded in detail as soon as possible after the accident took place. As a consequence of this method the sample of accident cases is usually small, includes few severe injury cases and is regional rather than national. The sample of riders serving as controls is more or less similar to the accident sample in terms of times and places of sampling and in any way consists of riders not involved in an accident at the time of sampling. By a comparison of accident and control samples it is possible to calculate relative accident rates i.e. the accident rate corrected for exposure of a subgroup of riders relative to that of all other groups of riders (or one specific other group).
Such studies have been done in Australia and New Zealand. In the New Zealand study 463 accident cases in 1993-1995 were compared to 1233 control cases. The results show a strong relation between relative accident rate and age as well as with familiarity with the specific motorcycle. There was no evidence of an effect of experience in terms of years riding a motorcycle after adjusting for age . The same data were used in a study on the effect of engine size on accident rate  and another study on motorcycle conspicuity .
The Australian study had 205 accident cases and 1225 controls . Relative accident rate was again found to be strongly related to age and this time a weak relation with years of riding was found. The results also indicated a higher relative accident rate for infrequent riding (less than three days a week) and for non work related riding. A sub group of riders without valid license had a high relative accident rate.
Another Australian study was designed as a survey in which both accident data and exposure data are obtained from 790 questionnaires completed by motorcyclists . The weakness of such a design is the response rate which usually is around 40%. This implies that the results found in the study do not necessarily apply to the whole group of PTW riders at which the study was originally aimed. Another point of concern is that accidents reported by the respondents are without injury or slight injury only. The advantage of such a study is that many questions can be asked on riding habits and psychological variables that seem to be relevant for accident involvement.
Among the results of this study is a relation between accident rate corrected for kilometres and kilometres per year, with lower rates for more kilometres per year. Three groups of riders were found with a high accident rate, all three with a pattern of recreational riding, either as off-road riding or as long distance riding on a sports motorcycle, or weekend riding in urban areas with low kilometres per year.
A large scale study was done in Great Britain in 2002  based on 11265 questionnaires from registered motorcyclists. These riders reported 1495 accidents of which over half occurred during commuting or work related riding. A statistical model was used to calculate the expected number of accidents per rider based on kilometres per year and other factors. This model showed that the age of the rider is the most important predictor, closely followed by kilometres per year. The experience of the rider was next most important, followed by riding conditions. The relation with kilometres per year was such that riders with more kilometres per year had lower accident rates corrected for kilometres per year. The effect of riding conditions was such that those who ride on a regular basis during the year irrespective of weather have the highest accident rate corrected for kilometres, age and experience. Riding experience may not be sufficient to overcome the extra problems created by adverse weather conditions during winter. Engine capacity was related with accident rate in the way that riders of motorcycles with an engine of more than 125cc had a 15% lower rate (corrected for other factors) than riders of 125cc motorcycles. The authors found no evidence that riders who returned to riding after a long break had higher accident rates compared to other riders of the same age.
The questionnaire in the Sexton study included a part on rider behaviour and a part on rider motivation. The latter part was based on several studies in Germany by Schulz and partners. Schulz  distinguished twelve different aspects of motorcycle rider motivation which were found to be closely related to age of the rider and type of motorcycle. In later work Schulz  used a questionnaire aimed at riding style. Analyses of the Great Britain questionnaire showed
- Three aspects to describe motivation: pleasure from riding, liking for speed, economic aspects
- Three aspects to describe riding style: careful vs. careless, tolerant vs. intolerant and slow vs. fast
- Five aspects describing behaviour: traffic errors, speeding, stunting, use of safety equipment, control errors
Rider motivation and riding style were found to be related to rider errors and violations, which were related to accidents. To be more precise: a part of the self reported accidents were the result of behaviour such as speeding, traffic errors and control errors, which were the result of riding style such as confident, fast and careless and/or the result of rider motivation such as a liking for speed. These authors conclude that an important part of the motorcycle safety problems stems from the motivations for choosing to ride motorcycles.
It might be expected that PTW riders differ from car drivers or the general population with regard to some psychological factors. There has not been much recent interest from researchers in this subject. A recent study with an interesting design compared motorcyclists and car drivers in Great Britain . Three groups of subjects were tested in a simulator with traffic situations and completed a questionnaire on behaviour and some psychological variables including social motives and sensation seeking. A first group of 47 motorcyclists had to react as a motorcycle rider but a second group of 47 riders were asked to react as if driving their car. The third group was 48 car drivers. The three groups were carefully matched with regard to age, sex and experience. The results of the car drivers and of the motorcyclists acting as car drivers were very similar except that the motorcyclists were better at detecting potentially dangerous situations. The group of motorcyclists reacting as riders travelled faster, pulled out into smaller gaps, overtook more often, but did not follow closer to a vehicle in front. These results do not necessarily apply to all motorcycle riders. The samples were not representative since the subjects volunteered themselves and all motorcyclists also held a car driving license. But it is clear that at least a group of motorcyclist is equally (or even more) competent and careful as a group of car drivers with similar demographics, except for the fact that the riders choose to ride a motorcycle in order to profit from the properties of the motorcycle such as being small and powerful.
Older Motorcyclists (aged 30+)
With increasing numbers of older motorcyclists in many countries there is a need for information on the safety aspects of this group. A rather detailed questionnaire study was made in Australia . Riders aged over 30 were divided in continuing riders (384) returned riders (240) and new riders (275, with a licence for only seven years or less). All three groups named touring as the most common reason for using their motorcycle but were different in other respects. Continuing riders were more likely to ride all year round and appeared to ride the most but in circumstances that are less likely to result in serious injuries. Returned riders rode less and new riders did more riding in urban areas on smaller motorcycles. The authors state that the accident rate corrected for kilometres for new and returned riders is probably higher than for continuing riders but there are no exact figures in the report.
Apart from speeding there are other examples of PTW rider behaviour in violation of the law, which may contribute to PTW accidents. Broughton  found that 12% of the motorcyclist fatalities died in a drink/drive accident in 1994-2001, compared with 22% for car drivers. The situation in Great Britain seems to be different from that in France. Filou  report 23% of motorcyclists and 32% of moped riders in a fatal accident as positive for alcohol, compared to 19% for drivers of a passenger car. The French authors also mention that 18% of the riders of a light motorcycle and 8% of the riders of heavier motorcycles who died in an accident did not have a valid rider license.
Tampering with a moped or 125cc motorcycle engine to make it go faster seems to be a problem in Germany. Raithel  obtained questionnaires from 137 young riders of PTW’s, half of whom admitted tampering. It is very likely that tampering is related to a higher accident rate. The “Handbook of road safety measures “ refer to a Norwegian study in which the relative rate for injury accidents was found to be about 50% higher for tampered mopeds.
The most extreme example of violations is from Greece in 1994, where only 15% of the injured motorcyclists had been wearing a helmet when this was compulsory . In more recent years the Greek wearing rates seem to be higher. In other countries where helmet wearing is compulsory the wearing rates of helmets by moped and motorcycle riders is usually reported as between 90 and 100%.
Perception by other road users
Accident studies show that the perception of PTW’s is problematic and a contributory factor to collisions with cars. Wulf et.al.  present a review of studies on these problems and discuss several explanations for them. The first explanation has to do with the physical properties of the PTW: the small size of the PTW compared to cars makes it less conspicuous and judging of distance and speed with a small frontal area or with only one headlight is difficult. Other explanations are of a psychological nature. Because of the small number of PTW’s in traffic, car drivers do not expect to meet a PTW and are therefore less prepared to notice or recognise a PTW. PTW’s seem to lack relevance to most car drivers. Wulf even suggests that a car driver is inclined to ignore the presence of a PTW since the impact of a collision with a PTW is less threatening than with a car.
To a certain extent these problems are made worse by the behaviour of the PTW riders such as overtaking in situations where cars can not do this and high speeds. Other road users may not look for PTW’s in places where they do not look for cars and they do not anticipate higher speeds and shorter approach times than for cars.
More fundamental research on the perception of objects has shown that attention may be drawn through physical properties and or through the relevance these objects have for the observer. The lack of relevance of PTW’s for car drivers has not been studied directly, but is illustrated in a study by Magazzu et. al. . The results of the MAIDS case control study were used to obtain 740 cases of a collision between a car and a PTW, together with an expert judgement if the rider or car driver was at fault. After correction for other factors such as age and experience of the car driver it was found that car drivers who also held a motorcycle licence were less likely to be at fault than car drivers without this licence. The possession of a motorcycle licence is an indication of interest in and experience with riding a motorcycle, which may have helped in the detection of an oncoming PTW and the prediction of its manoeuvre.