Logic dictates that the physical consequences of hazards are fixed objective. However perceptions of risks associated with those hazards are highly subjective and as a result risk has been defined in a number of ways. Risk can be viewed as the probability of undesired consequences, the seriousness of the maximum possible undesired consequences or the product of these two outcomes. Another definition of risk is the variances of all the consequences about their mean.In addition to research which aims to identify the factors which affect our perception of risk, studies have investigated the motives which affect our propensity to take risks as well as the best techniques for communicating those risks.

Both lay person and expert assessment of risk involves human judgement. Expert judgement is likely to be more accurate but both groups are subject to common distortions of risk which tend to favour certain criteria which reflect our subjective perception of risk rather than the reality of it (Sprent, 1988). Sprent identified ten common distortions of risk.Two such distortions of risk are the notions that concentrated risks are worse than diffuse risks and involuntary risks are worse than voluntary risks.

Slovic et al (1978) studied the common distortion of risk whereby concentrated risks are regarded as worse than diffuse risks even though the latter may have a much higher death toll. Participants in his study estimated the death rates from 40 hazards using the death rate from car accidents as a point of reference. Slovic found that participants overestimated deaths from infrequent causes but underestimated deaths from infrequent causes.When participants were given a statistical frequency of death and a choice between two hazards which it could represent, they attributed more deaths to the more salient hazard, for example more deaths to murder than to diabetes, when in fact death rates for both of these hazards are the same. Research has also found that we regard involuntary risks, such as poisoning from contaminated food as worse than voluntary risks such as smoking even though there is a much greater risk of death associated with the latter.This paradox was humorously expressed by Chauncey Starr (1969) who noted that "we are loathe to let others do unto us what we happily do to ourselves! " Slovic (1987) identified three risk appraisal factors, basic dimensions connected with the perception of risk.

The first is "Dread Risk" which is characterised by a perceived lack of control, catastrophic potential, fatal consequences and the inequitable distribution of risks and benefits.This dimension is closely related to the general public's perception of risk and is best illustrated by the risks associated with nuclear power stations or more recently the threat of chemical or biological attack. The second of Slovic's risk appraisal factors is "Unknown risks". These are hazards for which the risks are as yet unknown and delayed in their manifestation of harm. For example, the risks associated with mobile phone use or genetically modified crops. The third risk appraisal factor relates to the number of people exposed to a given risk.

The greater the number of people exposed to possible negative consequences, the greater the perception of risk. Experts and lay people attach differing degrees of significance to these separate appraisal factors. Slovic et al (1979) found that lay people are moderately good at estimating annual fatalities from various risks. However their overall perceptions of risk are more related to factors such as catastrophic potential and threat to future generations revealing a strong correlation between risk appraisal factors one and three.

This tendency to emphasise such features of risk may also reflect the high degree of media coverage they acquire. Experts are however very good at estimating annual fatalities from given hazards and their perceptions of risk correlate highly with concrete statistics. For example when asked to rate 30 different activities and technologies in terms of risk lay people placed nuclear power as the greatest hazard. However experts placed nuclear technology twentieth. They perceived it as less of a risk than riding a bicycle, which in objective measures of annual deaths, is an accurate perception.

Individual differences in the tendency to take physical risks have also formed the subject of extensive research. Zuckerman's Sensation Seeking Scale (1964) was an attempt by personality theorists to interpret physical risk taking as sensation seeking. Scale items were randomly selected by intuition and kept if found to be statistically valid. These items were then split into subscales by method of factor analysis. These subscales were seeking experience, disinhibition, thrills and adventure and relief from boredom.High sensation seeking groups ( and therefore high physical risk taking groups) have been identified as men, multi drug users, promiscuous people, smokers and young people.

Some studies have correlated sensation seeking behaviour with physiological responses and some have implicated the role of genetic factors (Telegen, 1988). Non-physical risk taking has also been studied. Atkinson (1957) related risk-taking to his theory that motivation depends on desires to achieve success and to avoid failure.People who measure highly on desire to achieve success but low on desire to avoid failure are those who take the most moderate risks in skilled tasks.

Individual differences mean that some people prefer to take more low probability risks which comprise of a high chance a small success and a low chance of failure. Others prefer to take more high probability risks with the underlying rationale being that failure will be attributed to the difficulty of the task and not to a perceived lack of skill on their part.This study has been used as a model for entrepreneurship in explaining why some people will achieve great success with "all or nothing" business strategies. Liverant ; Scodel (1960) investigated the effect of locus of control on propensity to take risks. They found that people with an external locus of control are more likely to opt for long shot bets than those with an internal locus of control. Such individuals may believe that lady luck has brought them rewards in the past and therefore they want to give good fortune a chance to work in their favour.

This could explain behaviours such as placing bets on the national lottery every week despite the extremely low probability of scooping the jackpot. Slovic (1982) highlighting the role of situation specific variables on an individuals likelihood to take risks. We do not have an overall propensity to take risks. For example the person who faithfully places their bet on the national lottery every week despite the high risk of receiving no return on their gamble, may be completely anti-smoking and refuse to take any risks when it comes to their physical health.However MacCrimmon ; Wehrung (1990) took risk taking measures from 300 executives and found that individual differences such as age, nationality and past success correlated highly with the propensity to take risks but situation specific variables did not.

We must however, be cautious about generalizing the results of this study. It seems intuitive that personality characteristics would play a significant role in achieving top positions in professions which are characterised by a constant high degree of situational risk (for example stockbroking).The risk homeostasis theory was proposed by J Adams (1995) claims that people try to balance out the costs and benefits involved in taking risks. This theory can be explored in relation to drivers and their propensity to take the risk of exceeding the speed limit. Higher speeds, argues Adams, are beneficial to drivers on one hand because it means that they save time but negative because speeding puts them at a higher risk of having an accident.

The optimum speed is achieved when the cost of the time they spend travelling and the costs associated with an accident are equal.If changes can be made by drivers that increase their safety, they will drive faster because the speed at which the two costs are equal will be higher. The safety measure will not have the desired effect and the danger to other road users may increase. Evidence for this theory comes from statistics that show that deaths of people in the back seats of cars as well as deaths of pedestrians and cyclists increased after the 1980s seatbelt laws came into effect. Drivers who had previously driven at a slower speed because they chose not to wear a seatbelt were now forced to buckle up.As a result they reassessed the costs involved with time versus safety and drove faster.

In contrast the number of people killed by heavy goods vehicles which were not affected by the introduction of the seatbelt laws decreased. These drivers now felt that for them, the speed at which costs of time and benefits of safety were balanced was lower. The risk homeostasis theory and its documented affects on speeding have serious implications for the communication of risk. Under the risk homeostasis theory the assessment of costs and benefits is carried out at an unconscious level.

Therefore simply telling people to drive more slowly will not work.This has been illustrated in the failure of campaigns which urge drivers to slow down around schools. In order to effectively slow down drivers, the relationship between the cost of speeding and the benefits of the time saved must be altered. We can either do this by harsh punishment of drivers caught speeding or making traffic jams a more socially accepted excuse to be late (Howarth,1988). When risks are taken only after weighing up the pros and the cons, (for example in finance) this is known as taking a risk. When there is no conscious analysis of risk, this is known as running a risk.

Often when we run a risk we are confronted by risks that we had not considered. Saarinen (1966) found that perception of risk grows more accurate and discriminating when people have greater direct experience of specific hazards. Obviously it is not advisable for people to experience the consequences of particular hazards in order to improve their perception of risk. Therefore in the field of accident prevention, accidents are studied, their risks are revealed and preventative action is taken to minimize their future effects. Wagenarr (1987) analysed 57 accidents at sea.

In 21% of cases information about the imminent danger was not available.In 27%, the information was received but not considered to be problematic, in 15% the error was in not considering the alternative that would have solved the problem and in 36% the error was in failing to evaluate the consequences of the action. Therefore 83% of accidents at sea occurred because there was no calculated acceptance of risk and people were thus taken by surprise. This study emphasizes how crucial accurate communication of risk is in the field of accident prevention. Understanding how we perceive, take and communicate risks is an important topic especially in the light of the recent terrorist threat.

Public perceptions of Slovic's (1987) risk factors, especially dread risk are at the forefront of our minds. Given the existence of other factors such as extensive media coverage, the vividness or availability of a threat and the recency of catastrophe which have all been shown to add to perception of risk we are living in a society in a constant state of anxiety over risks which are perceived as unknown, uncontrollable and of possible catastrophic impact. The global tightening of security measures illustrates the value placed on preventative behaviours when perceptions of risk are high.