• No se han encontrado resultados

Driver licensing tests are an integral and important aspect of road safety, although in many respects licensing procedures have become a political issue in our highly mobile society, where driving is regarded as a right rather than a privilege (Engel, 1 994; W right et al., 1 984 ). The underlying notion exists that almost every candidate presenting for a licence will eventually get one (Perkins, 1 984).

Practical driving evaluation can comprise various formal and informal methods of assessing an observable sample of driving. Formal on-road driving tests are typically used to assess learner competencies in the acquisition of a driver's licence which is mandatory in most countries. These make up the bulk of all practical driving evaluations. Shinar ( 1 978) points out that "all licencing programmes are basically tests that evaluate the potential driver's ability to negotiate safely on the road and in t he presence of other drivers" (p. l 3 1 ) . Whether this can be achieved, however, is a contentious issue, as the validity of driver testing is frequently compromised (Norcini, 1 994; Haladyna, 1 994 ). Other formal tests may avail for special licences and endorsements such as those required by taxi drivers, heavy transport licensees, and over 70's drivers. Informal methods of evaluation can include special circumstances such as an occupational therapist's evaluation of adaptive aids over an open- or closed­ road course. With few exceptions, the available literature on practical driver testing appears to be deficient of the rigorous validation studies typical of psychometric testing instruments (Gianutsos, 1 994 ).

Criterion-related validity of driving tests. Inadequate operational defi ni tion underlies many of the problems faced with establishing rel iable criteria for measurement of driving. As already discussed (Chapter Two), there are few guidelines for identifying, instructing and promoting certain levels or standards of driving. Thus, establishing adequate criteria which define driving competence is difficult and multifaceted (Engel, 1 994, Evans, 1 99 1 ; Haladyna, 1 994; Norcini, 1 994) .

44

As part of criterion-related validity, content validity of practical driving tests is limited by measurement of a small range of driving skills which are directly observable in a limited e nvironment. Consequently, perceptual motor tasks are predominantly measured by standard driving tests, while internal states such as motivation, attitude, attention, decision making, and other psychological processes, are not adequately covered ( Ash, B aehr, Joy & Orban, 1 988). This limitation calls into question the validity of driving test criteria as a sole predictor of the driving ability of individuals who have sustained neurological damage (Kaufert, 1 988).

Similarly, the construct validity of standard driving tests is reduced by measures which neglect the extremes of a persons driving capability. In this context, Little ( 1 970) noted that driving tests "cover only the basic minimum knowledge involved in operating a motor vehicle and in no sense attempt to measure the ability to cope with emergency situations or even with normal traffic problems" (p.265). Ethical, social, and practical implications, also prevent evaluation of many driving behaviours critical to driver safety in high risk situations. On the other hand, evidence from training and defensive driving courses does not necessarily suggest that drivers equipped for emergency situations are more able to respond effectively and safely (Evans, 1 99 1 ).

Additional driving test criteria. In conjunction with practical driver testing, many formal evaluations require general knowledge of road laws and a test of visual acuity. Some also require concurrent medical examinations, although there 1s inadequate evidence to suggest this is valid for the general population (Little, 1 970). General knowledge of road laws is typically measured by written and oral tests. The efficacy of pen and paper tests as a component of driver licencing has been investigated using traffic violations as a criterion. However, there is a question mark over the validity of such tests for predicting safe driving (Conley & S miley, 1 976). Written tests are poorly correlated with subsequent measures of road safety, partly due to acquisition of road knowledge during the course of driving (Ash et al. , 1988).

In a practical driving evaluation, vision testing typically involves a test of visual acuity only (Medical Aspects of Fitness to Drive: A Guide for Medical Practitioners, 1 990).

Despite this, there is evidence to suggest that other visual factors are relevant, and maybe more important in the driving process. In general, little research has examined

DRIVING MEASUREMENT the effectiveness of other physiological and medical criteria, as used in conj u nction with practical driver testing . B oth visual and medical factors, as individual characteristics of drivers, will be discussed in forthcoming Chapters.

In the case of the neuropsychologically-impaired driver, Simms ( 1 987) points out that it is crucial that existing medical and visual bars to holding a licence are diagnosed early in the assessment picture. Simms ( 1 987) stresses that these need to be separate criteria not to be confused with other assessment issues.

Predictive validity and reliability of driving tests. The principle aim of a

driving test is to prescribe a level of driving skill or driving competence. However, Wright et al. ( 1 984) state that there is no clear evidence to support that a standard licencing test score can predict performance post l icence, particularly in the c ase of learner drivers. They argue that "a driving test may perform the valuable function of setting a basic criteria for skill but we cannot expect the pass/fail judgement made in the test to convey anything more that at the time of testing the applicant did, or did not perform safely. Whether the driver so licenced will continue to behave that way is a matter which the test cannot predict" (Wright, Hatten & Perkins, 1 984, p. 1 83).

Reliability of driving tests is compromised by a lack of repeatable measurement due to variable testing situations, routes and assessors. As with other measures of functional ability, it is almost impossible to control for situational and motivational variables as well as other performance-related factors inherent in the testing of individuals (Kaufert,

1 988). Cross validation of test results is, therefore, a major problem.

Inter-rater reliability has clearly been found to differ as a function of the type of driving test used. In particular, inter-rater reliability ratings of tests based on clearly defined specific tasks tend to be higher than ratings for more global and continuous driving evaluations. In this respect, Perkins ( 1 984) stresses that possible improvements to increase the reliability ratings of some tests may be at the expense of validity as it pertains to the interrelatedness of behaviours and the traffic environment. A study by West et al. ( 1 993) observed driving over a predefined urban and motorway test route, and found good interrater reliabilities for overall skill and safety ratings. However, the level of agreement differed for individual variables which made up these global ratings. That is, agreement was higher on some of the more clearly defined variables such as

C H A PT F: I{ T H R E E

---·---·--·---

speed ( determi ned from odometer readings) and lower for more subjective measures

such as 'preferred di stance to car in front'.

Exa m i ners can have slightly different criteria for rating dri ving behaviours . I n add i t i o n . the re l iability o f the assessment may be affected by t ime frames and

defi n it ions of when behav iours begin and end. Conti nuous rating runs the risk of examiners missing some behav iours in the process of recording others. For some

prac t ical tests this is overcome by having defined rating and recording periods (Wright et a l . . 1 984 ). A n examiner's position in the vehicle and the slightly different visual ang les that result have been considered to affect inter-rater reliability. One examiner is

usual ly positioned in the front and one examiner in the rear of the vehicle when ratings of the same sample of driving behaviour are made. Further, the potential threat to the

safety o f vehicle occupants has also been found to have an effect on e xaminer re liab i l i ties. For example, Forbes et al. ( 1 975) found a significant difference between examiners on aspects of skill, especial ly when "observers experienced difficulty in focusing on psychomotor skill behaviour only and ignoring potential hazard" (p. 269) . S i m ilarly, while variables such as cal mness and attentiveness had good inter-rater agreement in the West et al. ( 1 993) study, the variable 'aggressiveness' , which suggests a risk factor while driving, did not reach such a high level of agreement. In the l i terature . the use of trained examiners or observers has consistently shown to i mprove the reliabi lity of practical driving measurement.

Documento similar