Abstract

The Roadwise Review has been reported to provide an effective means of self-assessing and predicting driving difficulties in older adults. We administered it to 73 community-dwelling older drivers ( M = 73 years ) and besides gathered data on self-reported drive difficulties, 2-year retrospective collisions, and moving violations. The acuteness tests and useful Field of View exhibited substantial ceiling effects that limit predictive utility, and there was a high gear failure pace on the head and neck flexibility test. additionally, the Roadwise Review did not predict self-reported repel problems or collision gamble. thus, in current form, it does not appear to be a useful joyride for assessing older drivers. Future research efforts should assess predictive robustness in a more heterogenous sample of older adults and with a broader stove of outcomes, including on-road force performance .
many older drivers are safe but, when adjusted for distance traveled, there is an increase in collision gamble that begins around 50 years of long time ( see Evans, 2004 for an excellent review ). Although recent data qualify this generalization ( Hakamies-Blomqvist, 2004 ; Tay, 2008 ), older adults ’ greater risk of injury or death and their burgeon numbers have motivated a growing interest in issues of licensure and assessment. There has been some success in the attempt to predict at-risk older drivers. Owsley, Ball, McGwin, Sloane, Roenker, White, and Overley ( 1998 ) investigated prospective collision involvement in accredited older drivers and found a significant correlation with the utilitarian Field of View ( UFOV ). Contrast sensitivity deficits are associated with crash interest in older people who have cataract ( for example, Owsley, Staveley, Wells, Sloane, & McGwin, 2001 ).

Anstey, Wood, Lord, and Walker ( 2005 ) pointed out that accurate self-monitoring is required for dependable drive. Accurate self-monitoring can be aided by good evaluation tools, and, to this end, a assortment of resources have been developed. Of particular relevance to the present research, the American Automobile Association and the Canadian Automobile Association presently distribute the Roadwise Review. It grew out of exercise by Staplin, Lococo, Gish, and Decina ( 2003 ), who carried out a large-scale study of older U.S. drivers to determine the predictive validity of a battery of tests that included measures of physical, centripetal, and cognitive functions. Most of the measures were able to identify collision-involved older drivers. All the measures in the Roadwise Review come from Staplin, Lococo, et alabama. ( 2003 ), with adaptations ( for example, standardized synergistic instructions ) to allow testing without a test administrator. There have been some recent efforts to evaluate the Roadwise Review. Myers, Blanchard, MacDonald, and Porter ( 2008 ) reported that older adults are broadly favorable to the test. Ball and colleagues ( 2006 ) used data from a subsample of older adults in Staplin and colleagues study and found Subtest 2 of the UFOV, Trails B, and the Motor Free Visual Perception Test predicted prospective at-fault collisions. Edwards and colleagues ( 2008 ) administered the Roadwise Review to licensed drivers at least 65 years of senesce. The UFOV and Trails B predicted self-reported 2-year retrospective collision engagement but did not predict a single-item self-assessment of driving “ quality. ” The objective of the salute study was to administer the Roadwise Review to a sample distribution of older drivers to determine its ability to predict self-reported repel difficulties and driving history. The research builds on previous workplace by incorporating a valid measurement of self-reported drive along with collision data. We expected the subtests of the Roadwise Review to be correlated with each other and to predict drive, collisions, and moving violations.

M ETHODS

Participants

The 73 participants whose data are reported ( data from two people were lost due to software errors ) were recruited from organizations in the Calgary area. Each received $ 20 ( canadian ) or had that sum donated to their rear constitution.

Materials and Procedure

The number of self-reported moving violations and at-fault collisions were recorded for the former 2 years ( see besides Edwards et al., 2008 ), along with demographic data such as distance drive, old age, and therefore forth. next, the Roadwise Review, in the form of the research-friendly DrivingHealth Inventory, was given. It consists of 11 tests administered in the follow : fixed order, four tests of acuteness, rapid footstep walk, head and neck flexibility, delayed recall, visualizing missing information, Subtest 2 of the UFOV, Trails A, and Trails B. Details can be found in Staplin and colleagues ( 2003 ). Following the Roadwise Review, participants completed the Driver Behavior Questionnaire ( DBQ ; Reason, Manstead, Stradling, Baxter, & Campbell, 1990 ), which measures how frequently the participants experience difficulties and engage in diverse behaviors while driving. There are four factors ( Lajunen, Parker, & Summala, 2004 ) : aggressive violations, ordinary violations, errors, and lapses. The DBQ predicts collision engagement in divers populations, including older adults ( Parker, McDonald, Rabbitt, & Sutcliffe, 2000 ). adjacent, they completed the Mini-Mental State Examination ( MMSE ; Folstein, Folstein, & McHugh, 1975 ). It had a maximum score of 30, with lower scores indicating more cognitive deterioration. The predictive robustness of the MMSE is discrepant, but it is normally used to screen for cognitive disability ( for example, Margolis et al., 2002 ; Owsley, Sloane, Ball, Roenker, & Bruni, 1991 ).

R ESULTS

Participants ranged from 50 to 88 years ( M = 73 years, SD = 7 years ), and 67 % were women. approximately 20 % were under a doctor ’ south concern for a good medical illness or condition, but 85 % rated their health as thoroughly or excellent. On average, they drove 14,221 km/year ( SD = 10,275 kilometer ). Twelve percentage reported having one or two at-fault collisions, and 27 % reported one or two moving violations in the last 2 years. descriptive statistics for most variables are provided in board 1. about 50 % of the participants failed the head and neck flexibility trial. Three of the four acuteness tests showed pronounce ceiling effects. There is more unevenness in performance on the low-contrast quiz of acuity, but still, 77 % of the participants had no errors. This is to be expected because all license drivers must have high-contrast acuteness levels better than is assessed in the Roadwise Review. even, ceiling effects limit the predictive validity of the tests.

Table 1.

Test items  SD 
Driver Health Inventory 
    High-contrast 20/80 Acuity (0–4)  0.15  0.36 
    High-contrast 20/40 Acuity (0–4)  0.22  0.42 
    Low-contrast 20/80 Acuity (0–4)  0.16  0.41 
    Low-contrast 20/40 Acuity (0–4)  0.36  0.71 
    Rapid pace walk (s)  5.64  1.42 
    Head–neck flexibility (0–1)  0.58  0.50 
    Visualizing Missing Information (0–11)  3.49  2.91 
    Trails A (s)  48.65  21.56 
    Trails B (s)  116.17  45.58 
    The Delayed Recall Test (0–3)  0.60  0.94 
    Useful Field of View (ms)  230.00  140.00 
Driver Behavior Questionnaire 
    Aggressive violations (0–6 point scale)  7.20  1.87 
    Ordinary violations (0–6 point scale)  13.36  2.58 
    Errors (0–6 point scale)  25.99  5.20 
    Lapses (0–6 point scale)  18.72  3.40 
    All items (0–6 point scale)  80.86  12.19 
Mini-Mental State Examination 
    Test score (0–30)  28.18  2.10 
Driver Experience Questionnaire 
    Average driving difficulties (1–5 point scale)  24.02  4.95 
    Number of moving violations in 2 years  0.30  0.52 
    Number of collisions in 2 years  0.14  0.38 
Test items  SD 
Driver Health Inventory 
    High-contrast 20/80 Acuity (0–4)  0.15  0.36 
    High-contrast 20/40 Acuity (0–4)  0.22  0.42 
    Low-contrast 20/80 Acuity (0–4)  0.16  0.41 
    Low-contrast 20/40 Acuity (0–4)  0.36  0.71 
    Rapid pace walk (s)  5.64  1.42 
    Head–neck flexibility (0–1)  0.58  0.50 
    Visualizing Missing Information (0–11)  3.49  2.91 
    Trails A (s)  48.65  21.56 
    Trails B (s)  116.17  45.58 
    The Delayed Recall Test (0–3)  0.60  0.94 
    Useful Field of View (ms)  230.00  140.00 
Driver Behavior Questionnaire 
    Aggressive violations (0–6 point scale)  7.20  1.87 
    Ordinary violations (0–6 point scale)  13.36  2.58 
    Errors (0–6 point scale)  25.99  5.20 
    Lapses (0–6 point scale)  18.72  3.40 
    All items (0–6 point scale)  80.86  12.19 
Mini-Mental State Examination 
    Test score (0–30)  28.18  2.10 
Driver Experience Questionnaire 
    Average driving difficulties (1–5 point scale)  24.02  4.95 
    Number of moving violations in 2 years  0.30  0.52 
    Number of collisions in 2 years  0.14  0.38 

Open in new tab

Table 1.

Test items  SD 
Driver Health Inventory 
    High-contrast 20/80 Acuity (0–4)  0.15  0.36 
    High-contrast 20/40 Acuity (0–4)  0.22  0.42 
    Low-contrast 20/80 Acuity (0–4)  0.16  0.41 
    Low-contrast 20/40 Acuity (0–4)  0.36  0.71 
    Rapid pace walk (s)  5.64  1.42 
    Head–neck flexibility (0–1)  0.58  0.50 
    Visualizing Missing Information (0–11)  3.49  2.91 
    Trails A (s)  48.65  21.56 
    Trails B (s)  116.17  45.58 
    The Delayed Recall Test (0–3)  0.60  0.94 
    Useful Field of View (ms)  230.00  140.00 
Driver Behavior Questionnaire 
    Aggressive violations (0–6 point scale)  7.20  1.87 
    Ordinary violations (0–6 point scale)  13.36  2.58 
    Errors (0–6 point scale)  25.99  5.20 
    Lapses (0–6 point scale)  18.72  3.40 
    All items (0–6 point scale)  80.86  12.19 
Mini-Mental State Examination 
    Test score (0–30)  28.18  2.10 
Driver Experience Questionnaire 
    Average driving difficulties (1–5 point scale)  24.02  4.95 
    Number of moving violations in 2 years  0.30  0.52 
    Number of collisions in 2 years  0.14  0.38 
Test items  SD 
Driver Health Inventory 
    High-contrast 20/80 Acuity (0–4)  0.15  0.36 
    High-contrast 20/40 Acuity (0–4)  0.22  0.42 
    Low-contrast 20/80 Acuity (0–4)  0.16  0.41 
    Low-contrast 20/40 Acuity (0–4)  0.36  0.71 
    Rapid pace walk (s)  5.64  1.42 
    Head–neck flexibility (0–1)  0.58  0.50 
    Visualizing Missing Information (0–11)  3.49  2.91 
    Trails A (s)  48.65  21.56 
    Trails B (s)  116.17  45.58 
    The Delayed Recall Test (0–3)  0.60  0.94 
    Useful Field of View (ms)  230.00  140.00 
Driver Behavior Questionnaire 
    Aggressive violations (0–6 point scale)  7.20  1.87 
    Ordinary violations (0–6 point scale)  13.36  2.58 
    Errors (0–6 point scale)  25.99  5.20 
    Lapses (0–6 point scale)  18.72  3.40 
    All items (0–6 point scale)  80.86  12.19 
Mini-Mental State Examination 
    Test score (0–30)  28.18  2.10 
Driver Experience Questionnaire 
    Average driving difficulties (1–5 point scale)  24.02  4.95 
    Number of moving violations in 2 years  0.30  0.52 
    Number of collisions in 2 years  0.14  0.38 

Open in new tab
Data from the delayed recall and visualizing missing data tests were positively skewed. For case, approximately 60 % of respondents made two or fewer errors on the latter. Trails A and B completion times were both normally distributed, but the UFOV scores exhibited a tag deviation from normality. Almost one half of the participants were able to achieve 75 % accuracy at the minimum duration of 100 thousand, even 6 % of the sample could not perform the test at the maximal duration of 500 mississippi. table 2 provides zero-order correlations between forecaster variables and the most important result measures, including subscale scores on the DBQ, self-reported collisions, and moving violations. To predict the continuous measures of the DBQ scales, we used ordinary least squares regression. For the dichotomous variables of moving violations and collisions, we used logistic regression. For both types of analyses, we proceeded in two steps : The first entered all the predictors from the Roadwise Review, and the second model added age and MMSE scores.

Table 2.

Predictor variable  DBQ-ordinary violations  DBQ-errors  DBQ-lapses  DBQ-aggressiveness  Collision (yes or no)  Moving violations 
Walking speed  −.174  −.039  −.158  −.143  −.003  .025 
Head–neck flexibility  .147  .175  .081  .060  −.055  .072 
High-contrast acuity (20/80)  −.130  −.115  −.205  −.086  .049  .051 
High-contrast acuity (20/40)  .074  .209  .139  −.181  −.103  −.053 
Low-contrast acuity (20/80)  −.037  −.102  −.051  −.044  .120  .025 
Low-contrast acuity (20/40)  .125  .016  .038  .095  .174  −.031 
Visualizing Missing Information  −.067  −.100  −.129  −.054  .111  .066 
Trails A  −.081  −.114  −.221  −.083  .023  −.066 
Trails B  .236*  .147  −.041  −.006  .075  −.082 
Working Memory  −.186  −.095  −.173  −.181  .078  .025 
UFOV  −.033  .047  .054  −.049  .117  .133 
MMSE  −.118  .001  .134  −.047  .124  .039 
Age  −.129  .069  .015  −.201  −.016  .024 
Predictor variable  DBQ-ordinary violations  DBQ-errors  DBQ-lapses  DBQ-aggressiveness  Collision (yes or no)  Moving violations 
Walking speed  −.174  −.039  −.158  −.143  −.003  .025 
Head–neck flexibility  .147  .175  .081  .060  −.055  .072 
High-contrast acuity (20/80)  −.130  −.115  −.205  −.086  .049  .051 
High-contrast acuity (20/40)  .074  .209  .139  −.181  −.103  −.053 
Low-contrast acuity (20/80)  −.037  −.102  −.051  −.044  .120  .025 
Low-contrast acuity (20/40)  .125  .016  .038  .095  .174  −.031 
Visualizing Missing Information  −.067  −.100  −.129  −.054  .111  .066 
Trails A  −.081  −.114  −.221  −.083  .023  −.066 
Trails B  .236*  .147  −.041  −.006  .075  −.082 
Working Memory  −.186  −.095  −.173  −.181  .078  .025 
UFOV  −.033  .047  .054  −.049  .117  .133 
MMSE  −.118  .001  .134  −.047  .124  .039 
Age  −.129  .069  .015  −.201  −.016  .024 

Open in new tab

Table 2.

Predictor variable  DBQ-ordinary violations  DBQ-errors  DBQ-lapses  DBQ-aggressiveness  Collision (yes or no)  Moving violations 
Walking speed  −.174  −.039  −.158  −.143  −.003  .025 
Head–neck flexibility  .147  .175  .081  .060  −.055  .072 
High-contrast acuity (20/80)  −.130  −.115  −.205  −.086  .049  .051 
High-contrast acuity (20/40)  .074  .209  .139  −.181  −.103  −.053 
Low-contrast acuity (20/80)  −.037  −.102  −.051  −.044  .120  .025 
Low-contrast acuity (20/40)  .125  .016  .038  .095  .174  −.031 
Visualizing Missing Information  −.067  −.100  −.129  −.054  .111  .066 
Trails A  −.081  −.114  −.221  −.083  .023  −.066 
Trails B  .236*  .147  −.041  −.006  .075  −.082 
Working Memory  −.186  −.095  −.173  −.181  .078  .025 
UFOV  −.033  .047  .054  −.049  .117  .133 
MMSE  −.118  .001  .134  −.047  .124  .039 
Age  −.129  .069  .015  −.201  −.016  .024 
Predictor variable  DBQ-ordinary violations  DBQ-errors  DBQ-lapses  DBQ-aggressiveness  Collision (yes or no)  Moving violations 
Walking speed  −.174  −.039  −.158  −.143  −.003  .025 
Head–neck flexibility  .147  .175  .081  .060  −.055  .072 
High-contrast acuity (20/80)  −.130  −.115  −.205  −.086  .049  .051 
High-contrast acuity (20/40)  .074  .209  .139  −.181  −.103  −.053 
Low-contrast acuity (20/80)  −.037  −.102  −.051  −.044  .120  .025 
Low-contrast acuity (20/40)  .125  .016  .038  .095  .174  −.031 
Visualizing Missing Information  −.067  −.100  −.129  −.054  .111  .066 
Trails A  −.081  −.114  −.221  −.083  .023  −.066 
Trails B  .236*  .147  −.041  −.006  .075  −.082 
Working Memory  −.186  −.095  −.173  −.181  .078  .025 
UFOV  −.033  .047  .054  −.049  .117  .133 
MMSE  −.118  .001  .134  −.047  .124  .039 
Age  −.129  .069  .015  −.201  −.016  .024 

Open in new tab
The results of these modeling efforts are seen in table 3. No model performed adequately in predicting either moving violations or collisions. We encountered more success in predicting self-reported drive difficulties. The Roadwise Review, either alone or in junction with age and MMSE scores, accounted for 25 % –30 % of the variance in DBQ ordinary violations, lapses, and errors. however, the signs of the arrested development weights ( see table 2 for zero-order correlations ) for the most authoritative predictors are inconsistent and often in an unexpected direction. For example, when predicting DBQ Ordinary Violations, Visualizing Missing Information, Trails A, and Trails B were individually significant in the model. The signs of the arrested development weights were negative for the first two predictors ; more disability was associated with fewer difficulties. The alone forecaster that performed well over these three DBQ scales was Trails B, which was meaning and incontrovertible in two of the three scales. As well, no model would be significant after adjustments for multiple tests.
As an alternate analysis scheme, we used the cut-points of Roadwise Review tests published by Staplin, Gish, and Wagner ( 2003 ) to predict retrospective collision status in our sample. These are the individual test scores they found to maximally separate collision-involved and uninvolved drivers in a population-based study of more than 1,800 drivers. As such, they should be the most stable values for prediction of gamble. The model was nonsignificant ( Nagelkerke R2 of approximately .10 ) and predicted all drivers to be collision free.

D ISCUSSION

The tests comprising the Roadwise Review have undergo several late investigations to determine if they predict driving behavior or collision risk in older adults. The present shape builds on previous work in significant ways. Ball and colleagues ( 2006 ) did not administer the Roadwise Review but rather examined the operation of like tests from the Maryland Pilot Older Driver Study ( Staplin, Lococco, et al., 2003 ). Compared with the presently marketed Roadwise Review, there are important differences in test content and government that may impact predictive cogency. Edwards and colleagues ( 2008 ) used the Roadwise Review and self-reported collisions but did not incorporate a validated measure of driving difficulties as we have done. Descriptive data ( i.e., averages and distributional properties ) from our cogitation are by and large coherent with data reported previously from both public toilet samples ( Edwards et al., 2008 ) and population-based studies ( Ball et al., 2006 ). Our results agree with those of Edwards and colleagues ( 2008 ), who found that the Roadwise Review did not predict self-reported drive quality. Although self-reports may be criticized, the DBQ is a valid forecaster of collision engagement in older adults ( Parker et al., 2000 ). As well, if test results are not coherent with drivers ’ estimates of their drive, then user toleration of the test will be a challenge. Ball and colleagues ( 2006 ) and Edwards and colleagues ( 2008 ) were able to identify those older drivers involved in collisions using the Roadwise Review data. We found no relative between retrospective at-fault collisions and performance on the Roadwise Review. The reasons for this discrepancy are not related to low collision rates. approximately 12 % of our sample reported at least one collision. This compares with Ball and colleagues ( 2006 ) ( 5.5 % ) and Edwards and colleagues ( 2008 ) ( 19 % ). notably, the average operation of our sample on some of the tests is very unlike from that reported earlier. On the UFOV, group means were 133 megabyte in Edwards and colleagues ( 2008 ) and 177 mississippi in Ball and colleagues ( 2006 ), much lower than the average of 230 ms found here. additionally, across studies, collision-involved drivers do not show the lapp proportional stultification on either Trails B or the UFOV, the two most authoritative predictors. On Trails B, our data and Ball and colleagues ( 2006 ) show a relative disability of 3 % and 6 %, respectively. Edwards and colleagues ( 2008 ) report 49 % relative stultification. For the UFOV, our data and Ball and colleagues ( 2006 ) show a relative damage of 27 % and 22 %, respectively. Edwards and colleagues ( 2008 ) reported a 49 % loss. Because there have been several iterations of the tests used, some of an indecipherable and proprietorship nature, these discrepancies may not be easy to explain. There is a pressing motivation to develop assessment tools to evaluate older drivers. Although the Roadwise Review may be valuable in providing drivers with information on skills related to driving performance, in its stream kind it does not appear to be utilitarian in the prediction of self-reported driving difficulties or risk in older adults. Floor or ceiling effects on many of the tests limit predictive cogency, and high failure rates of some tests will work against drug user acceptance. low correlations with self-reported driving difficulty and retrospective collisions are baffling. Although predicting self-reported driving demeanor from the Roadwise Review was found, in some instances, to have modest success, a closer examination of the data indicate that more disability was associated with less trouble in driving. This discover challenges to acceptance for consumers and policy makers alike. There are two major limitations to the salute study to be addressed in future workplace. Our sample consisted of self-selected elders who were presently driving. It would be beneficial to collect data on older drivers who have recently given up driving or who have been referred for checkup reasons ( for example, gloomy mental condition scores ), as was done in Staplin and colleagues ( 2003 ). As well, it would be useful to determine if the Roadwise Review is able to predict driving performance through on-road tests.

F UNDING

This research was funded by grants to R.T. from the Alberta Motor Association and to C.S. from the Natural Science and Engineering Research Council of Canada. We would like to thank the members of the Confederation Park Senior Citizens Centre and the Greater Forest Lawn Seniors Centre for their participation in this learn. Results were presented, in part, at Driving Assessment, 2009 ( Big Sky, MT ).

References

Anstey K wood joule overlord second walker joule Cognitive, centripetal and physical factors enabling driving safety in older adults,  clinical Psychology Review,  2005, vol.  25 (pg.  45- 65), vol. ( pg . ball K Roenker vitamin d Wadley v Edwards joule Roth d McGwin gram raleigh r joyce j Cissell g Dube thyroxine Can high-risk older drivers be identified through performance-based measures in a Department of Motor Vehicles setting ?,  Journal of the american english Geriatrics society,  2006, vol.  54 (pg.  77- 84), vol. ( pg . Edwards j Leonard K Lunsman thousand Dodson joule Bradley sulfur Myers C hubble boron acceptability and robustness of older driver screening with the DrivingHealth inventory,  accident analysis and prevention,  2008, vol.  40 (pg.  1157- 1163), vol. ( pg . Evans fifty Traffic safety,  2004 Bloomfield Hills, MI science Serving club Folstein molarity Folstein sulfur McHugh phosphorus Mini-Mental State : A hardheaded method acting for grading the cognitive condition of patients for the clinician,  Journal of Psychiatric Research,  1975, vol.  12 (pg.  189- 198), vol. ( pg . Hakamies-Blomqvist liter safety of older persons in traffic,  fare in an aging society,  2004 Washington, DC Transportation Research Board Lajunen deoxythymidine monophosphate parker d Summala heat content The Manchester Driver Behaviour Questionnaire : A cross-cultural discipline,  accident analysis and prevention,  2004, vol.  36 (pg.  231- 238), vol. ( pg . Margolis KL Kerani RP McGovern p Songer metric ton Cauley JA Ensrud KE risk factors for motor vehicle collisions in older women,  Journal of Gerontology : aesculapian Sciences

,  2002, vol.  57 (pg.  M186- M191), vol. ( pg . Myers A Blanchard gas constant MacDonald fifty porter megabyte Process evaluation of the American Automobile Association Roadwise Review CD-ROM : Observed and reported experiences of older drivers,  Topics in Geriatric Rehabilitation,  2008, vol.  24 (pg.  224- 238), vol. ( pg . Owsley C Sloane maine ball K Roenker deciliter Bruni junior Visual—Cognitive correlates of fomite collisions in older drivers,  psychology and Aging,  1991, vol.  6 (pg.  403- 415), vol. ( pg . Owsley C ball K McGwin g Sloane thousand Roenker five hundred White meter Overley einsteinium ocular process stultification and gamble of motive fomite collision among older adults,  Journal of the American Medical Association,  1998, vol.  279 (pg.  1083- 1088), vol. ( pg . Owsley C Staveley BT Wells joule Sloane maine McGwin guanine ocular hazard factors for collision engagement in older drivers with cataract,  Archives of Ophthalmology,  2001, vol.  119 (pg.  881- 887), vol. ( pg . parker vitamin d McDonald lambert Rabbitt p Sutcliffe phosphorus aged drivers and their accidents : the Aging Driver Questionnaire,  Accident Analysis & Prevention,  2000, vol.  32 (pg.  751- 759), vol. ( pg . reason joule Manstead A Stradling randomness Baxter j campbell K Errors and violations on the roads,  Ergonomics,  1990, vol.  33 (pg.  1315- 1332), vol. ( pg . Staplin lambert Gish K wagner vitamin e MaryPODS revisited : Updated crash analysis and implications for screening program execution,  Journal of Safety Research,  2003, vol.  34 (pg.  389- 397), vol. ( pg . Staplin lambert Lococo K Gish K Decina liter Model driver screening and evaluation program : Final Technical Report : volume II : Maryland Pilot Older Driver Study ( Report Number DOT HS 809 583 ),  2003 Washington, DC National Highway Traffic Safety Administration tay roentgen fringy effects of increasing ageing driver on injury collisions,  accident analysis and prevention,  2008, vol.  40 (pg.  2065- 2068), vol. ( pg.

Author notes

© The author 2010. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail : journals.permissions @ oxfordjournals.org .

Leave a Reply

Your email address will not be published.