STATE OF CALIFORNIADEPARTMENT OF MOTOR VEHICLES
CLEARING A ROAD TO BEING DRIVING FIT BY BETTER ASSESSING DRIVING WELLNESS
DEVELOPMENT OF CALIFORNIA’S PROSPECTIVE THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
Technical Report
October 2009
Authors: David F. Hennessy, Ph. D. &
Mary K. Janke, Ph. D.
Research and Development Branch
© California Department of Motor Vehicles, 2009 Licensing Operations Division
RSS- 05- 216
REPORT DOCUMENTATION PAGE
Form Approved
OMB No. 0704- 0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202- 4302, and to the Office of Management and Budget, Paperwork Reduction Project ( 0704- 0188), Washington, DC 20503.
1. AGENCY USE ONLY ( Leave blank)
2. REPORT DATE
October 2009
3. REPORT TYPE AND DATES COVERED
Final Report
4. TITLE AND SUBTITLE
Clearing A Road to Being Driving Fit by Better Assessing Driving Wellness - Development of California’s Prospective Three- Tier Driving- Centered Assessment System ( Technical Report)
6. AUTHOR( S)
David F. Hennessy, Ph. D. & Mary K. Janke, Ph. D.
5. FUNDING NUMBERS
7. PERFORMING ORGANIZATION NAME( S) AND ADDRESS( ES)
California Department of Motor Vehicles
Research and Development Branch
P. O. Box 932382
Sacramento, CA 94232- 3820
8. PERFORMING ORGANIZATION
REPORT NUMBER
CAL- DMV- RSS- 05- 216
9. SPONSORING/ MONITORING AGENCY NAME( S) AND ADDRESS( ES)
Office of Traffic Safety
2208 Kausen Drive., Suite 300
Elk Grove, CA 95758- 7115
10. SPONSORING/ MONITORING
AGENCY REPORT NUMBER
11. SUPPLEMENTARY NOTES
12a. DISTRIBUTION/ AVAILABILITY STATEMENT
Unlimited
12b. DISTRIBUTION CODE
UL
13. ABSTRACT ( Maximum 200 words)
This report has two main purposes: ( 1) describe the development of California’s prospective 3- Tier driving- centered assessment system, and ( 2) present an “ ecological perspective” on driver licensing. Driving- centered is an ecological concept— it means taking into consideration when, where, why, and how individual drivers customarily drive. Rather than an endpoint in delicensing drivers assessed as unsafe, 3- Tier fundamentally alters the purpose of assessment to be a starting point, if feasible, for extending the safe driving years of functionally- limited licensed drivers.
The 3- Tier system integrates new assessment tools into those currently used by the Department of Motor Vehicles. All renewal applicants required to pass the department’s knowledge test are assessed on Tier 1, and those who are found to have a driving- relevant visual, mental, or physical limitation( s) are further screened on Tier 2. Based on these assessments, drivers are classified as driving well, somewhat functionally limited or extremely functionally limited; the extremely functionally- limited drivers are required to pass a Tier 3 road test to be licensed. The results of a small scale pilot study upon which the 3- Tier system was developed showed that somewhat- limited drivers, perhaps because they were less aware of their limitations, were more likely to be crash involved than extremely- limited drivers, who were probably more aware of their limitation( s) and compensated accordingly. In contrast, extremely- limited drivers were more likely to fail an office- based road test.
The report concludes with 22 recommendations for statewide implementation of 3- Tier, including recommendations that the department’s R& D branch evaluate the reliability and validity of the current area drive test, and if needed, develop a better one, that this test be available to extremely limited drivers as an option for their Tier 3 road test requirement, and that the department educate somewhat- limited drivers about compensating for their limitation( s).
15. NUMBER OF PAGES
217
14. SUBJECT TERMS
3- Tier, Driving Wellness, Driving Fitness, Licensing Tests, Driving Assessment System, Functionally- limited Drivers, Driving- Centered, Driver- Centered, Ecological Perspective on Driving and Driver Assessment
16. PRICE CODE
17. SECURITY CLASSIFICATION
OF REPORT
Unclassified
18. SECURITY CLASSIFICATION
OF THIS PAGE
Unclassified
19. SECURITY CLASSIFICATION
OF ABSTRACT
Unclassified
20. LIMITATION OF ABSTRACT
None
NSN 7540- 01- 280- 5500 Standard Form 298 ( 2- 89)
Prescribed by ANSI Std. Z39- 18
298- 102 CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
 
i
PREFACE
This report has two basic purposes. The first purpose is to describe the development of California’s prospective 3- Tier Assessment System based on research carried out in four Southern California Department of Motor Vehicles field offices in 2001 and 2002.
The second purpose of the report is to present an “ Ecological Perspective” on driving and driver licensing that the first author developed and that is especially relevant to drivers who have been identified as having a “ driving- relevant limitation( s).” This perspective views the challenges posed by the driving environment as constantly changing and, therefore, drivers as constantly adjusting the demands of their driving tasks. Compensating for a driving- relevant limitation( s) is always or continuously “ done,” though with varying degrees of adequacy and consistency. The definitions and relationships among key concepts, such as “ driver- centered,” “ driving- centered,” “ driving wellness,” “ driving fitness,” and “ limitation- naivety”, are presented in detail in the Introduction and Rationale Section and elaborated on throughout the remainder of the report.
The Research and Development Branch of the California DMV, in the interest of stimulating discussion about the licensing of functionally- limited drivers, presents the Ecological Perspective without comment. Sufficient empirical evidence has not yet been obtained to accept or reject the individual concepts or broad relationships hypothesized by this perspective. A pilot project currently under way will attempt to obtain that evidence.
For now, readers are encouraged to read this report with care and reflect on the 3- Tier Assessment System developed by both authors and the Ecological Perspective conceptualized by the first author. Readers are further encouraged to use the latter as a heuristic to advance the safe- driving impact of driver licensing by departments of motor vehicles and related services provided by others, especially for customers with driving- relevant limitations. CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
ACKNOWLEDGMENTS
This project was made possible through the support of the California Office of Traffic Safety and the National Highway Traffic Safety Administration. This report was prepared by the Research and Development Branch, Licensing Operations Division, of the Department of Motor Vehicles. The opinions, findings, conclusions, and recommendations expressed in this report are those of the authors and not necessarily those of the California Department of Motor Vehicles or the State of California.
The authors would like to acknowledge with appreciation the many individuals who contributed to this study. First and foremost we are grateful for the exceptional contributions that were made to the success of this study by all of the Motor Vehicle Field Representatives and Licensing/ Registration Examiners who identified and tested the study participants in the four Department of Motor Vehicles Field Operations’ study offices: Santa Monica, Costa Mesa, Van Nuys, and Pasadena. They are named in the table below. We are also grateful for the tremendous support of the study- office managers named in that table and the general support of all the staff working in the study offices. Pilot testing was carried out in the Carmichael field office.
Three- Tier Study Offices and Staff
Office managers
Study offices
attached to study
Study LREs
Study MVFRs
Phase I ( Jun– Dec 01)
Santa Monica
Michael Dillon IV
Elizabeth Hernandez III
Pete Carranza I
Lola Craven
Danny Mersiehazen
Loraine Gilliam
Antonia Montes
Costa Mesa
Marilyn Busell V
Lynn Sosa III
Rosa Casas I
Sharon Langerman
Dinah Heimos
Charise Slach
Frances Ward
Phase II ( Jan- Jun 02)
Van Nuys
Robert Nelson IV
Mark Dragan III
Rosie Romero I
Dolores Orrante I
Lee Carlson
Jake Duran
Elsa Gutierrez
Ana Martinez
Odilia Moreno- Zunigo
Pasadena
Dixon Jones V
Elizabeth Fenner III
Mary Millsaps I
Aaron Lee
Lashonda Thompson
Leila Giraldo
Debbie Pulley
ii
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
 
iii
Headquarters Field Operations managers, John Rooney ( III) and Babette Williams ( V), supported the study offices in carrying out various data collection activities. Hannah Lee ( University of California Los Angeles graduate student) and Harrison Tanji ( Van Nuys Driver Safety Office) served as local study coordinators for study phases I and II, respectively. They were both exceedingly competent in the day- to- day running of the study.
Clifford J. Helander, retired and former Chief of the Research and Development Branch and David J. DeYoung, Chief of the Research and Development Branch, provided general direction. David DeYoung and most especially Leonard Marowitz ( Research Manager II) reviewed earlier drafts of the report and provided valuable comments for improving the study report. Mary Bobo, an undergraduate student assistant from California State University, Sacramento, contributed to the development of two checklists for law enforcement use in informing the Department of Motor Vehicles of its reasons for considering any particular driver hazardous enough to warrant a reexamination. Douglas Rickard, Staff Services Analyst and Debbie McKenzie, Associate Governmental Program Analyst in the department’s Research & Development Branch, formatted the report.
And finally, we are especially grateful to Jane Stutts, retired and former Associate Director for Social and Behavioral Research, University of North Carolina and Judy Geyer, Research Associate, University of California Berkeley Traffic Safety Center, for their review and comments on an early draft of this report. CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
HIGHLIGHTS
Would you be surprised to learn that among older driver- license- renewal applicants the ones who we need to worry the most about would likely pass a road test? That’s what we predicted when we looked at driving from an ecological perspective. In an ecological perspective, drivers are viewed as actively adjusting the demands of their driving environments and the demands of their driving tasks in accordance with their perception of their driving- relevant abilities and limitations. We confirmed this and other ecology- driven predictions with our study of California’s prospective three- tier driving- centered assessment system. Driving- centered is an ecological concept— it means taking into consideration when, where, why, and how individual drivers customarily drive. The traditional approach to assessment, taken by most DMVs, is driver- centered. Traditional assessment is focused on the accurate identification of high- risk drivers. It is an endpoint in the controlling and delicensing of these problem drivers. Traditional assessment does not take into consideration when, where, why, and how individual drivers customarily drive. The driving- centered Three- Tier Assessment System ( 3- Tier) described in this study report represents fundamental changes in the approach and objectives of driver assessment.
3- Tier is offered as the answer to the following fundamental question posed at the beginning of the study:
How can the DMV better identify and assess licensed drivers of any age who have acquired a driving- relevant functional limitation( s) so that the DMV, together with physicians, driving- rehabilitation specialists, and others can aid such drivers, if feasible, in driving safely by referring for physician- based evaluation and treatment, educating about driving- relevant limitation( s), recommending behind- the- wheel training, restricting ( conditional licensure), and so on?
Rather than an endpoint, 3- Tier fundamentally alters the purpose of assessment to be a starting point in extending the safe driving years of functionally- limited licensed drivers. This repurposing requires screening licensed drivers for marginal as well as more severe driving- relevant limitations. Marginally- limited drivers need to be educated about recognizing and avoiding the periodic everyday convergence of factors ( for example, the darkly clothed pedestrian at dusk inside the upcoming faded crosswalk) that substantially elevate the crash potentiating effect of their particular limitation. Repurposing assessment as a starting point also requires assessing drivers identified as having acquired a severe limitation( s) for whether they can manage to drive safely despite their limitation( s) under reasonably- specifiable licensing conditions.
iv
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
 
Driving- relevant functional limitations are decrements in driving wellness. Driving wellness is a phrase designed to focus attention on preserving and proactively improving driving- relevant functional abilities. A driver assessed as “ driving- well” would be a driver deemed free from driving- relevant functional limitations. The degree of driving wellness varies in accordance with the number, kinds, and severity of limitations in driving- relevant abilities, such as contrast sensitivity— the amount of contrast a driver needs to quickly detect objects that are important to safe driving, like other cars and pedestrians. Assessing a driver’s contrast sensitivity is a necessary part of assessing the driver’s degree of driving wellness, since good contrast sensitivity is critical to noticing potential road hazards under numerous everyday low- contrast viewing conditions, such as a light car in the fog, a dark car in the shade, or a darkly- clothed pedestrian at dusk inside the upcoming faded crosswalk.
Based on the Tier 1 and Tier 2 assessment of driving- relevant visual ability, mental ability, and physical ability ( the locomotor/ manipulative abilities necessary for operating vehicle controls), 3- Tier sorts drivers into one of three driving- wellness categories:
 Driving- Well— free from driving- relevant functional limitations.
 Somewhat Functionally Limited— one marginally- limited driving- relevant functional ability.
 Extremely Functionally Limited— two or more marginally- limited driving- relevant functional abilities and/ or one or more severely- functionally- limited driving- relevant functional abilities.
Tier 1 is brief and very easily administered. Four assessment tools ( ATs) comprise Tier 1:
 DMV’s current Snellen test of visual acuity.
 Chart- based test of contrast sensitivity.
 Brief cognitive screen ( recalling social security number).
 Structured observations for physical limitations.
v
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
Driver license renewal candidates who are unsuccessful on Tier 1 must complete a computer- based assessment of perceptual- response time ( processing speed) on Tier 2. Tier 2 is also comprised of one other AT, the department’s current written knowledge test of safe- driving practices and the laws and rules of the road. The latter AT would seemingly have fit naturally on Tier 1 but had to be placed on Tier 2 because a departmental decision was made to automate the knowledge test; thereafter, it more naturally fit with the other Tier 2 computer- based AT. In an operational system, driver- license- renewal candidates successful on Tier 1 would go on to take the knowledge test on Tier 2. Licensure is contingent on passing the knowledge test. Drivers who are successful on Tier 1, but who fail the knowledge test two or more times before passing it, would also be required to complete the computer- based assessment of perceptual response time.
In order to renew their license, drivers who are sorted into the extremely- limited driving- wellness category are required to pass a Tier 3 road test. 3- Tier makes a distinction between driving wellness and driving fitness. Even an extremely functionally- limited driver may be assessed on Tier 3 as “ driving fit.” Driving fit means that the driver’s level of risk for making a critical driving error ( CDE) would be expected to be consistently small in that driver’s customary driving environments and conditions and for that driver’s customary driving practices. A CDE is an action or inaction which has a high probability of precipitating an adverse- driving event if the consequences of the CDE are not attenuated by such factors as a lack of conflicting traffic. Adverse- driving events include the following negative outcomes: forcing other road users to take evasive action in order to avoid a crash, actually colliding with them, or nearly colliding with them. The degree of driving fitness is the level of risk for making a CDE. The degree of driving fitness is contingent on how well the driver manages— by regularly searching/ scanning for hazards, slowing down or speeding up, and so on— the different ongoing everyday challenges to maintaining a low level of risk for making a CDE. These everyday challenges derive from the combined effects of four factors:
 The driver’s driving- relevant limitations ( decrements in driving wellness).
 The driver’s driving- relevant abilities— especially proficiency in compensating for the fitness- diminishing effects of decrements in driving wellness.
 Ongoing changes in the demands of the driver’s customary driving tasks ( making left turns, merging with high- speed traffic, and so on).
 Ongoing changes in the demands of the driver’s customary driving environments and conditions.
vi
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
 
T
he last three factors moderate— attenuate or exacerbate— the fitness- diminishing effect of a decrement in driving wellness. For example, a driver assessed in a DMV field office as having severely- limited contrast sensitivity should not be expected necessarily to be at an elevated risk for making a CDE. Such a driver may routinely avoid driving in low contrast viewing conditions such as dusk or dawn, residential streets at night, and when it is foggy or raining. Nor, as indicated above, should a driver assessed as having only a marginal decrement in contrast sensitivity ( and who has no other decrements in driving wellness) be expected always not to be at an elevated risk for making a CDE. Even though a driver’s level of risk for making a CDE is a function of his or her degree of driving wellness, the degree of driving fitness is not determined by the degree of driving wellness. There is potentially a great difference between a driver’s expected risk for making a CDE, given their driving- relevant limitations, and a driver’s actual risk for making a CDE.
U
nlike measuring driving wellness in the DMV office, determining whether a driver is “ driving fit” ( consistently small level of risk for making a CDE) requires an assessment of the driver, with whatever driving- relevant abilities and limitations he or she may have acquired, made relative to his or her:
 Customary driving practices ( especially compensating practices).
 Customary driving tasks.
 Customary driving environments and driving conditions.
Note that what is “ customary” can be changed either by the driver, for example, by getting behind- the- wheel training in fully compensating for limitations, or by the DMV, for example, by restricting the driver from driving at night, dusk and dawn, or when it is foggy or raining.
Lacking in knowledge of ways of fully compensating for a limitation is one of the many factors detailed in the study report that variably constrain drivers from consistently and adequately compensating. These constraints on consistently and adequately compensating include mistaken beliefs about aging and driving. Effectively extending the safe driving years of functionally- limited licensed drivers will require that the following three myths— illusions of knowledge— be debunked:
 Older drivers are a functionally unitary group.
 “ Old age” causes driving- relevant functional limitations.
 “ Older” can be used as a sign of diminished capacity for driving safely.
vii
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
F
or example, in the case of the second bullet, expecting an inevitable decline in all driving- relevant functional abilities with advancing age is one of the bases for researchers posing such age- based driver- centered questions as: Are older drivers a problem? or Do older drivers pose a threat to the health and safety of other road users? In merely posing age- based driver- centered questions, let alone purportedly researching and answering them, researchers contribute to many older drivers mistakenly expecting a decline in all of their driving- relevant functional abilities with advancing age. In a 2003 California DMV survey, over 85% of the sampled older drivers indicated that “ being over the age of 65” is a “ sign of diminished capacity for driving safely.” Expecting a gradual decline with aging often results in elders failing to seek help from health professionals for driving- relevant functional limitations that could, in fact, be corrected, controlled, or at least slowed in their progression.
T
his report emphasizes the point that age- based driver- centered questions do not make sense— aging- associated “ declines” in visual, mental, or physical abilities are neither aging- determined nor necessarily irreversible. Aging does not cause driving- relevant functional limitations. For example, the study report cites the finding that 37% of the sampled license- renewal applicants aged 70- to 96- years- old performed as well as the 19- to 39- year- old renewals in scoring the shortest perceptual response time ( processing speed) that is achievable under the conditions of the study’s computer- based test of processing speed.
California’s prospective three- tier driving- centered assessment system serves as the starting point for initiating various means of extending the safe- driving years of functionally- limited licensed drivers. By extending their safe- driving years, 3- Tier aids functionally- limited drivers in maintaining their safe mobility which is viewed as a resource for everyday healthy living in our aging driving population.
viii
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
 
ix
The study report concludes with 22 recommendations for statewide implementation of a 3- Tier driving- centered assessment system. They are divided into four parts:
 Recommendations directly supported by this 3- Tier study.
 Procedural and policy changes recommended for a 3- Tier system to be effective.
 Complementary recommendations that make general good sense.
 Future multi- phase pilot study.
As noted in the last bullet, a future pilot study is recommended to evaluate the operational feasibility, costs, and customer and staff acceptance of the 3- Tier assessment system described in the 22 recommendations.
The reader is advised that it will be difficult to fully understand the recommendations and their rationales without first reading the study report.
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
TABLE OF CONTENTS
PAGE
PREFACE........................................................................................................................ ............... i
ACKNOWLEDGMENTS.............................................................................................................. ii
HIGHLIGHTS..................................................................................................................... ......... iv
INTRODUCTION AND RATIONALE......................................................................................... 1
Overview....................................................................................................................... ............. 1
Background..................................................................................................................... .......... 10
Early Efforts to Conceptualize a Viable Assessment System................................................ 10
The Idea of Tiered Vision Testing......................................................................................... 13
Maryland Model Screening Program..................................................................................... 13
Monash University ( Australia) Project.................................................................................. 15
AGILE.......................................................................................................................... ......... 17
California DMV Projects Culminating in the Present 3- Tier Assessment System................ 19
Ecological Perspective on Driving............................................................................................ 20
Compensating................................................................................................................... ..... 21
Constraints on Consistently and Adequately Compensating................................................. 22
“ Lowering” Barriers to Safe Mobility.................................................................................... 22
Driver- Centered “ Compensation”.......................................................................................... 23
Critical Driving Error............................................................................................................. 24
Driving Wellness.................................................................................................................... 25
Driving Fitness....................................................................................................................... 26
How and Why the Driver Competently Manages.................................................................. 27
Driving Wellness Versus Driving Fitness.............................................................................. 27
Driving Fit............................................................................................................................ . 31
Crash Record Cannot Be Used Alone as a Reliable Indicator of a Driver’s Degree of Driving Fitness....................................................................................................................... 32
Barriers to Maintaining Safe Mobility...................................................................................... 33
Historical Backdrop............................................................................................................... 33
Abiding and Transient Factors............................................................................................... 35
Barriers to Maintaining Safe Mobility: Factors That Variably Constrain Consistently and Adequately Compensating for Functional Limitations................................................... 35
An Overriding Barrier to Maintaining Safe Mobility— Expecting a Gradual Decline in Functioning with Advancing Age.............................................................................................. 39
Among Drivers....................................................................................................................... 39
Among Physicians.................................................................................................................. 41
x
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
 
TABLE OF CONTENTS ( continued)
PAGE
Among Researchers............................................................................................................... 41
Follow- up to Age- Based Driver- Centered Questions & Answers: Two Kinds of Logical Errors......................................................................................................................... .. 43
Logical Errors......................................................................................................................... 43
Adverse Consequences of Lumping Together All the Older Drivers.................................... 44
Screening Drivers for Constraints on Consistently and Adequately Compensating................. 45
Better Assessing Driving Wellness — Driving- Centered Assessment...................................... 46
Fundamental Driving- Centered Question.............................................................................. 46
Theoretical Backdrop to DMV Making a Driving- Centered Assessment of Driving Wellness....................................................................................................................... ......... 46
DMV Making a Driving- Centered Assessment of Driving Wellness.................................... 50
Operationalizing “ Driving- Well,” “ Somewhat Functionally Limited,” and “ Extremely Functionally Limited”............................................................................................................... 53
Validating the Three Operationalizations for Making a Driving- Centered Assessment of Driving Wellness................................................................................................................... 56
Two Null Hypotheses................................................................................................................ 58
Recommendations................................................................................................................ .... 58
METHODS........................................................................................................................ .......... 59
Data Collection Phases & Locations......................................................................................... 59
Assessors...................................................................................................................... ............ 60
Non- Driving ATs....................................................................................................................... 60
Snellen Test........................................................................................................................... 61
Knowledge Test..................................................................................................................... 61
Experimental Non- Driving ATs............................................................................................. 62
Pelli- Robson Chart- Based Test of Contrast Sensitivity......................................................... 64
Functional Acuity Contrast Test ( FACT).............................................................................. 66
Smith- Kettlewell Institute Low Luminance ( SKILL) Card................................................... 67
PC- Based Perceptual Response Time ( PRT)......................................................................... 68
Schieber’s Auto- Trails........................................................................................................... 69
Driving Information Survey................................................................................................... 70
Participant Selection and Processing......................................................................................... 70
Tier 1 Assessment Procedures................................................................................................... 72
Tier 2 Assessment and Retention Procedures........................................................................... 75
xi
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
TABLE OF CONTENTS ( continued)
PAGE
Road Test Referrals................................................................................................................ 75
Renewal Tier 1 “ Passes”........................................................................................................ 75
Renewal Tier 1 “ Fails”........................................................................................................... 76
Tier 3 Procedure...................................................................................................................... . 76
Operationalizing: Plan for Developing Two Cut Scores for the ATs that Yield a Range of Scores......................................................................................................................... .......... 78
Operationalizing: Scoring Procedures and Combining the Results of the Novel ATs.............. 85
Assessing Physical Limitation............................................................................................... 85
AT Scoring Procedures.......................................................................................................... 85
Combining Tier 1 ATs........................................................................................................... 86
Tier 2 AT............................................................................................................................. .. 86
Comparing Alternative AT Combinations................................................................................ 87
Validating: Statistical Analyses of Outcome Pattern Matching................................................ 90
Study Limitations.................................................................................................................... . 92
RESULTS AND DISCUSSION................................................................................................... 93
Part I: Finalizing 3- Tier ATs..................................................................................................... 93
Participants................................................................................................................... ......... 94
Erroneously Aggregating ( Amalgamating) Older Drivers..................................................... 97
Face Validity of Non- Driving ATs........................................................................................ 99
AT Intercorrelations............................................................................................................. 100
Cut Scores for the Novel ATs.............................................................................................. 103
Relationship between DMV Cataract Policy and Cut Sores for Contrast Sensitivity Testing........................................................................................................................ ......... 104
SDPE Usually Failed Due To SDPE Structured- Critical Driving Error ( CDE).................. 104
Pass/ Fail Screening Characteristic Of Novel ATs Vis- à- Vis Performance on SDPE......... 105
Proposed Tiers 1 and 2 ATs for Making a Driving- Centered Assessment of Driving Wellness....................................................................................................................... ....... 110
Part II: Outcome Pattern Matching.......................................................................................... 118
Validating The Three Operationalizations For Making a Driving- Centered Assessment of Driving Wellness......................................................................................... 118
Outcome pattern matching................................................................................................ 118
Excluding XFails from calculating Pearson correlation coefficients................................ 122
Elderly Renewal PP, SFail, and XFail Performance on Selected SDPE Elements.............. 125
xii
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
 
TABLE OF CONTENTS ( continued)
PAGE
Evidence of Elderly Renewal XFails Differentially Compensating for Their Functional Limitations......................................................................................................... 127
Part III: Exploring the Positive Correlation Between Knowledge Test Errors and Failing the SDPE..................................................................................................................... 132
Part IV: Exploring the Negative Correlation Between SSN Recollection and Having Crashed........................................................................................................................ ........... 135
RECOMMENDATIONS............................................................................................................ 139
Part I: Recommendations Directly Supported by the 3- Tier Study......................................... 139
Recommendation # 1 ─ Use the Unobtrusive Structured Observations & Cognitive Screen............................................................................................... 139
Recommendation # 2 ─ Use the Pelli- Robson test as 3- Tier’s measure of contrast sensitivity......................................................................................... 141
Recommendation # 3 ─ Eliminate screening acuity in both eyes together............................. 142
Recommendation # 4 ─ Tier 2 should have two parts: Perceptual Response Time Test and the current written knowledge test............................................ 142
Recommendation # 5 ─ Processing steps for 3- Tier driving- centered assessment................. 142
Recommendation # 6 ─ Require the completion of a Report of Vision Examination for Pelli- Robson XFails......................................................................... 149
Recommendation # 7 ─ Require all renewal applicants who are required to pass the knowledge test to complete Tier 1 and, if Tier 1 is failed, the PRT on Tier 2................................................................................... 149
Recommendation # 8 ─ Require Road Test Referrals to pass the knowledge test before taking SDPE..................................................................................... 149
Recommendation # 9 ─ Do not issue a temporary license to a Visual Acuity Referral who does not first pass the knowledge test...................................... 150
Recommendation # 10 ─ Require referrals for whom a road test is indicated to take the full battery of 3- Tier ATs, including an SDPE or a content- valid ADPE.................................................................................... 150
Part II: Procedural and Policy Changes Recommended for 3- Tier to be Effective:................. 150
Recommendation # 11 ─ Law enforcement officers should continue using the DMV Form DS 427 checklists to refer a driver for a priority or regular re- examination................................................................... 150
xiii
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
TABLE OF CONTENTS ( continued)
PAGE
Recommendation # 12 ─ Educate drivers assessed as somewhat functionally limited about their limitation...................................................................... 151
Recommendation # 13 ─ Give Renewals who XFail Tiers 1 and 2 a choice between an SDPE and a content- valid ADPE................................................... 152
Recommendation # 14 ─ Develop a reliable and content- valid ADPE.................................. 154
Recommendation # 15 ─ Distribute “ Driving Safely While Aging Gracefully” to the RBM eligible drivers over 60........................................................ 154
Recommendation # 16 ─ Establish a task force to improve the training of DMV staff in field operations procedures............................................................ 155
Part III: Complementary Recommendations that Make General Good Sense:....................... 155
Recommendation # 17 ─ Rescind procedural memo DL 2004- 10......................................... 155
Recommendation # 18 ─ Rescind procedural memo DL 2004- 13......................................... 156
Recommendation # 19 ─ Revise departmental policy to bar functionally- limited drivers who do not decline, then subsequently fail, the freeway portions of the SDPE, from freeway driving................................. 156
Recommendation # 20 ─ Establish an inter- divisional task force to better meet the objectives of knowledge testing..................................................... 156
Recommendation # 21 ─ Restrict 3- Tier XFails from all cell phone use ( and advise them to minimize conversational distractions) while driving........ 157
Recommendation # 22 ─ Task DMV R& D with formalizing procedures for initiating safe- mobility interventions and developing a formal means of screening drivers for the specific factors that may keep them from adequately and consistently compensating........................... 157
Part IV: Future Multi- Phase Pilot Study................................................................................. 157
REFERENCES..................................................................................................................... ..... 159
APPENDICES..................................................................................................................... ...... 174
APPENDICES
NUMBER PAGE
A Selected Items from the Driving Information Survey..................................................... 175
B Unobtrusive Structured Observations & Cognitive Screen............................................. 176
xiv
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
 
TABLE OF CONTENTS ( continued)
APPENDICES ( continued)
NUMBER PAGE
C Structured Observations Checklist and Worksheet......................................................... 178
D Intersection Problem....................................................................................................... 179
E California Vehicle Code - Authority to Make Study Participation Mandatory............... 180
F Exclusionary Criteria for Renewal By Mail................................................................... 181
G SDPE Structured- CDE Scoring Criteria......................................................................... 182
H Study Test Acceptability Ratings.................................................................................... 184
H- 1A Sample: Renewals– Acceptability of Pelli- Robson Contrast Sensitivity Test................ 185
H- 1B Sample: Renewals− Acceptability of Perceptual Response Time ( PRT)........................ 186
H- 2A Sample: Road Test Referrals− Acceptability of Pelli- Robson Contrast Sensitivity Test........................................................................................................................... ...... 187
H- 2B Sample: Road Test Referrals− Acceptability of Perceptual Response Time ( PRT)........ 188
H- 3A Sample: Visual Acuity Referrals– Acceptability of Pelli- Robson Contrast Sensitivity Test................................................................................................................ 189
H- 3B Sample: Visual Acuity Referrals– Acceptability of Perceptual Response Time ( PRT)........................................................................................................................... ... 190
H- 4A Sample: Staff– Acceptability of Pelli- Robson Contrast Sensitivity Test........................ 191
H- 4B Sample: Staff– Acceptability of Perceptual Response Time ( PRT)................................ 192
I Additional Screening Characteristics for the ATs in Tables 14- 17 and the Written Knowledge Test.............................................................................................................. 193
I- 1 Written Knowledge Test................................................................................................. 194
I- 2 Physical Limitation ( Physical)........................................................................................ 195
I- 3 Recall SSN...................................................................................................................... 196
I- 4 Intersection Problem....................................................................................................... 197
I- 5 Pelli- Robson Contrast Sensitivity................................................................................... 198
I- 6 SKILL Card, Dark Letters Correct.................................................................................. 199
I- 7 SKILL Card, Light – Dark Letters Correct..................................................................... 200
I- 8 FACT Row 2................................................................................................................... 201
I- 9 Tier 1 ( Physical, SSN, Pelli- Robson Contrast Sensitivity.............................................. 202
I- 10 Tier 1 ( Physical, Intersection, Pelli- Robson Contrast Sensitivity).................................. 203
I- 11 Tier 1 ( Physical, SSN, Intersection, P- R Contrast Sensitivity)....................................... 204
I- 12 Tier 1 ( Physical, SSN, FACT Row 2)............................................................................. 205
xv
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
TABLE OF CONTENTS ( continued)
APPENDICES ( continued)
NUMBER PAGE
I- 13 Tier 1 ( Physical, SSN, SKILL Card- Dark Letters)......................................................... 206
I- 14 PRT............................................................................................................................ ..... 207
I- 15 Tier 1 ( Physical, SSN, Pelli- Robson Chart) & Tier 2 ( PRT).......................................... 208
I- 16 Tier 1 ( Physical, SSN, FACT Row 2) & Tier 2 ( PRT)................................................... 209
I- 17 Tier 1 ( Physical, SSN, SKILL Card- Dark Letters) & Tier 2 ( PRT)............................... 210
J DMV Form DS 427, Law Enforcement Request for Re- Examination of Driver, and examples of the data used to construct the Form DS 427 checklists: Observed Driving Behavior and Driver Condition......................................................................... 211
J- 1 Law Enforcement Request for Re- Examination of Driver.............................................. 212
J- 2 Examples of data used to construct the Observed Driving Behavior checklist............... 214
J- 3 Examples of data used to construct the Driver Condition checklist................................ 216
LIST OF TABLES
1 Salient Differences between Driver- and Driving- Centered Assessment............................... 8
2 Obtained Values for Evaluating Three Plausible Pass Cut Scores for the PRT................... 81
3 Obtained Values for Evaluating Three Plausible Pass Cut scores for the Pelli- Robson Contrast Sensitivity Test.......................................................................................... 81
4 Scoring Procedures for the Novel Assessment Tools........................................................... 87
5 Counts ( A, B, C, and D) for Calculating Pass/ Fail Screening Characteristics vis- à- vis SDPE Performance............................................................................................................... 88
6 Prospective Tier 1 and Tier 2 Results Combined in Two Different Ways for Two Different Purposes................................................................................................................ 90
7 Study Participants................................................................................................................. 95
8 Means and Standard Deviations for Sampled Renewals for the Three Age Groups on Perceptual Response Time.................................................................................................... 98
9 Means and Standard Deviations for Sampled Renewals for the Three Age Groups on Pelli- Robson Contrast Sensitivity......................................................................................... 99
xvi
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
 
TABLE OF CONTENTS ( continued)
LIST OF TABLES ( continued)
NUMBER PAGE
10 Pearson Correlation Coefficients between Performance on Possible 3- Tier ATs, SDPE Failure, and 3- year Prior Involvement in One or More Crashes with the Corresponding P- values for Renewals Aged 70- to 96- Years- Old ( N= 152)..................... 101
11 Pearson Correlation Coefficients between Performance on Possible 3- Tier ATs, SDPE Failure, and 3- year Prior Involvement in One or More Crashes with the Corresponding P- values for Renewals Aged 70- to 96- Years- Old ( N= 148)..................... 102
12 Cut Scores for the Novel Assessment Tools....................................................................... 103
13 Number of SDPE Fails and Percent Who Failed by Committing One or More SDPE Structured- CDEs................................................................................................................. 105
14 Pass/ Fail Screening Characteristics of Possible Tier 1 ATs vis- à- vis SDPE Performance for Renewals Aged 70- to 96- years- old......................................................... 106
15 Pass/ Fail Screening Characteristics of Possible Tier 1 AT- Combinations vis- à- vis SDPE Performance for Renewals Aged 70- to 96- years- old.............................................. 107
16 Pass/ Fail Screening Characteristics of Possible Tier 2 AT vis- à- vis SDPE Performance for Renewals Aged 70- to 96- years- old......................................................... 108
17 Pass/ Fail Screening Characteristics of Possible Tier 1 AT- Combinations Combined with Tier 2 PRT vis- à- vis SDPE Performance for Renewals Aged 70- to 96- years- old............................................................................................................................ .......... 108
18 Pearson Correlation Coefficients and Corresponding P- Values for Individual ATs and Tiers 1 and 2 ATs Combined for Renewals Aged 70- to 96- years- old ( N= 137)......... 112
19 Performance on Proposed Tiers 1 and 2 Combined by Age and Participant Status........... 113
20 Percentage Failing SDPE by Age Group, Participant Status, and Performance on Tiers 1 and 2 Combined...................................................................................................... 119
21 Percentage Crashing At Least Once in the 3 Years Prior to Study Participation by Age Group, Participant Status, and Performance on Tiers 1 and 2 Combined................... 121
22 Pearson Correlation Coefficients and Corresponding P- Values for ATs With SDPE Failure and Having Crashed— Including and Excluding XFails— for Renewals Aged 70- to 96- years- old ( N= 137)............................................................................................... 124
23 Percent Who Performed as Indicated on Selected SDPE Elements by Performanceon Tiers 1 and 2 Combined ( PP, SFail, and XFail) for Renewals Aged 70- to 96- years- old ( N= 137)......................................................................................................................... 126
xvii
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
TABLE OF CONTENTS ( continued)
LIST OF TABLES ( continued)
NUMBER PAGE
24 Percent of 70- to 96- years- old ( N= 137) Renewal PPs, SFails, and XFails Who Declined Freeway Portion of SDPE ( first row) or Reported Often/ Always Avoiding Various Maneuvers, Conditions and Situations.................................................................. 129
25 Percentage Failing Written Knowledge Test by Age Group and Participant Status.......... 132
26 Percentage Failing SDPE by Age Group, Participant Status, and Having Passed or Failed the Written Knowledge Test.................................................................................... 134
27 Percentage Failing to Recall SSN by Age Group and Participant Status........................... 135
28 Percent of Renewals Aged 70- to 96- years- old Who Performed Poorly on indicated Variables by Whether They Recalled SSN......................................................................... 137
29 Percent of Road Test Referrals Aged 70- to 96- years- old Who Performed Poorly on Indicated Variables by Whether They Recalled SSN......................................................... 138
LIST OF FIGURES
1 Ecological perspective on the nature of the relationship between the degree of driving wellness and the degree of driving fitness, and the nature of the relationship between the latter and the probability of an adverse driving event....................................................... 30
2 Ecological perspective on the nature of the relationship between the driver’s degree of driving wellness and the level of risk for making a structured- CDE on a structured road test........................................................................................................................... ....... 50
3 Generally observed changes in the shape of the Renewals’ frequency distribution of driving- relevant functional ability with aging........................................................................ 79
4 Distributions of Perceptual Response Time for the Youngest and Oldest Sampled Renewals....................................................................................................................... ......... 83
5 Distributions of Pelli- Robson Letters Correct for the Youngest and Oldest Sampled Renewals....................................................................................................................... ......... 84
6 Age distributions of Renewals versus Road Test Referrals.................................................... 96
7 Age distributions of Renewals versus Visual Acuity Referrals.............................................. 96
8 Age distributions of Renewal PPs vs. SFails........................................................................ 115
xviii
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
 
TABLE OF CONTENTS ( continued)
xix
LIST OF FIGURES ( continued)
NUMBER PAGE
9 Age distributions of Renewal PPs vs. XFails....................................................................... 115
10 Age distributions of Road Test Referral PPs vs. SFails........................................................ 117
11 Age distributions of Road Test Referral PPs vs. XFails....................................................... 117
12 Percentage of Renewals Aged 70- to 96- years- old failing the SDPE and the percentage who crashed at least once in the three years prior to study participation........... 123
13 Percentage of reported days driving in a week for 70- to 96- year- old Renewal PPs, SFails and XFails.................................................................................................................. 128
14 Overview of recommended 3- Tier driving- centered assessment system for Renewals who Pass or SFail Tier 1 and Snellen fails whose visual acuity, after referral to a vision specialist, is corrected to 20/ 40 or better.................................................................... 143
15 Overview of recommended 3- Tier driving- centered assessment system for Renewals who XFail Tier 1and Snellen Fails whose visual acuity, after referral, is not correctable to 20/ 40 or better................................................................................................ 147
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
INTRODUCTION AND RATIONALE
The greatest obstacle to discovery is not ignorance – it is the illusion of knowledge.
Daniel J Boorstin
Washington Post, " The Six O'Clock Scholar" by Carol Krucoff ( 29 January 1984)
Your ideas are your cage.
Seung Sahn Soen Sa Nim
Only Don’t Know: The Teaching Letters of Zen Master Seung Sahn ( 1982)
Overview
This report describes the development of a three- tier driving- centered assessment system and the validation of one way for the DMV to make a driving- centered assessment of driving wellness. What is meant by “ driving-” versus “ driver-” centered assessment is discussed below in this brief overview of the study rationale.
California’s prospective three- tier driving- centered assessment system has three primary objectives:
 Screening licensed drivers for marginal as well as more severe driving- relevant functional limitations.
 Serving as the starting point for initiating various means of extending the safe driving years of functionally- limited licensed drivers, if feasible.
 Assessing drivers identified as having acquired a driving- relevant functional limitation( s) for whether they can manage to drive safely despite their limitation( s) under reasonably specifiable licensing conditions.
1
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
As recommended by the Older Drivers Project, created in January 2002 by the American Medical Association in partnership with the National Highway Traffic Safety Administration ( Wang & Carr, 2004; see also Carr, 2000): “ Driving cessation is recommended only after the safety of the driver cannot be secured through any other means” ( Wang & Carr, 2004, p. 143).
E
ven so, the overall objective of “ 3- Tier” is in fact to improve road safety.
Historically state motor vehicle departments have served first as gatekeepers: licensing most, but not all, driver license applicants— usually teens. In regards to already licensed drivers, the authors of the 2004 report on the Older Drivers Project observed that:
Traditionally, traffic safety efforts for the older population have focused on methods to identify unsafe drivers to enforce driving cessation, but driving cessation deprives the majority of older Americans of their primary form of transportation and has been associated with an increase in depressive symptoms. ( Wang & Carr, 2004, p. 143)
3- Tier would expand the California Department of Motor Vehicles’ ( DMV) licensing role to one that includes a spotlight on aiding drivers in maintaining their safe mobility in our aging driving population. This is one of the U. S. Department of Transportation’s ( 2003) strategies for achieving “ safe mobility for a maturing society” ( p. viii). Eberhard et al. ( 2006) noted in this regard that:
… the preferred policies and practices among government agencies and professional societies, such as the American Association of Motor Vehicle Administration ( AAMVA), AARP, AAA ( formerly American Automobile Association), American Association of State Highway and Transportation Officials, National Governors Association, and the private sector, have been to find ways to allow older adults to continue to drive as late in life as possible while maintaining safety. ( p. 4)
Consistent with the policies and programs of other state transportation and highway agencies, especially those in Florida, Iowa, Maryland, Michigan, and Oregon ( Stutts, 2005), Bill Cather, then California DMV Assistant Director for Legislation, made the following pertinent remarks at the March 7, 2005 United We Ride Mobility Summit:
[ California] DMV’s mission is not to take away your driver’s license. Quite to the contrary, our mission is to keep safe drivers on the road and driving for as long as we possibly can.… There are lots and lots of things that we can do to try to keep you mobile and to keep
2
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
you in your own car and still have our primary goal of protecting you and the other drivers on the freeways and highways of our state.… One of the things that we’re looking at and constantly trying to do better is assess drivers and facilitate rehabilitative services.… we think that many folks with today’s modern medicine can benefit from those rehabilitative services. We want to help facilitate that interaction and make sure that you are aware of the
available options. ( p. 10, http:// www. 4tacc. org/ MobilitySummit/ Proceedings- MobilitySummit- Part1- Draft% 20_ 2_. pdf)
Safe mobility is an aspect of health as a resource/ means for everyday living. The constitution of the World Health Organization ( WHO, 2007, p. 1) defines health as
… a state of complete physical, mental and social well- being and not merely the absence of disease or infirmity.
The enjoyment of the highest attainable standard of health is one of the fundamental rights of every human being without distinction of race, religion, political belief, economic or social condition.
The 1986 WHO Ottawa Charter for Health Promotion ( WHO, 1986, p. 1) noted that in order
… to reach a state of complete physical, mental and social well- being, an individual or group must be able to identify and to realize aspirations, to satisfy needs, and to change or cope with the environment. Health is, therefore, seen as a resource for everyday life, not the objective of living.
Safe mobility as a means for everyday living is regularly, and therefore, safely achieving one’s travel ( mobility) objectives which can range from accessing life’s necessities, such as medical appointments and food, to accessing social and cultural experiences ( TRB, 2005, p. vi; Dickerson et al. 2007; Oxley & Whelan, 2008). Extending the safe driving years of functionally- limited licensed drivers is the 3- Tier DMV means of aiding drivers in maintaining their safe mobility. 3- Tier assessment is intended to serve as a basis for a comprehensive licensing program. Drivers identified as having acquired a driving- relevant limitation( s) would be aided in extending their safe driving years. How? Functionally- limited drivers would be aided in improving their driving- relevant functioning ( remediation) and in adequately and consistently compensating for the crash potentiating effects of the specific limitation( s) that they have acquired ( Oxley & Whelan, 2008). This aid could take the form of DMV initiating a
3
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
variety of possible safe- mobility interventions. 1,2 Listed immediately below are five examples of safe- mobility interventions. Please note that in initiating a safe- mobility intervention a DMV may or may not involve itself in actually administering the intervention.
 Referring drivers identified by the DMV as functionally limited for physician- based evaluation and treatment. Eberhard et al. ( 2006) note that licensed drivers who have recently acquired a driving- relevant limitation may benefit from being treated by an ophthalmologist, neurologist, orthopedist, neuropsychologist, occupational therapist, physical therapist, podiatrist, or exercise physiologist.
 Educating drivers about the implications of their specific driving- relevant functional limitation( s) for their driving safely under specific driving conditions and providing functionally- limited drivers with specific ways of compensating for the effects of their limitations.
 Encouraging functionally- limited drivers to get behind- the- wheel training in compensating consistently and adequately. In said encouraging, the driver should be informed that compensating consistently and adequately is best accomplished with the aid of a certified driving rehabilitation specialist ( e. g., Wheatley & DiStefano, 2008).
 Instructing functionally- limited drivers in ways of adapting the drivers’ vehicles to each driver’s particular limitation( s). Dickerson et al. ( 2007) note that, “ Some vehicle modifications, such as easy- locking seat belts, visor extenders, steering- wheel covers to improve grip, and seat and back support cushions to relieve back pain or improve line of sight, do not require special training” ( p. 582).
1 Like closing a window on a cold winter’s day first permits the heating system to achieve a comfortable room temperature and then the closed window aids in maintaining that room temperature, so do the listed interventions first permit the functionally- limited driver to achieve safe mobility and then the outcomes/ results of the interventions aid the driver in maintaining that safe mobility. For example, as will be discussed in detail in the section Ecological Perspective On Driving, safe mobility may be in part maintained by adjusting the demands of the driving environment/ conditions. Therefore, DMV might aid functionally- limited drivers in maintaining their safe mobility by restricting their driving to daylight hours or restricting their driving to familiar and well- practiced routes. In abiding with such restrictions, the functionally- limited driver would reduce the demands of their driving environment/ conditions, and thereby, more readily maintain their safe mobility.
2 Renewing the license of a functionally- limited driver should be conditional on the applicant demonstrating or providing DMV with certification of successful completion of the relevant interventions identified by DMV ( Brainin, 1980). 4
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
 Licensing but formally restricting3 functionally- limited drivers to corrective devices, for example, corrective lenses, and/ or restricting them to the driving environments/ conditions under which they have demonstrated an acceptable level of driving fitness. These would be tangible and reasonably specifiable licensing conditions such as driving only during certain hours of the day and/ or driving only on certain familiar and well- practiced routes4.
Other approaches to aiding drivers in maintaining their safe mobility, such as ones based on self- assessment are listed and described in AARP’s Promising Approaches for Promoting Lifelong Community Mobility ( Molnar et al., 2007), and in Strategies and Tools to Enable Safe Mobility for Older Adults ( Eberhard et al., 2006; also see Dickerson et al., 2007; Eby, Molnar & Kartje, 2009; Stutts & Wilkins, 2003, and Stutts, 2005). The website, www. eldersafety. org, is an especially valuable resource for “ facilitating safe mobility for seniors.”
By incorporating interventions such as these into a systematic licensing program, it is expected that a DMV would effectively extend the safe driving years of functionally- limited licensed drivers. Raedt & Ponjaert- Kristoffersen ( 2000, p. 517) refer to such an approach as 3- Tier as focusing on “ possibilities for maximum mobility,” in contrast to the traditional approach that focuses predominately on drivers’ limitations. Marottoli ( 2007) and Marottoli et al. ( 2007) make the following timely observations:
Much of the literature on crash risk and driving performance focuses on specific diseases or impairments and whether they affect driving safety and mobility. Only recently has attention turned to potential interventions to enhance driving performance. ( Marottoli, 2007, p. 5)
There is a need to identify individuals at increased risk for crashes or poor driving performance and to determine if driving performance can be enhanced and crash risk lowered, thereby, maximizing and prolonging safe driving by older persons. ( Marottoli et al., 2007, p. 591)
3 Restricted/ conditional license.
4 In this regard, Staplin and Hunt ( 2004, p. 87) note: “ One way licensing agencies accommodate drivers with age- related diminished capabilities while still carrying out the mandate for public safety is to impose restrictions that either ameliorate the functional deficits or restrict the exposure of these individuals, effectively and gradually retiring them from driving.”
5
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
Tuokko, Rhodes, & Dean ( 2007) echo Marottoli ( 2007) and Marottoli et al. ( 2007):
Little, if any, research with older drivers has gone beyond the identification of the association between health conditions ( or illnesses) and driving ( e. g. accident rates, cessation of driving). Understanding the manner in which health conditions relate to driving difficulties may prove useful for identifying strategies to promote continued mobility in older adults. ( p. 389- 390)
Our findings suggest that it may be more fruitful to examine for specific health symptoms [ rather than conditions], and relate these to specific forms of driving difficulties. ( p. 393)
The traditional DMV- assessment focus on controlling and delicensing unsafe/ high- risk/ at- risk/ problem drivers is here referred to as “ driver- centered”: DMV regulates the driver. A driver- centered DMV is focused on the “ accurate identification of high- risk drivers” ( Eby & Molnar, 2008, p. 4). The 3- Tier focus on extending the safe driving years of functionally- limited drivers is here referred to as “ driving- centered”: If off- road absolute standards for visual, mental, and physical functioning are met— that is, the driver is not “ medically impaired,” but nonetheless, “ medically at- risk” for unsafe driving ( Dobbs & Carr, 2005), then the driving- centered DMV would aid/ regulate the functionally- limited drivers’ compensating ( driving) by identifying and initiating one or more relevant safe- mobility interventions5: referring for physician- based evaluation and treatment, educating about driving- relevant limitation( s), recommending behind- the- wheel training, restricting ( conditional licensure), and so on. Relevant interventions would be initiated before, and after, on- road assessments. Relevant interventions may be identified before, during, and after, on- road assessments. With the driving- centered DMV’s focus on extending safe driving years, driving rehabilitation and driving assessment necessarily somewhat intertwine ( Wheatley & DiStefano, 2008). If drivers are ultimately assessed as unable to drive safely under any reasonably specifiable conditions then they would be delicensed. A driving- centered DMV’s focus also includes identifying drivers having only one marginally- limited driving- relevant functional ability ( typically not deemed “ high- risk drivers”). Said drivers offer an opportunity to intervene early before their driving- relevant limitation, if not corrected in its initial- stages or not consistently and adequately compensated for, eventually contributes to them and/ or other drivers crashing ( e. g., Carr, 2000).
5 As indicated, a driving- centered focus does not preclude the development and use of absolute standards in the delicensing of drivers who are highly unlikely to be able to drive safely under any reasonably specifiable conditions, for example, a driver having a visual acuity worse than 20/ 200 best corrected with both eyes together or a driver having been diagnosed as likely having moderate ( middle- stage) or worse Alzheimer’s disease.
6
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
Early intervention of course does not guarantee no limitation- related crashes, but rather should attenuate the possibility. Screening only for so- called “ at- risk” ( of crashing) drivers due to having developed severe deficits would be a poor strategy for maintaining safe mobility. Salient differences between driver- and driving- centered assessment are detailed in Table 1. Some of the terms used in Table 1 remain to be defined in the remainder of the Introduction and Rationale.
3- Tier’s objective of extending the safe driving years of functionally- limited drivers and achieving this objective by aiding/ regulating drivers’ compensating is rooted in looking at driving from an ecological perspective ( Hennessy, 1992; van Lier, 2004). In an ecological perspective, drivers are viewed as actively and continuously adjusting the demands of their driving environments/ conditions and the demands of their driving tasks in accordance with their perception of their driving- relevant abilities and limitations. This and other aspects of an ecological perspective on driving are discussed below in a subsequent section of the Introduction and Rationale. In reviewing various perspectives on driving, Ranney ( 1994) made the following pertinent observations that for the most part have remained the case in the ensuing 15 years:
In the highway safety field, priority has generally been given to identifying risk factors through epidemiological studies of accident causation. The result has been an overreliance on accidents and accident- causing behaviors, and a failure to consider driving behavior within the broader context of transportation for a particular purpose ( e. g. to get from home to work). ( p. 734)
… moving the focus of research away from the driver in isolation [ emphasis added] and focusing more on the interaction of the driver and driving situations would improve the ecological validity of roadway safety research. ( p. 747)
Road safety research is ecologically valid, and thus of value for DMV licensing- policy making, to the extent that it incorporates into its methods a consideration of when, where, why, and how individual drivers customarily drive ( e. g., Shinar, 2007). As indicated by Ranney ( 1994); Marottoli ( 2007); and Tuokko, Rhodes, and Dean ( 2007), road safety research has typically been driver- centered. Road safety research has typically not taken into consideration when, where, why, and how individual drivers customarily drive.
7
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
Table 1
Salient Differences between Driver- and Driving- Centered Assessment
Assessing
Approach
driving wellness
Determining
whether driving fit
Driver centered ( traditional approach)
Assessment is an endpoint in controlling & delicensing unsafe/ high- risk/ problem drivers
What does this mean for?
DMV staff: Easy to administer
Customer: Not aided in maintaining their safe mobility
Physician: ‘ Delicensing’ role. Resists working with DMV
Researcher: Easy to studya
Determine whether
‘ pass’ or ‘ fail’ off- road assessments of driving- relevant functional abilities.
Theoretically, the cut scores are crash- predictive.
Theoretically, determining whether a driver is driving fit is not necessary if off- road tests are crash predictive. Nonetheless, a failure of an off- road test may be followed by office- based on- road testing.
Driving centered ( 3- Tier ecological approach)
Assessment is a starting point in extending the safe- driving years of functionally- limited licensed drivers
What does this mean for?
DMV staff: Challenging & rewarding to initiate relevant interventions
Customer: Actively aided in maintaining their safe mobility
Physician: ‘ Intervention’ role. Willing to work with DMVb
Researcher: Challenging to study
If off- road absolute standards are met ( visual, mental, & physical), then functionally categorize as:
Driving well.
Somewhat functionally limited: one marginally- limited driving- relevant functional ability.
Extremely
functionally limited.
1) Initiate relevant interventions:
medical referral, educating,
behind- the- wheel training.
2) Conduct on- road assessment.
Primary objective for—
Somewhat functionally limited:
early lowering of barriers to safe mobility— educating about limitation.
Extremely functionally limited:
Driving- based determination of whether the level of risk for making a critical driving error is consistently small in customary driving environments/ conditions & for customary driving practicesc.
a Straightforward experimental design and statistical analyses.
b “… unlike the current physician role in assessment and reporting, which is perceived as having many negative effects on patient well- being including loss of license, loss of driving, and decreased out- of- home mobility and activity, interventions have potential psychological and practical benefits in enhancing safety and prolonging safe driving and mobility” ( Marottoli, 2008a, p. 135).
c For example, can the driver keep from making a critical driving error or a large number of important, but less critical errors, on a content- valid road test ( this would include infrequent though important regular destinations)? 8
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
The remainder of the Introduction and Rationale is divided into 11 major sections:
 Background.
 Ecological Perspective on Driving.
 Barriers to Maintaining Safe Mobility.
 An Overriding Barrier to Maintaining Safe Mobility— Expecting a Gradual Decline in Functioning with Advancing Age.
 Follow- Up to Age- Based Driver- Centered Questions & Answers: Two Kinds of Logical Errors.
 Screening Drivers for Constraints on Consistently and Adequately Compensating.
 Better Assessing Driving Wellness — Driving- Centered Assessment.
 Operationalizing “ Driving Well,” “ Somewhat Functionally Limited,” and “ Extremely Functionally Limited.”
 Validating the Three Operationalizations for Making a Driving- Centered Assessment of Driving Wellness.
 Two Null Hypotheses.
 Recommendations.
9
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
B
ackground
Early Efforts to Conceptualize a Viable Assessment System
As a matter of historical interest two early efforts to conceptualize a viable assessment system will be described. The descriptions are taken from the Janke ( 1994) review ( pp. 220- 221).
A model older driver licensing and improvement system
Pursuant to a contract with the National Highway Traffic Safety Administration ( NHTSA), Brainin ( 1980) suggested the following model system for all drivers above some arbitrary age and elderly drivers under that age who have reduced abilities for driving, sometimes because of medical conditions. The system involves distribution of an age- specific manual, consideration of driver history, medical screening, and assessment by nondriving and driving tests before a licensing decision is made.
People enter the model system, Brainin ( 1980) noted, in a variety of ways. Some states, of course, require road tests for drivers above a certain age. Absent such a requirement, if a state has an in- person renewal process for elderly drivers, license examiners can be trained to spot restricted- ability drivers. Other ways to enter the system are voluntarily; through accumulation of a sufficiently bad driving record; or upon referral from rehabilitation groups, health care personnel, relatives or friends, and others. Each individual entering the system is given a manual specifically geared to older drivers, upon which the later knowledge test will be based.
Prior to testing, each individual's driver record is reviewed. If the reason for any excess of accidents or violations can be determined, a recommendation is made for rehabilitation, corrective action, or license withdrawal. The rehabilitation programs, Brainin ( 1980) mentioned, can be administered by licensing agencies and may incorporate warnings, discussions with a driver improvement analyst, license restrictions, and/ or a specific driver improvement program. More commonly there is either no apparent driving problem or the reason for such a problem is not known, so the driver moves to the next stage.
In the next stage drivers may undergo medical screening and evaluation, although they are first checked by a driver licensing examiner to determine if this is obviously
10
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
necessary. Brainin ( 1980) noted that NHTSA has sponsored examiner- training programs to educate examiners in making this kind of determination. Medical evaluation, if necessary, can be accomplished in several ways— through an examiner's application of preexisting medical criteria, through scrutiny by a medical advisory board, or through an individual physician's examination. In any case, the driver is certified or not certified as being medically fit to drive.
License restrictions are considered if the driver is not medically fit, as are assistive devices and special training. This determination is made outside of the licensing agency. The driving privilege will be withdrawn in cases where no remediation is judged possible, but the individual may be referred to a social service agency for assistance in meeting mobility needs.
If the driver is medically fit, or if rehabilitative measures have been successful, a series of tests must be passed— traffic- law knowledge, an expanded vision test, and an in- car performance test specifically designed for older drivers to elicit unsafe behaviors characteristic of that age group ( left- turn difficulties, for example). At all of these testing stages, failure leads to reconsideration of restrictions and other means of reducing risk. Those who fail the performance test for suspected medical reasons ( and have not been medically screened before) now go through a second medical screening and evaluation process. Successful completion of this process will allow the driver to retake the performance test. Drivers for whom the conditions underlying their driving problems could not be diagnosed previously may be diagnosed in this stage, given the benefit of knowledge of their test performance.
As a result of the system described above, all drivers will be issued an unrestricted license, a restricted license, or no license. Former drivers who fail may reenter the system at a later date. While Brainin ( 1980) admitted that his model is relatively complex and costly, and will probably never exist completely, he expressed the hope that it will point licensing in the correct direction— that of maintaining the safe mobility of the elderly driver.
Remedial licensing – NPSRI
National Public Services Research Institute ( NPSRI; McKnight & Stewart, 1990) outlined a competency- based driver assessment system, distinguishing four stages of licensing— pre-, new, renewal, and remedial. Our concerns here are with the remedial
11
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
licensing stage, which deals with diminishing of competency and ways in which to help drivers recognize and adapt to this. McKnight and Stewart identified four strategies:
Reduce exposure by limiting the amount, time, and place of travel.
Reduce situational demands by using help from passengers ( e. g., navigational assistance), or through use of appropriate vehicle types, sizes, accessories, and special aids to driving.
Maintain physiological competence ( health) through exercise, rest, medicine, and diet.
Avoid conditions that cause deterioration in performance— e. g., fatigue, alcohol, and drugs.
Remedial licensing, they noted, can be handled by incorporating it into the renewal process. ( However, in the case of a driver reported to the department for possibly hazardous driving, handling may need to be more expeditious than this.) Licensees in the upper age ranges may be provided a manual and administered a test focusing upon those competencies identified as being pertinent to their age group. The material can be integrated into a special version of the renewal manual and test, or administered as a supplement.
Automated testing for psychophysical screening to identify drivers who have diminished competency was strongly recommended by McKnight and Stewart. ( The NPSRI test battery has been described in Part 3.) Automation, they felt, would enable use of a wide range of test stimuli, rapid change from one test situation to another in order to assess different competencies, and use of testing sequences that change as a function of ongoing test performance ( adaptive testing) in order to achieve maximum efficiency and minimum testing time. The technology is now available, as the authors wrote, to automate the testing of knowledge, vision, perception, and a broad range of psychophysical functions.
The two assessment plans outlined above, both focusing on older drivers, are different from many that were proposed in the past in that they can be incorporated into the regular renewal process, as contrasted with reexamination programs that apply only to drivers whose abilities have been called into question.
12
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
The Idea of Tiered Vision Testing
Not a fully- developed assessment plan but a thoughtful paper, written by Shinar and Schieber ( 1991), addressed visual requirements for licensing older drivers, and introduced the idea of tiered vision testing. Tiering was thought to be one of two alternative methods for effectively monitoring and controlling visually impaired drivers who cannot necessarily be assumed to restrict themselves appropriately. 6 The first tier, in their view, would consist of standard tests and emerging techniques administered at licensing stations. Those who failed this first tier would be referred to a centralized testing center or private practitioner for a comprehensive clinical evaluation which would include additional measures such as motion detection and effective field of view. The authors wrote that their first alternative is already operating in a rudimentary way. In California this would be through the DL62 process7. Shinar and Schieber, however, questioned the cost- effectiveness of “ periodic gross vision screening at the state licensing stations” ( p. 515), given that in California extremely large- scale studies ( on renewal by mail) had found no traffic- safety effect of renewal testing ( e. g., Kelsey, Janke, Peck, & Ratz, 1985). That finding, however, was interpreted by the California researchers as an indication that renewal testing, as it was then, was inadequate. This was one of the chief motives spurring development of an improved assessment system.
Maryland Model Screening Program
Perhaps the most developed functional capacity screening system is the “ Model Driver Screening and Evaluation Program” funded by NHTSA in 1996, and conducted by researchers from the Scientex Corporation who are now affiliated with TransAnalytics LLC ( Staplin, Lococo, Gish, & Decina, 2003; also see Staplin & Hunt, 2004). In its major effort, the “ Maryland Pilot Older Driver Study,” a battery of functional tests was developed and administered in Maryland Motor Vehicle Administration ( MVA) sites and in the community. Screening data were collected and analyzed for three separate samples of drivers aged at least 55. These data represented 1,876 drivers who visited MVA offices for license renewal or other transactions, 366 drivers referred by various sources to the MVA for medical evaluation because of suspected driving- related
6 The second alternative, apparently preferred by Shinar and Schieber, would involve requiring older drivers to present a certificate attesting to their good vision from a licensed vision specialist. They wrote, “ The inclusion of state- specified vision tests would then become part of a standard evaluation. This approach would ensure better screening of elderly drivers and provide them with professional help to improve their vision” [ p. 515].
7 The DL- 62 process is one of referral, by way of driver license form 62, to a vision specialist for applicants who fail the DMV screening standard for visual acuity. Upon coming back to DMV with the specialist’s report and, if prescribed, new corrective lenses, the applicants’ acuity is retested.
13
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
impairments, and 266 drivers in an affluent residential community for seniors who used the services of a mobile MVA office that made periodic visits to their facility.
Analyses were done to relate drivers’ test performance, representing their functional status at the time of testing, to prior and subsequent traffic crashes and moving violations as gleaned from state records. The odds of incurring a crash or committing a moving violation within the time frame of one year before testing to an average of two years afterward were calculated. The odds ratio for drivers who failed a particular screening test, as opposed to those who passed it, served to indicate the usefulness of the test in a licensing context.
Tests that were best able to predict undesirable road safety outcomes included:
 Motor- Free Visual Perception Test/ Visual Closure Subtest. This is a test of visuospatial abilities in which subjects are shown a pattern lacking some elements, and are asked to indicate what the pattern would be if complete. According to Staplin et al. ( 2003), the test measures the ability to visualize whole objects or patterns when there are missing elements and only partial information is available. Within the battery used, performance on this test showed the strongest relationships to safety outcomes, both crashes and traffic convictions.
 Trail- Making, Part B ( Trails B). This was significantly related both to crashes and to moving violations. It is based on a test with a long history as a valuable neuropsychological instrument. The original paper- and- pencil test, due to Reitan ( 1955, 1958), required the person being examined to search for and connect randomly arranged numbers 1- 13 and letters A- L in order but in alternating sequence, resulting in a 1- A, 2- B,… pattern. Speed is measured as the dependent variable. Trails B calls on an assortment of perceptual/ cognitive abilities, including complex conceptual tracking, directed visual search, visuomotor coordination, and short- term memory. In the Maryland study, a PC- based version called “ Dynamic Trails” was used in the battery. Dynamic Trails is a PC- based test that preserves the basic task described above, but the letters and numbers are presented upon the moving image of a freeway- driving scene, to introduce distraction. Both speed and errors were measured.
 Useful Field of View Subtest 2. This component of the PC- based Useful Field of View test yields a combined measurement of information processing speed and divided attention. Staplin et al. noted that differences in the size of the “ useful field of view” were
14
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
not measured. In contrast, Useful Field of View Subtest 1 ( PRT) was used in the present 3- Tier study, as it was in the Novato pilot study ( Janke, 2001). This first component of the computer- based Useful Field of View test yields only a measurement of information processing speed. See the Methods section for a description of the measurement of PRT.
In addition to the above, other measures showing significant promise were Delayed Recall, Rapid Pace Walk, and Head/ Neck Rotation, measuring respectively working memory, strength and mobility of the legs, and flexibility of the neck and upper torso.
Monash University ( Australia) Project
Stage 1 of this project was to develop a model license re- assessment program for older drivers ( see Fildes et al., 2000). In stage 2 of the project ( Fildes, et al., 2004), the first study evaluated the stage 1 model procedure in an Australian licensing environment, and a second study evaluated the effectiveness of four candidate- screening tests in predicting performance on a standardized road test as an indicator of potential crash risk. The discussion here will focus on the second study of stage 2. The goal was to identify effective licensing tools for “ older and functionally- disabled drivers,” using the screening tests to provide two thresholds. The lower one would identify drivers who could be considered unsafe, and would require a case review to consider such countermeasures as restrictions, temporary suspension, retraining, or rehabilitation. The upper threshold would identify drivers who could be considered safe. The area between the two thresholds would represent drivers who were of indeterminate safety and who, therefore, would be required to undergo another level of assessment. The tests studied were:
 Gross Impairments Screening Battery of General Physical and Mental Abilities ( GRIMPS), developed by staff of Scientex Corporation ( later TransAnalytics, LLC). It is composed of 11 subtests measuring abilities thought to be necessary for safe driving and at risk of decline in advanced age. The measures included in the GRIMPS battery, several of which have been mentioned above, were rapid- pace walk, foot- tapping, overhead arm reach, head/ neck/ upper body rotation, Motor- Free Visual Perception test ( Visual Closure subtest involving selecting the most probable completed versions of incomplete stimuli), cued recall, delayed recall, scan test ( categorizing examinees’ scanning patterns as systematic [ normal], erratic, and neglectful), Parts A and B of the Trail Making Test, and visual acuity ( high and low contrast).
15
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
 CALTEST. According to the report, two of these tests were taken from the battery developed for California’s cooperative venture with NHTSA in the mid- 1990s. However, it should be noted that the California studies, unlike the Monash study, used only the first UFOV subtest ( PRT). The tests included:
o Autotrails ( an automated version of Trail Making Part A, or Trails A, developed by Frank Schieber ( Heimstra Human Factors Laboratories, University of South Dakota). This is described below in the Methods section, because it was considered for the updated 3- Tier system that is the subject of this report. Involving visual search and attention, Autotrails requires examinees to touch, in numerical order, each of 14 randomly arranged numbered circles superimposed on a ( stationary) traffic scene. The number of errors and time to completion were recorded.
o Visual Awareness’ UFOV or Useful Field of View test ( Ball & Owsley, 1991; Owsley, McGwin, & Ball, 1998). The Monash investigators used the last two components of this three- part PC- based test: divided attention and selective attention. The selective attention task is the same as the divided attention task with the addition of distracters.
o HPT or Hazard Perception Test ( Congdon, 1999). The task here was an adaptation of a test, developed by VicRoads in Australia, requiring examinees to assess potentially hazardous video situations. It was never used in California. The adaptation contained six subtests, video sequences of traffic scenes selected from the full set of HPT items. Unlike the scoring system on the original test, the number of correct responses was defined as the number correct on either the first or second trial.
 DriveABLE ( DriveABLE Inc., 1997). As used in the Monash project, DriveABLE included six computer- based tasks. These yielded measures of motor speed, useful field of view, judgment of gap size, attention shifting, executive functions, and component driving abilities ( videos of traffic situations about which examinees were required to make judgments). Because the battery is proprietary, in place of scores for the individual tests the experimenters were furnished a confidence rating for each examinee indicating his or her probability of road- test failure ( i. e., failure of the DriveABLE standard road test).
16
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM

Elemental Driving Simulator ( EDS, Life Science Associates, New York). This test, as its name implies, was presented on a computer- based driving simulator. With the examinee seated at a computer and responding by means of steering wheel/ turn signal and accelerator controls, six scores were derived. These reflected steering control, reaction speed, field of vision, adjustment to changed circumstances, self- control ( number of errors on the most difficult task, a contingent differential response), and consistency ( the difference between mean and median response time).
T
hese tests tapped the realms of visual abilities, cognitive abilities, and other bodily abilities thought to be relevant to driving safely, any of which may decline in the course of aging.
The study was conducted in two steps. In the first, the ( volunteer) participants, all of whom were near their 80th birthday or older, completed one of the four screening batteries. This yielded almost 400 valid assessments. The EDS was discarded, in part because of unreliability, and in step 2 a further 560 elderly participants were assessed on the three remaining screening tests. Participants’ performance on a standard driving test–– the New Zealand road test required of drivers 80 or older–– was used as a measure of safety risk. Analyses showed that there were significant relationships between performance on this test and the following ones:
 GRIMPS ( overall score, rapid pace walk, foot tap, delayed recall, Trails B, and visual acuity ( high contrast and low contrast).
 CALTEST ( Autotrails, UFOV Selective Attention).
 DriveABLE.
The authors recommended implementation of the reassessment model and a much larger field trial of the more promising screening batteries with a prospective evaluation of the association between test results and crashes. This would both establish the predictive accuracy of a selected test or tests, and highlight any need for further test refinement.
AGILE
AGILE, which stands ( in English) for “ AGed people Integration, mobility, safety and quality of Life Enhancement through driving,” is a project, conducted within the European Union, which seeks to “ develop a new set of training, information, counseling and driving ability assessment
17
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
and support tools for the elderly, evaluating their full range of physical, cognitive, behavioral and interaction abilities . . . .” ( Breker, et al., 2003, p. 2). To this end, participants in the project have written a series of reports ( AGILE Deliverables) that are posted here: http:// www. agile. iao. fraunhofer. de/ deliverables. html. One of these, AGILE Deliverable 2.1 ( Middleton, et al., 2003), addresses the variability in procedures for assessing fitness to drive found in a survey of EU countries. AGILE urges that there be a pan- European driving- assessment process, containing a standardized set of tests and aimed particularly at elderly drivers ( age undefined), who would often enter the system on medical grounds. They have found a lack of standardization in procedure across sampled countries, with no agreed- upon criteria for invoking the assessment process, no agreement on the specific medical conditions that would require assessment, and little knowledge regarding guidelines and standards relating to specific driving- relevant functional defects. AGILE Deliverable 5.2 ( Arno and Boets, 2004) describes in detail their proposed three- tier assessment system:
… the assessment primarily aims at evaluating the cognitive functional status of the elderly driver. When cognitive decline is detected, the assessment further focuses on the detection of possible compensation mechanisms. ( p. 6)
The overall AGILE assessment procedure involves a medical pre- screening [ by a general practitioner, physical therapist, or occupational therapist] to identify early age- related conditions ( e. g. dementia) with implications for functioning. When a medical condition with possible implications for driving is diagnosed, an evaluation of functional skills is proposed via paper and pencil as well as computerised tests ( screening). When scores are below specified thresholds, referral for further in- depth assessment should be recommended at specialised driving assessment centers, where trained psychologists can assess specific functional deficits through neuropsychological assessment. Any symptom indicating a functional deficit considered important for safe driving has then to be further evaluated by means of a behind- the- wheel test, which can be performed in a driving simulator and/ or on the road in real traffic conditions. Assessment should not focus on driving skills only but on safe driving behaviour. The on- the- road assessment should indeed allow to evaluate whether age- related cognitive decline is - at least partly - compensated by better awareness of risks or improvement in hazard perception, which are skills developed through experience. ( p. 18)
18
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
California DMV Projects Culminating in the Present 3- Tier Assessment System
Two separate California DMV projects led up to the present 3- Tier assessment system. One of these projects was a cooperative venture, starting in 1993, between the California Department of Motor Vehicles ( DMV) and the National Highway Traffic Safety Administration ( NHTSA). Its aim was to identify functions important to test, and suitable tests, for a tiered licensing- agency assessment of older drivers. The goal was not to develop a complete, integrated assessment system like the recommended 3- Tier system of the present study, but the idea of 3 tiers of testing arose very early. Page iv of the Executive Summary of the first report of the NHTSA- DMV project ( Janke, 1994) states that the report “ offers suggestions for a three- tier assessment system which might be used by licensing agencies to ( 1) identify drivers with possibly driving- related impairments ( first tier), ( 2) assess those identified further, to estimate the degree to which any impairments would be likely to affect driving ( second tier), and ( 3) test their on- road driving performance in a standardized manner ( third tier).”
Janke ( 1994) thus identified, similar to other authors ( e. g., Shinar & Schieber, 1991), the usefulness of an assessment system containing three testing tiers that have a progressively increasing relationship to driving. Her 1994 report also identified common aging- related medical conditions and their effect on driving, described existing driving and non- driving assessment tools ( ATs) for measuring functional abilities judged necessary for safe driving, and discussed elderly driver- licensing programs and licensing provisions in selected states. From the important driving functions and ATs studied for the literature review, promising ATs were selected and piloted in a DMV field office ( the Santa Teresa office in San Jose). A preliminary report came out of this ( Janke & Eberhard, 1998) that also described independent work at that site by the Scientex Corporation, which studied intersection negotiation using as subjects the same older drivers ( Staplin, Gish, Decina, Lococo, & McKnight, 1998). A second study to identify suitable 3- Tier ATs was conducted in Novato’s ( Marin County, California) Buck Center for Research in Aging; participants were a group of elderly volunteers. The Santa Teresa and Buck Center studies are described in Janke & Hersch ( 1997) and Janke ( 2001). These two studies, for the most part, piloted different ATs, using road test performance as the principal criterion measure, and recommended their use as is, their further development, or their abandonment. The ATs studied and recommended here to be part of the final 3- Tier assessment system were among the most promising survivors of the Santa Teresa and Buck Center studies.
The ATs chosen here are to be administered by field- office staff, under field- office conditions, within a necessarily limited amount of time. In these circumstances too much complexity can be
19
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
a drawback. While a simulator test, for example, showed great promise in Santa Teresa when administered only by the second author, Janke and Hersch ( 1997) warned that:
Into the foreseeable future, the more complex simulator tests would probably be better administered by professionals like occupational therapists than by driver licensing technicians. Administering these tests is a demanding task. It is necessary not only to know how to “ run” the tests but also how to monitor the subject knowledgeably in case his or her performance is degraded because of an equipment problem, failure to understand instructions, or some other factor. If there is a simple misunderstanding of instructions, it is important for the test administrator to be able during initial practice trials to detect the problem and clarify the subject’s task. Long- term experience in administering tests to functionally- limited persons, and in particular [ administering] the specific test being used, would do much to assure valid results. ( p. 193)
The other project that led up to the present 3- Tier assessment system is Hennessy’s ( 1995) research into visual and perceptual ATs, conducted as a component of DMV’s driver competency enhancement program which called for an enhanced vision- test system. ATs studied were the Pelli- Robson chart- based contrast sensitivity test, the Smith- Kettlewell Low- Luminance ( SKILL) card, the Berkeley Glare Tester, standard and attentional visual field using a modified Synemed perimeter, and the Visual Attention Analyzer version of the Useful Field of View test ( UFOV, including all three modules, the first being what we call here, perceptual response time [ PRT]). Using prior crashes rather than road- test performance as a criterion, and self- reported levels of different forms of self- restriction as moderating variables, Hennessy found that the most promising ATs were the Pelli- Robson chart and PRT. These ATs–– which were also studied, respectively, at Santa Teresa and Buck Center as part of the agreement with NHTSA–– are a large part of the recommended 3- Tier system. Note that the two projects found the same ATs useful in spite of the fact that they used very different criterion measures.
Ecological Perspective on Driving
Ecology is concerned with the nature of the relationships and the interactions between organisms and their environment ( the physical and biological factors affecting them and affected by them). Taking an ecological perspective on driving means first keeping in mind that in
20
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
driving and in preparing for driving, drivers continuously actively adjust— in accordance with their perception of their driving- relevant abilities and limitations:
 The demands of the driving environment/ conditions— for example, by avoiding nighttime driving or by restricting driving to familiar and well- practiced routes.
 The demands of the driving task— for example, by slowing down or making three right turns rather than making a left turn.
In other words, drivers are actively and continuously compensating ( e. g., Powers, 1973; Ranney, 1994) in maintaining their safe mobility.
Compensating
In regards to the importance of researchers keeping in mind drivers’ continuously compensating, observations made by Näätänen and Summala’s ( 1976) over 33 years ago still remain generally relevant.
A profound misunderstanding of the basic nature of the driver’s task by many workers in the field has led research in fruitless directions: little attention has, for example, been paid to the driver’s ability to compensate for changes in the degree of difficulty of traffic situations by modifying his efforts ( attention, vigilance).… Driving indeed should not be understood as involving a forced- pace task in which the driver principally has only a responsive role in his interaction with the traffic situation; instead his active role and initiative most [ of the] time on the road should be given sufficient notice. ( pp. 36- 37)
Compensating is universal. Drivers of all ages and states of health continuously adjust their driving both in response to and in anticipation of challenges to maintaining their safe mobility. When youthful and relatively inexperienced drivers fail to detect a hazard in a timely manner–– commonly because of inadequate visual search or immaturity- based excessive speed–– they often are able to compensate by using their quick reflexes to brake or steer. When experienced drivers notice a decrement in any of their driving- relevant abilities ( noticing the decrement, at least at some level, is the necessary first step), they commonly minimize challenges to maintaining their safe mobility by avoiding problematic driving environments/ conditions like nighttime driving, freeway merging, or unfamiliar areas. Based on Michon’s ( 1985) hierarchical conceptualization of driving as concurrent activity at three different time scales of change ( strategic,
21
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
tactical/ maneuvering, and vehicle control), Ranney ( 1994) illustrated the “ continuously- adjusting” nature of compensating:
… changes in trip plans, such as the avoidance of rush- hour or nighttime driving by older drivers ( Planek & Fowler, 1971), are examples of strategic- level compensations. Adjustments to safety margins, such as the rejection of a higher percentage of gaps during on- road merging by older drivers ( Wolffelaar, Rothengatter, & Brouwer, 1987), or during conditions of poor visibility, are maneuver- level compensations. Momentary adjustments to steering and acceleration in response to slippery roads are examples of compensation at the vehicle- control level. ( p. 743)
Lundberg ( 2003) offers additional common ways of compensating and related schemes for classifying these different ways of compensating.
From an ecological perspective, compensating is not something the driver either does or does not do; compensating is always or continuously “ done,” though with varying degrees of adequacy and consistency ( Bäckman & Dixon, 1992; Hennessy, 1995). Please note that the varying degrees of adequacy and consistency in compensating are frequently changing attributes of a driver’s driving rather than fixed/ stable attributes of the driver.
Constraints on Consistently and Adequately Compensating
Lacking in awareness or understanding of a driving- relevant limitation( s), lacking in knowledge of ways of fully compensating for a limitation, and passively accepting a driving- relevant limitation as inescapably caused by aging are some of the factors detailed in the next major section ( p. 35) that variably constrain drivers from consistently and adequately compensating ( and from improving their driving- relevant abilities as well). In viewing driving from an ecological perspective, these constraints amount to “ barriers” to maintaining safe mobility; some barriers are higher than others.
“ Lowering” Barriers to Safe Mobility
In addition to viewing most of the safe- mobility interventions listed in the Overview ( referring for physician- based evaluation and treatment, educating about driving- relevant limitation( s), recommending behind- the- wheel training, restricting [ conditional licensure] and so on, see pp. 4- 5) as different ways available to DMV for aiding/ regulating functionally- limited drivers’ 22
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
compensating, these interventions may also usefully be viewed as mechanisms for “ lowering” barriers to safe mobility. That is to say, they are mechanisms for lowering barriers to consistently and adequately compensating. For example, in the case of a driver lacking awareness and understanding of a newly- identified driving- relevant limitation and lacking knowledge of ways of fully compensating, a DVD- based education intervention might be used to lower these two barriers to consistently and adequately compensating. A barrier need not be eliminated; it just needs to be lowered enough to where it’s just another bump in the road to being driving fit ( consistently small level of risk for making a critical driving error; Table 1, p. 8).
Driver- Centered “ Compensation”
Researchers sometimes pose as an empirical question, “ whether” older drivers compensate for their driving- relevant “ deficiencies.” Sometimes researchers simply assert that older drivers do compensate for their driving- relevant “ deficiencies.” For example, McKnight ( 2003) states:
… older drivers tend to compensate for their deficiencies by driving more slowly and more carefully than they once did and by avoiding the situations that present the greatest threat. As a result, they do not pose a substantially greater threat to the public than any other age group. ( p. 30)
Four years later Loughran, Seabury and Zakaras ( 2007) echoed McKnight’s assertions:
There is also evidence, however, that older individuals compensate for their impairment by changing their driving behavior… To identify the appropriate policy response to older drivers, such self- regulation must be taken into account. ( p. 2)
Because they are aware of their own limitations and adjust their driving patterns in response, older drivers pose only a slightly increased risk to other drivers. ( p. 14)
The above assertions are based on a driver- centered conceptualization of “ compensation” versus a driving- centered conceptualization of “ compensating.” When one’s focus is on regulating the driver rather than the driver’s driving, “ compensation” is not viewed as an ongoing driving process as it is in a driving- centered conceptualization of “ compensating”. Instead, driver- centered “ compensation” is viewed as a possible fixed/ stable attribute of the driver. “ Compensation” is viewed as neutralizing a driving- relevant deficiency( s). There is no sense of
23
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
8“ compensation” varying in its adequacy and consistency as it surely does, let alone differently for different drivers. Furthermore, attributing or not attributing driver- centered “ compensation” to all older drivers entails making two kinds of logical errors: an amalgamation error ( the aggregating of two or more functionally disparate groups) and an ecological fallacy ( attributing a characteristic of the aggregate at large, typically the mean, to one or more of the individuals making up the aggregate). These logical errors are examined in a later section of the Introduction and Rationale: Follow- Up to Age- Based Driver- Centered Questions & Answers: Two Kinds of Logical Errors.
Thinking of “ compensation” as a possible fixed/ stable attribute of the driver, and that if possessed would neutralize a “ physical defect,” would also seem to be the basis of routinely implementing California Vehicle Code section 12804.9 ( a) ( 3) ( quoted immediately below) by administering an office- based road test.
CVC § 12804.9 ( a) ( 3) A physical defect of the applicant that, in the opinion of the department, is compensated for to ensure safe driving ability, shall not prevent the issuance of a license to the applicant.
As indicated in the Overview, in the case of a driver having been identified as having a “ physical defect,” a fully driving- centered DMV, in contrast to a traditional driver- centered DMV, would first identify and initiate relevant safe- mobility interventions. Later the driving- centered DMV would make a driving- based determination of whether the level of risk for making a critical driving error ( see next subsection) is consistently small in that driver’s customary driving environments/ conditions and for that driver’s customary driving practices. 9 Relevant safe- mobility interventions would include, if feasible, providing a functionally- limited driver with specific ways of compensating for their limitations.
Critical Driving Error
Failing to consistently and adequately compensate for the effects of a driving- relevant functional limitation can result in the driver making a critical driving error ( CDE). A CDE is defined here as an action or inaction which has a high probability of precipitating an adverse driving event if
8 This may be due to a variety of factors, such as fatigue and distractions, which are described in the next major section: Barriers to Maintaining Safe Mobility.
9 For example, can the driver keep from making a critical driving error or a large number of important, but less critical errors, on a content- valid road test ( this would include infrequent though important regular destinations)? 24
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
the consequences of the CDE are not attenuated by such factors as a lack of conflicting traffic. Adverse driving events include the following negative outcomes: forcing other road users to take evasive action in order to avoid a crash, actually colliding with them, or nearly colliding with them ( cf. Marottoli, 1997).
Driving Wellness
When it comes to driving, maintaining safe mobility is first a matter of preserving a high degree of driving wellness. “ Driving wellness” is a phrase due to Emerman and Finn ( 2001); it was designed to focus attention on preserving and proactively improving driving- relevant functional abilities. A driver assessed as “ driving well” would be a driver deemed free from driving- relevant functional limitations. The degree of driving wellness will vary in accordance with the number, kinds, and severity of driving- relevant functional limitations. In an internet- based health- promotion module funded by a collaborative agreement between the American Society on Aging and the Centers for Disease Control and Prevention and called a “ Road Map to Driving Wellness,” its authors, Nancy Ceridwyn & Sandra Maldague ( 2002- 06), describe a variety of ways that drivers may preserve ( maintain) and/ or improve their driving wellness. For example:
 Actively working at keeping physically fit [ See Marottoli et al. ( 2007) for documentation of a physical conditioning program that aids in maintaining safe mobility].
 Getting regular physical examinations to assess and possibly improve visual, cognitive, and physical functioning.
 Abiding by the recommendations made by the driver’s doctor.
 Periodically reviewing the state’s Driver Handbook to refresh one’s knowledge of safe driving practices and the laws and rules of the road.
The Older Drivers Project has developed driving- wellness tools for physician use in “ optimizing the driver through health care” and published them in the Physician’s Guide to Assessing and Counseling Older Drivers ( Wang & Carr, 2004; Wang, Kosinski, Schwartzberg & Shanklin, 2003).
25
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
Driving Fitness
The level of risk for making a CDE is termed here as the degree of “ driving fitness.” The degree of driving fitness is contingent on how well the driver manages— by regularly searching/ scanning for hazards, slowing down or speeding up, and so on— the different ongoing everyday challenges to maintaining a low level of risk for making a CDE ( cf. Fuller, 2005) . As indicated in Figure 1 ( discussed below), these everyday challenges derive from the combined effects of four factors:
 Ongoing changes in the demands of the driver’s customary driving tasks.
 Ongoing changes in the demands of the driver’s customary driving environments/ conditions.
 The driver’s driving- relevant limitations.
 The driver’s driving- relevant abilities— especially proficiency ( consistency and adequacy) in compensating for the effects of any driving- relevant limitations that the driver may have acquired.
Therefore, the level of risk for making a CDE is not an attribute of the driver as is driving wellness. The degree of driving fitness is not some thing that could ever be measured in a DMV or doctor’s office. Instead, like the varying degree of adequacy and consistency in compensating, the degree of driving fitness is another frequently changing attribute of a driver’s driving rather than a fixed attribute of the driver. Just as a driver may not be assessed as always adequately compensating or not, a driver may not be assessed as having a high degree of driving fitness or not. However, a driver may be assessed as “ driving fit.” This would be a driver whose level of risk for making a real- world CDE is expected to be consistently small in that driver’s customary driving environments/ conditions and for that driver’s customary driving practices. For example, the driver did not make a structured- CDE ( see Appendix G for examples) during the course of a content- valid road test ( this would include infrequent though important regular destinations).
26
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
How and Why the Driver Competently Manages
Just as the safe- mobility interventions listed in the Overview ( referring for physician- based evaluation and treatment, educating about driving- relevant limitation( s), recommending behind- the- wheel training, restricting [ conditional licensure] and so on, pp. 4- 5) may be viewed as different ways that DMV may “ lower” barriers to drivers’ consistently and adequately compensating ( p. 23), they may also usefully be viewed as mechanisms available to DMV for aiding functionally- limited drivers in competently managing their everyday driving challenges and thereby maintaining a low level of risk for making a real- world CDE. This is because “ compensating for the effects of limitations” ( other than not driving) and “ managing everyday driving challenges” refer to different levels in the hierarchy of concurrent processes making up driving. “ Compensating for the effects of limitations” is embedded in “ managing everyday driving challenges” which in turn is embedded in “ maintaining safe mobility.” Regardless of the reference level, going down a level in this hierarchy answers “ how?” to regulate at a given level by in part using the next lower level; going up a level answers “ why?” regulate at a given level ( Powers, 1973). In addition to maintaining a low- CDE risk ( driving safely) by regularly searching/ scanning for hazards, slowing down or speeding up ( Fuller et al., 2006), and so on, consistently and adequately “ compensating for the effects of limitations” ( for example, focusing one’s attention exclusively on the drive task at hand) is also how a best- corrected functionally- limited driver may in part competently “ manage everyday driving challenges.” Why competently “ manage everyday driving challenges?” It is a “ lower level” means of “ maintaining safe mobility.” In addition to functionally- limited drivers regularly achieving their travel objectives by avoiding nighttime driving, restricting driving to familiar routes, and so on, competently “ managing everyday challenges” ( maintaining a low- CDE risk) is also how a functionally- limited driver “ maintains their safe mobility.” Maintaining a low- CDE risk, that is, driving safely, is integral to “ maintaining safe mobility.”
Driving Wellness Versus Driving Fitness
A driver’s degree of driving wellness is a driving- relevant description or assessment of the driver made irrespective of the driver’s customary driving environments/ conditions and irrespective of his or her customary driving practices. It includes aspects of functional health that have potential relevance for driving, as well as such cognitive acquisitions as knowledge of safe driving practices and knowledge of the laws and rules of the road. Assessing a driver’s contrast sensitivity ( the amount of contrast a driver needs to quickly detect objects that are important to safe driving, like other cars and pedestrians), for example, is a necessary part of assessing the
27
CALIFORNIA’S THREE- TIER DRIVING- CENTERED ASSESSMENT SYSTEM
driver’s degree of driving wellness, since good contrast sensitivity is critical to noticing potential road hazards under numerous everyday low- contrast viewing conditions such as the light car in the fog, a dark car in the shade, and faded lane- boundary markings. However, this does not mean that a driver assessed in a DMV field office as having severely limited contrast sensitivity should be expected necessarily to be at an elevated risk for making a CDE. Nor does it mean that a driver assessed as having only a marginal decrement in contrast sensitivity ( and who has no other decrements in driving wellness) should be expected not to be at an elevated risk for making a CDE. There is potentially a great difference between a driver’s expected risk for making a CDE, given their driving- relevant limitations, and a driver’s actual risk for making a CDE ( Hakamies- Blomqvist, 1994; Näätänen & Summala, 1976; Ranney, 1994; Shinar & Schieber, 1991). As already indicated, this is because even though a driver’s level of risk for making a CDE is a function of his or her degree of driving wellness, the degree of driving fitness is not determined by the degree of driving wellness. Figure 1 schematically summarizes the nature of the relationship between the degree of driving wellness and the degree of driving fitness, as well as the nature of the relationship between the latter and the probability of an adverse driving event. Both of these relationships may be strongly moderated. A moderated relationship is one in which the nature of the relationship between factors A and B depends on the value of another factor( s), C, the moderating factor ( Jaccard, Turrisi & Wan, 1990). In Figure 1, the arrows pointing to another arrow indicate a “ moderated relationship” ( Jaccard, Turrisi & Wan, 1990). Figure 1 indicates how the driving fitness- diminishing effect of a decrement( s) in driving wellness may be moderated in two fundamentally different ways:
 The driving fitness- diminishing effect of a decrement( s) in driving wellness depends on the driver’s