Driver Education A Review of its History and Effectiveness as a

Document Sample
Driver Education A Review of its History and Effectiveness as a Powered By Docstoc
					     Driver Education:
    A Review of its History and
Effectiveness as a Safety Program

         James L. Nichols Ph.D.
                 for the
  National Transportation Safety Board
The Youth Crash Problem
Magnitude and Characteristics
             Unintentional MV Traffic Deaths
                      by age group: Year 2,000
               Source: National Center for Health Statistics
           12000
                                  10323
           10000

           8000
# Deaths




                                          6716   6757

           6000                                         5210

           4000                                                3372


           2000           1647
                   563
              0
                   '1-4   '5-14   '15-24 '25-34 '35-44 '45-54 '55-64
             Key Factors
    Associated with Youth Crashes
 Youngest Drivers
 1st six months
 Alcohol
 Low Seat Belt Use
 Young Passengers
 Nighttime Driving
 Male
 Older Vehicle
                          Deaths per 100K Licensed Drivers
                           Age 16 vs. Older Groups, 1975-1996
                                      Extrapolated from IIHS, 1998
                          40
Deaths per 100K Drivers




                                                                             35

                          30                                      30
                                              28
                                                        26

                          20     19

                                 14           14
                                                        12.5      12.5       12
                          10


                          0
                               '75          '80       '85       '90        '96

                                         age 16    ages 17-19   ages 20+
                                Percent of Young Drivers Killed
                                     With BACs >= 0.08
                                            Source: NHTSA, 2001
                           40
                                                                                35
Percent with BAC >= 0.08




                                                                       30
                           30
                                                              24

                           20      17                16

                                            10
                           10


                           0
                                 age 15*   age 16   age 17   age 18   age 19   age 20
                               Percent Unbuckled
                          for Various Occupant Groups
                             Source: NHTSA, NOPUS, FARS
                    100
                     90
                                                                                    79
                     80
Percent Unbuckled




                                                        67            70
                     70                   62
                     60
                     50
                     40
                     30      27

                     20
                     10
                      0
                          Observed,   Fatalities,   Fatalities,   Fatalities,   Fatalities,
                           all ages    all ages     ages 16-20    ages 21-24    16-24 with
                                                                                 high BAC
                              Driver Crash Rates
                          and Presence of Passengers
                               Source: Williams, 2003

                      7
# crashes/10K trips




                      6
                      5
                      4
                      3
                      2
                      1
                      0
                           Ages 16-17        Ages 18-19           Ages 30-59

                                 0 pass.   1 pass.   2 pass.   3+ pass.
                              Daytime vs. Nighttime Crashes
                                per Million Miles Traveled
                                        Source: Williams, 2003
                         40
crashes per 100m miles




                         30


                         20


                         10


                         0
                              16   17     18   19    20-24   25-29    30-64   65-69   70+
                                                      age

                                               Nighttime    Daytime
Some Progress Has Been Made
  Youth Fatalities down by 1/3   rd

           Since 1978
                               Deaths per 100K Population
                              Motor Vehicle Occupants by Age Group
                                        Source: NHTSA, FARS
                         50
# Deaths per 100K Pop.




                                 45
                         40      41
                                               38
                                        35     34       34
                                        33
                         30                             31         31
                                                                              28      29
                                                                   26         25      26
                                 24
                         20             21     21       20
                                                                   16         16      16

                         10

                         0
                               '78    '82    '86      '90        '94        '98     '02

                                      ages 16-20    ages 21-24         ages 25-34
                         But …Much More Needs to be Done
                           Fatalities, by Year of Age: 0-16
                                     Source: NHTSA/FARS, 2001
                       1000
                                                                                           822
Number of Fatalities




                       800

                       600

                       400                                                           298
                                                                               183
                       200 106 94 99 97 101
                                            77 95 94 81 81 78 87 86 108
                         0
                              <1 1   2   3   4   5   6    7   8   9 10 11 12 13 14 15 16
                                                         Year of Age
    Driver Education:
as a highway safety program

          History
       Effectiveness
        Future Role
A Historical Review


Key Events, Expansion,
Issues, and Challenges
     Four (arbitrary) Phases:

 1930-1940 – The Beginning
 1941-1965 – Rapid Expansion
 1965-1980 – Federal Involvement
                and Challenges
 1980-2000 – A Period of Decline
 2000 – 2003 – A Re-Emergence?
     The Beginning (1930-1940)
   1932 -Dr. Herbert Stack, Columbia Univ.
   1934 - Dr. Amos Nyhart, Penn. State Univ.
   Early Involvement by NSC, AAA, ASF
   1st Classrooms, textbooks, simulators, driving
    ranges, 3-phase programs, etc.
   1937 - 1st city-wide program – Cleveland
   1941 - 1st evaluation – Cleveland
   Teaching centers at NYU & Penn State
   AAA Text “Sportsmanlike Driving”
   Ohio State and Iowa State (ranges and simul)
    Rapid Expansion (1947-1964)
   Quality Control, Curriculum, Teacher Prep
   Relatively Uncontrolled Development
   NCSE (1943 - ASF grant)
   Teacher Training at Penn State and NYU
   1st Conference (1949); “30+6” recommended
   Allstate Provides Premium Discounts (1952)
   2nd & 3rd Conferences (1953, 1958)
   Michigan Requires DE for Licensing Under 18
   Additional Centers at Maryland and Mich. State
   4th Conference (1963) “90 hours” recommended
   Growth In Enrollment (1947-64)
and Number of Schools Offering HSDE
     Source: Pubic Technology (AAA); Butler, 1982)
15
                                                   13
                                                   12
10



5
                 3
                 2
0
            1947-48                          1964-65
               Schools (x 1000)   Students (x 100,000)
     Federal Role & New Challenges
                        (1965-1980)
   Highway Safety Act of 1966
    – NHSB/NHTSA
    – HSDE is a Priority H.S. Program; Funding to the States
    – NCSE Discontinued
   Challenges to Effectiveness Claims
    – Moynihan Report (1968); McGuire & Kersh (1969)
    – Goldstein vs. McGuire (1969); Roberston & Zador (1978)
   R&D Efforts to Develop Model HSDE Program
    – Task analysis and Instructional Objectives (1971)
    – Instructional program (1973), Acquisition of Skills (1974)
    – Pilot Test and Curriculum Performance Measures (1977)
   Growth In Enrollment (1964-78)
and Number of Schools Offering HSDE
                   Source: Butler (1982)
35
                                                           32
30
                                                     28
25                                       24
                              23
20                 20
                                                     16    17
15       15        14                    15
         13                   13
10
5
0
     1964-65   1967-68   1969-70    1974-75    1976-78
                 Schools (x 1000)   Students (x 100,000)
              A Period of Decline
                     (1980-2000)
   Impact of SPC Evaluation
    – No clear evidence of impact
    – Evidence of effect on Licensing
   Change in Status as a Priority Program
    – Re-examination ordered by Congress
    – DE dropped from list of priority programs
   Changes in Enrollment and School Offerings
    – No solid data
    – Decline in State reporting
    – Anecdotal Data (90% enrolled  50% -> 40%)?
                                Decline in Enrollment
                   Estimated % Eligible who are Enrolled
                          Sources: Butler (1982); Abbate (1984)
                   100

                   90
                                         81
Percent Enrolled




                   80
                           73     75            74
                   70
                                                              64
                   60
                                                       56
                   50                                                                 50

                   40

                   30
                         '72    '74    '76    '78    '80    '82    '84   '86   '88 '90*    *
 Re-emergence with a New Role?
A Parallel Development in Licensing
                   (2000-present)
   1965-1980 Period of Expansion for HSDE
    – GDL concept developed (Teknekron, 1977)
    – Largely unsuccessful efforts to implement
   1980-2000 Period of Decline for HSDE
    – GDL successful in New Zealand, Australia, Canada
    – Recent successes in USA (post-1995)
   2000-2003 A New Emergence for HSDE?
    – DE still part of many State licensing laws
    – Becoming integrated into GDL systems
    – Frequently involves a time incentive (of Concern)
      Driver Education:
Evidence of Effectiveness


 Early Claims, Challenges, Controlled
     Studies, SPC, Other Studies
   Driver Education:
  What’s the Objective?

   to develop the skills,
 knowledge, and attitudes
necessary for safe driving?
    Driver Effectiveness Studies
     Four (arbitrary) Phases:


 1940-1965    –    Early Studies
 1965- 1980    –     Challenges, New Studies
 1980-1990    –     SPC Results & Challenges
 1980-2000    –     Other Studies
           Driver Effectiveness
       Early Studies (1940-1965):

   “DE Reduces Accidents One-Half” (AAA)
 30+ Studies (20-50% effect)
 Conducted by Advocates, DMVs, Univ.
 AAA, NCSE/NEA, ACSC, NSC, ADTSEA,
 “Self-Selection” factor not recognized
 Much of effect due to group differences
           Driver Effectiveness
         Challenges (1965-1980)
 Moynihan Report, 1968
 McGuire and Kersh, 1969
 Goldstein vs. McGuire & Kersh
 California DMV Studies
    – Coppin, et al, Conger et al., Ferdun et al.
    – Harrington, Jones, Goldstein
   A New Problem: Increased Licensing
    – IIHS Study/NHTSA Response
        Driver Effectiveness
    Summary of Results (1965-1980)
   McGuire and Kersh (MS) HSDE -
   Conger et al (CA) HSDE +
   Harrington (CA) HSDE +
   Coppin et al (CA) BTW -
   McGuire and Kersh (CA) BTW/Sim –
   Vernon & Phillips (TX) multi-phase > more basic
   Jones (CA) HSDE BTW = commercial BTW
   Goldstein (CA) HSDE BTW > commercial BTW
   Robertson & Zador; NHTSA DE  licensing
  Mayhew & Simpson (1996)

   The DeKalb Driver Education
Project ..”remains the most critical
test of the safety impact of driver
        education to date.”
         Driver Effectiveness
       SPC Results: 1980-1990

 Stock (1983) no effect/ SPC effect/ Lic
 Smith and Blatt (1987) PDL effect
 Lund et al. (1986) no effect/ Lic
 deWolf and Smith (1988) no effect
 Davis (1990) no effect/ Lic
         Initial SPC Results
     Percent Involved in a Crash
                 Stock et al. (1983)
35
30
25
20
15
10
5
0
      Assigned            Licensed         Completed and
                                             Licensed

                    SPC    PDL   Control
         Further SPC Analysis
Ave # Crashes in 1st Six Months Among Licensed
               Stock et al., 1983
0.14
                                          0.122
0.12                        0.107
             0.105
 0.1
0.08
0.06
0.04
0.02
  0
                           Licensed

                     SPC    PDL     Control
          Driver Effectiveness
    Other U.S. Studies: (1990-2000)

 TX: HSDE > Commercial (effect?)
 OH: HSDE > Commercial (effect?)
 OR: HSDE (no effect - more crashes)
 47 States: DE Law (effect) (Lic)
 PA: HSDE (no effect; crashes, SBU, etc.)
           Driver Effectiveness
      Foreign Studies: (1990-2000)
   Quebec: HSDE Law Effect (?)(+/-) (Lic)
   Denmark: new DE prog (pos effect)
   Sweden: prof DE train = parent train (no effect)
   Norway: DE (no effect)
   Austria: DE law (no effect)
   Australia: DE (no effect); on-road = off-road
   New Zealand: DE (no effect) (Lic)
   Tasmania: (small effect -> no effect)
       Driver Effectiveness
28 Studies Reviewed: (1970-2000)
   18 looked at effect of DE versus Control
    – 2 (11%) found evidence of effect
    – 4 (22%) found inconsistent effect or effect challenged
    – 12 (67%) found no evidence of effect
   Two studies of laws found effect but also licensing
   Ten studies reported effect on licensing
   Eight studies used random assignment
    – 6 found no evidence of effect
    – 2 found conflicting evidence (SPC)
   Four studies looked at HSDE vs Commercial
    – Mixed results
          Driver Effectiveness
    Other Reviews and Conclusions:
   Mayhew & Simpson (1986)
    – No conclusive evidence (exc. Licensing)
   Vernick et al. (1999)
    – No convincing evidence (exc. Licensing)
   Christie (2001)
    – Little support for claim of safety effect (disbenefit)
   Roberts et al. (2002)
    – No evidence of effect (exc. Licensing)
   Engstrom et al. (2003)
    – Much knowledge not yet applied; need experiment
       Driver Effectiveness Reviews
    Other Findings: Mayhew & Simpson

   Best learning environment is on-road
   Advanced skill training – mixed results, risk
   No compelling evidence of effect – MC rider
   Some evidence for remedial MC training
   Some evidence of effect for nighttime training
       Driver Effectiveness Reviews
    Other Findings: Christie et al. (2001)

   Little evidence of effect (prof vs parent trained)
   Lack of support for off-road training; disbenefit
   More supervised practice effective (up to 35%)
   Emergency training may increase crash risk
   Higher-order cognitive skills learned on-road
   No evidence of effect for mandatory training
    (but it does encourage licensing)
   Targeted deterrence and enforcement effective
       Driver Effectiveness Reviews
      Other Findings: Engstrom (2003)

   Young Driver Crashes (SV, lose control, speed,
    alcohol, no belts, fatigue, night, weekend)
   Acquisition of early experience is crucial
   HSDE, if introduced, no licensing incentive
   Short professional courses (DE) should not be
    exchanged for shorter periods of practice
   GDL restrictions have been effective
   Laws and enforcement are effective, but some
    youth are hard to reach
      Driver Education:
as a highway safety program

   What’s in the Future?
     Important Opportunities to
   Partner in an Effective Program
                   The Future
Problems and Solutions: NHTSA Report

   Problems
    – Focus on basic and higher-order skills (timing)
    – Students not motivated to learn
   Solutions
    – Teach skills first; safety later
    – GDL environment provides opportunity for
      phased learning and the motivation to learn
    – Two-Phase HSDE program proposed
                           The Future
Problems and Solutions: Mayhew & Simpson

   Problems
    –   Failure to teach critical knowledge & skills
    –   Students not motivated to learn
    –   Fostering overconfidence
    –   Failure to effectively address lifestyle issues
    –   Failure to tailor content
   Solutions
    –   Identifying “when” to teach
    –   Emergency skills (?); need insight training
    –   Tailoring to student; distribute over time
    –   Link with GDL; parental involvement in supervised practice
                            The Future
    Problems and Solutions: Christie (2001)

   Problems
    – False Assumptions
          Skill deficiencies are the problem
          Deficiencies can be remedied
    – May be impossible for DE to reduce crashes
          Set more reasonable objectives
          Depend on more effective approaches
   Solutions
    –   Set More Reasonable Objectives
    –   Depend on more effective programs (e.g., law enforcement)
    –   Focus on increased supervised learning
    –   Integrate with GDL systems
                   Lists
 Lifestyles not adequately addressed
 Instruction needs to be timed better
 Motivation to learn is an issue
 Effect of advanced skills training?
 1st 6-months most dangerous
 Supervised, on-road experience is priority
 Delayed licensure is desirable
 Restrictions (e.g. night, passenger) work
          Lists (continued)
 Youth “mature out” of high risk behavior
 GDL provides for some early experience in
  a “safer” environment
 GDL systems are effective
 DE is already tied to GDL in many States
 DE should be modified to fit better
 DE is not an effective substitute for
  supervised experience
     Driver Education:
    A Review of its History and
Effectiveness as a Safety Program

         James L. Nichols Ph.D.
                 for the
  National Transportation Safety Board