Research Studies & Reports
DMV’s Research & Development Branch has been conducting research and producing studies and reports since the 1950s. Research & Development reports help DMV to measure the impact of new laws on making drivers safer. We also identify areas where we can improve our processes, explore new approaches to solving existing problems, and branch out into new opportunities to serve you better.
Studies & Reports Sections
Studies and reports are assigned to a Section that best describes the type of report. Click on a section title below to see a short description.
I. Driver Education & Training Studies
II. Driver Licensing Screening Studies
III. Studies on Improvement and Control of Deviant Drivers
IV. Basic Research & Methodological Studies: Driver Performance, Accident Etiology, Prediction Models, and Actuarial Applications
V. Driver Licensing / Control Systems & Safety Management Studies
VI. Studies on Special Driver Populations
VII. Miscellaneous Studies & Reports
Request printed copies of studies and reports by mail at:
Department of Motor Vehicles
Research and Development Branch
2415 1st Ave. Mail Station: F-126
Sacramento, CA 95818
(916) 914-8125
Please include the report number, the number of copies requested, and your name, address, and phone number.
| Report ID | Date Published | Title | Section | Links |
|---|---|---|---|---|
| 187 | 2000/ 06 |
Using Traffic Conviction Correlates to Identify High Accident-Risk DriversThis study further explored previous research involving the viability of predicting accidents from equationsconstructed to predict convictions for the general driving population. Models that better identify drivers at increased risk of future accident involvement will increase the number of accidents prevented through post license control actions. Although the results do not support the hypothesis that equations keyed to citations do as well as or better than equations keyed to accidents in predicting subsequent accidents, the results suggest that identification of future accident-involved drivers can be improved by either of two approaches. The first is to construct equations based on a combination of prior accidents and citations. California’s neg-op system basically reflects such an approach since points are allocated to traffic convictions and culpable accidents. The second alternative is more elaborate, involving a truly multivariate approach in which the prediction equation consists of a two-variable vector of subsequent citations and accidents. The canonical correlation analysis performed for this study resulted in two orthogonal canonical functions or roots: A driving-incident function consisting of primarily citations and secondarily accidents and an almost exclusively accident function. The results reported in this study indicate that subsequent driving record can be predicted from prior driving record for groups of individuals; however, the error rates at the individual level are inherently large. The models derived from the canonical analysis, while superior to the simpler models, would be very difficult to implement operationally. The most obvious problem relates to its complexity. Canonical correlation is difficult to comprehend. Another problem is that the equations contain a number of variables (e.g., age and gender) that would not be legally defensible in taking license control actions. This problem could be rectified, with some sacrifice in predictive power, by deleting the unacceptable variables. In addition, use of variables such as age and gender might be permissible for triggering educational advisory interventions. |
III | |
| 186 | 1999/ 12 |
Development of a Conceptual Integrated Traffic Safety Problem Identification DatabaseThe project conceptualized a traffic safety risk management information system and statistical database for improved problem-driver identification, countermeasure development, and resource allocation. The California Department of Motor Vehicle Driver License (DL) and Vehicle Registration (VR) database systems, the California Highway Patrol Statewide Integrated Traffic Records System (SWITRS), and the National Highway Traffic Safety Administration (NHTSA) Fatality Analysis Reporting System (FARS) each provide valuable information on crashes in California for conducting problem identification analyses, developing and evaluating traffic safety programs, and allocating services. The project explored these four primary systems and investigated several additional data sources suggested by the project advisory committee, including the California Department of Transportation Traffic Accident Surveillance and Analysis System, the Department of Health Services Hospital Discharge Database, and the Department of Justice Criminal Justice Information System. Based on a review of the existing systems and inputs from a project advisory committee of representatives of leading national and state traffic safety research interests, the department has decided to develop a prototype ITSPID system that would integrate the DL, VR, SWITRS, and FARS databases. |
IV | |
| 185 | 2000/ 01 |
2000 ANNUAL REPORT OF THE CALIFORNIA DUI MANAGEMENT INFORMATION SYSTEMIn this ninth annual legislatively mandated report, 1997 and 1998 DUI data from diverse sources were compiled and cross-referenced for the purpose of developing a single comprehensive DUI data reference and monitoring system. This report presents crosstabulated information on DUI arrests, convictions, court sanctions, administrative actions and alcohol-involved accidents. In addition, this report provides an evaluation of the effectiveness of alternative court and administrative sanctions (including alcohol treatment programs and license actions) upon the 1-year postconviction records of first and second DUI offenders over a time period of eight years. The postconviction driving records of second DUI offenders arrested in 1995 and 1997 were evaluated for three- and one-year periods, respectively. |
V | |
| 183 | 1999/ 07 |
Strategies for Estimating Driver Accident Risk in Relation to California’s Negligent-Operator Point SystemA sample of approximately 140,000 records of licensed California drivers containing information on age, gender, and driving record variables was examined. The goal of this paper was to assess the accuracy of predicting future accident risk using various combinations of demographic and prior driving record variables as predictors in 17 regression models. All of the models were consistent in demonstrating that increased probability of subsequent accident involvement is associated with increased prior citation and prior accident frequencies, being young, and being male. Results from the regression models indicated the following: • Models that use prior total accidents as a predictor variable perform better than models that do not use prior total accidents as predictors. • Models that use prior culpable accidents as a predictor do not perform as well as models that use prior total accidents as a predictor. • A comparison of models in which 17 individual violation types are used as predictors to those in which only total citations is used as a predictor shows only a small advantage of using individual violation types. • Models that use as predictors the demographic variables of age, gender, and license class along with various combinations of citations and accidents perform better than California’s current neg-op system, which uses a weighted combination of countable citations and responsible accidents. It was concluded that if the goal of driver record adjudication systems is to identify and apply sanctions to high-risk drivers in order to intervene before this risk is realized, then the results presented in this report support the current pointcount strategy which attempts to optimize the identification of drivers having a high probability of subsequent accident involvement. |
IV | |
| 182 | 1999/ 03 |
EVALUATION OF THE CLASS C DRIVER LICENSE WRITTEN KNOWLEDGE TESTSThis report presents the results of an evaluation of the English DL 5 (Rev. 10/98), Spanish DL 5 (Rev. 5/98), and English DL 5T (Rev. 8/98) Class C license written knowledge examinations. Specifically, the study assessed the fail rate, mean number of errors, and internal-consistency reliability for each test form, as well as the pass rate, percentage of applicants selecting each answer choice, and item-total correlation for each item on each test form for the English tests. Also presented is an assessment of the randomness of the answer choice assignment for the English DL 5. |
II | |
| 181 | 1998/ 12 |
An Evaluation of the Impact of California’s Driving PerformanceEvaluation Road Test on Traffic Accident and Citation RatesThis evaluation is the final stage of a project to develop and evaluate the driving performance evaluation(DPE) drive test for possible statewide implementation in California. This study and earlier studies in theseries are part of a more extensive effort by the department to increase the competency of Californiamotorists by improving the driver licensing process. The DPE is currently being used in over 60 fieldoffices in southern California.The effect of the DPE on accident and citation rates was determined by comparing driving records for asample of applicants taking the DPE with those of a sample of applicants taking the current drive test bothbefore and after implementation of the DPE.Logistic regression analysis was used to compare the crash and citation rates of the groups during the 2years immediately following driver license application. The findings failed to substantiate any reduction inaccident involvements or traffic law violations resulting from implementation of the program. However,there is no question that the DPE is a more reliable and content-valid test. Therefore, it is recommendedthat the DPE road test be expanded to all offices in the state. |
II | |
| 180 | 1998/ 12 |
An Evaluation of the General Deterrent Effect of Vehicle Impoundment onSuspended and Revoked Drivers in CaliforniaWhile license suspension/revocation have been shown to be effective, it is also known that most suspended/revoked(S/R) drivers violate their S/R order and continue to drive, accruing traffic convictions and becoming involved incrashes. In an attempt to strengthen license actions and to better control S/R and unlicensed drivers, Californiaenacted two laws effective January 1995 which provide for the impoundment/forfeiture of vehicles driven by S/Rand unlicensed drivers. The current study evaluates the general deterrent impact of vehicle impoundment on thecrash rates for the general population of S/R drivers in California, regardless of whether they have been apprehendedand punished for driving-while-suspended.Crash rates for S/R drivers, as well as a control group of non-S/R drivers, were examined over a 5-year periodconsisting of 3 years prior to the implementation of vehicle impoundment/forfeiture and 2 years subsequent.Interrupted time series statistical models used to analyze the data showed that while there was a statisticallysignificant drop in crashes for S/R drivers at the point at which the laws became effective, there was also a significantdrop for control drivers, who shouldn't be affected by the laws since they are not S/R. Additional analyses whichjointly estimated crashes for both S/R and control groups revealed that the drop in crashes for S/R drivers was nolonger statistically significant once the crash rate for control drivers was statistically accounted for. Thus, this studyfailed to find compelling evidence of a general deterrent impact of vehicle impoundment/forfeiture in California. |
III | |
| 179 | 1999/ 01 |
1999 ANNUAL REPORT OF THE CALIFORNIA DUI MANAGEMENT INFORMATION SYSTEMIn this eighth annual legislatively mandated report, 1996 and 1997 DUI data from diverse sources were compiled and cross-referenced for the purpose of developing a single comprehensive DUI data and monitoring system. This report presents crosstabulated information on DUI arrests, convictions, court sanctions, administrative actions and alcoholinvolved accidents. In addition, this report provides an evaluation of the effectiveness of alternative court and administrative sanctions (including alcohol treatment programs and license actions) upon the 1-year postconviction records of first and second DUI offenders over a time period of seven years. The postconviction driving records of second DUI offenders arrested in 1994, and 1996 were evaluated three- and one-year periods, respectively. |
V | |
| 178 | 1998/ 12 |
Evaluation of the Delegated Drive Test Pilot Program: Technical AppendixThis study evaluated the safety impact of allowing driving schools to administer the Driving PerformanceEvaluation (DPE) to provisional license applicants. The results of the driver record comparisons betweenprovisional applicants tested by the driving schools and those tested by DMV did not indicate a statisticallysignificant difference in the 6-month post-licensure accident or citation rates for the groups. Unfortunately,inadequate sample sizes and the potential biases present in the study preclude drawing any firm conclusionsregarding the comparative safety impact of private versus DMV testing. However, the results of the scoringconsistency and reliability analyses are more interpretable and less subject to these problems. The comparisons ofscoring consistency between driving school and DMV examiners indicates that the driving school examinersfollowed the DPE scoring criteria less stringently than did the DMV examiners, and were far more lenient, havingpassed many applicants who subsequently failed the drive test at DMV. Although these findings also requirequalification, it is very unlikely that differences of the magnitude observed can be attributed to bias alone. The lowvolume of subjects, which was a major reason for the low statistical poser of the analyses, may indicate that themarket for delegated testing is small, both within the general public and the driver training industry itself |
II | |
| 177 | 1998/ 05 |
Evaluation of the Referral Driving Performance EvaluationProgram—Follow-Up ReportThis study evaluated the safety impact of the new Referral Driving Performance Evaluation (RDPE) drivetest program. The 3-year prior accident and citation rates for drivers taking the RDPE drive tests werecompared to the general driving population and to drivers who passed the Special Drive Test (SDT) in anearlier DMV study. The results indicated that in every age and gender category except one, drivers in theRDPE program had much higher prior accident and citation rates than did drivers in general. This findingsupported the department’s policy of testing drivers referred for medical and other reasons. The prioraccident rates for drivers who passed the RDPE tests were not significantly different from those for driverswho failed the tests. Hence, the validity of using RDPE test results as indicators of accident risk was notconclusively supported by the data. Contrary to expectation, drivers who passed the RDPE tests also hadaccident rates similar to those for drivers who passed the SDT, which indicated that the RDPE tests wereno better than the SDT at distinguishing between higher- and lower-risk drivers. However, because theRDPE tests fail a much higher percentage of referral drivers than does the SDT, the tests do result inaccident savings. |
II |