'——"——'—;-”— -.——-_- - —-—w— m—w‘ —" TI-IE FEASIBILITY OF USING SELECTED STUDENT - DATA BASES FOR THE ASSESSMENT AND EVALUATION OF DRIVER EDUCATION PROGRAMS IN THE STATE OF MINNESOTA ' Thasis for the Degree of Ph. D. v MICHIGAN STATE UNIVERSITY IEROME W. WITHERILL 1973 "3:01‘ 5- - Q; o. ."t' um - r g Rikki-3313 1-3 A Umver-sny ‘ This is to certify that the thesis entitled The Feasibility of Using Selected Student Data Bases for the Assessment and Evaluation of Driver Education Programs in the State of Minnesota presented by Jerome W. Witherill has been accepted towards fulfillment of the requirements for Doctoral degree in Educatlon WZJM/ I Ibbert O. Nolan I Major professor Date July 27, 1973 0-7639 ABSTRACT THE FEASIBILITY OF USING SELECTED STUDENT DATA BASES FOR THE ASSESSMENT AND EVALUATION OF DRIVER EDUCATION PROGRAMS IN THE STATE OF MINNESOTA BY Jerome w. witherill The search for a method of evaluating high school driver education nationally has been ongoing for some time. but without much success. Differences in prOgrams, instructors, available data and the environment from one location to the next have produced such variance that comparisons are nearly, if not totally, impossible. The prime difficulty in these evaluative efforts have been the develOpment and maintenance of adequate criterion data such as violations, accidents, and the circumstances surrounding each and both. This study used criterion data generated from five sources in an effort to provide an adequate data base for comparing the relative efficacy of driver education prOgram types and for determining the correlates of successful driving among late-adolescent drivers. The sources were: 1. School records 2. Local police records 3. State Motor-vehicle records Jerome w. witherill 4. Court records, and 5. Driver self-reports The students were randomly selected from those enrolled in summer driver education programs in Duluth. Minnesota. Twenty-five males and twenty-five females were selected each year from each of the three treatment types (traditional. range and simulation) offered for the years 1967-1969. Chi-square tests of significance were used to test hypotheses concerning treatment and sex effects. A correlation matrix was used to determine the strengths of association for correlative data. Cross tabulations and content analysis were used to study interagency data bases. student responses and accident records. The findings of the study were the following: Data Sources 1. Only 65 percent of the accident reports available from all sources were obtainable from state driver license records. Of the reports available from the state. however, 53 percent were not obtainable from other sources. 2. Eighty-three percent of the citation reports available from all sources were obtainable from state driver license records. 0f the citation reports available from the state, 40 percent were not available from other sources. 3. Local police records contain 68 percent of the accident reports available from all sources, and of the local police reports, 52 percent are not obtainable from other sources. Jerome w. witherill 4. The local court records contain 60 percent of the total of citation reports from all sources. but only 17 percent of the local court reports are not obtainable from other sources. 5. Most of the subjects (91%) were easily contacted and asked to answer the questions of the phone survey. Treatment Comparisons 6. The acquisition of citations was up; significantly affected by treatment type (traditional. range, and simulation). 7. Involvement in accidents was 59; significantly affected by treatment type (traditional. range, and simulation). Sex Comparisons 8. Females acquired significantly fewer citations than did males. This may be qualified by the significantly greater exposure: (1) number of days of driving per week. (2) number of miles driven per week. (3) percent of night time driving. (4) percent of weekend driving, and (5) motorcycle miles driven of the male driver. 9. Females acquired significantly fewer accidents than did males. This also may be qualified by the significantly greater exposure: (1) number of days of driving per week. (2) number of miles driven per week, (3) percent of night time driving. (4) percent of weekend driving. and (5) motorcycle miles driven of male drivers. Demographic Data Comparisons 10. Involvement in accidents was significantly related in a negative direction to driving experience as defined by days driven per week. percentage of night driving. and percent- Jerome w. witherill age of weekend driving (as these variables increase. accidents decrease), and was significantly related in a positive direction to motorcycle miles driven (as this variable increases. accidents increase). 11. Drivers from larger families had significantly more accidents than drivers from smaller families. 12. Involvement in accidents Egg significantly related in a positive direction to the acquisition of citations. This may be qualified by the inclusion of citations received at the accident with the citation data. 13. The acquisition of citations was significantly related in a negative direction to driving experience as defined by days driven per week. and percentage of weekend driving. and in a positive direction to motorcycle miles driven. Student Suggestions 14. The most frequently stated suggestions for classroom instruction were (a) more practical information. (b) use more "scary" films and (c) more information about accidents. 15. The most frequently stated suggestions for on-the-street instruction were directed toward more driving experience, specifically winter driving, parking, city driving. and the use of standard shift cars. Accident Records 16. In 25 (40%) of the accident reports. skidding was listed as an apparent contributing factor. 17. Inattentive driving (19%). illegal or unsafe speed (16%), and failure to yield (16%) were identified as the most common driver faults in accident reports. THE FEASIBILITY OF USING SELECTED STUDENT DATA BASES FOR THE ASSESSMENT AND EVALUATION OF DRIVER EDUCATION PROGRAMS IN THE STATE OF MINNESOTA BY I‘» C 6 Jerome WT'Witherill A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Education 1973 3 ACKNOWLEDGMENTS I wish to express my sincere appreciation to Dr. Robert 0. Nolan for his support. guidance. and patience throughout all phases of the doctoral program. Appreciation is also extended to Dr. Dale Alam. Dr. Robert E. Gustafson. and Dr. William A. Mann for their time and constructive criticism. A note of thanks is extended to the University of wisconsin and the Instructional Research Center. A special thanks to Dr. Robert Clasen without whose patience and assistance the statistical analysis and interpretation of this study would have been impossible. And finally. a very special appreciation to my wife. Sheila. for her patience. encouragement. and sacrifice. ii TABLE OF Chapter I. II. III. IV. CONTENTS THE NATURE OF THE PROBLEM . Introduction . . . Background of the Need Importance of the Study . Purpose and Objectives Definition of Terms . Organization of Remaining SELECTED REVIEW OF LITERATURE Goals and Objectives of Driver Education . Utility of Accident and Citation Records . Status of Driver Education . Effectiveness of Driver . National Highway Traffic Safety Administrat Summary s s s s s s 0 DESIGN AND METHODOLOGY Subjects . . . . . Data Collection . . Analysis . . . . . summary 0 O O O O 0 ANALYSIS OF DATA . . s Feasibility Study of Interagency Data Chapters Education 0 O . 0 ion Sources . Comparability of Traditional. Range and Simulation Efficacy Comparability of Sex Effects Correlative Data . . Frequency Data of Student Suggestions . Accident Report Data--Apparent Contributing FaCtorS s a s s s o summary a s s a s o 0 iii 32 33 40 41 41 47 49 54 58 6O 61 Chapter V. SU REFERENC BIBLIOGR MMARY, MAJOR FINDINGS, CONCLUSIONS AND RECOMMENDATIONS . . . . . . Summary 0 s s a s s s a s a Major Findings of the Study Conclusions . . . . . . . . Recommendations . . . . . . Recommendations for Further Discussion . . . . . . . . ES CITED 0 O O O 0 O O O O APHY O O O O O I O O O O 0 APPENDIX A: DATA REPORT FORMS . . APPENDIX 88 ACCIDENT REPORT DATA 0 O O O C O O O I I O O O O 0 Research Page 64 64 67 7O 74 76 76 79 83 89 95 LIST OF TABLES Table Page 1. State-Level and Total Accident and Citation Records Available 0 O O O O O O O O I O O O O O O O O O O 42 2. Local Police Department and Total Accident Records Available 0 C C O O O O O O O O O O O O O O O O O 44 3. Local Court and Total Citation Records Available . 45 4. Summary of Data Availability at Schools . . . . . . 46 5. Summary of Individual Data Availability . . . . . . 46 6 . Frequency Counts and Chi-Square Values for Citations by Treatment Types and Year . . . . . . 48 7. Frequency Counts and Chi-Square Values for Accident by Treatment Type and Year . . . . . . . 50 8. Frequency Counts and Chi-Square Values for Citations by Sex and Year . . . . . . . . . . . . 51 9. Means. Standard Deviations. and t-Test Results for Exposure Variables by Sex for Each of Three Years and Total . . . . . . . . . . . . . . . . . 53 10. Frequency Counts and Chi-Square Values for Accident by Sex and Year . . . . . . . . . . . . 55 11. Correlation Matrix Showing the Pearson Product Moment Correlation Among Selected Demographic Variables (N e 346) . . . . . . . . . . . . . . . S7 12. Frequency Count of Suggestions from Students for Improvement of Classroom and On-the-Street Aspects of Driver Education . . . . . . . . . . . 59 13. Accident Apparent Contributing Factor Data Summaries 62 's 1.’ CHAPTER I THE NATURE OF THE PROBLEM Introduction Historically,driver education is one of the newest and fastest growing disciplines in the school curriculum. Since the first course was offered at Gilbert. Minnesota. in 1923 (29). driver education has developed to its present state where Bl percent of the potential schools in the 50 states offer at least a minimum program of 30 clock-hours of class- room instruction and 6 clock-hours of on-the-street instruction or the equivalent substitution of range or simulation instruction (42). In fact. driver education has become so accepted as a part of the school curriculum that. in contrast to the usual policy of strict church-state separation. many states provide financial support to driver education in both private and parochial schools (24). Throughout this explosive growth of driver education. as now. little research has been conducted to provide methods or techniques for program evaluation. Educators have relied on the past. economics. "common sense.” or whatever seemed applicable from their experience in other phases of education to determine what driver education should be. Today. as never before. vast quantities of driver behavior data are being collected by state agencies and 1 it, stored in computers. These data are available to the schools and can be retrieved from the computer in a matter of minutes or seconds. Schools themselves have improved their record keeping systems in an effort to document student demOgraphic information. These tremendous reservoirs of data can serve as meaningful inputs to direct schools in their prOgram improvement. In addition to these data. each year hundreds of thousands of students complete driver education and shortly thereafter are called upon to perform in the real world the basic skills taught them. Their perceptions. ideas. and feelings about the adequacy of the driver education instruction they received. along with their personal experience in the performance of the driving task, could be an invaluable aid for prOgram improvement. Unfortunately. few. if any. schools use the above data sources for improving their programs. Because Duluth, Minnesota is no exception to this situation. it was important that a comprehensive study be made to determine the value of existing data and student responses as a technique for evaluating and improving existing prOgrams. Background of the Need The gravity of the highway safety problem is without parallel. Each year more than 50,000 people are killed in the United States and 2,000,000 people receive disabling injuries as the result of highway crashes. Related costs approximate 12 billion dollars annually (1). In Minnesota in 1972 1.031 persons lost their lives and more than 39.000 people were injured in traffic crashes (1). The Duluth police depart- ment reported 24 peOple killed and more than 970 people injured in their city that year (4). Education. more specifically driver education. remains among the major countermeasuresfor solving the highway crash problem (28). Since the 1930's. driver education programs have expanded rapidly in the nation's school systems. They have grown at a rate 2-1/2 times that of any other subject introduced into the school curriculum. From the school year of 1947-48 to that of 1970-71. the number of public schools providing driver education has increased from about 3.000 to an estimated 15.000. During this same time. the number of students enrolled in driver education courses has increased from 200.000 to an estimated 2.500.000 (42). As the number of programs and students increased. new approaches and techniques of instruction were sought to improve the efficiency and the quality of instruction. The multiple-car range. born in the City of Chicago in the mid-1930's. has become popular in the last 10 years and is presently being used in more than 464 schools throughout the country. Simulation. developed in 1953. has grown in papular- ity even faster than range programs and is presently being used in 1.011 schools in the country (42). Duluth's driver education prOgram has experienced a growth rate similar to that of the nation. From a modest beginning of 250 students in 1948, the prOgram has grown to include more than 2,000 students per year. Moreover. the instructional techniques have changed from the traditional pragram of classroom and on-the—street instruction to include multiple-car range and simulation phases. This rapid expansion of driver education in Duluth and other cities has created numerous and varied "growing pains." These growing pains have resulted in a quantitative response; it is now time to address the issue of quality. Given the problems of growth and the national demand for accountability of education. the need for investigating more effective driver education has never been greater. There has never been a better time to assess driver education programs and implement improvements where needed. Importance of the study A question (which is frequently debated) about driver education is. “Should driver education be included in the school curriculum?" Studies concluding that school-trained drivers are better drivers than non-school-trained drivers are cited by those who support driver education. Studies that conclude just the Opposite--that nonschool-trained drivers are better drivers than school-trained drivers--are referred to by 5 the critics of driver education. McGuire and Kersh have found. after reviewing the studies supporting each side. that the lack of scientific methodology makes the findings suspect on both sides of the issue (24). . Intuitively it can be concluded that all people are trained in learning to drive. Some people receive their training from a teacher. some from a relative. some from a friend or private school. but everyone does receive training. The question now becomes. "Where and how can driver education best be accomplished?“ The answer to the first part of the question-~where-- has been answered by the incentives presently being offered for selecting school driver education. The average driver education student is 16 years old. Since most states have a provision for a lower licensing age if the applicant has successfully completed an approved driver education program. the lG-year old is eligible to be licensed 1 or 2 years earlier than would otherwise be possible. Another decided advantage for the student who takes driver education is that. in virtually all the states. students who have successfully completed the driver education course are eligible for reduced insurance premiums. This economic feature also motivates parents to encourage their children to enroll in driver education. and seems to be one of the sustaining reasons for local support of driver education programs. Special financial support of driver education is also provided by most states. This has been a major incentive for schools to continue and expand their programs. These incentives alone would seem to guarantee a place for driver education in the school curriculum. Its position has been made even more secure. however. by its nationwide rec0gnition in the Highway Safety Act of 1966 (Public Law 89-564). which is a national attempt to promote and encourage schools to increase their efforts to highway safety. which states. “There is a national need for the improvement of public and private driver education and for making them more available." The act further provides for financial help to the states for implementation of these prOgrams (28). The position of driver education in the school has been established. Now efforts need to be directed toward providing the bext prOgrams possible. Ongoing assessment and evaluation are musts for every driver education prOgram, not only to provide direction for program improvement. but also to answer the demands for accountability that have accompanied the local. state. and national support. Purpose and Objectives The purpose of this study was to determine the feasibility of using selected student data bases for the assessment and evaluation of driver education programs in the State of Minnesota. A random sample of Duluth driver education students who enrolled in the laboratory instruction phase of 7 driver education in the summers of 1967. 1968. and 1969 served as subjects. The students' education records were obtained from the Duluth public school system; student driving records were obtained from the Minnesota Department of Public Safety, the St. Louis County Probate Court. and the Duluth Police Department: and a phone survey was conducted to obtain student demographic data, driving experience. and opinions. The specific treatment and criteria variables are discussed in greater detail in Chapter III. The six primary objectives were: 1. To complete a feasibility study by using interagency data sources for studies of the relative efficacy of driver education. 2. To compare the relative efficacy of range. traditional and simulation driver education practices as traffic safety countermeasures. Hypotheses: H01 There is no significant difference in the number of citations attributable to treatment type. H02 There is no significant difference in the number of accidents attributable to treatment type. 3. To compare the driving behavior of males and females. Hypotheses: H03 There is no significant difference in the number of citations attributable to sex differences. H04 There is no significant difference in the number of accidents attributabfle to sex differences. 4. To secure data which would contribute to the assessment and evaluation of driver education in Minnesota by ferreting out predictive relationships between demographic variables and accident and citation data. Hypotheses: Ho5 There is no significant relationship between citations or accidents and demographic variables such as. attitude. grade-point average. driving days per week. miles driven per week. percentage of night driving. percentage of weekend driving. model year of car driven. year of completion of classroom instruction. number of brothers. number of sisters. and number of cars in family. 5. To present suggestions offered by students for improving classroom and on-the-street instruction. 6. To investigate accident records for information that could be used for prOgram improvement. Age: Attitude: Classroom phase: Crash/Accident: Driving experience: Driving experience, days/week: Driving experience. night: Definition of Terms The age of the subject in months as of September 15. 1970. The on-the-street instructor's rating of the subject at the completion of instruction using a five-point scale: exceptional. good. average. below average. poor. The portion of the driver education prOgram that is taught in the classroom setting. An unplanned event resulting in death. injury. property damage. or inconvenience involving the use of a motor vehicle. An estimate of the average number of miles driven per week since being licensed. An estimate of the average number of days per week that some driving is done. An estimate of the percentage of driving that is done at night (7 p.m.--7 a.m.). Driving experience, weekend: Driving experience, motorcycle: Exposure: Laboratory phase: Minimum standards: Multiple-car range: On-the-street phase: Scholastic average: Simulation: 9 An estimate of the percentage of driving that is done during the weekend (Friday, 7 p.m.--Monday, 7 a.m.). An estimate of the total miles driven operating a motorcycle. That small segment of the total exposure data dealt with in this investigation: number of days of driving per week. number of miles driven per week, percent of night time driving, percent of weekend driving. and motorcycle miles driven. The phase of driver education employing "real" driving exPeriences. Included within this definition are simulation, off-street driving ranges. and on-the- street driving experiences conducted singly or in conjunction with each other. The minimum number of hours accepted by the State of Minnesota for classroom and laboratory instruction. Present driver education standards conform to the national standards: 30 hours of classroom and 6 hours of on-the-street instruction. Minnesota also provides for the substi- tution of at most 3 hours of on-the-street instruction on the ratio of 1:2 with multiple-car range instruction and on the ratio of 1:4 with simulation instruction. The multiple-car method permits several automobiles to be Operated simultaneously on a special off-street facility, under the direction of one or more teachers positioned outside the vehicles. The teacher typically communicates with students by radio. The phase of instruction conducted on public streets using a dually controlled automobile to provide actual traffic experience. The grade-point average of all subjects taken 9th grade through highest grade achieved as of September 15, 1970. A teaching-learning device using electronic driving components, programmed motion pictures, and an instructional response system in a classroom setting. 10 Traditional driver A prOgram that uses the standard 30 education: hours of classroom instruction and 6 hours of on-the-street instruction. Traffic accident: The traffic accidents (crashes) reported in this study refer to those reported to the Minnesota Department of Public Safety. which include all accidents resulting in injury to or death of any person or total property damage to an apparent extent of $100 or more and/or any accident reported to or investigated by the Duluth Police Department. Unique data: Data available at only one source. Organization of Remaining Chapters Chapter II contains a review of the literature. The review of the literature consists of descriptions of the goals and objectives of driver education. utilization of accident and citation records. the status of driver education. effectiveness of driver education. and research grants sponsored by the National Highway Traffic Safety Administration for evaluating driver education. Chapter III contains a detailed description of data collection process as well as the techniques used for analyzing these data. Chapter IV contains the findings based on the statistical test results and student interviews. Presented in Chapter V are the summary. major findings. conclusions. recommendations. recommendations for further research. and a discussion. CHAPTER II SELECTED REVIEW 0!“ LITERATURE Five general sources of literature relevant to this study were reviewed: (1) literature that describes the goals and objectives of driver education. (2) literature examining the utility of accident and citation records. (3) literature on the status of driver education. (4) literature on the effectiveness of driver education. and (5) literature on research contracts for the evaluation of driver education that have been sponsored by the National Highway Traffic Safety Administration (formerly Highway Traffic Safety Bureau). gggls ang_9bjectiyes of Driver Education The goals and objectives of driver education have been described in the Driver Education Standard 4.4.4 by the National Highway Traffic Safety Administration (NHTSA). This standard states that a student enrolled in a driver education pragram should be provided with “. . . a course of instruction designed to train his to drive skillfully and as safely as possible” (28). This definition is of particular relevance because states are required to meet this standard to qualify for federal highway funds . A popular driver education textbook. Let's Drive Ri ht. has similarly defined the goals of driver education. The 11 12 authors state that a course of instruction in driver education must be ”. . . dedicated to the idea of helping young people save their lives“ (18. Preface). A more general definition of the goals and objectives of driver education was set forth by the National Education Association (NEA) in their publication. Practices ang Policies. In this publication. driver education is described as a program bent toward '. . . achieving a desirable pattern of behavior in our society” (30. p. 3). This more general description was also used by the Automotive Safety Foundation in its recent publication. Ajgesogggg Curriculum in Driver and Traffic Safety Education (35), This publication refers to secondary school driver education as ”. . . one of the direct forces influencing operator behavior . .-." (35. p. 4). Since students of driver education are being educated to participate as motor vehicle operators in the total transportation systems. the goals of the system have relevance for driver education. The Institute of Traffic Engineers in its publication. An Introduction to Highway Transpgrtation En ineerin . has defined the goals of our transportation system as ”. . . safe. efficient. and convenient movement of goods and peeple' (23. p. 9). Utility of Accident and Citation Records The definitions of driver education state either explicity or implicity that accidents and citations are measures of success or failure of the programs. The result 13 is that evaluation of driver education should. if not must. include assessment of accident and violation experience of the graduates of the program. There are. however. limitations in using the number of accidents and the number of violations a person has had as a measure of program success because of differences in enforcement indices. variation in adjudication of cases. and differences in record keeping from One locale to another. In addition. complicationSexists from other factors such as the multiplicity of enforcement agencies and their differences. the insurance rate structure. and the definition of what constitutes a reportable accident from one state to another. For example. an accident in which there is $100 or more property damage is required to be reported in Minnesota. while in Wisconsin there must be at least $200 of property damage for an accident to be reported. 'As a result of these and other questions concerning accident and violation records. a number of studies have been undertaken to determine the utility of these records as criteria of driver performance or for identifying potentially hazardous drivers. Hakkinen (17) analyzed accident records to investigate the problem of criterion reliability. He concluded that accident records for an eight-year period are desirable for use as a reliable criterion for validating performance tests. Forbes (15). re-analyzing data from a study of over 29.500 drivers. found that by comparing accident 14 records of one three-year period to a succeeding three-year period. conclusions differed markedly from previous interpreations of accident records analyzed by the total time period only. The instability of accident records has been reported by Burg (8) who found that a large proportion of drivers with accidents in a succeeding three-year period were different from those with accidents in the preceding three years. He did. however. state that driving record information becomes more reliable as the period of time over which it is accumulated increases. Campbell (9) also reported instability of accident records of the North Carolina Department of Motor Vehicles when he found 80.7% of the drivers who had an accident in one two-year period were accident free during the second period. This study further revealed that many people who do experience accidents have no record of traffic violations in the prior two years. In other studies to identify potentially hazardous drivers Tarrants (41) pointed out that several studies have indicated accident and violation rates may be influenced by exposure. age. sex. socioeconomic status. education. and intelligence. Harrington (20) used a sample of 13.415 drivers in the 16-17 year old age group in a longitudinal study to determine the safety advantage gained by raising the minimum licensing age from 16 to 18 years of age. It was reported that no matter how the accident trend was adjusted for mileage. 15 the resulting rate showed a steady decrease across years (accidents decrease as mileage increases). The question cf accident proneness is the topic of a summary (Pech. McBride. and Cnppin. (33)) of nine studies conducted under the auspices of the California Department of Motor Vehicles. The initial focus of the studies was on accident and citation stability. In conclusion. the authors argue that accident frequencies are not direct measures of an individual's driving behavior. Rather. they are highly dependent on the behavior of others and various random contingencies. Bishop (7) reported a study in which he attempted to group “errors“ rather than individual categories of accident causation. One hundred-nineteen drivers under the age of 30 who had experienced accidents were interviewed to obtain descriptions of the accidents. Accidents were classed under eight error categories. the general findings being that most accidents were caused by failures in psychological areas (attention. attitude toward chances of accident involvement. drinking and driving. emotion. perception-judgment) rather than by physical or operational skills (emergency driving situations. factors involved in operating a motor vehicle. fatigue). In reviewing a large number of studies using accident and violation records. Leon Goldstein (16) pointed out one of the chief problems of using these data was the very low 16 occurence probabilities-~most drivers never have more than one reportable accident. and the largest group of them have none during periods of time ordinarily available for research programs. The use of accident and citation records as evaluative criteria is limited. but as was found by Berg (8) the value of these data improve over time. Also recognizing the limitations of accident and citation data each of the five National Highway Traffic Administration studies recommend the use of accident data as one of the measures of evaluating driver education programs. Status of Driver Education The literature indicates that current driver education programs for the nation's schools are far from standardized. and in some cases. exist in a form far from optimum. There is great diversity in course content. in the use of instructional aids. and in the nature and arrangements of various program elements. The scope of an individual course may range from fulfillment of only the minimum requirements necessary to prepare students for licensure. to a four-phased pragram which integrates classroom. simulation. multiple-car range. and on-the-street instruction. Seals states that currently the most common offering in driver education is ”30 and 6”: 30 hours of classroom instruction and 6 hours of ”laboratory” instruction. The latter consists of opportunities for actual or simulated 17 driving experiences. The "30 and 6" time allocation is generally regarded as an acceptable minimum standard course, although increased time is more and more frequently being recommended (36). According to the literature. course content and teaching techniques vary greatly among conventional "30 and 6" programs (5). Even greater disparity exists when the use of simulators and multiple-car ranges are taken into consideration. It is unlikely that this situation will change much in the immediate future. Currently. almost no data exist regarding the relative effectiveness of specific course content. of various teaching techniques. or of the various laboratory methods. Nor are data available that specify the optimum combination of the various laboratory techniques. In addition. and perhaps one of the major reasons for the existing course differences. is the range of professional preparation programs for driver education teachers (21). State requirements differ vastly, and many instructors are required to meet only minimum standards. Most states require a minimum three to six semester hours of course credit in driver education. but the overall range is wide (13). The NBA. in an attempt to strengthen this area. has recommended a teacher preparation curriculum for those intending to teach driver education. In 1968. however, only 68 of all universities and colleges with driver education programs conformed to these recommendations (32). NBA has also 18 established standards for university and college instructors of driver education. In 1968. slightly more than half of the active instructors met this standard (32). Despite the many problems besetting driver education. and despite its present ambiguous state. the continued existence of driver education as part of the secondary school curriculum seems assured. Anticipated national enrollment in driver education for the 1971—72 school year was about two and one-half million students. It was expected that close to 100 percent enrollment of all eligible students would be attained in the 1972-73 school year (27). This prediction not only suggests that driver education is here to stay. but also points out the urgency of the need for all programs to engage in pregram evaluation to assure the best pregram possible for the students. _Effectiveness of Driver Education The American Automobile Association (AAA) sponsored in 1945 one of the first effectiveness studies of driver education (12). In its study. the driving records of two groups of Cleveland. Ohio. high school students were compared. One group received high school driver education: the second group did not. The study concluded that driver education students have half as many accidents as do those students who have not had driver education. This study received national recognition and frequently has been cited by the advocates of driver education. Several 19 other studies (Minnesota, 1955 (31); New York. 1963 (28); Illinois. 1963 (37); Oregon. 1963 (39); and Wisconsin. 1970 (40)) found that driver education students have better driving records than students who have not had driver education. Coppin. Ferdun. and Peck (11) and McGuire and Kersh (25) have reported findings that are at variance with those described above. The conclusions of theSe studies raises serious questions about the effectiveness of driver education as a positive force in accident and violation reduction. In studies conducted in California and Mississippi (25). McGuire and Kersh concluded that little or no difference can be found between driver-educated and nondriver-educated groups in either accident or violation frequency. "Data concerning the effects of high school driver education are negative" (25). Where does the answer to the question of driver education effectiveness lie? Perhaps at some midway point between the divergent findings reported by various studies. A close scrutiny of much of the past and present research in driver education has exposed inadequacies in the design and in the reporting of the data in a majority of the studies. Their conclusions. therefore. may be held suspect (13). National Highway Traffic Safety Administration Recognizing the need and demand for evaluation of driver education. the National Highway Traffic Safety Administration (formerly Highway Traffic Safety Bureau) in 1967 20 awarded driver education research grants to four independent groups: (1) The American University. Washington. D.C. (3); (2) The Center for Safety. New York University. New York, New York (32); (3) Dunlap and Associates. Incorporated. Darien. Connecticut (13): and (4) Institute for Educational Development. El Segunde. California (22). These grants used funds established under the Highway Safety Act of 1966. The groups. working independently but concurrently. investigated the problems of evaluating driver education and driver training programs. A final report, submitted to the bureau by each group. suggested plans for short-term and long-term evaluation. as well as recommendations for instruments to be used in specific areas of evaluation. After the four reports were submitted to the Highway Traffic Safety Administration. a fifth contract was undertaken by another independent group. the National Academy of Sciences of the Highway Research Board. who subcontracted to the Educational Testing Service. Princeton. New Jersey (14). The task of this group was to synthesize the plans and instruments developed by the first four contract groups. and to develop plans for evaluating driver education proqrams on a national basis. The National Highway Traffic Safety Administration hOped that the national plan could and would be adOpted by all schools in the country. They felt that driver education programs would in this way be improved and 21 would then be able to demonstrate their ability as a traffic safety countermeasure. Because of the tremendous work that went into these five projects and the anticipated use of the final plan by schools throughout the country. a synopsis of each of the plans will be presented. A Study on Evaluation of Driyer Education. Washington. D.C.. The American University. Development Education and Training Research Institute. July. 1968. The American University report states that the objectives of its study was ”to develop and recommend a plan or plans for evaluating the effectiveness of driver education programs. not only as they are currently being taught. but also as they might be taught” (3. p. 7). The study deals with three major concerns-~an analysis of the evaluation problem. a proposed driver performance analysis. and the evaluation recommendations. In the discussion of the evaluation problem. the authors include a brief exposition of the currently accepted. fundamental principles of evaluation. The study continues with a brief discussion of criteria and measures. Then the authors outline what they consider an ”ideal evaluation process.” In essence. 22 this design would involve a large group of students who would be pre-tested to reveal pre-prOgram characteristics and capabilities. ”The students should then be randomly assigned to the various programs in such a way as to have matched groups of students in each program” (3. p. 19). The American University researchers conclude their study with two recommended planSo-a short-term evaluation plan and a long-term evaluation plan. The short-term plan is a survey of driver education courses and includes a proposed questionnaire for teachers. This questionnaire. "prOperly refined and pre-tested.” would offer "subjective evaluation of the probable accident countermeasure relevance of the learning experiences provided students in existing courses" (3. p. 156). A preliminary questionnaire is included in the report. The best feature of the plans presented in the study is rec0gnition of the need to assess driver education programs on the nature of the driving process and on the course's behavioral objectives. The teacher questionnaire has some useable features. but for the most part it offers only a rather quick and ready method for cataIOguing various driver 23 education courses throughout the country. Driver Education and Training. New York City. New York. New York University. The Center for Safety. May. 1968. The Center for Safety's proposed plan for evaluating the effectiveness of driver education deals first with the selection of variables: (1) evaluation criterion variables and (2) major related variables. The evaluation criteria are classified as short-term. intermediate. and long-term. .Short-term criterion variables are defined as those that can be used to measure the immediate effects of a driver education course. Two of these variables are driving performance via a simulator and driving performance via a road test. To develop the road test. the Center suggests three subprojects: determination of the criteria of good traffic driving performance. identification of the number and proportion of critical elements involved in good traffic driving performance. and a road test course layout to test the correctness of a driver's performance of the identified critical elements. The Center's other short-term criterion variables include self-rated driving performance. a test of driving knowledge. and a test of driving attitudes. The second phase would be another series of questionnaires and tests and would immediately follow course completion. There would be student rating of the teacher's ability. a self-rating of his own driving ability. a knowledge test. a driving attitude scale. a personality e I 24 inventory. a test of his ability on the driving simulator. and tests of driving performance by means of a road test and of driving performance under stress conditions. The third phase of testing would be done about two years after completion of the course. Essentially. tests given at this time would be the same as those given immediately after completing the course. except the teacher would not be rated. In addition. through the media of questionnaires and interviews. the students would supply biographical data. the number of accidents they were at least partially responsible for. the number of moving violations accumulated. and the number of ”near misses” experienced. The fourth phase of testing would be administered three to five years after completing the course and would be the same in content as that of the third phase. A second evaluative possibility reported is a comparison of the relative cost effectiveness of alternative techniques employed within various types of traffic safety programs. It is suggested that there be a comparison of “various methods used with comparable classes by comparably educated and experienced teachers“ (32. pp. 3-47). One major weakness of the Center's design is that it calls for the evaluation of a domain that has not been defined except in the most general terms. In the main design. it is proposed that the critical elements of the traffic driving task be identified. but this is to be done only in a substudy 25 of the major design. In addition. since results are to be measured by the road test. presumably the critical elements to be identified would include only motor skills and no related knowledge and attitudes. Driver Education and Training. Darien. Connecticut: Dunlap and Associates. Incorporated. May. 1968. Dunlap and Associates began their research by establishing a driver education information base obtained through a survey of driver education and educational research literature and by field trips and visits to various driver education programs and recognized authorities in the field. Using this data base. the group undertook an investigation of selected previous studies. The Associates' conclusion: While the overwhelming weight of “published studies are typically interpreted as strong endorsement for driver education. most of the studies had weaknesses in one or more areas . . . the findings across all studies are by no means consistent (13. pp. 16-17). And since they found the same weaknesses present in advocates' and critics' studies. they conclude that "it is difficult to state with any confidence that driver education has had a demonstrated effect on accident experience” (13. p. 17). 26 Dunlap and Associates make recommendations for evaluating driver education. Primarily. they prOpose a short—term evaluation approach involving self- evaluation by the schools. The evaluation device suggested by the Dunlap group is the driver education section of gyaluative Criteria. an evaluation scheme used by the National Study of Secondary School Evaluation and one that is lauded in the study as an ”excellent example of a carefully conceived and continuously revised evaluation program” (13. p. 6). The gyaluative Criteria is a check list plan and is done by a staff member or members of a school and by members of a visiting committee. The check list requires five letters: E (extensive coverage). 5 (moderate coverage). L (limited coverage). M (missing). and N (coverage not desirable or applicable). Space is allowed for notes or qualifications. Dunlap and Associates do not suggest a long-term evaluation plan. Due to present research re- garding data bases and subsequent changes in all traffic-related systems. the group believes that: 27 the only step which can be recommended with respect to a long-term evaluation plan is to develop a recommended driver education information base in the several states. at which time it would be possible to develop comparative evaluative studies (13. p. 49). Driver Educatign and TrainingflProject for the National Highway Safety Bureau. El Segundo. California: Institute for Edu- cational Development. June. 1968. During the study undertaken by the Institute for Educational DevelOpment. the group became convinced that the evaluation program of driver education must have a broader range than that which would be involved in determining the effectiveness of driver preparation programs. They believe that evaluation must be ”directed at the generic question of the influences on the acquisition. maintenance and updating of driver proficiency” (22. pp. 1-4). Consequently. the institute proposes to ask: Given an explicit definition of driver performance variables. how. when and where can influences be brought to bear on improving driver proficiency that will provide an appropriate return on investment in the effort (22. pp. 1-4). This approach makes identification of driver performance variables necessary. and the institute sees three distinct issues revolving about the variables: (1) What are the driver performance variables that contribute to effective driving in the real world? (2) To what extent and how can performance of these variables be influenced? (3) What interventions to influence driver proficiency are feasible? The Institute provides three alternative plans: Plan I: Evaluating Program Characteristics: Plan II: Evaluating Driver 28 Proficiency: Plan III: Validating PrOgram Effectiveness. After comparing the strengths and weaknesses of the short- and long-term implications of each of the plans. the group concluded that the three plans are not autonomous. but instead are complementary and iterative phases. Thus. they recommend that the three plans be implemented concurrently. Evaluation of Driver Education and Training Programs. Princeton. New Jersey: Educational Testing Serv1ce. March. 1969. Following the submission of the reports completed by the American University. Dunlap and Associates. New York University. and the Institute for Educational DevelOpment. work was begun on a fifth contract issued by the National Highway Traffic Safety Administration. This fifth report was granted to the National Academy of Sciences Highway Research Board. which appointed a project advisory committee and subcontracted to the Educational Testing Service (ETS). Princeton. New Jersey. It was the task of ETS to analyze the reports made by the four independent contractors. to synthesize their recommendations and plans. and to integrate the strengths of each into a national evaluation design. 29 Among major findings reported by ETS in regard to the four contracted studies were the following: (1) No definitive statements can be made about the effectiveness or ineffectiveness of driver education. (2) All concurred on the difficulty of evaluating the effectiveness of driver education on accidents. especially in view of present accident recording systems. The usefulness of intermediate criteria as surrogates for real-world driving performance was recognized: however. all four reports emphasized that "driver education must aim at the ultimate development of driver proficiency in the real world. as reflected in efficient traffic flow as well as accident reduction” (14. p. 5). (3) There was general agreement on the vital need of the driving task analysis in the real world and of the subsequent develOpment of more objective measures of driver proficiency and attitudes. Six major activities are identified by ETS as elements of the overall long-term evaluation: (1) measures of driver performance: (2) measures of program characteristics: (3) research studies of program characteristics and driver proficiency tests: (4) different levels of criteria: (5) research studies of driver proficiency tests and real world 30 driving: and (6) overall evaluation of driver education and training prOgrams. In each of these areas detailed suggestions are outlined. Following the completion of the driver education evaluation research grants. the National Highway Traffic Safety Administration awarded a contract to develop and evaluate a safe driving performance curriculum for secondary school driver education. This contract was awarded on a competitive basis to Human Resources Research Organization (HumRRo) and Central Missouri State University in June. 1972. The project is structured in two phases. Phase I requires the development of a safe performance curriculum and the procurement of instructional materials for pilot testing. The curriculum is being designed as an accident countermeasure for novice drivers. Phase I was completed in mid-February. 1973. Phase II requires pilot testing the safe performance curriculum under conditions that allow for the best possible research controls. This phase was started in June. 1973. in the Kansas City Public Schools. Kansas City. Missouri. The pilot test is currently scheduled for completion in September. 1974. Summary The review of the literature reveals that the goal of driver education is safe driving behavior: that driver education pragrams vary tremendously throughout the country: and that the 31 evaluation of driver education programs has produced inconsistent findings. Several studies indicate that accidents and citation records need to be a part of driver education evaluation. A number of studies. however. point out the weaknesses of using accident and citation data. A summary of studies to evaluate driver education prOgrams concluded that a national standard of evaluation is nearly. if not totally. impossible. In spite of the problems. the need for evaluation is expressed in the following statement from the Report of the Secretary's Advisory Committee on Traffic Safety, U.S. Department of Health. Education and Welfare: . . . there is inadequate scientific information on the nature of the driving process and the factors affecting it. and on the ways in which information obtained from research in this area can be employed in making driver education as effective as possible for the different kinds of individuals. (27) CHAPTER III DESIGN AND METHODOLOGY This study was designed to determine the feasibility of using selected student data bases for the assessment and evaluation of driver education programs in the state of Minnesota. The procedures used in this study, following the approval of Mr. Richard Pearsen. Assistant Superintendent and Mr. Harry Brown. Director of Driver Education in the Duluth. Minnesota Public Schools. are explained as they relate to subjects. data collection and data analysis. Subjects The population for this study consisted of all students taking driver education in the Duluth public schools during the years 1967. 1968. and 1969 (about 2,000 students each year). The samples of this study were 50 males and 50 females selected on a stratified random selection basis for 1967 and 1968. In 1969. 75 males and 75 females were selected, also using a stratified random selection. The sample size for the smaller samples was determined by using the formula N aND2+1 (19) 32 n 33 n 8 Optimum sample. N = population. D a the proportion of the standard deviation of the population chosen at the 95 percent confidence level. In this case. N was 2.000 each year. Arbitrarily. it was decided to sample statistics of minor population parameters within .10 of a standard deviation. N - ND2+1 n 2.000 2.000(.10)2+1 2.999.. ' 20.0+l n='95 Accordingly. a sample greater than 95 should give the desired confidence. Data Collection The data for the study were obtained in the summer and fall of 1970 from five sources: (1) the Duluth public schools. (2) the Duluth Police Department. (3) the St. Louis County juvenile court. (4) the Minnesota Department of Public Safety. and (5) the students. Duluth Public Schools The subjects for the study were selected from the records kept by the Department of Safety and Driver Education. 34 For the years of 1967 and 1968. the students were randomly assigned to one of two treatment groups: range or traditional. From each of these groups. 25 boys and 25 girls were randomly selected for the study. The 1969 students receiving driver education were randomly assigned to one of three treatment groups: range. traditional or simulation. The subjects. 25 boys and 25 girls. were then selected randomly from each of the three treatment groups. The records kept by the Department of Safety and Driver Education provided the driver's license number. treatment type. attitude (as defined by the on-the—street instructor). the year the subject received classroom instruction. age. address. and high school attended. Each subject's high school record. which is housed at the high school attended. was previewed to obtain the telephone number and the student's grade-point average. Then data were placed on the student profile sheet. Duluth Police Department The Duluth Police Department has a special accident investigation section that investigates and reports all accidents attended. The name of each driver involved is placed on a three-by-five card and alphabetized. The accident reports are then filed according to the location of the accident. Each subjects records were checked and the accident reports were obtained for those who had been involved in an accident. These data were placed on the student profile sheet. 35 Sty_LQgi§_County Juvenile Court Since many of the students were under age 18 at the time of receiving a citation. the adjudication of the charges and records are kept by the juvenile court. Permission was obtained from the court to review the files for subjects' traffic citation records. These data were also placed on the student profile sheets. State Department of Public Safgiy The driver's license record for each student was obtained from the Department of Public Safety. These data included accidents involved in and traffic violation convictions reported to the state. Each record was checked with the data from the Duluth Police Department and the juvenile court and all new data were then recorded on the student profile sheet. Student Subjects were contacted by phone to obtain driver experience data. the subject's feelings about the instruction received and the subject's recommendation for improving the program. These data were recorded on the student profile Sheet 0 Analysis The student profile sheets were coded and keypunched for computer analysis. Data analysis was undertaken at the 36 University of Wisconsin Computing Center* using standard library programs. Those analyses that did not require statistical programs were done manually (cross tabulation. content analysis) during the 1971-72 academic year. Dr. Robert Clasen. Director of the Instructional Research Center at the University of Wisconsin served as a consultant for analyzing the data. Feasibility Study of Interagency Data Sources The data necessary to develop student profiles is located at five of society‘s major institutions: (1) the school. (2) the police department. (3) the courts. (4) the state government, and the home. Data on individual students (those selected at random from the total population of students taking driver education in the Duluth public schools during 1967. 1968. and 1969) were sought in all five institutions. The following are data factors obtained. School-~driver license number. treatment type. attitude (as defined by the on-the-street instructor). the year the subject completed classroom instruction. age. address. high school attended. grade-point average. Duluth Police Department—~accident involvements and reports. *Part of the cost for computer time was furnished by the University of Wisconsin Graduate Research Committee. 37 Courts--violations of subjects who were under age 18. Sigi§--accident and violation data for each subject as recorded on their driver license record and the individual accident reports. Subject-~driving days per week. miles driven per week. percentage of night driving. percentage of weekend driving. miles of motorcycle driving. model year of car driven. number of brothers. number of sisters. number of cars in family, and the subjects recommendations for program improvement. To determine the accessibility and quality of the data available. cross tabulation and comparability studies were madEe Comparability of TraditionaiJL Range and Simulation Efficacy To compare the relative efficacy of traditional. range. and simulation driver education practices as traffic safety countermeasures. two analyses were performed with respect to the relevant hypotheses. Ho1 There is no significant difference in the number of citations attributable to treatment type. Ho There is no significant difference in the number of accidents attributable to treatment type. 38 Since these data were basically dichotomous citation-noncitation. accident-nonaccident, and treatmentl- treatment -treatment they were analyzed using a chi-square 2 3’ statistic to determine levels of significance. The level of significance was set at .05. Comparability of Sex Effects To compare the driving behavior (accidents and citations involvements) of male and female driver education students. two analyses were performed with respect to the relevant hypotheses. H03 There is no significant difference in the number of accidents attributable to sex differences. Ho There is no significant difference in the number of citations attributable to sex differences. Since these data were basically dichotomous male-female. citation-noncitation. and accident-nonaccident, they were analyzed using the chi-square statistic to determine levels of significance. The level of significance was set at .05. Correlative Data To assess the impact of intervening variables: (1) attitude, (2) grade-point average. (3) driving days per week. (4) miles driven per week. (5) percentage of night driving. (6) percentage of weekend driving, (7) miles of motorcycle driving, (8) model year of car driven, (9) year 39 of completion of classroom instruction (10) number of brothers. (11) number of sisters, (12) number of cars in family. (13) number of citations. and (14) number of accidents. a correlation matrix (Pearson Product Moment) was calculated. The coefficient of correlation was used to determine the strength of association between continuous demographic data and citations and accidents. The level of significance was set at .01. Ho5 There is no significant relationship between citations or accidents and demographic variables such as attitude. grade-point average. driving days per week, miles driven per week. percentage of night driving. percentage of weekend driving. miles of motorcycle driving. model year of car driven. year of completion of classroom instruction. number of brothers. number of sisters. and number of cars in family. Frequency Data of Student Suggestion This section presents a frequency tabulation of the suggestions offered by the students for improving classroom and on—the-street instruction. At most. two different responses were recorded for any one subject for each of the two categories. 4O Accident Report Data--ContributingyFactors In this section an analysis of the apparent contributing factors listed on the accident reports is presented. The other accident report data--time of day. time of year. citations. injury. road type. traffic control. weather. road surface. road defects. and the use of seat belts--are reported in Appendix B. Summary In this chapter the design and methodology used to conduct the study were presented. The sample pOpulation was selected from the students enrolled in summer driver education at Duluth. Minnesota. Data was collected from the school district. local police files. local court files. the state driver license record file and the student responses to a phone survey. A chi-square test of significance was used to test hypotheses concerning treatment and sex effects and a correlation matrix was used to determine the strength of association for correlative data. Cross tabulation and content analysis were used to study interagency data sources. student responses and accident records. In the following chapter the findings of this study are presented. CHAPTER IV ANALYSIS OF DATA The purpose of this study was to determine the feasibility of using selected student data bases for the assessment and evaluation of driver education programs in the state of Minnesota. The purpose was divided into six areas of study: (1) feasibility study of interagency data sources. (2) comparability of traditional. range and simulation efficacy. (3) comparability of sex effects. (4) correlative data. (5) frequency data of student suggestions. and (6) accident report data-~apparent contributing factors. In this chapter. data will be presented and each of the hypotheses recapitulated which either reject or fail to reject eaCh hypothesis. These results will be discussed in Chapter V. Feasibility Studyyof Interagency Data Sources A driver education evaluation project involving interagency data sources for data collection in Duluth. Minnesota. is feasible. There are. however, fundamental limitations and qualifications to that feasibility. 41 19 19 19 To In Ste no: 42 Table 1 contains a 3 year summary of state-level and total accident and citation records available. TABLE 1 STATE-LEVEL AND TOTAL ACCIDENT AND CITATION RECORDS AVAILABLE L m;— ‘L 4—.” Ind. Total Data Data Unique Records Type of Available Available Data Year Checked Record Local & State at State to State No. % No. % No. % 1967 100 Accident 21 100 7 33 5 24 Citation 33 100 30 91 18 55 Accident 21 100 15 71 6 29 1958 100 citation 31 100 27 87 13 42 Accident 20 100 18 9O 10 50 1969 148 Citation 23 100 15 65 4 17 Total 348 Accident 62 100 40 65 21 34 Citation 87 100 72 7 83 35 4O Of the 62 total accidents for which data were obtained. only 40 accidents (65%) had been recorded at the state level. This may seem insignificant. but of these 40. 21 or 53 percent of these data were unique at the state level. In other words. 53 percent of the accident data kept at the state level are unique accident data. Obviously. then. it is not feasible to do an accident study without checking state data sources. 43 Of the 87 citations on which data were collected. 72 (83%) of the citation data were available at the state level. Thirty-five of these pieces of data (49%) were unique to state files: that is. they were not obtainable from any other source. Citation studies. then. are not feasible without checking state sources. Conversely. in no instance (accident or citation) was the state a complete data source. Therefore. accident and citation studies demand that local data sources be checked as well. If cost is not an issue. certain additional data are available at the state level with respect to licensing. accidents. and citations. Such data are revealing from a summative as well as from an individual driver point of view. 1. Between 5 and 10 percent of drivers checked were involved in crashes each year. 2. Between 10 and 15 percent of drivers checked were involved in citation experiences each year. Local Police Department Data. The accident data are generally well organized and kept by the local police departments. Unfortunately. they are data which require hand tabulation. Table 2 presents data available from local police resources on accidents (citation data are available in the courts. locally. but not in the police records). 44 TABLE 2 LOCAL POLICE DEPARTMENT AND TOTAL ACCIDENT RECORDS AVAILABLE in: Data Ind. Total Data Available Data Unique Records Type of Available Local Police to Local Year Checked Record Local & State Dept. Police Dept. No. % No. % No. % 1967 100 Accident 21 100 17 81 14 67 1968 100 Accident 21 100 15 71 6 29 1969 148 Accident 20 100 10 50 2 10 Total 348 Accident 62 100 42 68 22 35 In total. 35 percent of the crash data available are available uniquely at the local police station: this data obviously must be used in complete accident studies. Qgggi. The juvenile court data can be obtained only on a limited basis. Juvenile court records are contained in files that must be individually hand-screened for each individual in the study. Table 3 presents the summary of citation data available from the courts and uniqueness of data found there. The 17 percent data available in the courts suggest that court data are also indispensable to accident and citation studies. 45 TABLE 3 LOCAL COURT AND TOTAL CITATION RECORDS AVAILABLE Ind. Total Data Data Data Records Type of Available Available Unique to Year Checked Record Local & State Local Court Local Court No. % No. % No. % 1967 100 Citation 33 100 15 45 3 9 1968 100 Citation 31 100 18 58 4 13 1969 148 Citation 23 100 19 83 8 35 Total 348 Citation 87 100 52 6O 15 17 School. As suggested by Table 4. school data are available and readily accessible. but most of such data must be reviewed and transferred by hand. These data are kept at the high school attended. Obviously. these are not accident and citation data. but many studies suggest that grade-point average is an important contributing variable where highway safety is concerned.(ll). StudentsZHome. Individuals. of course. carry unique data sets with them. particularly where attitudes. experiences. and perceptions are concerned. When these variables are at issue. data on them are readily available. Table 5 presents a summary of the data availability at Duluth schools. 46 TABLE 4 SUMMARY OF DATA AVAILABILITY AT SCHOOLS School Record Data ngg 1967 1968 1969 Total Records Checked Records Available 100 100 100 88* 148 148 336 348 100 88 100 97 *Some of the school records were burned in a school fire and could not be used. TABLE 5 SUMMARY OF INDIVIDUAL DATA AVAILABILITY Individual Data Year 1967 1968 1969 Total Contacts Attempted Contacts Made 100 95 100 92 148 129 348 316* Le 95 92 87 91 a pending court case. telephone number of married daughter. service. *One student refused to answer questions because of One parent refused to provide the Seven males were in 47 These results, while emanating from a Minnesota setting nevertheless can be generalized in the sense that they suggest that accident and citation studies that do not under- take several levels of data collection are very likely to be incomplete. The feasibility of collecting data from all of these sources depends on money and time available. Comparability of Traditionaly Range and Simulation Efficagy An attempt was made to compare the relative efficacy of range. simulation. and traditional driver education practices as traffic safety countermeasures. Ho There will be no significant difference 1 in the number of citations attributable to treatment type. Since these data were frequency counts. it was necessary to employ the chi-square statistic* (test of independence) to analyze them. Table 6 contains the chi-square results for violations—nonviolations by treatment type (range. traditional. simulation) for a random sampling of students completing driver education in 1967. 1968. and 1969. Perusal of Table 6 reveals that there is no significant dependent relationship (interaction) between treatment type and violations: therefore. the null hypothesis must be accepted. *George Ferguson. Statistical Anaiysis in PsycholOgy and Education (St. Louis. McGraw-Hill. 1966). pp. ZOO-204. and 407. 48 .oocomoeefip pcdowhwswwm oc memos .Q.m.z .wwma was head cw oomommo won as: codeHsswm::podofiaeeo nos memos .<.z .a.m.z .a.m.z .n.m.z monaosueamem do Hm>ms mosad> mm.m sw.m sm.m massam-eso Headsets m a a .u.a me.m em.o aw.m mafia» cassam:uso was cos cos anew seesaw musea> Haumpmspapm e: m .<.z .<.z .<.z .<.z cospaassem H: m mm as m: m saucepeaate m: s m: a am we amass .mmwmmmmm censuses masseuse sausages convenes .mmmmmmmm oz oz oz meme meme some sesame so ambasz MmmmmmMHMw m mmoq most> Sm.m aw.m :w.m musedmlflno Hdowkuo N H H .M.d mm.o om.o mm.m osad> ohmsqmlfino mad OOH ooa omen magnum mussa> Haoepmspapm mg m .<.z .<.z .<.z .<.z cosuassaem as m we e m: m assessesahe me e H: OH em me amass .mmmmmmmw pcoeeooa squeegee .mmmmmmmm .mmmmwmmm .mmmmwmmm oz oz oz some seas some oedema to honasz .mmmmmmmmw m mmoH mosad> oo.e so.» so.» so.m mausomuHso HsOHHHuo H H H H .o.o oom.om oo.HH mmm.o ss.: mmsHa> measomsHso mam wad 00H ooa omen maessm mous> HaoHpmeaHm moH NH me m A; m ms e oHaSom :NH m: em 3 am 3 mm PH «Hos coHpapHo ooHoaHHo coHpooHo coHpapHo coHpopHo ooHHopHo aoHpapHo aoHHapHo oz oz oz oz Hepoe mmma mwmfl bmma admoom no sobasz Rom m mmoH oocdofiuwcmHm : a mm.aaa Hw.mm Ana oadfioz co>Hma moawz Hoo. hm.m oaoonOpoz m:.-m hm.mam NmH ofldz m~.mw mm.m: :ma madame wsH>HHQ osoxooz .Q.m.z m:.o: mo psoomom 0H.mm mm.m: mma mad: Hm.:m mo.wm MMH madame maH>Huo pamHz .Q.m.z :2.H mo vcoomom mm.mm ma.m: mma was: om.mm mm.Hm mma oddsom zoos mom Hoo. mm.m so>fima moaez :m.moa ~m.moa mma was: wm.w mm.: MMH madame zooz Hem Ho. mw.m wsw>fiua when Ho.m mm.m mma was: .m osad> p sofipmfi>on memo: .m .mmm odbmfimd> pudendum mm:hwma nude» oo-eooH omszzoo mom saw so mmsm mmomomxm mom meHommm some p oz< onemo omaoz mo.» so.m so.m so.» ousoomuHso HaoHpHuo H H H H .o.o mH.oH mm.m mo.m om.m mosHa> orgasmsHoo osm osH ooH ooH ouHm mHoaam mooHo> HaOHpmeapm mwa ma ow m m: m >4 m madame MMH H: Hm NH mm :H mm ma was: pcooHoo< poooHoo< poooHooa pcmoHooa poooHooa HoooHooa poooHooa poooHoo< oz oz oz oz Hosea mmma coma Pwma mamoom Mo monasz .mmm m mm