Publication Details

Accuracy of drivers' estimates of vehicle mileage driven

Type: Paper

Author(s): House, Elizabeth G.; Waller, Patricia F.

Pages: 23

Publisher: UNC Highway Safety Research Center

Url: https://finding-aids.lib.unc.edu/40114/

Publication Date: Nov-1971

Address: Chapel Hill, NC

Abstract: The primary purpose of this study was to determine the accuracy of mileage estimates provided by a sample of drivers. The study was conducted on the campus of the University of North Carolina in Chapel Hill. Odometer readings were obtained from a large sample of cars registered with the university and parked on the campus. Readings first were obtained at the beginning of the fall semester of 1969 and again at the end of the spring semester of 1970. Thus, for those cars observed both in the fall and in the following spring, it was possible to determine how far a particular car in the sample had been driven during a specific period of time. The Traffic and Motor Vehicle Registration Department of the University made available the names and addresses of the persons to whom the automobiles were registered. Toward the end of the spring 1970 semester each of these persons was sent a questionnaire requesting information about the car they usually drove and about how far the automobile was driven (including mileage put on by other people). Hence, it was possible to compare the estimates given by the drivers with the actual mileage as determined by the difference in odometer readings. A second purpose of the study was to determine whether there are any driver characteristics, such as age, sex, amount of responsibility assumed for the car, and degree of interest in the car, that appear to be associated with greater accuracy of mileage estimates. Therefore, in addition to the exposure estimates, the questionnair included this type of information and provided a basis for a responsibility index and an interest index. An 80 percent return of the questionnaire was obtained. Based on the 505 complete cases, there is a significant correlation coefficient of 0.65 between the estimated and the actual monthly mileage. Hence, when the actual mileage is high the estimate is high and vice versa. About half of the drivers estimated their mileage within 200 miles per month. High mileage was associated with lower scores on the responsibility index and higher scores on the interest index. However, neither index was related to accuracy of mileage estimates. Males tended to overestimate their mileage, while females tended to underestimate theirs. Overestimates were also associated with a large accumulation of the mileage by someone other than the respondent and with low car mileage. Thus, it was found that the male driver who has low car mileage and for whom much of the mileage is accumulated by someone else tends to overestimate the vehicle's mileage. The only driver characteristics which appear to be associated with accuracy of estimated mileage are sex and amount of mileage put on the car. Males are more accurate than females in their estimates, and the low mileage driver is more accurate than the high mileage driver. It appears that the male who puts low mileage on his car provides the most accurate mileage estimates.