Strauss, M., Carnahan, J.
Publication F2010-D-043. FISITA World Automotive Congress Proceedings, Budapest, Hungary
(International Federation of Automotive Engineering Societies)
View on FISITA.com
The purpose of this study was to investigate the accuracy of the typical adult’s ability to view, and then estimate distances along a roadway. A survey was conducted where 124 subjects were asked to look at, and then estimate the distance to common objects. The entire population tended to underestimate distances to objects that were from about 20 to 400 feet (6 to 122 meters) away. The average estimation error was –8.6%; here negative error indicates underestimation. The variation in performance among individuals was extremely large, with error extremes ranging from –96% to 811%, including the outliers identified in the data. With outliers removed, the average estimation error changed to –23.4% because the outliers removed were large overestimation (positive) errors.
With outliers included in the data, the distribution of error did not conform to a Gaussian (normal) distribution because of those same large overestimation errors. Box plots were used to identify 9 respondents who gave 15 error estimates which were outliers – extraordinary in their difference from the rest of the data. Tests for independence indicated that the likelihood of an error estimate being an outlier was not influenced by gender, age, or whether the subject was a police officer. Nonparametric statistical methods compared the median performance of males to females for the data with outliers included; the female subjects’ distance estimation error was larger, in a statistically significant way.
Once outliers were removed from the data, analysis of variance methods could then be used to explore differences in average estimation error. In particular, gender and a self-assessment of distance estimation ability were statistically significant, while age and police were generally not. The survey of the participants also included the documenting of their education, occupation and hobby interests. A surprising result was that many survey subjects with apparently high quantitative ability did poorly at estimating distance. Other analyses indicated that estimation accuracy was better for subjects (such as skilled trades, and certain sports) who were especially cognizant of distances. Some additional results are provided for differences in accuracy related to the kind of object being viewed.