Citation

Forbes TW, Katz MS. Highw. Res. Board bull. 1957; 161: 18-29.

Copyright

(Copyright © 1957, National Research Council (U.S.A.), Highway Research Board)

DOI

unavailable

PMID

unavailable

Abstract

A specialized segment is reported from research sponsored by the pennsylvania turnpike commission on human and physical factors as causes of turnpike accidents. State police accident reports are coded and analyzed continuously by ibm punch-card technique, thus furnishing indications of trends and clues and data for special studies. Analysis by 3-month periods of the various physical and driver behavior factors showed 'inadequate coping with road conditions' involved in a higher percentage of passenger-car-responsible than in truck-responsible accidents, especially in the fall and winter quarters. Further comparisons indicated that many of these accidents occurred on snowy and icy highway after the weather had cleared. These results indicated the importance of immediate cindering and the elimination of winter road conditions. The pennsylvania turnpike commission instituted improved maintenance and enforcement procedures to accomplish reduction of the hazard. A special study of driver behaviors and locations was undertaken to isolate possible combinations of driver behavior, weather, and road characteristics which contributed more than their share of winter road condition accidents. It was necessary to correct for traffic volume and exposure to winter weather conditions. Estimates of traffic volumes in each direction and of percentage of time exposed to winter road conditions were developed for each 10-mile strip of highway. Contingency analyses tested for greater-than-expected relationships between various factors, such as grade and curve combinations and accidents. Certain 10-mile segments showed a rate per million vehicle miles from five to ten times as high as other 10-mile segments. East-bound and west-bound accidenth on the same segmant often differed, and certain stretches in the mountain area showed the highest rates. On the basis of exposure to winter road conditions, such stretches showed higher than expected values under icy conditions but not necessarily under other conditions. Use of the correction factors for exposure showed the hazards for the ordinary motorist to be much greater under the winter road conditions, even though the total number of accidents was often greatest for dry roadway on an over-all basis. Relationships of different driver behaviors and other factors in accidents under the different road conditions were analyzed. A further study is under way to investigate further underlying causes for the relationships indicated.