Citation

Rock S. Transp. Res. Rec. 1997; 1581: 89-92.

Copyright

(Copyright © 1997, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

10.3141/1581-12

PMID

unavailable

Abstract

Instrumentation is one of the threats to the validity of experiments. Four possible cases of instrumentation in a time series of traffic accident statistics in Illinois since the mid-1970s were tested, primarily by using autoregressive integrated moving average methods. Two of these cases, a 1977 change in the reporting threshold for property-damage-only (PDO) accidents and a 1989 change in the definition of a fatality, were not found to be significant. A 1989 change in the method of tabulating monthly data and a 1992 change in the reporting threshold for PDO accidents were statistically significant. These two cases combined could account for a more than 15 percent decline in PDO accidents.

Language: en