No abstract

Traditional travel diary surveys collect 1 or 2 days of travel data from participant households. Although useful in determining the overall average travel behavior of a regional population, cross-sectional travel diary surveys provide little insight into intrahousehold and intraperson trip variability. Longitudinal surveys are generally preferred for examining travel variability. The objective of the research is to examine the intrahousehold travel variability in the Commute Atlanta, Georgia, study, a Global Positioning System–based instrumented-vehicle monitoring study that collected vehicle trips from approximately 500 vehicles in 260 representative households. The research effort uses 2004 travel data collected for the Commute Atlanta study, in which the average variability or deviation in the number of trips by a household was 3 trips/day. Demographic variables (e.g., household size, household income, vehicle ownership, number of children, number of workers, and number of students) significantly affect the day-to-day variability in the total number of household trips per day. The variability due to seasonal effects is controlled by separately analyzing travel data during specific months in spring, summer, and fall. Results indicate that demographic variables have a significant effect on the day-to-day variability of the household number of trips when the variability associated with seasonal effects is excluded. Vehicles identified by participants as being used always or occasionally for business or commercial purposes have travel patterns different from those of other vehicles, and their presence in the sample will significantly bias analytical results in the analysis of longitudinal data. Commercial use vehicles are excluded from travel variability analysis, and the argument is made that households with such vehicles must be treated as an independent sample in future travel diary data collection and longitudinal studies.

The Commute Atlanta program is an instrumented vehicle research program funded by the Federal Highway Administration (FHWA) Value Pricing Program and the Georgia Department of Transportation. One major objective of the multi-year program is to assess the effects of converting fixed automotive costs, such as vehicle registration and insurance fees, into variable operating costs. The main research hypothesis is that given a per-mile pricing system, participants will modify their driving patterns in an effort to reduce their total insurance and other payments, pocketing the savings. The Commute Atlanta project includes the parallel collection of instrumented vehicle data, household socio-demographic surveys, two-day travel diaries, and employer commute options surveys. The research team will monitor the changes in driving patterns and will use statistical analyses of household characteristics, vehicle travel, and relevant employer survey data to examine the relationships between the incentives offered and subsequent travel behavior changes. This paper focuses on the recruitment methods and travel diary response rates for the two-day diary surveys conducted in February/March 2004. As in other instrumented vehicle studies, researchers collected data that allow the comparison of reported diary travel with monitored vehicle travel. However, this paper focuses on a new type of comparison. Because the households had been recruited into the study 8-months prior to the diary study and their vehicles were transmitting activity data during the travel diary period, the research team had their first opportunity to examine whether there were differences in household vehicle activity across those 77% of households that completed the diary data collection and the 23% of those that did not. The differences were significant at both the high and low ends of the travel-reporting spectrum and may have some major implications for evolving household travel survey methods.

No abstract

The effectiveness of centerline rumble strips in reducing cross-over-the-centerline crashes and improving the safety of undivided roadways was evaluated. Twenty U.S. states, along with several Canadian provinces, are currently using centerline rumble strips. A detailed analysis of crashes on State Routes 2,20, and 88 in Massachusetts before and after installation of centerline rumble strips showed no significant change in crash frequencies; however, no fatal crashes have occurred on State Routes 2 and 88 since the installation of centerline rumble strips. Three cross-over-the-centerline fatal crashes did occur on State Route 20 after the centerline rumble strips were installed; centerline rumble strips were not a countermeasure to these specific crash types. Driver behavior at shoulder and centerline rumble strips was evaluated with a full-scale driving simulator. Drivers were found to react and correct the vehicle trajectory more quickly with centerline rumble strip encounters than with shoulder rumble strip encounters. About 27% of drivers made an initial leftward correction of their vehicles when encountering centerline rumble strips. Although this percentage may be inflated because of laboratory conditions, there is some probability of a driver confusing centerline rumble strips with shoulder rumble strips and reacting improperly. No improper (rightward) corrections were experienced with shoulder rumble strip scenarios. Centerline rumble strips were found to be effective at gaining drivers' attention and therefore to be an effective traffic control device and safety countermeasure in areas where a history of cross-over-the-centerline fatal and injury crashes occur.