A PRELIMINARY VALIDITY STUDY OF THE TEXAS AUTO VISITOR PROFILE. FINAL REPORT

The purpose of this study was to conduct a preliminary examination of the validity of the methods used for the TxDOT Auto Visitor Profile. There were two sources of bias investigated: (1) bias resulting from systematic differences in characteristics between respondents and nonrespondents to the TxDOT survey, referred to as non-response bias; and (2) bias resulting from systematic differences in characteristics between visitors who stop at the Travel Information Centers and those visitors who do not stop, referred to as sampling bias. The methods of this study were directed at out-of-state auto visitors traveling in Texas along Interstate 10 (at Orange near the Louisiana border) during November-January of 1989-1990 and May-July 1990. Both non-responses and sampling biases were found on many items of the Texas Auto Visitor Profile. The most important source of bias was that travelers from the neighboring state, when compared to other out-of-state travelers, were not only less likely to stop at the Travel Information Centers, but when they did, were less likely to respond to the TxDOT survey. The severe bias on point-of-origin of auto travelers influenced the magnitude and direction of bias on most other variables.