ABSTRACT

Scientific decisions should be based on sound analysis and accurate information. This chapter provides the theory and interpretation of confidence intervals, hypothesis tests, and population comparisons, which are statistical constructs (tools) used to ask and answer questions about the transportation phenomena under study. Despite their enormous utility, confidence intervals are often ignored in transportation practice and hypothesis tests and population comparisons are frequently misused and misinterpreted. The techniques discussed in this chapter can be used to formulate, test, and make informed decisions regarding a large number of hypotheses. Such questions as the following serve as examples. Does crash occurrence at a particular intersection support the notion that it is a hazardous location? Do trafficcalming measures reduce traffic speeds? Does route guidance information implemented via a variable message sign system successfully divert motorists from congested areas? Did the deregulation of the air-transport market increase the market share for business travel? Does altering the levels of operating subsidies to transit systems change their operating performance? To address these and similar types of questions, transportation researchers and professionals can apply the techniques presented in this chapter.