Many of the processes of interest to the transportation engineer exhibit a great deal of natural variability and uncertainty. The traffic stream, for example, does not have a single characteristic speed but a "distribution" of individual vehicle speeds. The transportation engineer needs a working knowledge of basic statistical procedures to account for the variability inherent in this "distribution of speeds" and to develop average or typical values that may be used to characterize the speed of the traffic stream as a whole. This section presents a review of some key statistical concepts and methods commonly used in transportation engineering. The review focuses on the following aspects of statistics and probability: descriptive statistics, probability, confidence bounds, and hypothesis testing.

The reader should consult a basic statistics textbook for additional background. [The Handbook of Statistical Methods for Engineers and Scientists (Wadsworth, 1990) is suggested.]

Descriptive Statistics

Descriptive statistics is concerned with recording and summarizing data. One way of summarizing data is to compute a value about which the data are centered. Three commonly used measures of central tendency are the mean, the mode, and the median. The mean is computed as

where n is the sample size (number of observations), x _i, are the individual observation of x. The mode is simply the most frequently occurring value in the data set, and the median is the middle number (50th percentile) in an ordered set of numbers.

In addition to summarizing a data set in terms...