The correct analysis of engineering data is crucial for the understanding of the subtleties of any engineering system. Incorrect analysis could result in errors and the misallocation of scarce resources. Appropriate analysis requires that the engineer has a good understanding of the character of the data. Statistical techniques that hide the structure of the data may result in the engineer arriving at the wrong conclusions. Recent developments in the statistical procedures along with computer graphics technology provide the environment and mechanism for obtaining this understanding. This paper reviews some graphical methods that aid in the understanding and presentation of engineering data. Emphasis is given to the interactions between the graphics procedures and statistical methods, and the analysis environment created by combining exploratory data analysis and statistical tests. The graphics methods are illustrated with reference to data on vehicle speed, origin-destination, average daily traffic flows, and parking lot performance. The associated statistical procedures are descriptive statistics, hypothesis testing, distribution fitting, and relationship determination.
|Number of pages||11|
|Journal||Microcomputers in Civil Engineering|
|Publication status||Published - 1 Jan 1987|