Abstract
The process of analysis of a set data aims to detect patterns, especially relationships between variables, in the data. Statistical techniques can be described broadly as techniques to detect particular types of patterns, and to test whether such patterns ‘really exist’. The use of graphs to aid in pattern detection is long established. A picture can be used to make a point quickly and easily. In the context of data analysis, it is making the point more powerfully than does a table of data. A well chosen and well drawn graph makes its point - reveals a pattern - very powerfully. Equally, a badly chosen or dishonestly drawn graph makes a point incorectly - but equally powerfully.