ABSTRACT
In many studies the calculation of a correlation coefficient, in any of the ways outlined in the previous chapter, may be sufficient in itself. This may indicate the most profitable lines for further research, or provide the data from which maps of iso-correlation may be drawn. In other cases, however, it may be desirable to take the analysis a stage further by calculating the value that might be expected for one set of data if some given value occurs in the other set. This could be done by separate calculations each time, but it is more effective to draw on a graph the line that represents the relationship between the two sets of data. The requisite values can then be read off as required. Such a graphical representation is equally valuable as a descriptive device to illustrate the form of the relationship between the two variables.
