ABSTRACT
Relationships between variables are very important in real-life decision-making. It is often vital to know, when the value of one variable is changed, how other variables are likely to be affected. The purpose of regression analysis, which is only carried out on intervally scaled variables, is to enable us to predict the value of a particular variable, given the value(s) of one or more other variables. Various conditions need to be met if one is to carry out a least-squares linear regression analysis and obtain the Pearson product moment correlation coefficient, including the requirement of interval scale measurement. It is important to scrutinise the scatter diagram carefully, as it can very often suggest the type of relationship and hence the form of the regression equation. Data on the number of casualties in road accidents and the number of currently licensed vehicles in Wales were taken from the Digest of Welsh Statistics.
