ABSTRACT
Multiple regression is a means of predicting one variable from measurements of
other variables that are associated with it. These other variables are either easier to
measure than the variable to be predicted or they are antecedent to it. For example,
a crucial measure of the quality of concrete used in structures is the compressive
strength after 28 days (28-day strength). However, a contractor needs to detect any
sub-standard concrete when it is delivered and will test random samples for slump
and density that are used to predict the eventual 28-day strength.
