Partial-least squares regression (PLS) is a statistical tool specifically designed to cope with problems like small datasets, missing values and the presence of multicollinearity . The paper proceeds first with a concise review of the PLS method.