4.2 Scientifically Valid

By scientifically valid, we mean measurements and measures produced with mathematical operations that are theoretically sound. As we will discuss below, measures of likelihood, importance, consequence, and impact must possess the proportionality or ratio scale measurement property.  For example, evaluating the importance of reputation and short-term financial impact on an ordinal scale of High, Medium, …. and using ordinal values of 5, 4, 3, … in computing risk is not mathematically meaningful and not scientifically valid.  In spite of this, most risk assessments being used today multiply ordinal measures of likelihood and/or impact, resulting in risk measures that are mathematically meaningless and, in fact, can be worse than having no risk measures at all!