It is essential to distinguish between responses produced by a test compound and the
changes which occur in plasma and urine as normal biological variations, particularly
where statistical differences are found for data in the absence of any other findings. For
each study, we can consider several variables or experimental steps: these variables can
be divided into (1) biological, (2) analytical and (3) pharmacological or physiological
effects due to the test compound: these three components may interact in a synergistic or
antagonistic manner. Sometimes these variables are expressed as: