attempt continually to determine how dependent your conclusions are likely to be because of (1) the particular criterion measures used, (2) the time of measurement, (3) the conditions outside the control of an individual, and (4) the distortions and biases inherent in the situation or the measuring instrument (human or otherwise)