Conditionally random sampling in relational analysis

For a statistical model where joint probabilities originate from random sam pling procedures st times with, at times without, replacements, the computer affords means to handle the implications of this distinction. They often are unma-nageable for hand computation and therefore neglected, even though they intervene, sometimes significantly, in the evaluation of interpersonal feelings and perceptions as given by a group experiment. Namely, in computing the variance of the number of occurrences - - assumed to be randomly distributed in a model that matches the group experimeut partially - - for a given diadie relationship whose observed frequency value in the experiment is to be tested for significance.