Page:CIAdeceptionMaximsFactFolklore 1980.pdf/12

C00036554 : Second, the tactical engagements contained in the data base constitute what is termed a convenience sample rather than a random sample (inter alia a function of data availability) and may not be fully representative. Thus, statistical tests of hypotheses could be misleading — but then so too would be counts, cross-tabulations, etc.

Thus, the view taken in this paper is that statistical tests are appropriate provided the results are interpreted with due caution considering the inherent data limitations. The analyses here should be termed exploratory rather than adjudicatory. Absent the selection and analysis of a truly random sample, an activity perhaps impossible in principle, this data base is sui generis, one of a kind. It would be imprudent to fail to consider such conclusions as may follow from analysis of these data.

The Deception Maxims
The following section contains ten principles or maxims that are relevant to deception. No claim is offered that this is a minimal, sufficient set, that these principles are entirely self-consistent or that