| Abstrakt: |
As a branch of network science, social network analysis widely uses graph techniques. Only in rare cases are results obtained from the graph models validated against "ground truth" and are directly applicable to objects in the investigated domain. Like extraneous solutions in mathematics, ungrounded mechanistic analogies, incorrect interpretation of indirect ties for intransitive relations and use of the "path" concept for social networks may lead to noninvertible results with no evidence outside the used graph model. The author investigates unimodal networks with dyadic ties, provides several examples of correct and incorrect applications and recovers the roots of incorrectness. [ABSTRACT FROM AUTHOR] |
