As computing devices become ubiquitous, devices are expected to encounter and associate with one another spontaneously to form ad hoc networks for sharing resources. One of the challenges in research remains in the task of making a device association scheme secure and, at the same time, easy to use. There have been numerous proposed solutions from literatures, with each solution designed for specific purposes and scenarios. At present, there is no clarity of differences and similarities amongst those association methods. In this paper, we present a classification of device association methods based on categories that influence the usability of an association.