The distinctions among interaction patterns that we offered in Chapter 6 carry attendant operationalization considerations. Here we examine measurement and design implications. In Chapter 8, we tackle statistical analysis issues.

EXCHANGE PRINCIPLES

Directed and contingent responses

Matching and complementarity patterns do not require demonstrating that one person's behavior is directed toward or contingent upon the other. All that is needed is evidence that interactants' behaviors are highly similar or dissimilar. Sometimes this is evident through visual inspection of the data. It may be readily apparent, for example, that interactants are using formal forms of address when talking or that they are eschewing touch except for an initial handshake. Other times, statistical measures of association or comparisons between means are used to confirm empirically that patterns are the same or different.

One peculiar problem that can arise when claiming that matching is due to people orienting to social norms is the absence of a statistical test to verify similarity. If, for example, interactants are expected to show a high degree of pleasantness and politeness during initial interactions with strangers, and if we believe people adhere closely to these norms, then we should expect to see highly similar linguistic and nonverbal patterns within any given pair of people and little variance between pairs of people. This presents a dilemma because the statistical hypothesis to be tested is the null (i.e., no differences), which cannot itself be tested, and other measures of association discussed in Chapter 8 require some variance for any association to be detected.