Three groups of 60 subjects, divided by age and sex, completed scales for rating the applicability of rules to 22 different relationships; 33 rules were used for all relationships together with a number of relation-specific rules. It was predicted that there would be universal rules applying to all relationships. Three of the 33 potentially universal rules were universal, or nearly so, while five more applied to half or more of the 22 relationships. These were less about the exchange of rewards than about regulating privacy, sex, public conduct and keeping confidences. The prediction that relationships would fall into intimate and non-intimate clusters was confirmed. The reverse of the same rules, or quite different rules, applied to the two clusters.

Intimate relationships had more rules, including more of the rules common to many relationships, and rules about social and emotional exchange. Non-intimate relations had rules that keep people at a distance. Particular relationships had rules that enable people to avoid sources of conflict unique to those relationships, e.g. the proscription of sex between kin, rules about the conduct of business in work relationships.

It was concluded that rules function to maintain interaction so that goals can be achieved—by coordinating behaviour, regulating the level of intimacy and avoiding relationship-specific sources of conflict. There were a number of age and sex differences: females endorsed rules about intimacy and self-disclosure more than males did, and formal rules were endorsed more by older subjects.