Measurement of intimacy: conceptual and methodological issues of studying close relationships

In the broad context, an understanding of the role of intimacy has been discussed as essential to the development of a science of interpersonal relationships (Hinde, 1978). Specifically, the absence of an intimate confiding relationship has been identified as a vulnerability factor in the onset of depression under adverse circumstances (Brown & Harris, 1978 a; Costello, 1982; Solomon & Bromet, 1982). Aneshensel & Stone (1982) suggest that, assuming that lack of perceived social support is not just a manifestation of depression itself, lack of intimacy may contribute to the creation of depressive symptoms independent of life events. Henderson et al. (1981) suggest that the perceived availability and adequacy with which significant others meet an individual's requirements for attachment play a small but significant role in the onset of non-psychotic emotional illness under adverse circumstances. Waring et al. (1981a) found that deficiencies of marital intimacy were significantly associated with the presence of symptoms of non-psychotic emotional illness in one or both spouses. The significance of these important studies is dependent on the reliability and validity of the methods of measuring intimacy and the conceptual definition of intimacy. Several conceptual and methodological issues raised by these studies merit a closer examination. The studies described above have largely employed structured sociological interviews, containing questions about the quality of personal relationships. These questions allow a rating of the depth of'intimacy' by a trained interviewer. One question which arises about the concept of intimacy evaluated in these interviews i s : ' Whose intimacy is being measured? Hers? His? Or theirs?' The possibility exists that the perception of relationships as unavailable or inadequate might be a product of an individual's attitudes or moods. On the other hand, they might be a valid expression of how others have behaved towards the respondent. Data which record the spouse's or partner's perceptions of the relationship might be essential to evaluate the theoretical possibility that perceived differences in intimacy may have a different impact on vulnerability to emotional illness from congruent perception. A second question is whether narrow operational definitions of intimacy or broad definitions of the concept are to be preferred (Schaefer & Olson, 1981). Several authors have addressed this second conceptual and methodological question (Tennant & Bebbington, 1978; Tennant et al. 1982; Shapiro, 1979). They assert that the variables of marital status and the quality of the marital relationships have been confounded. Brown & Harris (1978 ft) reply that, although these variables are interlinked, they are not confounded: many unmarried women had intimate relationships with their boyfriends and many married women were unable to confide in their husbands. Separate analysis of the data for the unmarried and the married might be instructive. Thirdly, these sociological measures involved no discussion as to whether scores for partners should be analysed separately or as joint scores on their combined level of interpersonal intimacy in comparison with general population means or standards. Finally, recent evidence suggests that intimacy can be defined as a multifaceted dimension of interpersonal relationships which may or may not have been defined too narrowly by several questions regarding confiding in a spouse and availability and/or adequacy of close relationships (Schaefer & Olson, 1981; Waring et al. 1980). Recent studies suggest that a stronger correlation between depression and deficiencies of marital intimacy than that originally reported by Brown & Harris may be found by broader definitions of intimacy (Waring & Patton, 1984). Although the