90 Methodological Issues in Developing a Scale to Measure Social Support

In recent years there has been con-siderable research and clinical interest in developing instruments to assess social supports available to individuals. There is, however, a notable deficiency of attempts to evaluate the psychometric prop-erties of these questionnaires. The present article describes efforts made to evaluate the properties of a Social Relationship Scale (SRS) that was developed as part of a prospective study of the psychoso-cial influences on the health status of a population. Some descriptive scale statistics are also presented. As part of a 2-year prospective study, a Social Relationship Scale (SRS) was developed to measure the role of social support in cush-ioning the effects of life stressors on health (McFarlane et al. 1980). The notion that social support acts as a buffer or cushion against disease has been well de-scribed by Weiss (1969), An-tonovsky (1974), and Cassel (1974), as well as by Caplan (1974), who has drawn attention to its impor-tance in community mental health. Henderson (1977) postulated that social bonds are necessary for per-sons to maintain a reasonable de-gree of affective comfort and to operate effectively in the face of adversity. Brown, Bhrolchain, and Harris (1975) found that lack of an intimate confiding relationship is one of four factors associated with increased morbidity in the pres-ence of escalating adverse events. In the development of the SRS, we chose to focus on both the quantitative and qualitative as-pects of the network of relation-ships available to our subjects. Our reading of the literature and our own experience suggested that people who can form trusting and reciprocally helpful relationships have access to coping resources. Thus, our task was to develop a questionnaire that would cover all categories of life stress and the number, type, and quality of re-lationships available for discussion in each of these categories. The questionnaire is designed to pro-vide information about the total number of relationships available, as well as the class or classes of relationships associated with dif-ferent kinds of stress. In addition, it can be used to determine the number of different categories of stress in which a particular re-lationship plays a role