Social Support: Methodological Issues in Design and Measurement

Despite plentiful evidence of a link between social support and adaptation to stress, we still know virtually nothing about why this connection exists. This paper is designed to help the researcher plan a study that offers the optimal test of any given explanation. We assume that the research is founded on a conceptual model of the association between stress, support, and strain. The specificity of such a model will depend upon what has been established from existing knowledge. The paper focuses on key issues in measurement, design, and analysis. A fundamental problem in measurement of social support is how to select from a host of possible strategies. The measurement section introduces important considerations that should govern that choice. The design section alerts the reader to sources of bias that commonly constrain the inferences that can be drawn from social support research. It also addresses the thorny problem of timing measurement to capture a process that transpires over time. The analysis section illustrates how structural equations can be used to model specific social support effects and discusses the circumstances that would make inclusion of additional variables advisable. Finally, we consider the limitations of relying exclusively on quantitative data and illustrate ways in which qualitative research can be used to develop unique information about the role of social support in circumventing the effects of stress.