Network Sampling: Some First Steps

Social network research has been confined to small groups because large networks are intractable, and no systematic theory of network sampling exists. This paper describes a practical method for sampling average acquaintance volume (the average number of people known by each person) from large populations and derives confidence limits on the resulting estimates. It is shown that this average figure also yields an estimate of what has been called "network density." Applications of the procedure to community studies, hierarchical structures, and interorganizational networks are proposed. Problems in developing a general theory of network sampling are discussed.