This article explores problems caused by nonrespondents in sociometric studies of organizational communication and describes how networks that include nonrespondents can be analyzed. An illustrative example is used to conceptualize the problems and issues in analyzing such networks. An empirical study is described that operationalizes the decision criteria for choosing a method of analysis. Suggestions are offered for the design of communication network studies that may enhance response rates and provide the information needed to justify how incomplete network data sets may be analyzed.

No abstract

This article examines the seasonal pattern of television viewing. A mathematical model that describes the frequency of television viewing as a function of time is presented. We hypothesize that it is composed of three components: an oscillation that describes the seasonal variation in viewing; a logistic translation, to account for the increase in viewing due to the diffusion of television; and a time-independent term. The model is evaluated with monthly data from the Nielsen Television Index (N = 460). It fits the television-viewing data, explaining about 99.9% of the variance in viewing. All three components are statistically significant. Next, the seasonal nature of television viewing is explored. It may be accounted for by environmental factors—temperature, precipitation, and daylight. The relations among these variables are demonstrated through spectral analysis. The implications for other cyclic processes in communication are discussed.

No abstract

This article documents the 30-year history of communication network research at Michigan State University (M.S.U.), providing a case study of the evolution and diffusion of an academic innovation. Three past and continuing issues for network scholars are identified: a lack of professional reward for developing user-friendly computer programs, unresolved methodological problems, and a need for better theoretical and conceptual frameworks. The narrative also illustrates the difficulty communication as a discipline has in impacting broader intellectual traditions. The story begins with the first doctoral dissertation (Schwartz, 1968) and the first network analysis software program in 1970 (Richards' Negopy), continuing to the last dissertation (Susskind, 1996), and ending in 1998 when J. David Johnson left the M.