(1978). On Facilitating Networks for Social Change ... - INSNA

- 4 7 -. . . Abstracts, cont'd .Richard V . Farace and Rolf T . WigandDept . of Communication, Michigan State University, East Lansing, Michigan 48824, U .S .A .The Communication Industry in Economic Integration :The Case of West GermanyPaper presented to the International Communication Association,Organizational Communication DivisionChicago, Ill ., April 23-26, 1975National economies can be viewed as networks of interdependent relations among the firms, industriesand sectors that make up the total economic system . Within this economic system, the communication industryplays a central role by providing message and media products and services, i .e . the linking andcoordinating infrastructure .Some limited research has been done on the nature and importance of the communication industry inthe total economic network of a nation . The study reported here, however, represents the first applicationof currently existing network analysis techniques to this problem on a large-scale basis . The dataused for this study represent the interorganizational relationships among 365 West German firms, with a1970-based stock value of $333,000 or more . Interorganizational linkages are a function of (1) directionof ownership, and (2) magnitude of ownership (percentage of stock owned weighted by the value of theowned firm) .The data analysis reveals that there are a number of dominating industrial groups, including acommunication industry . The inter-firm relationships of the communication industry have been mapped andmeasured . The analysis demonstrates that there is relatively little dominance in and widespread integrationwith other major economic clusters of the economy . These findings are considered as being preliminarysince only publically reported data were available, i .e . privately controlled firms are not includedin the analysis . A number of measures for connectedness and integrativeness are presented that suggestthemselves in addition to the technique of network analysis per se as potential alternative measures .Ove FrankDept . of Statistics, University of Lund, Lund, SwedenEstimation of the Number of ConnectedComponents in a Graph by Using a Sampled SubgraphThe number of connected components of arbitrary of fixed node size in an unknown parent graph is tobe estimated by using a sampled subgraph . This problem is first discussed for two kinds of parent graphs :a transitive graph and a forest . Some approaches pertaining to a general parent graph are then illustratedby simple computer experiments .A Note on Bernoulli Sampling and Horvitz-Thompson EstimationForthcoming in Scandinavian Journal of StatisticsThe problem of estimating a total T = £ f y ij using observations y ij for (i,j)ES2 , wherei€V j (tVS is a random sample selected from V , is dealt with . The random sample S is chosen as a Bernoullisample with unequal or equal selection probabilities, and some earlier results by the author concerningHorvitz-Thompson estimators and variances are generalized .