Melbourne School of Psychological Sciences - Research Publications
Permanent URI for this collection
Search Results
1 - 7 of 7
-
ItemComparison of three methods for ascertainment of contact information relevant to respiratory pathogen transmission in encounter networks(BMC, 2010-06-10)BACKGROUND: Mathematical models of infection that consider targeted interventions are exquisitely dependent on the assumed mixing patterns of the population. We report on a pilot study designed to assess three different methods (one retrospective, two prospective) for obtaining contact data relevant to the determination of these mixing patterns. METHODS: 65 adults were asked to record their social encounters in each location visited during 6 study days using a novel method whereby a change in physical location of the study participant triggered data entry. Using a cross-over design, all participants recorded encounters on 3 days in a paper diary and 3 days using an electronic recording device (PDA). Participants were randomised to first prospective recording method. RESULTS: Both methods captured more contacts than a pre-study questionnaire, but ascertainment using the paper diary was superior to the PDA (mean difference: 4.52 (95% CI 0.28, 8.77). Paper diaries were found more acceptable to the participants compared with the PDA. Statistical analysis confirms that our results are broadly consistent with those reported from large-scale European based surveys. An association between household size (trend 0.14, 95% CI (0.06, 0.22), P < 0.001) and composition (presence of child 0.37, 95% CI (0.17, 0.56), P < 0.001) and the total number of reported contacts was observed, highlighting the importance of sampling study populations based on household characteristics as well as age. New contacts were still being recorded on the third study day, but compliance had declined, indicating that the optimal number of sample days represents a trade-off between completeness and quality of data for an individual. CONCLUSIONS: The study's location-based reporting design allows greater scope compared to other methods for examining differences in the characteristics of encounters over a range of environments. Improved parameterisation of dynamic transmission models gained from work of this type will aid in the development of more robust decision support tools to assist health policy makers and planners.
-
ItemNew specifications for exponential random graph models(WILEY-BLACKWELL, 2006)The most promising class of statistical models for expressing structural properties of social networks observed at one moment in time is the class of exponential random graph models (ERGMs), also known as p* models. The strong point of these models is that they can represent a variety of structural tendencies, such as transitivity, that define complicated dependence patterns not easily modeled by more basic probability models. Recently, Markov chain Monte Carlo (MCMC) algorithms have been developed that produce approximate maximum likelihood estimators. Applying these models in their traditional specification to observed network data often has led to problems, however, which can be traced back to the fact that important parts of the parameter space correspond to nearly degenerate distributions, which may lead to convergence problems of estimation algorithms, and a poor fit to empirical data. This paper proposes new specifications of exponential random graph models. These specifications represent structural properties such as transitivity and heterogeneity of degrees by more complicated graph statistics than the traditional star and triangle counts. Three kinds of statistics are proposed: geometrically weighted degree distributions, alternating k-triangles, and alternating independent two-paths. Examples are presented both of modeling graphs and digraphs, in which the new specifications lead to much better results than the earlier existing specifications of the ERGM. It is concluded that the new specifications increase the range and applicability of the ERGM as a tool for the statistical analysis of social networks.
-
Item
-
ItemStructural Logic of Intraorganizational Networks(INFORMS, 2010-05-01)In this study we examine the structural logic underlying complex intraorganizational networks. Drawing on different propositions about structural regularities in networks and using a comparative case study, we empirically investigate the structural logic of collaborative networks for the strategic decision process in two German corporations. In both organizations, data were gathered on cooperative relationships between all managers belonging to the top two management levels. We model structural regularities at the dyadic and the extradyadic level by applying a class of multivariate exponential random graph models. Our findings contribute to the existing literature in three ways: (1) Although networks are particularly likely to exhibit some types of structural regularities (e.g., reciprocity and transitivity), there are other relational forms such as cycles that seem to be of limited relevance. (2) Structural regularities are not limited to a single type of relation but may comprise instrumental and affective relational ties simultaneously. (3) An organization's formal cooperation structure has surprisingly limited influence on the structural patterns of cooperation, whereas friendship ties are embedded in managers' regular interaction patterns in many different ways.
-
ItemRecent developments in exponential random graph (p*) models for social networks(ELSEVIER, 2007-05)
-
ItemAn introduction to exponential random graph (p*) models for social networks(ELSEVIER, 2007-05)
-
ItemManufacturing relations: An empirical study of the organization of production across multiple networks(INFORMS, 2006)Organizational communities present two generic features that are recurrently documented in empirical studies, but only imperfectly accounted for in current models of interorganizational relations. The first is the tendency of participant organizations to construct observed macrostructure locally, through relational activities that involve only a small subset of possible network ties. The second is the tendency for different types of ties to overlap, concatenate, and induce a variety of local structures - or relational motifs - across network domains. A critical task in the analysis of organizational communities is to specify appropriate local dependence structures across multiple networks, starting from detailed observation of dyadic interaction among participants. In this paper we illustrate one way in which this analytical task might be accomplished in the context of a study of interorganizational networks. We use data that we have collected on different types of relationships among 106 organizations, located in Southern Italy, involved in the production of means of transportation to test hypotheses about patterns of local network ties and paths across multiple networks. Our empirical analysis is guided by the general claim that the formation of network ties is subject to endogenous and exogenous processes. We specify statistical models for random graphs that allow us to examine this claim, and to formulate and test specific hypotheses about the form that such network-based processes might take. The results that we report provide clear empirical support for the relational motifs implied by our hypotheses. We also find strong empirical support for the proposition that interorganizational dependencies extend across multiple networks.