Mitchell Centre Seminar Series
|Dates:||23 November 2016|
|Times:||16:00 - 17:30|
|What is it:||Seminar|
|Organiser:||School of Social Sciences|
Marijtje A J Van Duijn, University of Groningen
Perceptions in Social Network Analysis
Social network data are usually collected as self-report dyadic data: actors assess the presence, absence or strength of one or more relationships with other actors in a well-defined group. The only information available about agreement on the ties comes from reciprocity in directed networks.
In some applications, however, dyads are reported by others than the actors involved in the data. Well-known examples are cognitive social structures where actors in the network not only report about their own ties, but also assess the ties between all other actors. The analysis of the resulting triadic data is not straightforward, as there are few models for triadic data, and many ways for reducing the data to dyadic data. The case of ego-centric data where ego reports on the ties between the alters in his/her social network may be one of the simpler cases to deal with.
In other applications, self-report data can be combined with information from external sources on the social network, for instance when parents and teachers assess the positive or negative relationships between children, or with information on other ties between the actors in the network, leading to a multiplex social network analysis.
In the talk I will present models and applications on how to deal with perceptions in social network analysis and would like to discuss their problems and possible solutions.
Travel and Contact Information