Tuesday, July 5th, 2011 - 2.30 pm
Building E - Room 1

ABSTRACT

=========

The recent development of non supervised methods has allowed to have
very detailed information about human mixing patterns. Whilst social networks were formerly investigated through self administered surveys or through  individual interviews with the few participants, this has allowed to have detailed information about contact patterns within populations of several hundreds of people and with an unprecedented time resolution.
 

We propose an application of some statistical methods to investigate a 2-day dataset collected by the SocioPatterns project in a primaryschool. First, we will compare the validity of several models proposed in the literature for modeling contact durations with an adequation test.

Second, we will quantitatively investigate what in sociology is named as gender homophily among children. Homophily is a term that refers to how people use a certain number of attributes to define their identity, and often choose to interact with others who share thesame attributes.

We will focus on describing how children tend to interact differently with their mates according to their gender and how this preference changes with age.