Taller y Charla sobre Análisis de Redes

Fecha y Hora: Martes 22 de enero, 9:30 – 13:30
Lugar: Sala Zócalo Norte, Subterráneo Casa Irma Salas, CIAE.
Inscripciones Aquí

Humans are connected in various ways, forming social networks. The quality of social ties and the structure of social networks strongly influences individual-, group- and society-level outcomes. However, social networks cannot be modelled using traditional techniques of statistical analysis. This is because observations, that is, social ties, are not independent, violating the basic assumptions of standard statistical methods (e.g. regression-type analyses). In the last few decades, several types of models have been developed to deal with this problem. The course will include a short introduction to the statistical modelling of social networks and the most important model families. Three model types will be specifically discussed: quadratic assignment procedure (QAP), exponential random graph models (ERGM), and stochastic actor-oriented models (SAOM). The course will include short demonstrations of these methods in R. Those who wish to follow these analyses should bring laptops with R and RStudio (or their preferred alternative for using R) installed.

Fecha y Hora: Jueves 24 de enero, 15:00 – 16:30
Lugar: Sala Seminario, Segundo Piso, Casa Irma Salas, CIAE.
Inscripciones Aquí

Peers have a crucial effect on the academic outcomes of students. Especially in adolescence, the importance of peers becomes substantial. Analysing peer effects, however, comes with conceptual and methodological challenges. First, not all peers are equally important sources of influence: for instance, friends are more relevant in this regard. In spite of this, the majority of studies conceptualise peer effects as aggregates of the characteristics of the classroom context. In my talk, I will focus on the influence of subjectively important others on various academic outcomes, using information about students’ social networks. Second, people do not select their friends randomly; instead, they usually prefer befriending those similar to them. Such selection processes lead to similar cross-sectional outcomes as processes of social influence: friends being similar to each other. Therefore, I apply a longitudinal social network approach to disentangle selection and influence processes: this way, we can decide whether students with similar academic characteristics became friends in the first place, or whether students become more similar to each other over time (or, potentially, both). I present results showing existing friend effects on academic aspirations and preferences of adolescents, as well as on the study specializations they choose. In some cases, additional influence of same gender peers is also present.

Translate »