Abstract

We used Exponential Random Graph Models (ERGMs) to determine how important is the role played by homophily, transitivity and preferential attachment effects in the formation of friendship networks obtained from data collected in 3 elementary schools; two are located in the rural area and the other is in the urban area of Yucatan, Mexico. Structural terms were considered, as well as individual attributes (gender and scholar grade) of each student in the network. We hypothesize three network effects thought to contribute to the observed structure: homophily, triad closure and preferential attachment. Assessment model was done using p-value, Akaike information criteria (AIC) and a graphical goodness of fit (GOF). Results from exponential random graph models support our hypothesized homophily and triad closure effects. All friendship networks in our investigation had positive and significant triad closure and homophily effects. On the other hand, we do not find evidence for preferential attachment processes. Well connected students are not more prone to gain additional links than sparsely connected students.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call