Flashman, Jennifer (2008). Change and stability in friendship networks. 2008 Add Health Users Conference.
Bethesda, MD: University of North Carolina at Chapel Hill, Carolina Population Center.
In this paper, I use network analysis to model the relationship between friendship network characteristics, structure, and change. I show how an individual’s race, parental education, and academic achievement affect friendship network change and how differences in these characteristics between individuals predict change, net of the structural characteristics of the network. This analysis uses data from the high schools in the Add Health saturated sample of schools. Because all students in these 10 schools were followed up across waves and asked to nominate their in-school friends, I can track their friendship networks across time using the In-School, Wave I In-Home, and Wave II In-Home surveys. I use the AHAA transcript data to capture students’ academic achievement. Academic achievement is observed across students’ high school careers, allowing me to model the relationship between friendship network change and change in academic achievement. I model network change using an actor oriented Markov chain model. This model is carried out at the school level, for each of the 10 high schools included in the saturated sample. School level results are then combined in a meta analysis to describe the general relationship between individual characteristics, network characteristics, and network change. Through this analysis, I show 1) how homophily in friendship networks affects network change, 2) whether similarity in race is more important to network change than similarity in socioeconomic background or academic achievement, and 3) how individual academic achievement responds to the compositional change in a network’s academic achievement.
2008 Add Health Users Conference
University of North Carolina at Chapel Hill, Carolina Population Center
City of Publication