SearcharxivSearch

arXiv subjects

Christian Steglich

Publications and source records attributed to Christian Steglich.

2 recordsLinked to original sources

Why echo chambers form and network interventions fail: Selection outpaces influence in dynamic networks

Are online networking services complicit in facilitating social change for the worse? In two empirically informed simulation studies, we give a proof-of-concept that the speed of networking and the amplification of network actors' relational preferences can have a profound impact on diffusion dynamics on social networks, essentially counteracting the benefits that should accrue from networking according to the strength of weak ties argument. Our findings can help understand variations in homogeneity of network neighbourhoods, i.e., in the degree to which these neighbourhoods act as "echo chambers", as well as the high context-dependency of success rates for a certain type of network intervention studies. They suggest that the general facilitation of connectivity like it today happens on the internet, combined with the use of personalisation algorithms, has strong and insufficiently understood effects on dynamic processes unfolding on the affected social networks.

physics.soc-ph

Statistical Power in Longitudinal Network Studies

Longitudinal social network studies can easily suffer from insufficient statistical power. Studies that simultaneously investigate change of network ties and change of nodal attributes (selection and influence studies) are particularly at risk because the number of nodal observations is typically much lower than the number of observed tie variables. This paper presents a simulation-based procedure to evaluate statistical power of longitudinal social network studies in which stochastic actor-oriented models (SAOMs) are to be applied. Two detailed case studies illustrate how statistical power is strongly affected by network size, number of data collection waves, effect sizes, missing data, and participant turnover. These issues should thus be explored in the design phase of longitudinal social network studies.

stat.AP