Modelling Heterogeneous Effects in Network Contagion: Evidence from the Steam Community

Loading...
Thumbnail Image
Degree type
Graduate group
Discipline
Subject
Information
Contagion
Cascades
Diffusion
Hawkes
Applied Statistics
Business Analytics
Business Intelligence
Dynamic Systems
E-Commerce
Marketing
Statistical Models
Survival Analysis
Funder
Grant number
License
Copyright date
Distributor
Related resources
Contributor
Abstract

This study considers heterogeneous effects of reviews and social interactions on diffusion or contagion of new products in a networked setting, using a sample of interconnected public user profiles from the Steam Community. Ownership and reviews of two cult hit independent games – The Binding of Isaac: Rebirth, and To the Moon – are analyzed over a period of four years. This data was fit with a Hawkes Process Hazard Regression Model with exponential decay kernels for each game, yielding estimates of scale and duration of incremental heterogeneous actions within the network. This analysis finds strong, short term, additive, and marginally decreasing, social contagion effects from other users buying games, with much smaller, but also far more durable and highly significant, effects from review posting behavior in the network, independent of review quality. This seems to suggest that review influence, while still distinguishable from network homophily, is unlikely to lead to cascade effects.

Advisor
Lorin Hitt
Date of degree
2018-01-01
Date Range for Data Collection (Start Date)
Date Range for Data Collection (End Date)
Digital Object Identifier
Series name and number
Volume number
Issue number
Publisher
Publisher DOI
Journal Issue
Comments
Recommended citation