Trabajo 1 by Aníbal Olivera is licensed under CC BY 4.0
Diffusion of Innovations with Individual Preferences: The Role of Social Reinforcement and Homophilic Ties
This research explores the dynamics of innovation adoption using an agent-based model that integrates individual rational choice with homophilous social influence. We aim to understand why some innovations achieve widespread success ("all-or-nothing" patterns) while others fail, by looking beyond purely structural network effects.
- Hybrid Agent Model: Simulates adoption based on an innovation's intrinsic utility (
Γ), individual preferences (q_i), social adoption thresholds (τ_i), and the scope of homophilous influence (h). - Realistic Network Base: Employs ATP-net (N=1000), a simulated network with socio-demographic attributes derived from the American Trends Panel.
- Extensive Parameter Sweep: Analyzes over 4 million diffusion scenarios by varying:
- Intrinsic Innovation Utility (
Γ) - Scope of Social Influence (
h) - Mean (
μ_τ) and Standard Deviation (σ_τ) of social adoption thresholds. - Five distinct initial seeding strategies.
- Intrinsic Innovation Utility (
Our simulations highlight that successful, widespread adoption often emerges from a critical interplay of factors:
- The "tipping point" for mass adoption is not solely dependent on an innovation's inherent appeal (
Γ) but is significantly modulated by the reach of social influence (h). - Increased heterogeneity in the population's social adoption thresholds (
σ_τ) consistently promotes both higher overall adoption and the likelihood of abrupt, large-scale adoption events (phase transitions). - The model identifies non-structural conditions under which diffusion can be significantly blocked or rapidly accelerated.
/simulation_scripts: R scripts for the agent-based model./analysis_scripts: R scripts for data processing and heatmap generation./results_data: Raw and processed simulation output (.rdsfiles)./plots: PDF heatmaps visualizing key findings.