How Conformity Drives Social Trends | Kaleda Denton | TEDxNewEngland

Key Concepts

• Conformity: The tendency to adopt a trait or behavior with a probability greater than its observed frequency in a sample.
• Anti-conformity: The tendency to adopt a trait or behavior with a probability less than its observed frequency.
• Disproportionate Tendency: The core mechanism of conformity where individuals copy the majority at a rate higher than the majority's actual prevalence.
• Mathematical Modeling: A framework used to simulate individual-level behaviors to predict population-level dynamics (e.g., fads, cycles, polarization).
• Population Structure: The division of a population into subgroups or networks, which significantly influences how behaviors spread.
• Cycling Dynamics: A phenomenon where population preferences oscillate between two states (e.g., skinny jeans vs. baggy jeans) rather than stabilizing.

---

1. Main Topics and Theoretical Framework

The video explores the drivers of social trends and group behavior through the lens of evolutionary biology and mathematical modeling. The speaker argues that complex group behaviors—such as fashion cycles, mob mentality, and political polarization—can be understood by scaling up simple individual decision-making rules.

• The 1985 Conformity Model: This foundational model posits that individuals choose between two options (A or B) by sampling a subset of the population. If an individual observes three people, they are statistically likely to adopt the majority trait with a probability higher than the observed frequency.
• The Role of Sample Size: The speaker discovered that the number of people sampled is critical. While small samples (n=3) lead to the dominance of one trait, larger samples (n≥5) allow for cycling, where populations oscillate between two traits, explaining why trends like fashion styles eventually return.

2. Methodologies and Modeling Processes

To study these behaviors, the speaker outlines a three-step framework for building predictive models:

1. Specify Parameters: Define the hypothetical world, including population size, number of subgroups, and the rate of movement (migration) between those groups.
2. Define Individual Behavior: Write an equation representing how a single person makes a choice based on their sample of others.
3. Scale to Population: Apply the equation to every individual in the simulation to observe emergent patterns over time.

3. Key Arguments and Research Findings

• The "Anti-Conformity" Paradox: Contrary to early assumptions that anti-conformity would stabilize a population at a 50/50 split, the speaker’s research shows that with sufficient sample sizes, anti-conformity creates perpetual cycles of popularity.
• The Impact of Population Structure: The speaker challenges the assumption that conformity always maintains in-group/out-group differences. Using a model with two subgroups and minor migration, they found that if one group exhibits "strong" conformity, it can force the entire population (including the other group) to converge on a single trait, effectively erasing group differences.
• Model Utility: Citing statistician George Box, the speaker notes: "All models are wrong, but some are useful." Models are not meant to perfectly replicate human complexity but to identify the underlying mechanisms that lead to norms, fads, and polarization.

4. Real-World Applications

• Fashion Trends: Explains the cyclical nature of styles (e.g., skinny vs. baggy jeans) as a result of anti-conformity and large-scale social observation.
• Social Dynamics: Provides insight into why people adopt specific political views or health behaviors (e.g., mask-wearing) based on their immediate social networks.
• Artificial Intelligence: The speaker notes that conformity biases have been detected in Large Language Models (LLMs) like ChatGPT, suggesting these mathematical patterns transcend biological organisms.

5. Synthesis and Conclusion

The study of conformity and anti-conformity reveals that individual choices are rarely made in a vacuum. By combining individual-level biases with population structure, we can predict how trends emerge and why they sometimes stabilize or cycle. The speaker concludes by urging the audience to reflect on their own social networks and the "groups" they belong to, as these structures fundamentally shape the choices we believe are our own. The ultimate takeaway is that understanding these mathematical biases allows for greater agency in navigating the social pressures of conformity and the desire to stand out.