715: Make Better Decisions with Data, with Dr. Allen Downey
Sep 19, 2023
auto_awesome
Renowned author and professor, Dr. Allen Downey, discusses Survival Analysis, Preston's Paradox, Overton Window, base rate fallacy, and protecting against biased samples. He explores statistical pitfalls, rare event preparation, and gradual belief shifts in decision-making using data.
Survival Analysis enhances decision-making with data insights.
Understanding statistical paradoxes like Preston's Paradox enriches analytical skills.
The Overton Window influences political ideologies and societal opinions.
Analyzing age, period, and cohort effects is crucial for gauging public attitudes.
Deep dives
Understanding Black Swan Events & Their Impact on Planning
Black swan events are unexpected, high-impact occurrences that deviate from past models, emphasizing the importance of using accurate statistical models for disaster planning. Long-tailed distributions like student t distributions reveal that large rare events happen more often than expected, reshaping approaches to assessing risks and preparing for extreme scenarios.
Exploring the Influence of Log-Normal Distributions on Perception
Log-normal distributions, characterized by exponential growth models, impact how our perceptions of outliers and exceptional achievers vary. These distributions, like the weight distribution curve, portray gradual shifts towards extreme boundaries, illustrating the essence of compound changes over time in defining societal norms and individual advancements.
Unveiling the Overton Window & the Influence of Public Opinion
The Overton window, rooted in assessing mainstream political ideas, highlights the range of acceptable opinions in society. The Overton paradox delves into age, period, and cohort effects on shifting attitudes, revealing how individuals and societal groups navigate evolving ideologies and political landscapes over time.
Analyzing Age, Period, & Cohort Effects in Public Opinion Evolution
Age effects, reflecting individual life stage influences, intertwine with period influences from shared societal events like shifting views on social issues. In tandem, cohort effects tied to birth cohorts shape collective attitudes over time, contributing to the nuanced interplay of changing public opinions and political dynamics.
Discussing Overton Window & Influence on Political Discourse
The Overton window concept elucidates the spectrum of socially acceptable political ideologies, guiding politicians in aligning their positions with public sentiments. By acknowledging age, period, and cohort effects in shaping attitudes and beliefs, the Overton paradox underscores the continual evolution of public opinion and its impact on political discourse.
Times Have Changed: From Liberal to Conservative Over Generations
Over time, political ideologies have shifted significantly. A study showed that a liberal from 1970 would align more with a moderate in 2000 and a conservative in 2023. This generational shift highlights how individuals may perceive themselves differently in varying political climates.
Moral Progress and the Moral Circle: Extending Rights and Respect
The podcast delves into the concept of moral progress and the expansion of the moral circle. Scholars such as Peter Singer and Steven Pinker have observed that societal moral values tend to broaden over time. Rights and respect have increased from narrow groups to encompass more diverse individuals, including animals. This positive evolution reflects a shift towards inclusivity and ethical considerations.
Understanding Statistical Fallacies: Base Rate Fallacy and Paradoxes
The discussion covers statistical fallacies like the base rate fallacy and paradoxes like the low birth weight paradox. The base rate fallacy exemplifies how different base rates can skew test interpretations. Additionally, paradoxes such as the low birth weight paradox challenge assumptions and highlight the complexities of analyzing data within distinct contexts.
Join us as Dr. Allen Downey, renowned author and professor, shares insights from his upcoming book 'Probably Overthinking It,' breaking down underused techniques like Survival Analysis, explaining common paradoxes, and discussing the dynamic Overton Window.
This episode is brought to you by the Zerve data science dev environment, by Modelbit, for deploying models in seconds, and by Grafbase, the unified data layer. Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.
In this episode you will learn: • Why interpreting data is not always easy [06:21] • What is Survival Analysis [15:32] • Preston's Paradox [22:09] • Are you Normal? [36:52] • How to better prepare for rare “Black Swan” events [42:48] • What is an Overton Window? [53:06] • What is the base rate fallacy? [1:23:31] • How to protect yourself from biased samples [1:33:39] • Simpson’s Paradox [1:42:43]
