The Spread of Evidence-Poor Medicine via Flawed Social-Network Analysis

@article{Lyons2011TheSO,
  title={The Spread of Evidence-Poor Medicine via Flawed Social-Network Analysis},
  author={Russell Lyons},
  journal={Statistics, Politics, and Policy},
  year={2011},
  volume={2}
}
  • R. Lyons
  • Published 16 July 2010
  • Economics
  • Statistics, Politics, and Policy
The chronic widespread misuse of statistics is usually inadvertent, not intentional. We find cautionary examples in a series of recent papers by Christakis and Fowler that advance statistical arguments for the transmission via social networks of various personal characteristics, including obesity, smoking cessation, happiness, and loneliness. Those papers also assert that such influence extends to three degrees of separation in social networks. We shall show that these conclusions do not follow… 

Figures and Tables from this paper

Correcting for the Dependence Structure in Social Networks
TLDR
This work finds that ignoring network dependence has devastating consequences for the validity of inference, and become more severe with increasing correlation: estimated coverage of 95% confidence intervals dropped as low as 33% when the sample exhibited high dependence.
Social contagion theory: examining dynamic social networks and human behavior
TLDR
The regularities that led us to propose that human social networks may exhibit a ‘three degrees of influence’ property are described and the statistical approaches used to characterize interpersonal influence with respect to phenomena as diverse as obesity, smoking, cooperation, and happiness are reviewed.
Network Dependence and Confounding by Network Structure Lead to Invalid Inference
Researchers across the health and social sciences generally assume that observations are independent, even while relying on convenience samples that draw subjects from one or a small number of
Sensitivity Analysis for Contagion Effects in Social Networks
TLDR
Sensitivity analyses suggest that contagion effects for obesity and smoking cessation are reasonably robust to possible latent homophily or environmental confounding; those for happiness and loneliness are somewhat less so; and the methodology that has been used in past studies for contagions in social networks, when used in conjunction with sensitivity analysis, may prove useful in establishing social influence.
Why and When "Flawed" Social Network Analyses Still Yield Valid Tests of no Contagion
TLDR
It is shown that if the alter's state is lagged by an additional period, then under the null of no contagion, the problems of model inconsistency and statistical dependence effectively disappear which allow for testing for contagion.
Challenges to Estimating Contagion Effects from Observational Data
TLDR
This chapter will focus on the challenges and the open problems and will not weigh in on the dilemma of researchers hoping to learn about contagion from observational social network data, except to mention here that the most responsible way to use any statistical method is with a healthy dose of skepticism.
Sensitivity analysis for network observations with applications to inferences of social influence effects
TLDR
This study proposed a simulation-based sensitivity analysis method that can evaluate the robustness of inferences made in social network analysis to six forms of selection mechanisms that can cause biases in network observations, and derived two sets of analytical solutions that can account for biases innetwork observations due to random, homophily, and anti-homophily selection.
Observational Social Network Studies Homophily and Contagion Are Generically Confounded in
The authors consider processes on social networks that can potentially involve three factors: homophily, or the formation of social ties due to matching individual traits; social contagion, also
Historical and Comparative Research on Social Diffusion: Mechanisms, Methods, and Data
Abstract Historical and comparative social scientists are increasingly interested in explaining the spread of innovations—which social scientists commonly refer to as diffusion and, broadly
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 169 REFERENCES
Commentary: Understanding social network analysis
TLDR
The authors cleverly use the Framingham heart study’s existing database that includes, fortuitously rather than by design, information that can be used for social network analysis, and broadly confirm the new and intriguing hypothesis that people's happiness is influenced by, among other things, the happiness of their acquaintances.
Social contagion theory: examining dynamic social networks and human behavior
TLDR
The regularities that led us to propose that human social networks may exhibit a ‘three degrees of influence’ property are described and the statistical approaches used to characterize interpersonal influence with respect to phenomena as diverse as obesity, smoking, cooperation, and happiness are reviewed.
The collective dynamics of smoking in a large social network.
TLDR
Despite the decrease in smoking in the overall population, the size of the clusters of smokers remained the same across time, suggesting that whole groups of people were quitting in concert.
Homophily and Contagion Are Generically Confounded in Observational Social Network Studies
TLDR
The authors demonstrate, with simple examples, that asymmetries in regression coefficients cannot identify causal effects and that very simple models of imitation can produce substantial correlations between an individual’s enduring traits and his or her choices, even when there is no intrinsic affinity between them.
The spread of obesity in a large social network over 32 years.
TLDR
Network phenomena appear to be relevant to the biologic and behavioral trait of obesity, and obesity appears to spread through social ties, which has implications for clinical and public health interventions.
Detecting implausible social network effects in acne, height, and headaches: longitudinal analysis
TLDR
Whether “network effects” can be detected for health outcomes that are unlikely to be subject to network phenomena is investigated and correlations in health outcomes of close friends to social network effects are attributed, especially when environmental confounders are not adequately controlled for.
Alone in the crowd: the structure and spread of loneliness in a large social network.
TLDR
Understanding of the broad social forces that drive loneliness is advanced and efforts to reduce loneliness in society may benefit by aggressively targeting the people in the periphery to help repair their social networks and to create a protective barrier against loneliness that can keep the whole network from unraveling.
...
1
2
3
4
5
...