Researchers outline bias in epidemic research and offer a new simulation tool to guide future work

Credit: Pixabay / CC0 Domain Public

The research team has uncovered a series of unknowns in epidemiological research, from clinical trials to data collection, and offers game theory to address them, in a new study. The project sheds new light on the pitfalls associated with technology upgrades and push-ups to combat global crises such as the COVID-19, with a view to future disasters.

“Even today, the sophisticated methods used by epidemiologists are lacking in design and implementation,” said Bud Mishra, a professor at New York University’s Courant Institute of Mathematics and lead author. , which appeared in the journal. Technology & Renewal. “In our work, we enlighten everyone, but surprisingly often unnoticed, the problems that lead to research methods — and the introduction of simulation tools that we think can improve the decision-making process.”

Even in an age when vaccines can be successfully developed in a few months, and tackling the complexities of the past in unimaginable ways in the past few decades, scientists can still have obstacles without their knowledge. because of errors in their methods.

In the paper, Mishra and her authors, Inavamsi Enaganti and Nivedita Ganesh, NYU undergraduates in computer science, discover some measure of balance, incomprehension, and incompetence in the context of thinking and demonstrating how they are suitable for work aimed at treating diseases. . These include Grue Paradox, Simpson’s Paradox, Self-Evidence, and more:

The Grue Paradox

The authors note that there are often obstacles to the analysis and errors associated with critical thinking, falling under what is known as the Grue Paradox. For example, if all emeralds observed during a given period are green, then all emeralds must be green. However, if we define “grue” as the property of being green for a period of time in time and then blue later, inductive evidence supports the conclusion that all emeralds are “real” and supports the assertion that all emeralds are green, preventing one. from achieving a definite decision on the color of emeralds.

“While constructing and comparing perceptions in a pandemic context, it is important to identify the temporal dependence of the predicate,” the authors write. These include perceptions of viral mutations, attracting herd shields, or frequency of infection.

Simpson’s Paradox

“Simpson’s Paradox is a phenomenon where the observations in the data when classified into different groups are reversed when combined,” the authors write. “This influence is abundant in academic literature and is well known to distort the truth.”

For example, if in a clinical trial 100 people took Treatment 1 and 100 Treatment 2 with a success rate of 40 percent and 37 percent, respectively, one would assume Treatment 1 was more effective. Of course, if you share this information with her genetic markers-ce, Genetic Marker A and Genetic Marker B – the quality of treatment can lead to different results. For example, Treatment 1 may be higher when you look at the population density, but its value may decrease for some smaller groups.

Proof of Bias

The most popular Bias Verification, or behavioral retrieval and retrieval data with great priority when supporting a researcher’s prediction, and harming pandemic research, the authors say.

“This phenomenon can already be seen in the context of COVID-19 in the data collection option to draw a picture that supports popular belief,” they wrote. “For example, evidence of the support of countries that are implementing stricter measures and public health measures is weighed against the evidence that countries that are relaxed with their measures are and the same reductions in factors. are of contextual nature and are different for different sectors, which may be neglected, such as population or prevention history. “

In addressing these challenges, the organization has created an episimmer platform that seeks to develop. support decision making to help answer user questions related to policies and restrictions during a pandemic.

Episimmer, which the researchers tested under several simulators public health emergencies, conducting “quarterly” analyzes, measuring what will happen to the environment if there are no interventions and policies, thus helping users identify and maximize their potential for COVID-19 strategy ( Note: platform python package. (Available on this page). These may include decisions such as “What are the days to be far away or in person” for schools and workplaces and “Which immunization program is most appropriate given the nature of home interactions?”

“Faced with a highly volatile virus, innovators must test, replicate, and push both innovative and innovative solutions while avoiding the pitfalls that plague society. clinical trials and related projects, ”Enaganti said.

In the wrong hand, immunization statistics can prove fatal. Simpson’s Paradox shows why

Learn more:
Inavamsi Enaganti et al, Creating Problems: Strategies for Data Retrieval in Diagnosis and Management, Technology & Renewal (2022). DOI: 10.21300 / 22.2.2021.12.… 22/00000002 / art00012

Its formation
New York University

hint: Researchers design bias in epidemiological research and provide new simulation tools to guide future work (2022, March 31) retrieved March 31, 2022 from -03-outline-bias-epidemic-simulation-tool.html

This document is copyrighted. Apart from any genuine transaction for the purpose of personal analysis or investigation, any part may not be reproduced without our written permission. Content is provided for informational purposes only.

Researchers outline bias in epidemic research and offer a new simulation tool to guide future work Source link Researchers outline bias in epidemic research and offer a new simulation tool to guide future work

Related Articles

Back to top button