Further Information

Meta-analysis Spurious precision? Meta-analysis of observational studies | Matthias Egger, Martin Schneider, George Davey Smith
Article Published: 1998

Type: Articles and Dissertations
Published By: BMJ 1998;316:140-144 (10 January)

Further Information

Meta-analysis can be generically defined as the study of a collective of studies on the same topic – a “study of the studies,” if you will. Those who are interested in epidemiological frauds against a particular product or lifestyle may at least be familiar with the word. Meta-analyses are often used in propaganda for the purpose of alarmism, or to confirm the impression that there is a “consensus” of studies or people that product or behaviour X is “hazardous.”

How good are meta-analyses, really? This article by Matthias Egger, reader in social medicine and epidemiology, Martin Schneider, research fellow, and George Davey Smith, professor of clinical epidemiology, educates us in what the propaganda of “public health” absolutely does not tell us; if it were honest and did, people would not take what it says seriously, thus would not modify their behaviour as desired. Although theoretically meta-analysis should be guided by strict rules, the rules are commonly ignored to achieve desired political results. Meta-analysis is mostly junk science that combines and amplifies the errors that plague the base studies in the first place.

FORCES is supported solely by the efforts of the readers. Please become a member or donate what you can.

Contact Info
Forces Contacts
Media Contacts
Evidence Categories
Quick Look-Up
List of Methodological Errors in ETS Studies
Hill's Criteria and Authoritative Citations
What Must an Epidemiologic Study Warrant?
Table of All Studies on ETS and Lung Cancer up to 2006

Pie Charts of ETS/Lung Cancer Studies
How many cigarettes must be smoked to create an ETS danger?

Passive Smoking: an Institutional Problem
A 13-minute video to understand the fraud

If you like to read rather then listen, download
Now available for free