Widespread Science Fraud in Covid19 Studies - Sampling Scandals

Numerous recent studies claiming high infection rates and low death rates – all make use of the same experiment manipulation technique, sampling bias.

Manipulated Studies

The covert manipulation power of sampling bias in mind, we notice that this manipulation tool has been used prolifically in recent Covid19 projections, by a range of scientists, for reasons unknown.

One interesting observation is that all of these studies appear to have vastly-overestimated the infection rate, leading to the possibility that the studies were intentionally manipulated in order to promote the belief that herd immunity has already been reached, facilitating a reopening of economies and a coverup of further deaths from the virus.

(1) Swedish Herd Immunity Study:

Study that was widely promoted by Sweden, as evidence that Sweden had reached herd immunity – retracted after it became apparent that the numbers in the study would require Sweden having a population many times larger than it does, just to match the lower-bound on the well-proven death rates from other countries. Sampling bias appears to be involved.

Exposing:

Prior press:

(2) Stanford University Study:

  • Stanford study, which had claimed a high infection rate – numerous methodological errors, including the wife of a scientist recruiting highly-biased experiment subjects via facebook:

Exposing:

Prior press:

New York City Random Study:

  • 28 April 2020 - NYC “random” sampling study, which finds a high infection rate, reported to not be particularly random:

Exposing:

Background: An Experiment Manipulation Primer

Sample size, according to popular belief, is the best tool for manipulating a study toward a particular outcome. Specifically, one should make sample size small, and with a little bit of luck or large number of experiments – the desired outcome can be achieved.

According to this popular belief, no other manipulation tool, other than outright fraud, could be easier to accomplish. If true, this would mean that it’s incredibly hard to manipulate studies which have non-trivial sample sizes – leading to the conclusion that large sample size, especially via multiple independent studies = highly likely to be credible.

A More Powerful Experiment Manipulation Tool

Fortunately for the evil scientist, this popular belief is not true. There is a far cheaper, powerful, and more covert tool with which to manipulate experiment outcomes – sampling bias.

Unlike other forms of results manipulation, where large amounts of adjustment produce a relatively small manipulation, e.g. substantially reducing your sampling size will only modestly increase your odds of a study luckily producing the outcome you want – In the case of sampling bias, a seemingly tiny nudge in manipulation of the sampling procedure can have a large outsized effect in the manipulation of the experiment outcome.

Sampling methodology manipulation has unlimited degrees of freedom, while sample size manipulation, pure luck, and other components of experiment design have far less effectiveness.

Illustrations of the power of sampling bias to mislead are numerous, and include:

What’s Next

Will this bad science be the justification for the removal of restrictions that triggers the second wave?

Are these studies being promoted by politicians and speculators who are attempting to delay the realization of the full scope of the disaster ahead?

If a think tank were to design protocols for delaying dissent, this would be how they’d write the script.