Stanford's president is not the only researcher pushing unfounded and incorrect conclusions

Bad science gets rewarded

Everyone gets excited when a top researcher was found to oversee a trainee who photoshopped a western blot.

“Yes! we found an example of fraud”

“OMG, the culture in his laboratory was so horrible.”

“Yay, science is self-correcting!”

Some of these things were said recently, when the Stanford University president resigned to spend more time doing research (which, ironically, was the problem).

Yet, these same people routinely use false studies to push for conclusions they like. In fact, the majority of scientists conduct research that is false, not useful or not helpful. Only a tiny fraction of scientists stick to true and useful.

While fraud is the worst conduct— using false studies to support your views is hardly better. The Figure illustrates the spectrum of bad science from Fraud to Truth. Let me give some examples of people using bad science to further their agenda, and argue for strategies to solve the problem.

Just in the last two years, we have seen bad science repeated widely.

Black doctors have half the risk of death when they care for black babies vs white doctors. That is an incredible claim, and the underlying PNAS paper is completely flawed, as I explain in the link. Sadly the flawed study was cited in the dissenting opinion of the Supreme Court.

During the pandemic, public health ‘experts’ amplified flawed mask studies and, those who routinely promoted flawed studies or flawed interpretations were bizarrely promoted to public health dean. Learn why this study was flawed in the link.

Finally, remember that paradigm HF was not an appropriate trial of Entresto. The control arm was half maximal dose enalapril vs. full dose valsartan plus a new drug (not a fair comparison), and worse there is no evidence for actually doses used in American today, which are far below the trial doses.

In all these cases, and many more— cataloged on Sensible Medicine and my Substack— scientists use false, unsupported, misleading studies to advance an agenda— it can be socio-political or it can be to support a for-profit medical product or device. But it is always the same: the study proves nothing, and yet the ‘scientist’ falsely claims it is sufficient evidence to draw a conclusion.

We pat ourselves on the back when we punish one university president, but we do not reward or appreciate good science. We fund people who are good at selling themselves and often making false claims in their grant applications.

Public health experts who promoted the harmful and flawed policies of lockdown, masking toddlers, and boosting young men who already had COVID have been rewarded by the current administration and by public health schools.

Public health has proven itself to be a field, not governed by true and useful science, but one that is nakedly partisan. Finally, careers are made, and billion dollar drugs are supported by bad trials.

Bad science exists on a spectrum and there is little practical difference between photoshopping a Western blot or using a field of ‘physician on record’ in a Florida dataset to make a claim of mortality by race— if you do not fully understand how that data is coded and the limitations of your analysis.

In both cases, other scientists are relying upon a claim that may not be true. Whether the data is forged or the analysis is manipulated— it doesn’t change the central deception: this finding has no actual support.

What is the solution?

If you want to be a real scientist, you cannot be a partisan. No political party has the monopoly on true and useful. Inevitably a good scientist will anger people in both parties. Sadly, the reality is that currently most scientists lean left, and support unscientific and harmful policies of Democrats— such as masking toddlers or closing schools. Better to be a-political.

Second— if you want to be a real scientist, you cannot take money from a company that sells a product in the biomedical space. These payments influence your behavior, and these companies have a uni-directional profit motive. They make money from selling something, but not the converse. Its hard to take seriously an analysis of Entresto from someone paid by Novartis.

Third, people who promote scientists should incentivize people who do good work. True and useful should be the metrics, not flashy and sensational. To do this, grant funding needs to be given for reproducing work, and given to reproducible labs. An impartial group should device a clinical trials agenda to test assorted dogma.

Fraud is rare in science, but even if you detected every instance of it, most science would still be false. We have to address the root of the problem. Forcing a university president to resign is hardly evidence science, the institution, is self correcting.