The so-called science claiming significant damage to the Great Barrier Reef has been hugely exaggerated, and in some instances, is just plain wrong. How could this happen? How could research conducted under the auspices of well-funded and important institutions become so unreliable without somebody pointing out that things were going awry? Are there structural weaknesses in the way that science institutions are organized? Are there assurance protocols to ensure the integrity and quality of their findings and conclusions?
Although not well known to the general public, the deep-seated problems in our science institutions have become a major topic of discussion within the scientific community during the past decade. The problem, far broader than work on the Great Barrier Reef, has been termed the “Replication Crisis.” It is the biggest thing in science in half a century. It is all bad news except that there is now a growing determination to fix the problem.
The Replication Crisis is the term that describes the fact that an exceptionally large part of the science literature, perhaps half of it, is seriously flawed. Professor John Ioannidis, a Stanford University mathematician who specializes in statistics of health and medical research, has a solid claim to being the first to properly recognize the Replication Crisis. Ioannidis published a now classic paper that has been cited by thousands of other scientists with the striking title “Why most published research findings are false.” In this paper, Ioannidis stated:
“There is increasing concern that in modern research, false findings may be the majority or even the vast majority of published research claims. However, this should not be surprising. It can be proven that most claimed research findings are false.”
This is a remarkable finding and is certainly not going unnoticed. It has been addressed by all the big-name science journals for more than a decade, and within some, but certainly not the majority, of science institutions. It is remarkable that the media, with their interest in the controversy, are largely unconscious of the Replication Crisis despite it being well-known among scientists. Or perhaps they know about it but don’t report on it since it would reveal most of their sensationalist climate change and pandemic coverage to be dangerous nonsense.
A fifty percent failure rate of reported science results would be better termed a replication scandal. Does any other profession have such a huge failure rate? Imagine if 50% of Elon Musk’s SpaceX rockets failed. You can be sure his company would never get another contract to launch payloads into space. Yet, even with a 50% failure rate for science papers, the media and politicians still treat published research as the gold standard of credibility to back policy decisions affecting the health and well-being of entire nations.
Replication is when a scientist tries to repeat the work of another scientist. It is fundamental to science. If research cannot be repeated by another scientist and produce equivalent results, there is a problem. When multiple attempts to replicate fail, the original work, the original findings, must be regarded as wrong. In contrast, if multiple attempts to replicate succeed, a solid basis for relying on the work has been laid. The work becomes genuine “science.”
Some fields of science are massively replicated and are utterly dependable. For example, the world relies upon Newton’s laws of motion and gravitation every day when people travel in a car, walk across a bridge, or fly in a plane. Einstein’s laws of relativity are tested every time the GPS on a phone is activated and when aircraft navigate the globe. These basic laws of physics are totally reliable within an uncertainty margin of a tiny fraction of a percent. Importantly for this well-replicated science, this uncertainty margin is accurately known so that the boundary between reliability and unreliability is established.
But most science is not massively replicated. Most science published in specialist journals and often reported in the press is not replicated at all and is thus not strictly science at all.
Ioannidis should be given the Nobel prizes for Physics, Chemistry, and Medicine because the implications of his work will ultimately clean up scientific research and make it dependable again. Compared with most Nobel prizes, Ioannidis’ work is of far greater importance because it focuses on quality assurance systems of all science.
Ioannidis highlighted the inadequate understanding of statistical methods that are often used in the health and medical sciences. Of these, possibly the best known is called P-hacking or data-dredging. With enough data and different parameters, it is almost always possible to find some correlation between two of the parameters. For example, to find relationships regarding human health, hundreds of different types of measurements, such as blood pressure, ethnic background, weight, hair color, length of the thigh, dozens of chemical tests of the blood, social data, and economic data can be used.
Almost inevitably, by pure luck, there will be some correlation between two of them. Yet, that proves nothing about causation. But climate and pro-vaccine activists often do this. We must call them on it every time!
Australia-based scientist Dr. Peter Ridd, who has published an entire book on this topic, will be our guest this weekend on both Saturday and Sunday at 8 pm on ET on The Other Side of the Story on America Out Loud Talk Radio.