Science Replication: An Ideal in Crisis
Journal editors are uniting to confront a crisis of confidence: lack of reproducibility of science results.
According to naive conceptions of science, replication is a source of confidence in scientific results – or as Finagle put it, “Experiments should be reproducible; they should all fail the same way.” Unfortunately, what sounds good in theory is not always true in practice. That’s why Nature and Science both posted special editorials titled, “Journals unite for reproducibility.” After all, “public trust in science is at stake.”
Just how bad is this ‘replication crisis,’ as Jonathan W. Schooler calls it in Nature? “In disciplines such as medicine, psychology, genetics and biology,” he says, “researchers have been confronted with results that are not as robust as they originally seemed.” This is in addition to many sciences that concern historical events that are not reproducible even in principle, such as big bang cosmology, the origin of life, and universal common ancestry. Any science that relies on singularities or contingent events is not reproducible.
Nature’s editors state the ideal of replication:
Reproducibility, rigour, transparency and independent verification are cornerstones of the scientific method. Of course, just because a result is reproducible does not make it right, and just because it is not reproducible does not make it wrong. A transparent and rigorous approach, however, will almost always shine a light on issues of reproducibility. This light ensures that science moves forward, through independent verifications as well as the course corrections that come from refutations and the objective examination of the resulting data.
It’s interesting that they should speak of “course corrections that come from refutations and the objective examination” of data, when the journal routinely screens refutations and objective examinations of Darwinian evolution. Be that as it may, the editors reported that journal editors met at the AAAS in June to formulate Guidelines in Reporting Preclinical Research. It’s part of “quality control” that Science Magazine says leads to “best practices” in science; “the important point is that a large number of scientific journals are standing together in their conviction that reproducibility and transparency are important issues.” The public needs this assurance; “The hope is that that these guidelines will not be viewed as onerous, but as part of the quality control that justifies the public trust in science.”
You’ve just been Gruberized. You were lied to; your science teacher told you that science was self-correcting and reproducible, and that’s what made it so trustworthy against all other forms of knowledge generation. They didn’t want to reveal how inaccurate that claim is in the real world, because the American public is too stupid to understand it without a nuanced presentation. It was important to get the hype passed by the media, because the elites wanted their exalted reputation to be swallowed without debate. Now that they’ve been caught red-handed with their deceit, they are in damage repair mode. “Don’t worry,” they tell us too late; “you can still trust us.”
Replication is a simplistic ideal that is rarely followed in practice. Except in rare high-profile cases, like the cold fusion claim of the 1980s, scientists are simply too busy to try to replicate someone else’s work. And who is going to replicate the Higgs boson discovery without a CERN supercollider to use, and hundreds of researchers? Obviously, some of the most worldview-dependent claims about big bang cosmology and evolution’s “great transformations” cannot be replicated.
In the philosophy of science, only a few branches of science, like chemistry and mechanics, even lend themselves to replication. Some are replicable in principle, like medicine, but are often too plagued by complexity to obtain solid reproducible results in the lab. Psychology is hopelessly confused by the largely unpredictable actions of free moral agents. The worst sciences at reproducibility are the evolutionary historical sciences. How is one to replicate the emergence of a particular species, especially when there is no fossil record? In cosmology, how is one to replicate the big bang, or the multiverse? Different methods are needed, one argues; well, the same is true for non-science fields, like history or theology.
If we can just get past the false dogma that a “scientific method” exists, we should focus our attention less on “scientific reasoning” and more on “logical reasoning” about evidence. There is no one scientific method, but a series of disparate fields of inquiry, all grasping at the honor of the word “Science,” each using different methods, tools, and types of evidence. Their explanations also vary widely in quality and evidential support.
This entry reminds us of how important integrity is to science. Integrity is not made of particles, nor does it evolve. Without a reliable, changeless standard of honesty and love of the truth, science cannot exist. The Bible provides that standard. Secular materialism does not. When the leading journal editors wring their hands about the necessity of transparency, public trust and convictions, tell them to kick the Darwin habit and read the Ten Commandments.