Posted on 12/07/2021 11:33:16 AM PST by nickcarraway
Eight years ago, a team of researchers launched a project to carefully repeat early but influential lab experiments in cancer research.
They recreated 50 experiments, the type of preliminary research with mice and test tubes that sets the stage for new cancer drugs. The results reported Tuesday: About half the scientific claims didn’t hold up.
“The truth is we fool ourselves. Most of what we claim is novel or significant is no such thing,” said Dr. Vinay Prasad, a cancer doctor and researcher at the University of California, San Francisco, who was not involved in the project. It’s a pillar of science that the strongest findings come from experiments that can be repeated with similar results.
(Excerpt) Read more at apnews.com ...
Ya, we seem to have the same problem with these supposed Covid “Vaccines”....
Ya, we seem to have the same problem with these supposed Covid “Vaccines”....
Anyone who wants on a Replication Crisis ping list, please let me know.
Settled science.
Follow the science.
When it comes to careful experiments in laboratory settings, one group of scientists will get a set of results, but another group of scientists may have difficulty replicating those results. Because — brace yourself — science is hard.
Meanwhile, if you have any doubts about Global Warming, you’re just an idiot. Because we KNOW what the temperature is going to be in 100 years.
Likewise, if you think about Evolution and find yourself with doubts, you’re just an idiot. Because we KNOW what happened 200 million years ago. We have no hope of trying to replicate any of it, but we don’t need to. We KNOW what happened.
Science is fine, as far as it goes. But I see it as a religion, and you can have faith in science if you want. It’s not a bad belief system or worldview. But just don’t be too sure of yourself.
50% for cancer research is not that bad when you know that replication for behavioral studies is 0%.
https://elifesciences.org/articles/71601
What does a failure to replicate mean?
A single failure to replicate a finding does not render a verdict on its replicability or credibility. A failure to replicate could occur because the original finding was a false positive. Indeed, there is accumulating evidence of the deleterious impacts of low power and small sample sizes, ignoring null results, failures of transparency of methodology, failing to publish all experimental data collected, and questionable research practices such as p-hacking on the inflation of false positives in the published literature (Casadevall and Fang, 2012; Chalmers et al., 2014; Gelman and Loken, 2013; Greenwald, 1975; Ioannidis, 2005; John et al., 2012; Kaplan and Irvin, 2015; Landis et al., 2012; Macleod et al., 2014; Macleod et al., 2015; van der Naald et al., 2020; Rosenthal, 1979; Simmons et al., 2011). This evidence suggests that published findings might often be false positives or have exaggerated effect sizes, potentially adding noise and friction to the accumulation of knowledge.
A failure to replicate could also occur because the replication was a false negative. This can occur if the replication was underpowered or the design or execution was flawed. Such failures are uninteresting but important. Minimizing them requires attention to quality and rigor. We contracted independent laboratories with appropriate instrumentation and expertise to conduct the experiments. This has the virtue of ostensibly removing biasing influences of self-interest (pro or con) on the outcomes of the experiments. A skeptic, however, might suggest that original authors are essential for conducting the experiments because of particular skills or tacit knowledge they have for conducting the experiments. Indeed, this was part of a critique of the Reproducibility Project: Psychology (Gilbert et al., 2016; Open Science Collaboration, 2015). In that case, follow-up investigations did not support the claim that replication failures were due to deficiencies in replication quality. For example, Ebersole et al. repeated the potentially flawed replication protocols and developed revised protocols that were peer reviewed in advance by domain experts and original authors; they found that the replicability of original findings did not improve with the revised vs. original replication protocols (Ebersole et al., 2020; see also Anderson et al., 2016; Nosek and Gilbert, 2017). Nevertheless, the possibility of flaws in research is always present.
For the present replications, we attempted to minimize the likelihood of replication errors by using original materials whenever possible, employing large sample sizes, engaging expert peer review of the methodology in advance, and by preregistering the experiment design and analysis plan. Despite all this, we cannot rule out the possibility of methodological error in the replications. To facilitate further review, critique, and follow-up investigation, all replications in this meta-analysis are reported transparently with digital materials, data, and code openly available on OSF.
A failure to replicate could also occur even when both original and replication findings are ‘correct’. Experimental contexts inevitably differ by innumerable factors such as the samples used, reagent and instrument suppliers, climate, software version, time of year, and physical environment. An experiment is a replication if the many differences between original and new experimental context are theoretically presumed to be irrelevant for observing evidence for the finding (Nosek and Errington, 2020a). The replication experiments underwent peer review in advance to arrive at a precommitment that they were good faith tests based on present understanding of the phenomena and the conditions necessary to observe evidence supporting them (Nosek and Errington, 2020b). However, differences that were deemed inconsequential during a priori peer review may be more critical than presently understood. After the replication results were known, some reviewers and commentators offered hypotheses for why the findings might have differed from the original (Errington et al., 2021b). This generative hypothesizing can be productive if it spurs additional investigations to test the new hypotheses. It can also be counterproductive if it is just rationalizing to preserve confidence in the original findings. Hypothesizing ideas to test is easily conflated with providing evidence to explain. Without follow-up investigation to test the hypotheses, that mix-up can promote overconfidence in original findings.
Real Science Fact: laboratory mice cause cancer. Just kidding.
If eaten in large enough quantities.
This is also true in the so called social sciences.
Yum ... I like mind deep fried in fully hydrogenated oil to maximize trans fats, until crispy all the way thru. You can eat the bones that way.
bookmark
I am being treated for metastatic prostate cancer at MD Anderson. I can confirm the Lupron an Xtandi work really well.
When they stop working, there are many additional options plus others in the clinical trial phase. Hopefully, I am many more years before I need to make the next decision.
There has been at least one drug for advanced prostate for which I wonder if the Big Pharma company cooked the books. That drug would be Provenge. No one conclusively understands how it works since it doesn’t lower PSA and and the pharma company’s own research claims only a 4 month increase average life extention; from 44 to 48 months. That’s hardly evidence of success.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.