Top all-time · #2962
DC's Improbable Science
Personal blog · dcscience.netscience
DC's Improbable Science is a personal blog at dcscience.net about science. Since 2012, 26 different Hacker News readers have linked it 30 times, most often “Regression to the mean is the main reason ineffective treatments appear to work”. It is still writing: the last post was on 2025-02-11, and you can follow it by feed.
Too few sampled comments (17) to say how often HN recommends it.?
Why HN links it
HN commenters link DC's Improbable Science mainly for its posts on placebo effects and regression to the mean, and on statistics and p-values. They cite it as support for their points rather than praising the writing.
Summarised by Claude from a sample of 17 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Truth, falsehood and evidence: investigations of dubious and dishonest science
Most-linked posts?
- Regression to the mean is the main reason ineffective treatments appear to work6 people · 6 mentions · first linked 2015-12-14
- Publish and Perish at Imperial College London: The Death of Stefan Grimm3 people · 3 mentions · first linked 2014-12-02
- Placebo effects are weak (2015)2 people · 3 mentions · first linked 2023-12-13
- Do Certain Countries Produce Only Positive Results? Systematic Review of RCTs2 people · 2 mentions · first linked 2014-02-22
- Why empirical calibration is needed to correct p-values (2013) [pdf]2 people · 2 mentions · first linked 2015-04-07
What Hacker News says?
> Unfortunately the problem is that University's promote professors based on publications. Not just that. At some universities professors can lose their jobs if they don't get enough funding within a year. “I am of the opinion that you are struggling to fulfil the metrics of a Professorial post at Imperial College which include maintaining established funding in a programme of research with an attributable share of…
Comment linking /2014/12/23/some-experiences-of-life-at-imperial-college-london-an-external-inquiry-is-needed-after-the-death-of-stefan-grimm/
on “Sci-Hub Proves That Piracy Can Be Dangerously Useful”Submitted Publish and Perish at Imperial College London: The Death of Stefan Grimm as a story · 141 points
HN discussionWhat scientific studies would those be? And by what standards are they scientific? Until a randomized controlled study can demonstrate real differences between acupuncture conducted with real needles and sham needles, I will happily call it a placebo. Ditto for acupuncture conducted per the way that acupuncturists think it should be conducted and acupuncture conducted per the way that they think it should not. Those…
Comment linking /2013/05/30/acupuncture-is-a-theatrical-placebo-the-end-of-a-myth/
on “Online activists are silencing us, scientists say”Submitted Regression to the mean is the main reason ineffective treatments appear to work as a story · 137 points
HN discussionHere is a great paper that shows that more than 99% of Chinese clinical trials show a positive result. Either null results are being suppressed or data is being fabricated. http://www.dcscience.net/Vickers_1998_Controlled-Clinical-Tr... Good science certainly comes out of China, but a lot bad science comes along too.
Comment linking /Vickers_1998_Controlled-Clinical-Trials.pdf
on “NASA presents evidence that microwave thrusters seem to work”Submitted Why philosophy is largely ignored by science as a story · 90 points
HN discussionThat's actually a very interesting one, and has been written about at length by a number of medical stats people. Hopefully I do them justice to their arguments by saying that the benefits from supplements (if any) are tiny - as in a hazard ratio of death is pretty much 1 unless you are actually deficient. Equally however the potential downside of them is pretty much zero - expensive wee is what they can mostly be s…
Comment linking /2013/11/18/we-know-little-about-the-effect-of-diet-on-health-thats-why-so-much-is-written-about-it/
on “How pharmaceutical industry financial modelers think about rare diseases”Submitted On the hazards of significance testing: the false discovery rate as a story · 25 points
HN discussionThe placebo effect itself isn't real (at least in the vast majority of cases where it has been claimed to exist), when people measure a "placebo effect" what they are actually measuring is simply a regression toward the mean, not a causal effect. https://slatestarcodex.com/2018/01/31/powerless-placebos/ https://www.dcscience.net/2015/12/11/placebo-effects-are-wea... https://pubmed.ncbi.nlm.nih.gov/6369471/ https://w…
Comment linking /2015/12/11/placebo-effects-are-weak-regression-to-the-mean-is-the-main-reason-ineffective-treatments-appear-to-work/
on “The Delusion of the Polygraph”Submitted Why empirical calibration is needed to correct p-values (2013) [pdf] as a story · 17 points
HN discussionThen that is incorrect. The situation is simultaneously better and worse than your interpretation: > while a low P value indicates that your data are unlikely > assuming a true null, it can’t evaluate which of two > competing cases is more likely: > * The null is true but your sample was unusual. > * The null is false. You're conflating the two populations, and indeed we don't know which is which. But the P-value is…
Comment linking /Sellke-Bayarri-Berger-calibration-of-P-2001.pdf
on “The Irreproducibility Crisis of Modern Science”Submitted Are women still at a disadvantage in science? as a story · 4 points
HN discussionIf a randomized controlled trial includes just two groups, one which gets a treatment and another which gets placebo, it cannot attribute any results to the placebo, because it's missing a control for the placebo, i.e. a group which gets no intervention at all. Many studies have made this error, but that doesn't mean there's never a placebo effect. http://www.dcscience.net/2015/12/11/placebo-effects-are-weak...
Comment linking /2015/12/11/placebo-effects-are-weak-regression-to-the-mean-is-the-main-reason-ineffective-treatments-appear-to-work/
on “Perceived Physical Activity and Mortality: Study”Submitted Implausible results in human nutrition research [pdf] as a story · 2 points
HN discussion>If used widely, advocates say, substituting some of the drugs we take for placebos could save billions of dollars in healthcare costs. Drugs are not the 'cost' per pill. It's all in development, so this is a silly point. Cost will go up, or part of the population will get slammed. Placebo's are over rated. As such you have to be very sceptical of this article, it screams of alternative medicine. I wouldn't base muc…
Comment linking /2015/12/11/placebo-effects-are-weak-regression-to-the-mean-is-the-main-reason-ineffective-treatments-appear-to-work/
on “You can train your body into thinking it’s had medicine”Submitted Surrogate Science: The Idol of a Universal Method for Scientific Inference [pdf] as a story · 2 points
HN discussionAs far as I'm aware, they have, and once you take that into account placebo only has a very limited effect in controlling self-reported pain and nausea, everything else was just bad statistics. http://www.dcscience.net/2015/12/11/placebo-effects-are-weak... Though as that article explains, shysters, whether practicing some kind of standard woo, or working for big pharma, are keen to continue to pretend they don't un…
Comment linking /2015/12/11/placebo-effects-are-weak-regression-to-the-mean-is-the-main-reason-ineffective-treatments-appear-to-work/
on ““The effects of lifelong exposure to a perverse contingency” (2002)”Submitted Do Certain Countries Produce Only Positive Results? Systematic Review of RCTs as a story · 2 points
HN discussionI believe that most discussions about the placebo effect are muddled by whether regression-to-the-mean effects are included or not. [1] When one is comparing a treatment with a "placebo", it makes sense to include "regression-to-the-mean" effects, since you're obviously looking to do better than that with your treatment. However, when people ask "What are the biological causes of the placebo effect?", they should re…
Comment linking /2015/12/11/placebo-effects-are-weak-regression-to-the-mean-is-the-main-reason-ineffective-treatments-appear-to-work/
on “Reiki can’t possibly work, so why does it?”Submitted Placebo effects are weak (2015) as a story · 1 points
HN discussionI think this is largely a case of 'moloch' as described by Scott Alexander: http://slatestarcodex.com/2014/07/30/meditations-on-moloch/ In that long (but excellent essay) Scott points out that self organizing systems like human society can sometimes arrange themselves in horrible local minimas which can be very difficult to escape. I think the way modern science is organized is decidedly suboptimal. Note that I'm ta…
Comment linking /2014/12/01/publish-and-perish-at-imperial-college-london-the-death-of-stefan-grimm/
on “Scientific Regress”Submitted Publish and Perish: The Death of Stefan Grimm as a story · 1 points
HN discussion>I have no idea on how they do those clinical studies which say that there is no difference between a placebo and actual homeopathic medicines. So read up on them. Here's a good place to start: http://en.wikipedia.org/wiki/Homeopathy it is very well researched and you can review the papers yourself. Also: http://www.quackwatch.org/01QuackeryRelatedTopics/homeo.html http://skeptics.stackexchange.com/questions/6860/ha…
Comment linking /Smith-response.pdf
on “Homeopathic remedies recalled for containing real medicine”Submitted On the hazards of significance testing. Part 1: the screening problem as a story · 1 points
HN discussionSee the conclusion in https://www.dcscience.net/Hrobjartsson-Gotzsche-Cochrane-pla... They find no strong placebo effect, and when there is a placebo effect, it is difficult to distinguish statistically from no-treatment.
Comment linking /Hrobjartsson-Gotzsche-Cochrane-placebo.pdf
on “Uncomfortable truth: How close is "positivity culture" to delusion and denial?”Submitted Blogs lead in critical thinking, but newspapers still matter as a story · 1 points
HN discussionInteresting reading: 54% of findings with p < 0.05 not statistically significant: http://www.dcscience.net/Schuemie-Madigan-2012.pdf easy stats paper as to why: http://www.stats.org.uk/statistical-inference/Lenhard2006.pd...
Comment linking /Schuemie-Madigan-2012.pdf
on “Scientists Perturbed by Loss of Stat Tool to Sift Research Fudge from Fact”The placebo effect is to a large extent due to regression to the mean, i.e. a statistical glitch that has nothing to do with health outcomes. [1] http://www.dcscience.net/2015/12/11/placebo-effects-are-weak... [2] https://www.ncbi.nlm.nih.gov/pubmed/6369471
Comment linking /2015/12/11/placebo-effects-are-weak-regression-to-the-mean-is-the-main-reason-ineffective-treatments-appear-to-work/
on “Galapagos finches caught in act of becoming new species”I try not to stop reading too early. Drudge is not a journalist. Indeed, I have seen at least definition of "journalist" (a "reporter who writes for an editor") that was described in the source I saw it in as an explicit exclusion of Drudge from the ranks of journalists. A blog post, "Blogs lead in critical thinking, but newspapers still matter," http://www.dcscience.net/?p=6083 submitted to Hacker News earlier toda…
Comment linking /?p=6083
on “Why Aren't Top Journalists Rich? ”Also, I think the placebo effect is popularly misunderstood: https://www.dcscience.net/2015/12/11/placebo-effects-are-wea...
Comment linking /2015/12/11/placebo-effects-are-weak-regression-to-the-mean-is-the-main-reason-ineffective-treatments-appear-to-work/
on “Every homeopathic eye drop should be pulled off the market, FDA says”
