Skip to main content

Lies, damned lies, and frequentist statistics

https://benlynn.blogspot.com/2013/11/lies-damned-lies-and-frequentist_12.html
Lies, damned lies, and frequentist statistics

Earlier this year I rekindled an interest in probability theory. In my classes, Bayes' theorem was little more than a footnote, and we drilled frequentist techniques. Browsing a few books led me to question this. In particular, though parts of Jaynes' "Probability Theory: The Logic of Science" sounded like a conspiracy theory at first, I was soon convinced that the author’s militant condemnation of frequentism was justified.

Today, I had the pleasure of reading a Nature article about a paper by Valen E. Johnson directly comparing Bayesian and frequentist methods in scientific publications, who suggests the latter is responsible for a plague of irreproducible findings. I felt vindicated; or rather, I felt I had several more decibels of evidence for the hypothesis that Bayesian methods produce far better results than frequentist methods when compared against the hypothesis that the two methods produce equivalent results!

This post explains it well. In short, frequentist methods have led to bad science.

An apologist might retort that it’s actually the fault of bad scientists, who are misusing the methods due to insufficient understanding of the theory. There may be some truth here, but I still argue that Bayesian probability should be taught instead. I need only look at my undergraduate probability and statistics textbook. On page 78, I see the 0.05 P-value convention castigated by Johnson, right after recipe-like instructions for computing a P-value. If other textbooks are similar, no wonder scientists are robotically misapplying frequentist procedures and generating garbage.

Johnson’s recommended fix of using 0.005 instead 0.05 is curious. I doubt it has firm theoretical grounding, but perhaps the nature of data that most scientists collect mean that this rule of thumb will usually work well enough. Though perhaps striving for the arbitrary 0.005 standard may require excessive data: a Bayesian method might yield similar results with less input. I guess it’s an expedient compromise. Those with poor understanding of statistical inference can still obtain decent results, at the cost of gathering more data than necessary.

The above post also mentions a paper describing how even a correctly applied frequentist technique leads to radically different inferences from a Bayesian one. The intriguing discussion within is beyond me, but I’m betting Bayesian is better; or rather, the prior I’d assign to the probability that Bayesian inference will one day shown to be better is extremly close to one!

 

Comments

Popular posts from this blog

Might Biden be a Liar & Predator like McCarrick?

September 15, 2020   Everyone knows that sexual predator ex-cardinal Theodore McCarrick is a liar. His whole life was a lie of betrayal of the most sacred vows he took and the violation of the moral tenets of the Catholic faith which he desecrated. Most people don't realize that part of this desecration of lies included lying for "gravely sinful" Democrats like Joe Biden. McCarrick protected Biden when then head of the Congregation for the Doctrine of the Faith Cardinal Joseph Ratzinger (later to be Pope Benedict XVI) wrote that bishops were not to admit to Communion politicians like "gravely sinful" Biden who supports the killing of unborn babies. McCarrick lied for politicians like Biden by ignoring the important parts of the Ratzinger letter and told bishops not to ignore the Catholic Church law.  Last year, Fr. Robert Morey denied Holy Communion to the “gravely sinful” Biden following a "2004 decree signed jointly by the bishops of ...

The Biben Lying Machine: "Joe , do you know what else is a Sin besides Killing Babies? Lying... "

October 09, 2020   It appears that Joe Biden was even a lying machine in 2008 according to the post " Media Ignores Biden Repeatedly Lies During 'Meet the Press' Interview" on the Weasel Zippers website: Joe Biden Repeatedly Lies During "Meet the Press" Interview, Claims he Doesn't Support Taxpayer Funded Abortions.....   Joe, do you know what else is a sin besides killing babies? Lying... ... Joe Biden repeatedly made the claim in a Sunday interview on the NBC political show "Meet the Press" that he opposes taxpayer funding of abortions. However, a look at his voting record over the years reveals numerous instances where Barack Obama's pro-abortion running mate did exactly that. "I don't support public, public funding. I don't, because that flips the burden. That's then telling me I have to accept a different view," he said on the program. As recently as February, Biden voted against an amendmen...

Vox Cantoris vs. Aqua

The Catholic Monitor commenter Aqua had this to say to the Vox Cantoris website: Aqua said… Fred, your topic here reminds me of a dust-up, a few days ago, on Vox Cantoris. He asserted that it is our duty as Christians to wear masks to the Holy Sacrifice of the Mass if the government tells us we must, or they will close our Churches. My response to him was that I find it inconceivable that an orthodox Catholic, such as himself, would ever submit to unjust dictates from secular government over how we approach Our Lord in Holy Mass. My response to him was that the Mass belongs to Catholics and we decide, within the bounds of Tradition, and in accord with the Word of Jesus, how we conduct ourselves in Holy Mass. Only one authority prevails over Mass and that is our God and the Sacred Tradition given by Him to guide us in all times and places. Understand, there is nothing inherently wrong with wearing a mask to Mass. But there is EVERYTHING wrong with wearing a symbol...