[HN Gopher] Bayesians moving from defense to offense
       ___________________________________________________________________
        
       Bayesians moving from defense to offense
        
       Author : Tomte
       Score  : 131 points
       Date   : 2023-12-23 14:48 UTC (8 hours ago)
        
 (HTM) web link (statmodeling.stat.columbia.edu)
 (TXT) w3m dump (statmodeling.stat.columbia.edu)
        
       | tomrod wrote:
       | Huh. Wouldn't publication bias dramatically impact a Bayesian's
       | conclusions?
        
         | ta988 wrote:
         | Depends how you model it. I would say not taking into account
         | the publication bias if you have ways to measure or at least
         | estimate its distribution is a bigger mistake to make.
        
         | ChadNauseam wrote:
         | Yes, it should affect anyone's conclusions unless they know
         | exactly how much publication bias there is and then account for
         | it
        
         | pacbard wrote:
         | Yes, possibly. Publication bias is a known problem when using
         | prior work (see [1] for some references on the problem and
         | potential solutions).
         | 
         | One of the strengths of Bayesian methods is that you can use a
         | less informative prior to model your skepticism of prior work.
         | 
         | [1]: https://www.sciencedirect.com/topics/medicine-and-
         | dentistry/....
        
         | epgui wrote:
         | Yes, but no more than a non-Bayesian's. The difference is
         | mainly that the Bayesian approach makes more of its assumptions
         | explicit.
        
           | jvans wrote:
           | This. Bayesian assumptions are explicit and you'll be forced
           | to defend your priors. Frequentist assumptions like
           | homoscedasticity are implicit and rarely challenged.
           | 
           | I wouldn't be surprised if many people using frequentist
           | methods have no idea what assumptions need to be true for the
           | methods to be valid. In a Bayesian framework you can't avoid
           | the assumptions
        
         | ajkjk wrote:
         | That's why it's for constructing a prior that you change later;
         | your study itself should update the prior in the direction of
         | being more correct. But it's not like you can do better than
         | "the best available prior" (by definition).
        
       | mjfl wrote:
       | I've never heard anyone promote frequentist statistics. I've only
       | ever heard Bayesians on the 'offense' talking about why their
       | framework is better. To be clear I am someone who doesn't really
       | understand the faultlines.
        
         | analog31 wrote:
         | Indeed, when I took statistics in college, nearly 40 years ago,
         | we proved Bayes theorem, and worked through several
         | applications, so it could hardly have been regarded as
         | controversial.
        
           | adastra22 wrote:
           | It's not. Statistics is statistics. This controversy is new
           | and contrived.
        
             | fbdab103 wrote:
             | Tell that to Ronald Fisher, who made it a point to deride
             | Bayesian stats wherever there was an opportunity.
        
             | jvans wrote:
             | The controversy is decades old and largely resolved. Both
             | methods work well on large enough data sets, Bayesian
             | methods are far superior on small datasets
        
               | adastra22 wrote:
               | This statement doesn't really make sense to me. A
               | frequentist approach assumes a random (or noisy)
               | underlying process and characterizes its probability. A
               | Bayesian approach assumes a fixed but unknown process and
               | updates the probability a given model matches that
               | reality.
               | 
               | It is ultimately the same math, if you are comparing
               | apples to apples. Just different ways of looking at a
               | problem. Sometimes one is better suited than the other.
               | 
               | It's like as if physicists were arguing over whether
               | Cartesian or Polar coordinates were better. It's the same
               | damn physics, just expressed differently. In some
               | problems one approach is easier to work with than the
               | other, and can even make seemingly intractable problems
               | solvable. But that doesn't mean the other approach was
               | "wrong."
        
           | tnecniv wrote:
           | Bayes theorem is fairly elementary probability. It isn't the
           | part that is questioned in Bayesian statistics. Someone
           | disagreeing with Bayes theorem would be like disagreeing with
           | addition.
           | 
           | The controversial part is the methodology for statistical
           | analysis built on top of it (e.g., you need a prior but where
           | did that come from?)
        
           | Tomte wrote:
           | Bayes theorem shares a name with Bayesianism, but it is not
           | specific to it. No frequentist would question Bayes theorem,
           | it is on the level of 1+1=2.
        
         | SpaceManNabs wrote:
         | > I've never heard anyone promote frequentist statistics
         | 
         | I will do it:
         | 
         | 1. Computationally easier
         | 
         | 2. often analytical theory available for most use cases so
         | interpretability is high
         | 
         | 3. more literature available so you can get unstuck faster if
         | you mess up
         | 
         | 4. no accusations of subjective bias in your prior (the con is
         | clear, no ability to leverage subjective expertise)
         | 
         | 5. In the asymptotic regime, MLE and bayesian MAP often
         | converge anyways
         | 
         | 6. king of hypothesis testing
         | 
         | For most people, it doesn't matter. It matters when you are
         | doing treatment for small sample sizes or other situations that
         | would cause low power.
        
           | bogtog wrote:
           | > king of hypothesis testing
           | 
           | This advantage can't be understated. Researchers, at least in
           | psychology, almost never seem to care about the magnitude of
           | an effect. It's simply enough to show that some effect
           | happens in some direction. For this (generally acceptable)
           | purpose, frequentist stats are great.
        
             | urschrei wrote:
             | A statistically significant result without something like a
             | Hedges G (or Cohens D if comparing means) measure would be
             | called out by even a moderately attentive reviewer as a
             | matter of course (especially if small sample sizes are
             | involved), and many reputable journals actually require
             | effect sizes to be stated, so you run the risk of being
             | desk rejected for leaving it out.
        
             | travisjungroth wrote:
             | Add some sampling bias, a significance threshold of 0.05.
             | Baby, you've got a replication crisis going!
        
         | zozbot234 wrote:
         | Frequentist statistics is largely based on the maximum
         | likelihood principle, which is a proper Bayesian analysis
         | whenever the prior is a flat distribution. And since inference
         | is mostly done on model _parameters_ in frequentist statistics,
         | the distribution you assume for them will also depend on how
         | you parameterize the model. So it can often be justified as a
         | convenient approximation to a  "properly" Bayesian result.
        
         | jdewerd wrote:
         | I've probably heard 1000 Bayesians rant about the alleged
         | Frequentist consensus and 0 Frequentists complain about
         | Bayesian analysis. I'd need some pretty steep priors to
         | interpret this as anything other than the academic equivalent
         | of "one WEIRD trick THEY don't want you to know" marketing.
        
           | paulsutter wrote:
           | If we were allowed one super-upvote per month, I would use
           | mine right now on your comment
        
             | selimthegrim wrote:
             | Dang is going to file this under the Bumble model of HN
             | monetization in his desk drawer
        
               | causality0 wrote:
               | There are times I'd be willing to pay a fiver to un-flag
               | my comment.
        
               | jowea wrote:
               | Or just HN Gold
        
           | throwawaymaths wrote:
           | There's always room for a first time.
           | 
           | Bayesian statistics is sound but I suspect it's often just
           | used to justify biases. It is technically valid to use a
           | prior and _de facto_ never update it, because you know I 'll
           | get around to updating my prior next week, or... eventually,
           | cough cough _let 's be honest, never_
        
             | marcosdumay wrote:
             | Bayesian statistics is the one where you must state your
             | bias explicitly, and justify them one by one under the
             | light.
             | 
             | It's the frequentist one that gets your biases implicitly,
             | on the form of corrections and hypothesis formulation, so
             | that people don't notice them.
        
           | exe34 wrote:
           | Every single statistics class I've ever taken or been sent to
           | has been 95% frequentist with a rushed Bayesian digression
           | near the end. I've submitted papers with Bayesian work and
           | the reviewer asked for p-values and would not budge. I gave
           | him his bloody p-values, because I could not afford not to
           | get the paper published that early in my career.
        
             | jdewerd wrote:
             | There's always a more complicated model.
        
               | zozbot234 wrote:
               | The "more complicated" version of frequentist statistics
               | is called robust statistics. There's most likely a way to
               | rephrase your favorite Bayesian analysis so as to make it
               | fully kosher from a "robust+frequentist" point of view,
               | even keeping the math unchanged. It just goes to show how
               | silly the "controversy" is.
        
               | PheonixPharts wrote:
               | Bayesian statistics is fundamentally _less_ complicated
               | than Frequentist statistics since everything can be
               | derived from a very simple set of first principles,
               | rather than complex frameworks of ad hoc testing
               | methodologies.
        
           | light_hue_1 wrote:
           | Yes, because Frequentists won a long time ago. And are
           | responsible for producing garbage science ever since.
           | 
           | Of course the winners don't rant about anything. But any time
           | we probe the consequences of Frequentist statistics they turn
           | out to be horrific for science, our health, and our planet.
        
         | travisjungroth wrote:
         | Frequentist is the default. People don't promote it. They just
         | teach it _as_ statistics, use p-values, and look funny at
         | anyone who does otherwise.
         | 
         | It's like how you wouldn't hear people in the US promote
         | imperial measurements for home baking. When you dominate, you
         | don't have to advocate.
        
           | analog31 wrote:
           | I'd identify a larger problem, which is teaching anything as
           | a bunch of recipes.
           | 
           | My college offered stats in two tracks: There was a one-
           | semester course for science majors, which was mostly plugging
           | numbers into formulas. The course was utterly baffling for
           | most of the students who took it.
           | 
           | There was also a two-semester course for math majors. It was
           | mainly about proofs, but also had time to go into more depth.
           | You have to know the assumptions underlying a formula, if
           | you're expected to prove it. ;-) But it was only taken by the
           | math majors.
        
           | mjfl wrote:
           | But I don't see p-values as connected to frequentist
           | statistics. Rather, p-values are a concept that applies to
           | the _problem_ of hypothesis testing, which both frequentist
           | and bayesian statistics should apply to, right? And in both
           | cases you would use a p-value. The unique thing that Bayesian
           | statistics can do is fit parameters can calculate uncertainty
           | in those parameters, which does not contradict any of the
           | concepts of hypothesis testing.
        
             | travisjungroth wrote:
             | A p-value is the frequency you'd expect to see a sample
             | with an effect this size or greater, given the null
             | hypothesis. It's inherently a frequentist approach to
             | inference.
             | 
             | You can squish it into Bayes by considering it a uniform
             | prior on the real number line that you never update, but
             | you're not really doing things "in the spirit" of Bayes,
             | then.
        
               | mjfl wrote:
               | What is a "prior" in the context of hypothesis testing?
               | That seems to me to be a concept used in parameter
               | estimation, with Bayes rule. But hypothesis testing is
               | not parameter estimation.
        
               | travisjungroth wrote:
               | As the linked post points out, you can regularize a
               | p-value based on prior experiments.
               | 
               | IMO, null hypothesis testing is way overused. We should
               | be quantifying the comparison of the null versus
               | alternative hypotheses. People already compare them.
               | Might as well bring some math into it.
        
               | mitthrowaway2 wrote:
               | How is hypothesis testing different from parameter
               | estimation? They seem conceptually the same to me, with
               | one a special case of the other.
        
               | PheonixPharts wrote:
               | In Bayesian terms there is no distinction between
               | hypothesis testing and parameter estimation.
               | 
               | If your hypothesis is that A is greater than B, then
               | you're test boils down to "The probability that A is
               | greater than B", which is arrived at through parameter
               | estimation via Bayes' Theorem.
               | 
               | You can consider the distribution of estimates for a
               | parameter to represent a space of possible hypothesis for
               | the true value of that parameter and their absolutely or
               | relative likelihood based on the information available.
               | 
               | The also has the pleasant consequence that hypothesis
               | testing using Bayesian methods nearly always is closer to
               | the actual question someone wants to answer rather than
               | replying with confusing statements such as "the
               | probability of rejecting the null hypothesis", which
               | contains an implicit double negative.
        
             | tingletech wrote:
             | I think with a Bayesian test one would use credible
             | intervals rather than p values?
        
             | PheonixPharts wrote:
             | > But I don't see p-values as connected to frequentist
             | statistics.
             | 
             | P-values are fundamentally Frequentist because this
             | framework see that _observed data_ as random, whereas
             | Bayesian statistics believe our observations are the only
             | thing that is actually known, and all the other parameters
             | are what is random in our experiments. That is: the data is
             | the only part of an experiment that is real, everything
             | else that we can 't directly observe is something we must
             | hypothesize about.
        
         | mapmeld wrote:
         | I went to a week-long Probabilistic AI seminar, and I got the
         | impression there was an era when enough stats and math
         | professors did not take a Bayesian framework seriously, that
         | there was a risk someone's thesis or dissertation defense would
         | be derailed by someone on faculty.
        
         | Avshalom wrote:
         | It's worth pointing out that the vast majority of "Bayesians"
         | you see talking shit have no actual training in bayesian
         | statistics, or any statistics training beyond the college-intro
         | level.
        
       | hackandthink wrote:
       | Link to the first quoted paper:
       | 
       | http://www.stat.columbia.edu/~gelman/research/published/pval...
        
       | pil0u wrote:
       | What resources would you recommend to implement an A/B test
       | evaluation framework using a Bayesian approach?
       | 
       | I would love to embrace (or at least try to) such a new approach,
       | but it feels like without a PhD in stats it's hard to get it
       | started.
        
         | jvans wrote:
         | Statistical rethinking by McElreath is an exceptional resource
         | for Bayesian modeling. It's very accessible and gives you some
         | good examples to work through.
         | 
         | Some very basic Bayesian models can go a long way towards
         | making informed decisions for a/b tests
        
         | plants wrote:
         | Specifically for A/B or A/B/N testing, you can use a beta-
         | bernoulli bandits, which give you confidence about which
         | experience is best and will converge to an optimal experience
         | faster than your standard hypothesis test. Challenges are that
         | you have to frequently recompute which experience is best and
         | thus, dynamically reallocate your traffic. They also only works
         | on a single metric, so if your overall evaluation criterion
         | isn't just something like "clickthrough rate", this type of
         | testing becomes more difficult (if anyone else knows how
         | multiple competing metrics are optimized with bandits, feel
         | free to chime in).
        
         | HuShifang wrote:
         | Here's a recent webinar from PyMC Labs' Max Kochurov (with some
         | slides and code too):
         | 
         | https://www.youtube.com/watch?v=QllfKQH-yQ4
        
         | fifilura wrote:
         | I am not the expert and humbled by all the experts in this
         | thread.
         | 
         | But I wanted to say that when I looked A/B testing few years
         | ago I started off with this book.
         | 
         | https://github.com/CamDavidsonPilon/Probabilistic-Programmin...
         | 
         | And somehere down the line he will introduce the Beta/Binomial
         | method for A/B testing.
         | 
         | In my (again humble) understanding the benefit of doing this in
         | the Bayesian way is both that you can actually get an
         | understandable answer, but also that the answer does not have
         | to be the final result you can continue by adding a loss
         | function for example.
        
         | esafak wrote:
         | It really is not; Bayesian statistics is way more intuitive.
         | For a practical guide see
         | https://dataorigami.net/Probabilistic-Programming-and-Bayesi...
         | 
         | For Bayesian ML see https://probml.github.io/pml-
         | book/book1.html and https://www.bishopbook.com/
         | 
         | For Bayesian statistics see
         | http://www.stat.columbia.edu/~gelman/book/
        
         | tmoertel wrote:
         | One straightforward approach is "Test & Roll: Profit-Maximizing
         | A/B Tests" https://arxiv.org/abs/1811.00457
        
       | clircle wrote:
       | I don't really see Frequentism and Bayes to be in conflict. You
       | use them to answer different questions. I wouldn't test a point
       | null hypothesis with a Bayesian model.
        
         | kneel wrote:
         | It honestly just seems like an abstract debate to claim that
         | your logical framework is somehow superior. Different tools for
         | different datasets
        
         | mitthrowaway2 wrote:
         | I think the Bayesians argue that you _would_ , so there does
         | seem to be a conflict.
        
       | zestyping wrote:
       | I don't know when Bayesians have ever been on defense. They've
       | always been on offense.
        
       | paulsutter wrote:
       | Bad reasoning trumps arcane methodology debates every time
        
       ___________________________________________________________________
       (page generated 2023-12-23 23:00 UTC)