Archive for hypothesis testing

abandon ship [value]!!!

Posted in Books, Statistics, University life with tags , , , , , , , , , on March 22, 2019 by xi'an

The Abandon Statistical Significance paper we wrote with Blakeley B. McShane, David Gal, Andrew Gelman, and Jennifer L. Tackett has now appeared in a special issue of The American Statistician, “Statistical Inference in the 21st Century: A World Beyond p < 0.05“.  A 400 page special issue with 43 papers available on-line and open-source! Food for thought likely to be discussed further here (and elsewhere). The paper and the ideas within have been discussed quite a lot on Andrew’s blog and I will not repeat them here, simply quoting from the conclusion of the paper

In this article, we have proposed to abandon statistical significance and offered recommendations for how this can be implemented in the scientific publication process as well as in statistical decision making more broadly. We reiterate that we have no desire to “ban” p-values or other purely statistical measures. Rather, we believe that such measures should not be thresholded and that, thresholded or not, they should not take priority over the currently subordinate factors.

Which also introduced in a comment by Valentin Amrhein, Sander Greenland, and Blake McShane published in Nature today (and supported by 800+ signatures). Again discussed on Andrew’s blog.

a Bayesian interpretation of FDRs?

Posted in Statistics with tags , , , , , , , , , , on April 12, 2018 by xi'an

This week, I happened to re-read John Storey’ 2003 “The positive discovery rate: a Bayesian interpretation and the q-value”, because I wanted to check a connection with our testing by mixture [still in limbo] paper. I however failed to find what I was looking for because I could not find any Bayesian flavour in the paper apart from an FRD expressed as a “posterior probability” of the null, in the sense that the setting was one of opposing two simple hypotheses. When there is an unknown parameter common to the multiple hypotheses being tested, a prior distribution on the parameter makes these multiple hypotheses connected. What makes the connection puzzling is the assumption that the observed statistics defining the significance region are independent (Theorem 1). And it seems to depend on the choice of the significance region, which should be induced by the Bayesian modelling, not the opposite. (This alternative explanation does not help either, maybe because it is on baseball… Or maybe because the sentence “If a player’s [posterior mean] is above .3, it’s more likely than not that their true average is as well” does not seem to appear naturally from a Bayesian formulation.) [Disclaimer: I am not hinting at anything wrong or objectionable in Storey’s paper, just being puzzled by the Bayesian tag!]

a null hypothesis with a 99% probability to be true…

Posted in Books, R, Statistics, University life with tags , , , , , , , , , , , on March 28, 2018 by xi'an

When checking the Python t distribution random generator, np.random.standard_t(), I came upon this manual page, which actually does not explain how the random generator works but spends instead the whole page to recall Gosset’s t test, illustrating its use on an energy intake of 11 women, but ending up misleading the readers by interpreting a .009 one-sided p-value as meaning “the null hypothesis [on the hypothesised mean] has a probability of about 99% of being true”! Actually, Python’s standard deviation estimator x.std() further returns by default a non-standard standard deviation, dividing by n rather than n-1…

admissible estimators that are not Bayes

Posted in Statistics with tags , , , , , , on December 30, 2017 by xi'an

A question that popped up on X validated made me search a little while for point estimators that are both admissible (under a certain loss function) and not generalised Bayes (under the same loss function), before asking Larry Brown, Jim Berger, or Ed George. The answer came through Larry’s book on exponential families, with the two examples attached. (Following our 1989 collaboration with Roger Farrell at Cornell U, I knew about the existence of testing procedures that were both admissible and not Bayes.) The most surprising feature is that the associated loss function is strictly convex as I would have thought that a less convex loss would have helped to find such counter-examples.

p-values and decision-making [reposted]

Posted in Books, Statistics, University life with tags , , , , , , , , , , on August 30, 2017 by xi'an

In a letter to Significance about a review of Robert Matthews’s book, Chancing it, Nicholas Longford recalls a few basic facts about p-values and decision-making earlier made by Dennis Lindley in Making Decisions. Here are some excerpts, worth repeating in the light of the 0.005 proposal:

“A statement of significance based on a p-value is a verdict that is oblivious to consequences. In my view, this disqualifies hypothesis testing, and p-values with it, from making rational decisions. Of course, the p-value could be supplemented by considerations of these consequences, although this is rarely done in a transparent manner. However, the two-step procedure of calculating the p-value and then incorporating the consequences is unlikely to match in its integrity the single-stage procedure in which we compare the expected losses associated with the two contemplated options.”

“At present, [Lindley’s] decision-theoretical approach is difficult to implement in practice. This is not because of any computational complexity or some problematic assumptions, but because of our collective reluctance to inquire about the consequences – about our clients’ priorities, remits and value judgements. Instead, we promote a culture of “objective” analysis, epitomised by the 5% threshold in significance testing. It corresponds to a particular balance of consequences, which may or may not mirror our clients’ perspective.”

“The p-value and statistical significance are at best half-baked products in the process of making decisions, and a distraction at worst, because the ultimate conclusion of a statistical analysis should be a proposal for what to do next in our clients’ or our own research, business, production or some other agenda. Let’s reflect and admit how frequently we abuse hypothesis testing by adopting (sometimes by stealth) the null hypothesis when we fail to reject it, and therefore do so without any evidence to support it. How frequently we report, or are party to reporting, the results of hypothesis tests selectively. The problem is not with our failing to adhere to the convoluted strictures of a popular method, but with the method itself. In the 1950s, it was a great statistical invention, and its popularisation later on a great scientific success. Alas, decades later, it is rather out of date, like the steam engine. It is poorly suited to the demands of modern science, business, and society in general, in which the budget and pocketbook are important factors.”

estimation versus testing [again!]

Posted in Books, Statistics, University life with tags , , , , , , , , , , on March 30, 2017 by xi'an

The following text is a review I wrote of the paper “Parameter estimation and Bayes factors”, written by J. Rouder, J. Haff, and J. Vandekerckhove. (As the journal to which it is submitted gave me the option to sign my review.)

The opposition between estimation and testing as a matter of prior modelling rather than inferential goals is quite unusual in the Bayesian literature. In particular, if one follows Bayesian decision theory as in Berger (1985) there is no such opposition, but rather the use of different loss functions for different inference purposes, while the Bayesian model remains single and unitarian.

Following Jeffreys (1939), it sounds more congenial to the Bayesian spirit to return the posterior probability of an hypothesis H⁰ as an answer to the question whether this hypothesis holds or does not hold. This however proves impossible when the “null” hypothesis H⁰ has prior mass equal to zero (or is not measurable under the prior). In such a case the mathematical answer is a probability of zero, which may not satisfy the experimenter who asked the question. More fundamentally, the said prior proves inadequate to answer the question and hence to incorporate the information contained in this very question. This is how Jeffreys (1939) justifies the move from the original (and deficient) prior to one that puts some weight on the null (hypothesis) space. It is often argued that the move is unnatural and that the null space does not make sense, but this only applies when believing very strongly in the model itself. When considering the issue from a modelling perspective, accepting the null H⁰ means using a new model to represent the model and hence testing becomes a model choice problem, namely whether or not one should use a complex or simplified model to represent the generation of the data. This is somehow the “unification” advanced in the current paper, albeit it does appear originally in Jeffreys (1939) [and then numerous others] rather than the relatively recent Mitchell & Beauchamp (1988). Who may have launched the spike & slab denomination.

I have trouble with the analogy drawn in the paper between the spike & slab estimate and the Stein effect. While the posterior mean derived from the spike & slab posterior is indeed a quantity drawn towards zero by the Dirac mass at zero, it is rarely the point in using a spike & slab prior, since this point estimate does not lead to a conclusion about the hypothesis: for one thing it is never exactly zero (if zero corresponds to the null). For another thing, the construction of the spike & slab prior is both artificial and dependent on the weights given to the spike and to the slab, respectively, to borrow expressions from the paper. This approach thus leads to model averaging rather than hypothesis testing or model choice and therefore fails to answer the (possibly absurd) question as to which model to choose. Or refuse to choose. But there are cases when a decision must be made, like continuing a clinical trial or putting a new product on the market. Or not.

In conclusion, the paper surprisingly bypasses the decision-making aspect of testing and hence ends up with a inconclusive setting, staying midstream between Bayes factors and credible intervals. And failing to provide a tool for decision making. The paper also fails to acknowledge the strong dependence of the Bayes factor on the tail behaviour of the prior(s), which cannot be [completely] corrected by a finite sample, hence its relativity and the unreasonableness of a fixed scale like Jeffreys’ (1939).

John Kruschke on Bayesian assessment of null values

Posted in Books, Kids, pictures, Statistics, University life with tags , , , , , , , , on February 28, 2017 by xi'an

John Kruschke pointed out to me a blog entry he wrote last December as a follow-up to my own entry on an earlier paper of his. Induced by an X validated entry. Just in case this sounds a wee bit too convoluted for unraveling the threads (!), the central notion there is to replace a point null hypothesis testing [of bad reputation, for many good reasons] with a check whether or not the null value stands within the 95% HPD region [modulo a buffer zone], which offers the pluses of avoiding a Dirac mass at the null value and a long-term impact of the prior tails on the decision, as well as the possibility of a no-decision, with the minuses of replacing the null with a tolerance region around the null and calibrating both the rejection level and the buffer zone. The December blog entry exposes this principle with graphical illustrations familiar to readers of Doing Bayesian Data Analysis.

As I do not want to fall into an infinite regress of mirror discussions, I will not proceed further than referring to my earlier post, which covers my reservations about the proposal. But interested readers may want to check the latest paper by Kruschke and Liddel on that perspective. (With the conclusion that “Bayesian estimation does everything the New Statistics desires, better”.) Available on PsyArXiv, an avatar of arXiv for psychology papers.