This paper by Andrew Gelman and Christian Hennig calls for the abandonment of the terms objective and subjective in (not solely Bayesian) statistics. And argue that there is more than mere prior information and data to the construction of a statistical analysis. The paper is articulated as the authors’ proposal, followed by four application examples, then a survey of the philosophy of science perspectives on objectivity and subjectivity in statistics and other sciences, next to a study of the subjective and objective aspects of the mainstream statistical streams, concluding with a discussion on the implementation of the proposed move. Continue reading
Archive for academic journals
Our paper with Diego Salmerón and Juan Cano using integral priors for binomial regression and objective Bayesian hypothesis testing (one of my topics of interest, see yesterday’s talk!) eventually appeared in Statistica Sinica. This is Volume 25, Number 3, of July 2015 and the table of contents shows an impressively diverse range of topics.
The Journal of the Royal Statistical Society, Series B, has a new cover, a new colour and a new co-editor. As can be seen from the above shots, the colour is now a greenish ochre, with a picture of pedestrians on a brick plaza as a background, not much related to statistical methodology as far as I can tell. More importantly, the new co-editor for the coming four years is Piotr Fryzlewicz, professor at the London School of Economics, who will share the burden with Ingrid van Keilegom professor from UCL (Louvain-la-Neuve) who is now starting her third year… My friend, colleague and successor as Series B editor Gareth Roberts is now retiring after four years of hard work towards making Series B one of the top journals in Statistics. Thanks Gareth and best wishes to Ingrid and Piotr!
Another email linked to the meta-spamming of pseudo-academic publications:
I feel pleasure to introduce our publications entitled “Oriental Journal of Computer Science and Technology” which is being published since last 6th years, under the banner of “Oriental Scientific Publishing Company”, Bhopal, India. T For maintaining the high standard of the journal, we always remain in search of researchers of repute of their respective fields, and with the same purpose I am inviting you to grace the Editorial Cum Advisory Board of our journal. You are supposed to review three-four articles per year and as per your will and convenience. We are eager to receive your acceptance and valuable suggestions for the improvement of the journal. I am pleased to announce that our reputed Editorial Cum Advisory Board member will enjoy concession rebate on the publication charges of their manuscripts.
It is amazing that they did not even check I am not in computer sciences! (Just considering the Editorial Board…)
*Springer* on a new *peer-reviewed, open access journal*
whose sole aim seems to generate more revenue for Springer. Indeed, papers published in this journal are charged $1025 each. Which is about the cost for a single subscription to the overpriced if scientifically excellent Statistics and Computing. (It takes a serious effort to discover the subscription rate of a Springer journal on their website!)
Indeed, I am quite surprised at a journal focussing on statistical distributions. What is a statistical distribution, exactly? The era when one would discover a new probability distribution in connection with a statistical estimation or testing problem and call it t, F, or Beta, seems long long gone! Just as gone as the production of statistical tables. (This is also why I wrote such a negative review of The Handbook of Fitting Distributions.) The webpage of the journal indicates that
The scopes include, but are not limited to, development and study of statistical distributions, frequentist and Bayesian statistical inference including goodness-of-fit tests, statistical modeling, computational/simulation methods, and data analysis related to statistical distributions. Significant and well-written articles on theory and methods in areas of statistical distributions and their applications will be considered for publication.
but this sounds so broad as to cover almost any statistical paper. So I am wondering at the purpose of this journal, except as an experimentation in “open access” commercial journals that are fully supported by the authors, in essence making grants pay twice for research.
I just arXived a survey entitled Bayesian computational tools in connection with a chapter the editors of the Annual Review of Statistics and Its Application asked me to write. (A puzzling title, I would have used Applications, not Application. Puzzling journal too: endowed with a prestigious editorial board, I wonder at the long-term perspectives of the review, once “all” topics have been addressed. At least, the “non-profit” aspect is respected: $100 for personal subscriptions and $250 for libraries, plus a one-year complimentary online access to volume 1.) Nothing terribly novel in my review, which illustrates some computational tool in some Bayesian settings, missing five or six pages to cover particle filters and sequential Monte Carlo. I however had fun with a double-exponential (or Laplace) example. This distribution indeed allows for a closed-form posterior distribution on the location parameter under a normal prior, which can be expressed as a mixture of truncated normal distributions. A mixture of (n+1) normal distributions for a sample of size n. We actually noticed this fact (which may already be well-known) when looking at our leading example in the consistent ABC choice paper, but it vanished from the appendix in the later versions. As detailed in the previous post, I also fought programming issues induced by this mixture, due to round-up errors in the most extreme components, until all approaches provided similar answers.
Following a discussion within the IMS publication committee and the coincidental publication of a central double page in Le Monde, weekend science&techno section [not that it was particularly informative!], here are some thoughts of mine on open access and publications:
First, the EU is philosophically inclined toward Open Access and has been putting some money into the game towards that goal:
As of 2014, all articles produced with funding from Horizon 2020 will have to be accessible: articles will
either immediately be made accessible online by the publisher (‘Gold’ open access) – up-front publication costs can be eligible for reimbursement by the European Commission;
or researchers will make their articles available through an open access repository no later than six months (12 months for articles in the fields of social sciences and humanities) after publication (‘Green’ open access).
This means that putting IMS publications on arXiv or on HAL (which is compulsory for CNRS and AERES evaluations, hence for most French public researchers, contrary to what Le Monde states) is fine and sufficient for EU funded research. It seems to be the same in other countries (ok, EU is not yet a country!) like Australia…
My personnal position on the issue is that I do not understand the ‘gold’ open access perspective. Since tax-payers are supporting public-funded research, why should they support the journals that publish this research if it is available on a public depository like arXiv for free? Simply because the publication in the journals gives a validation of the scientific contents? The argument was that it would save money on public libraries subscribing to expensive journals like Elsevier‘s, but paying the ‘gold’ open access is another way of redirecting tax-payers money towards publishers’ pockets, so this sounds like a loophole… I would thus be very much in favour of keeping the arXiv solution as is, since it is the greenest one, as long as we comply with local national regulations.