Science By Press Release: where I find myself in agreement with Dr. Gavin Schmidt over PR entropy

Over at RealClimate, Dr. Schmidt has written something that I’ve found myself wholly in agreement with – the sad state of science reporting and the use of embargoes that aren’t adhered to by some journalists.

Readers may recall that I’ve made similar complaints for sometime about the sloppy press releases that are often issued from university and organization science departments that don’t even include the title of the paper or a citation, forcing you to go looking for it, sometimes unsuccessfully. Dr. Schmidt writes:

Finally, while I am sort of on the topic of science journalism, can’t we please, please, please mandate the citation of the DOI tag in any online story about a new paper? Tracking down the actual paper being described is all-too-often far more difficult than it needs to be. (For example, the doi for the study referred to above took four links and a targeted google search to find. Why?).

I fully agree.

Here, in my opinion as 30 year TV/radio/web media reporter on science is what should be in any professionally produced science press release:

  • The name of the paper/project being referenced
  • The name of the journal it is published in (if applicable)
  • The name of the author(s) or principal researcher(s)
  • Contact information for the author(s) or principal researcher(s)
  • Contact information for the press release writer/agent
  • The digital object identifer (DOI) (if one exists)
  • The name of the sponsoring organization (if any)
  • The source of the funding for the paper/project
  • If possible, at the minimum, one or two full sized (640×480 or larger) graphics/images from the paper/project that illustrate the investigation and/or results.

Yet, if you go on the world’s leading science press release aggregation service,  Eurekalert, right now and examine the press releases there, you’ll find few if any that have all these features.

In their help section, Eurekalert talks about embargoes, but doesn’t offer any guidelines on what should be included in a science press release, leaving that up to individual PIO’s and press agents. Hence, we get more entropy in the process.

Even worse, for example, ESA’s Hubble group offers a Press release guidelines for scientists and with the exception of photos/videos doesn’t cover the basic points I list above but instead focus on style.

In a post on CompassBlogs story “Press Release Peeves” the topic is:

Journalists Comment on Science Press Releases – In a recent post, Liz asked science writers to share their pet peeves about press releases and advice for scientists on how to improve them.

WHAT IS THE ONE THING SCIENTISTS COULD DO TO MOST IMPROVE RELEASES ABOUT THEIR WORK? Clear callouts to usable art, multiple phone numbers and other contact info for the relevant researchers, inclusion of titles or links to drafts of the papers, particularly if they’re not being published in Science/Nature/PNAS.

That’s pretty much my entire list, independently derived.

And that’s only the tip of the iceberg. In a January article From the Writer’s Desk: The Dangers of Press Releases By Charles Q. Choi he writes:

In cases where the scientists are not contacted about their research, we have “churnalism” — news released based largely if not totally on press release alone. We also have press-release farms such as PhysOrg and ScienceDaily that seem to me to do little else but repackage press releases one can find on science press releases sites such as EurekAlert.

The discovery that Maggie and I made was not that churnalism happens. There are a fair number of opinions regarding press releases among science journalists — reporters are free at some outlets to use material from them while they are prohibited from doing so at others. In many ways, press releases are kind of like the dark matter of the science news universe — invisible to the public for the most part, but they exert a tremendous force on science journalism.

The problem, as we talked with scientists, was that apparently researchers don’t often get to vet press releases before they are published. That profoundly shocked and disturbed me and Maggie. Journalists have a number of ethical considerations with whether or not they let sources read articles before publication, but the press officers who write the press releases should have no such restraints. But what I heard that night corroborated an experience of mine in the last week, where I talked with a scientist for a story who told me he didn’t see the press release on his work.


I think a revolution in science PR is needed, and some standards should be drafted and applied. It is mind blowing that in the exacting world of science, the basic informational issues I’ve outlined are not covered by some sort of standard for submission. Most journals have strict standards for submission on the input side, so why is the output side left to the whims of the writer who may not be familiar with the material, leaving gaps in the basic info (like the name of the paper), or  even share the PR with the authors before setting it loose on the world?

Having submission standards for press releases would in fact reduce the churnalism and make for better science reporting overall, no matter who writes about it. There’s no point at all in holding back that basic information about the paper/study which seems to be regularly missing, and all it does is force reporters who are deadline oriented to waste more time tracking that information down, which could be better spent on writing about the science itself.

Since the American Association for the Advancement of Science (AAAS) runs the world’s largest science press release outlet, Eurekalert, the drafting of some standards for press release crafting and submission would go a long way towards solving these issues.

If you are a member of AAAS, or a PIO/PR agent reading this, I urge you to contact AAAS and/or Eurekalert and ask them to look at this essay and to consider implementing some submission standards at Eurekalert. Once that is done, the rest of the science world will follow that example.

Thank you for your consideration. – Anthony Watts

* top artwork adapted from original art at Bradley University


Related: (h/t to Tom Nelson)

Iranian mathematicians latest to have papers retracted for fake email addresses to get better reviews « Retraction Watch

It’s tempting to start calling this a trend.

Three Elsevier math journals are among the latest scientific publications to be retracting papers because fake email addresses were used to obtain favorable peer reviews.

0 0 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
September 24, 2012 8:48 am

And right on queue, an overly wordy missive from Mann on the Huffington Post:
“FiveThirtyEight: The Number of Things Nate Silver Gets Wrong About Climate Change”

September 24, 2012 8:53 am

For mining stocks, press releases have to be approved by an independent “qualified person” i.e. an independent consulting geologist. Typically, press releases are signed off on by directors.
Coauthors of a science article should not only be permitted to review the press release, but REQUIRED to sign off on it.

Louis Hooffstetter
September 24, 2012 9:05 am

It’s nice to find something that we can agree on with Dr. Schmidt. I just wish he would express similar frustration with the way authors archive (or refuse to archive) data, code, and algorithms for the sake of reproducibility. It should be just as easy to check an author’s work as it is to read his paper.

September 24, 2012 9:10 am

Completely agree with your own proposals here Anthony. I’d actually go further –
– Following my own research into media churnalism, I’d also like to see a requirement that journalists MUST explicitly cite any press releases used in their copy; the problem of PR churnalism is HUGE and climate science is one of the worst areas for this.

Bill Parsons
September 24, 2012 9:20 am

RE: “The name of the author(s) or principal researcher(s)”
I see there are 348 authors for the recent article on break-through Breast Cancer research, released in the Journal Nature. Is the convention usually the first-named scientist?
The practice of naming lots of people seems to be typical of climate research as well.

September 24, 2012 9:45 am

re: researchers vetting press releases
Well that seems highly useful, but also not a panacea, as evidenced in the notorious “11 degrees” episode with a PR signed off on by Oxford’s Myles Allen:
notorious “11 degrees” press release furor
Still, improvements to the preparation of PRs are welcome. As in all human endeavors there is no perfection, but people must try to end the worst practices and continue to improve in all areas.

September 24, 2012 9:59 am

Amen. This wouldn’t have helped 10 years ago, but now the DOI system is well established and will usually lead to at least an abstract. Beyond that, it’s often worth paying for a paper that seems interesting or important.

September 24, 2012 10:11 am

“Science PR” is an oxymoron of sorts. Science, at times, can be very unpopular, as we have seen through time immemorial. PR is “spin”. Kepler, Galileo, Newton….didn’t have any PR to slide them into the good graces of those who nowadays would write blasphemy laws (and did back then). PR has spawn’d an entire industry of McKibbens and Gleicks whose machinations have resulted in a sick and twisted confusorium full of deceit and subterfuge. Science has crumbled to dust in the onslaught as a new generation of sycophantic students rises.
True, A science press release should be just as you outlined, Anthony, but that will desiccate it’s sensational bent. The DOI is DOA….because the press release is not for scientists, it is for agendaists.

September 24, 2012 10:15 am

With more than 4-5 authors they need to adopt a convention.
Usually the 1st author is important but depending on the field, the last author is often the mentor in whose lab the research took place (and the one who got the grant to fund it).
But the Press release from the institution, with the main authors signing off can make sure the press release gets the key details right on which names should be mentioned.
I’m sure I would agree with Gavin Schmidt on many, many things, especially in climate research. Just not the degree of certainty and alarmism in some aspects. I don’t remember Gavin as being one of the worst in this regard anyway.

September 24, 2012 10:44 am

Here’s a really bad science press release.
In the Huffington Post?

September 24, 2012 10:47 am

That’s all great for the journalists. As a reader, the problem is that newspapers and other web-based media refuse to put links to the paper in or with articles. God forbid you should leave their page, and maybe forget to come back!

September 24, 2012 10:57 am

All of the stories I’ve read about this problem, including the ones from the worlds leading science journalism departments, indicate a strong correlation between the toil and ennui of conducting thorough reporting and the absense of meaningful detail in science reporting. The most recent work that drew my attention was penned by a Pulitzer prize winner who claimed that it was often disadvantagious to cite sources by author or publication because it fouled the flow of the article, encouraged readers to seek information from competing media outlets, and besides the casual reader was not interested in much beyond the headline anyway. In his excellent paper he also shared that many of the authors shied from reviewing press releases on their own work, especially in cases where the journalist erred in a favorable policy direction – getting the benefit of honey without being responsible for the angry bees.

September 24, 2012 10:58 am

I’m with you on the first three. If all are available they should be used. In the absence of that or by preference, then the DOI. The rest are entertaining noise I think. I can understand the argument about possibly illustrating biases, should the author’s be in the ‘results for pay’ crowd, but it hardly seems to rise to the level of critical information necessary to identify the subject that is being written about.

September 24, 2012 11:01 am

Other studies indicate that reporting standards have increased 22.6 percent over the last decade, and an astounding 400 percent since 1917. A top advisor to the president said “It’s not just science articles. Political reporting has improved as well”.

September 24, 2012 11:03 am

I posted a comment on one of Mann’s points in the Huffington Post article on Nate Silver. Who knows if the moderator there will allow it or not?
Mann links to a May 2007 Real Climate article by its blogger, who works for Hansen, in support of Hansen’s 1988 projections. In that year, land & sea surface temperatures were still close to the lowest projection, which assumed no additional CO2 after 2000. Since then, the temperature data have continued to diverge from the two higher scenarios & now remain under the lowest, although CO2 concentrations have continued rising. The situation is even worse when comparing atmospheric measurements, instead of the land & sea stations, which are cooked books, not only by Hansen’s “adjusted” GISS data, but by NOAA & the British HadCRU number oven, too.

September 24, 2012 11:40 am

On the subject of embargoes and science papers, Revkin had a doozy in a GMO paper that embargo’d fact checking !!

george e smith
September 24, 2012 12:03 pm

“””””…..Bill Parsons says:
September 24, 2012 at 9:20 am
RE: “The name of the author(s) or principal researcher(s)”
I see there are 348 authors for the recent article on break-through Breast Cancer research, released in the Journal Nature. Is the convention usually the first-named scientist?
The practice of naming lots of people seems to be typical of climate research as well…..”””””
Well it would be nice if in the actual text of the paper, when concepts, or data, or other “key” statements are made; they have an attached reference number; exactly as other papers are referred to, and then at the end of the paper in the “bibliography” section, they list those numbers and the name of the specific named author, who was responsible for that piece of information.
I know that in the patent literature, it is very common (too common) for “supervisors” or “managers” of inventors, to add their name to the list of named inventors, in the patent filing.
In one of my patents (now expired) filed while consulting, workers at my client company, added totally unrelated ideas related to practicalities of packaging and manufacture, to the text of the patent, and of course added their names (and the manager’s) to the list of named inventors. By the time the patent examiner got through with the patent, he had stripped out everything of that nature, as having nothing whatsoever to do with the inventive concept, which was entirely my own work, so the patent as issued contains nothing besides my original work. The Company lawyers failed to remove the names of the others who added stuff the examiner removed, so the patent issued with my name down an alphabetical list of about 8 “inventors”. If it had ever been contested in court, the patent would have been thrown out for fraudulent authorship.
Any person named as an inventor in a patent, must be able to point to at least one claimed required element in at least one claim of the patent, that they alone were responsible for unassisted, by anyone else; otherwise they are not an “inventor”.
Easiest way to kill a patent, is to show false authorship; and if there are more than three named inventors, it’s a fairly good bet, that they are not all inventors.
A peer reviewed paper with 348 named authors, sounds like total rubbish to me. I can’t imagine 348 different individuals having collaborated on a research paper. The paper would be unreadable, if it in fact contained original individual input or data from 348 different individuals.
If readers of a peer reviewed paper, can’t figure out which named author contributed what information; they shouldn’t be named as authors.

September 24, 2012 12:33 pm

Your proposals would be a disaster. Implementing them would weaken the value that these Press Release can add by spinning the science so that it properly maintains the narrative.

September 24, 2012 12:37 pm

Laughing stock! The sad thing is they are dragging down some very intelligent people along with ’em. It’s the lies, if your source of science is from the media then you’re a bystander who’s walked away from an accident, meaning that you know and are a witness of the events but you left the scene of the crime. My regards to all those scientists who are building with solid bricks of facts, logic and the scientific way.

September 24, 2012 1:00 pm

What does DOI mean? in English?

Paul Coppin
September 24, 2012 1:17 pm

I have a better idea. Make the journal authors responsible for producing their own press release as part of the publishing process. It’ll take a little bit of training in style differences and to see the releant hot points get covered, but then there should be no quibbles about content. If they can write an abstract, they can write a press release (hint, hint…). I’d suggest that the press release be made by the senior author, but then he have to actually read the paper his students produced…

September 24, 2012 1:42 pm

As far as Gavin goes, I agree with him. Great. But I need to add:
(a) concessional recognition that blogs importantly represent the public’s attempts to re-establish balanced reporting, where they have perceived imbalance. Of course there will be as many notions of “balanced” as there are members of the public, nevertheless, blogs should be a first check for balance.
(b) more evidence of true scientific attitude and free, intelligent, imaginative questioning of status-quo science, in science reporters. The success here shows that these qualities ARE valued by both ordinary and scientifically-trained readers.

and I don’t think he would like either of these, somehow.

September 24, 2012 1:42 pm

I can’t tell you the number of times I’ve gone off to Google Scholar to try to track down a paper. I use clues like the name of a scientist mentioned though it is not always clear he / she was involved in the research. Many times the title of the paper is not mentioned. No DOI link etc. At the very minimum give me the title and journal publication and it would make searching easier and faster.

R Barker
September 24, 2012 1:55 pm

Including a link or word-for-word copy of the reformed press release with any news story that gets published would help counter the usual hype and spin put into the story….but I guess that would be too much to ask. It surely would help those who rely soley on the mainstream media for their science information to develop a more balanced viewpoint. Some might even develop a healthy skepticism.
It is hard for me to place all the blame on the journalists. They are doing what comes naturally and indeed are under pressure to get the most attention and readership for their publication. I do place a lot of blame on people who just believe whatever they are told without question. You don’t have to be a “scientist” to ask a question.

John M
September 24, 2012 4:44 pm

A couple of comments…
TFNJ says:
September 24, 2012 at 1:00 pm

What does DOI mean? in English?
It’s a unique identifier that can be used to specifically identify a publication, regardless of the publisher. Basically a Social Security Number for published articles. If done right, there should at least be a link to a free abstract, with complete reference, authors and affiliations.
Also, although it’s alluded to throughout the post and comments, worth pointing out specifically that most press releases are directly from the researchers’ own institution.
Rembember this masterpiece from MIT?
The researchers themselves should view their role as the primary enforcers of quality control. It is their “brand”. When their institution issues press releases that look like they were written by Grade School science students, it’s the resercher’s reputations that suffer.
And finally, there’s an additional issue with science by Press Release. Often, a researcher gets a “free shot” at gilding the lily in a PR blurb, and makes comments that go way beyond what would have passed peer review (such as it is). These comments then become assoicated by some with an agenda to push with the actual scientific publication, and take on the aura (or stench in some cases) of “Peer Reviewed Science”.

September 24, 2012 10:56 pm

I agree with this
Steve McIntyre says:
September 24, 2012 at 8:53 am
For mining stocks, press releases have to be approved by an independent “qualified person” i.e. an independent consulting geologist. Typically, press releases are signed off on by directors.
Coauthors of a science article should not only be permitted to review the press release, but REQUIRED to sign off on it.

Old Ranga from Oz
September 25, 2012 5:05 am

I’ve just forwarded the link for this post to a mate who’s a journalism lecturer at one of our Oz universities. Invaluable advice for any budding journo, particularly the tart thread comments from those who’ve copped it.
You want professional respect as a journalist? Then earn it.

Ian Blanchard
September 25, 2012 6:40 am

“Coauthors of a science article should not only be permitted to review the press release, but REQUIRED to sign off on it.”
One thing that this might stop is the addition of an excessive number of co-authors named on papers. I was talking to a University lecturer who is an expert on SEM/EDX analysis last week – his initial specialism was geology and geochemistry, but over recent years he’s done SEM work across a number of disciplines. He was saying that he was recently sent a final draft of a paper on something (biology / plant science related, can’t remember exactly what) where he took a few SEM photomicrographs for the research group, and he has been named as a co-author despite:
1 – Not having had anything to do with the writing of the paper or liaising with the authors after providing the SEM work
2 – Largely not understanding the paper (well away from his specialisation)
3 – None of the SEM images he provided actually having been used in the paper.

John T
September 26, 2012 1:00 pm

I’ve unsuccessfully tried to find journal articles based on press releases myself. It is frustrating. I though the whole idea of the press release was to promote the article. A press release that doesn’t at least mention the title of the article is like the press release for a new product that doesn’t mention what the product is called.
I thought basic journalism meant covering the “who, what, where, when, why…”, but too often the press releases for journal articles leave off all of that. They don’t tell who did the research, what the paper is titled, they usually get the where (have to promote the University) but not the “where” it was published. They simply don’t cover the basic facts.

September 26, 2012 10:10 pm

I had this issue the other day looking for a particular nutrition paper. It’s ridiculous to write an article on a paper and not name the paper.

Verified by MonsterInsights