On 28th September, I woke up to look at Twitter and find Pete Etchells fulminating about a piece in the Guardian.
It was particularly galling for him to read a piece that implied research studies had shown that voice-responsive devices were harming children’s development when he and Amy Orben had provided comments to the Science Media Centre that were available to the journalist. They both noted that:
a) This was a Viewpoint piece, not new research
b) Most of the evidence it provided consisted of anecdotes from newspaper articles
I agreed with Pete’s criticism of the Guardian, but having read the original Viewpoint in the Archives of Disease in Childhood, I had another question, namely, why on earth was a reputable paediatrics journal doing a press release on a flimsy opinion piece written by two junior medics with no track record in the area?
So I wrote to the Editor with my concerns, as follows:
Dear Dr Brown
Viewpoint: Effects of smart voice control devices on children: current challenges and future perspectives doi 10.1136/archdischild-2022-323888 Journal: Archives of Disease in Childhood
I am writing to enquire why this Viewpoint was sent out to the media under embargo as if it was a substantial piece of new research. I can understand that you might want to publish less formal opinion pieces from time to time, but what I cannot understand is the way this was done to attract maximum publicity by the media.
The two people who commented about it for the Science Media Centre both noted this was an opinion piece with no new evidence, relying mainly on media reports.
Unfortunately, despite this warning, it has been picked up by the mainstream media, where it is presented as ‘new research’, which will no doubt give parents of young children something new to worry about.
I checked out the authors, and found these details:
These confirm that neither has a strong research track record, or any evidence of expertise in the topic of the Viewpoint. I can only assume that ADC is desperate for publicity at any cost, regardless of scientific evidence or impact on the public.
As an Honorary Fellow of the Royal College of Paediatrics and Child Health, and someone who has previously published in ADC, I am very disappointed to see the journal sink so low.
Yesterday I got a reply that did nothing to address my concerns. Here’s what the editor, Nick Brown*, said (in italic), with my reactions added:
Thank you for making contact . My response reflects the thoughts of both the BMJ media and publication departments
Given my reactions, below, this is more worrying than reassuring. It would be preferable to have heard that there had been some debate as to the wisdom of promoting this article to the press.
It is a key role of a scientific journal to raise awareness of, and stimulate debate on, live and emerging issues. Voice control devices are becoming increasingly common and their impact on children's development is a legitimate topic of discussion.
I have no quarrel with the idea that impact of voice control devices on children is a legitimate topic for the journal. But I wonder about how far its role is ‘raising awareness of, and stimulating debate’ when the topic is one on which we have very little evidence. A scientific journal might be expected to provide a balanced account of evidence, whereas the Viewpoint presented one side of the ‘debate’, mainly using anecdotes. I doubt it would have been published if it had concluded that there was no negative impact of voice control devices.
Opinion pieces are part of a very wide range of content that is selected for press release from among BMJ's portfolio of journals. They are subject to internal review in line with BMJ journals´overall editorial policy: the process (intentionally) doesn't discriminate against authors who don't have a strong research track record in a particular field
I’ve been checking up on how frequently ADC promotes an article for press release. This information can be obtained here. This year, they have published 219 papers, of which three other articles have merited a press release: an analysis of survey data on weight loss (July), a research definition of Long Covid in children (February) and a data-based analysis of promotional claims about baby food (February). Many of the papers that were not press-released are highly topical and of general interest – a quick scan found papers on vaping, monkey pox, transgender adolescents, unaccompanied minors as asylum seekers, as well as many papers relating to Covid. It’s frankly baffling why a weakly evidenced viewpoint on a topic with little evidence was selected as meriting special treatment with a press release.
As for the press release pathway itself, all potential pieces are sent out under embargo, irrespective of article type. This maximises the chances of balanced coverage: an embargo period enables journalists to contact the authors with any queries and to contact other relevant parties for comment.
My wording may have been clumsy here and led to misunderstanding. My concern was more with the fact that the paper was press-released, which is, as established above, highly unusual, rather than with the embargo.
The press release clearly stated (3 times) this article was a viewpoint and not new research, and that it hadn't been externally peer reviewed. We also always include a direct URL link to the article in question in our press releases so that journalists can read the content in full for themselves.
I agree that the press release included these details, and indeed, had journalists consulted the Science Media Centre’s commentaries, the lack of peer review and data would have been evident. But nevertheless, it’s well-known that (a) journalists seldom read original sources, and (b) some of the less reputable newspapers are looking for clickbait, so why provide them with the opportunity for sensationalising journal content?
While we do all we can to ensure that journalists cover our content responsibly, we aren't responsible for the manner in which they choose to do so.
I agree that part of the blame for the media coverage lies with journalists. But I think the journal must bear some responsibility for the media uptake of the article. It’s a reasonable assumption that if a reputable journal issues a press release, it’s because the article in question is important and provides novel information from recognised experts in the field. It is unfortunate that that assumption was not justified in that case.
I just checked to see how far the media interest in the story had developed. The Guardian, confronted with criticism, changed the lede to say “Researchers suggest”, rather than “New research says”, but the genie was well out of the bottle by that time. The paper has an Altmetric ‘attention’ score of 1577, and been picked up by 209 news outlets. There’s no indication that the article has “stimulated debate”. Rather it has been interpreted as providing a warning about a new danger facing children. The headlines, which can be found here, are variants of:
“Alexa and Siri make children rude”
“Siri, Alexa and Google Home could hinder children’s social and cognitive development”
“Voice-control devices may have an impact on children’s social, emotional development: Study”
“According to a study, voice-controlled electronic aides can impair children’s development”
“Experts warn that AI assistants affect children’s social development”
“Experts warn AI assistants affect social growth of children”
“Why Alexa and Siri may damage kids’ social and emotional development”
“Voice assistants harmful for your child’s development, claims study”
“Alexa, Siri, and Other Voice Assistants could negatively rewire your child’s brain”
“Experts warn using Alexa and Siri may be bad for children”
“Parents issued stark warning over kids using Amazon’s Alexa”
“Are Alexa and Siri making our children DUMB?”
“Use of voice-controlled devices ‘might have long-term consequences for children’”
And most alarmingly, from the Sun:
“Urgent Amazon Alexa warning for ALL parents as new danger revealed”
Maybe the journal’s press office regards that as a success. I think it’s a disaster for the journal’s reputation as a serious academic journal.
*Not the sleuth Nick Brown. Another one.