Fact-check the hyperlinks

Megan McArdle recently linked to a story in Slate about why Americans have stopped spanking.  Perhaps one of these days I’ll write about corpoal punishment itself, but that’s not what I’m posting about today.  I’d like to hone in on one paragraph that raised my hackles.  Writes Darshak Sanghavi (apparently a fellow pediatrician):

Primary care physicians tacitly approve of corporal punishment. According to well-designed surveys, 70 percent of family physicians and 60 percent of pediatricians think “striking of the child’s buttocks or hand with an open hand … leaving no mark except transient redness” is fine. In a hypothetical scenario of an 8-year-old who refuses to go to bed at the usual time, for example, one in five family physicians think the child should be spanked. Interestingly, even 40 percent of academic child abuse specialists think “spanking is appropriate sometimes.” A key committee of the American Academy of Pediatrics debated spanking for more than 10 years before they decided not to condemn it categorically. (Daniel Armstrong, the director of the Mailman Center for Child Development in Miami and the AAP policy’s main writer, told me “there was a clash of beliefs” and in the end, the committee condemned hitting kids with objects and in the face, but felt whacks on the buttocks were OK.)

That is one shitty paragraph.  Certainly its shittiness is sufficient to undermine its thesis statement about primary care physicians’ present-day approval of corporal punishment.  What this sloppy little paragraph does demonstrate beautifully is how hyperlinks can give the appearance of good research and support for on online article where little exists.

First of all, it is impossible for me to determine how well-designed that first cited survey is, because the link doesn’t provide you with anything more than an article title and a sign-in page for subscribers to JAMA: the full-text article is not available for non-subscribers.  Not being a member of the AMA, I’m not a subscriber myself.  This is presumably also true of most Slate readers.  I’ve tried to access the article through my institution’s online accounts, but doing so remotely from my office is cumbersome and difficult, and I gave up after a few tries.  Presumably even this amount of effort is more than most readers would go to, and would simply take Sanghavi’s word for it that the survey was well-designed.

But no matter how well-designed that survey is, attentive readers will note that the journal article was published in 1992.  Given how long it takes to conduct any kind of clinical research, write the article, submit it and revise it for publication, one can reasonably guess that the survey itself was conducted before 1992.  Which means, whatever the findings of the survey and no matter how well-designed it was, its results are two decades old and woefully out of date.  It tells us nothing particularly useful about how primary care physicians feel today.  Given how much has changed in our approach to disciplining children and how likely attitudes are to change from one generation to the next, it is not unreasonable to assume that the authors would get very different results if the same survey were to be conducted today.

Not one of the three citations Sanghavi includes in this paragraph about how medical providers view (present tense) corporal punishment was published within the past dozen years.  The second link, which again merely gives you an article title and a sign-in page, and which supposedly tells you what 40% of academic abuse specialists (writ large) think, is to a survey conducted on a convenience sample of attendees at a pediatric conference who went to a special interest group meeting on the subject.  And even assuming that this convenience sample is a good representation of what American academic abuse specialists thought as a whole fifteen years ago (I’m skeptical), it’s even more of a stretch to get from the article’s conclusion:

Most academic child abuse professionals believe that spankingis inappropriate and their beliefs are influenced by the context in which spanking occurs.

back to Sanghavi’s statement about providers’ tacit approval.

After noting briefly that there’s a big gap between declining to condemn something outright and tacitly approving of it, and also that Sanghavi plays pretty loose with how he presents his facts, this excerpt highlights how easy it is to misuse hyperlinks.  A reader sees an assertion of fact in blue type, and assumes that the linked material verifies the fact as presented.  But those who want more information or to verify for themselves should be directed to a source that is actually accessible to them.  Two of the three links are to subscription-required articles, and either Sanghavi wants us to trust his analysis of the pieces (in which case the links are really only there for show) or he only read the abstracts himself (in which case he is lazy, and also in no position to cite the material authoritatively).  None of the articles are recent enough to be worth anything as an indicator of medical providers’ current attitudes.

It’s interesting to note that McArdle herself does something similar in her own piece.  Tying the decline in spanking to entitled behavior among young employees (unconvincingly, in my opinion), she reports that employers are complaining about these new, presumably unspanked workers.  In support of this, she links to a blurb about a book on Amazon.  I guess we should take it on faith that she’s read the book, and that it is of sufficient quality that what it says is reliable?  The link is essentially window-dressing, there to make McArdle’s statement appear more grounded in fact than I believe it is.

One of the wonderful things about posting an essay on the Web is that one can insert access to supporting or supplementary material right into one’s text.  I am probably going to be the last of the die-hard pulped-tree book fans, but this is an undeniable advantage of writing online.  I try to use hyperlinks thusly (and am obviously inviting further scrutiny to make sure I do so by writing this post).  While I wouldn’t go so far as to say that Sanghavi is being dishonest with his hyperlinks, he’s certainly not using them to the reader’s advantage, either.  Either insert a link that will be genuinely helpful for readers seeking a deeper understanding of what you’re writing, or leave them out altogether.

Russell Saunders

Russell Saunders is the ridiculously flimsy pseudonym of a pediatrician in New England. He has a husband, three sons, daughter, cat and dog, though not in that order. He enjoys reading, running and cooking. He can be contacted at blindeddoc using his Gmail account. Twitter types can follow him @russellsaunder1.

7 Comments

  1. I have read the book, and it does say what I said–indeed, there’s a whole nascent management book literature on managing the current generation. And I hyperlinked it because I thought perhaps some of my readers would be interested in reading more deeply on the topic. It’s a decent book, as far as management books go, on the topic.

    However, I believe I also used the word “anecdotally”, because that’s what the quality of the evidence is in these books. It is far from conclusive, and I was not hanging the weight of proof on this statement that “fact check” implies. It just mirrors what I’ve heard from employers I’ve interviewed–who, again, are not a representative sample.

    • Fair enough. You do note that the evidence is anecdotal, and so you’re right that it doesn’t require the same burden of proof. While I can see why you would be inclined to include it, I wonder if there might have been something that offered better support for your observations.

      But then, perhaps I’m nit-picking. It made me irritable to read Sanghavi make statments about primary care providers grossly out of keeping with my own practice and observations, with citations that were published before I graduated from medical school.

  2. This reminds me of when, about 15 years ago, a baptist minister (a father of a friend of mine) gave me a book purporting to prove that God had a special vision for America, evident from Columbus’ voyages onward. His reasons for why I should consider the book credible is that “it has footnotes,” as if the mere fact that an author provides citations proves that the author’s assertions are true, without any question of how appropriate the citations were to what they were supposed to prove. I confess, I kept the book but never read it.

    More directly to hyperlinking: I like the potential that hyperlinking has, although I would like to see more online bloggers use real footnotes. I will say, however, that on my own blog, I rarely use footnotes, or even hyperlinks. In other words, I just rely on assertions.

  3. I read the linked articles and you come across as a nitpicker who deliberately misses the forest for the trees. Both articles link to numerous primary and secondary sources, some of which are firewalled because not all sources offer up their content for free to anybody who wants it. (Nothing wrong with Slate linking to a top tier journal like JAMA, I think, though this seems to piss you off. Would you prefer links to crap journals that are free?) But back to the point: You’re royally pissed that 2 of the dozens of studies cited in the pieces are from the 1990s. But you offer nothing of substance to back up your feelings that today’s doctors are all different and enlightened, presumably like you are based on your “own practices and observations.” That’s all you’ve got? (In a quick Medline search, I couldn’t find any newer studies, by the way, of doctor’s attitudes. Do you have any?) I guess your brand of pissy, jealous carping is what’s be expected from a doctor blogger that cowers behind a pseudonym.

    • Let’s start with the last thing you say and work upward, shall we?

      I suppose I could write under my own name, but that would constrain me somewhat in writing honestly about the institutions with which I am affiliated. I can only hope to learn how to be a courageous soul such as yourself, Anonymous. In the meantime, I will console myself with my clear understanding of the word “irony.”

      You read the full text of all the linked articles? Good for you! You clearly have access through some kind of subscription, possibly through an institution at which you are employed. Slate, maybe? Unfortunately, I would guess most Slate readers do not, thus the hyperlinks are useless to them. Non-subscribers to those journals can’t even check the references, so they (like I) will have to take your word for it that they are as numerous and comprehensive as you say.

      I couldn’t possibly be less bothered that there is a link to JAMA in the article. Sadly, as I’ve already said, by doing so they provide a link to an inaccessible article, thus rendering the link moot.

      Whether today’s doctors are all different and enlightened is wholly irrelevant. I said nothing of the sort in the main post, though I did mention in a follow-up comment how the thesis of the paragraph differed from my own observations. But then, I also don’t go on to use my own observations as the basis for any kind of broader statement, do I?

      Circling once again to the point of the post, the excerpt in question presents data that are a dozen years or more old as being representative of current provider attitudes. “Twenty years ago” is not current. The author further garbles his point by misrepresenting what is available for even cursory review. I don’t know what I’m “deliberately” missing, but I would like to see data that are more recent than when I was in high school, and I’d also like to see it represented correctly. If (as you say) there aren’t data more recent than that, then the author can simply report that there is need of updated study.

    • Hey, fella–lay off the honorable Doctor–he has more character and intelligence in his little baby finger than you have in your entire body. And how funny it is that you, “Anonymous”, chime in with, ” a doctor blogger that cowers behind a pseudonym.” Just a hunch, but I suspect “Anonymous” is not your birth name.

      Considering that physicians, who so much as look as sideways at a patient, can be subject to lawsuits in the millions, would ever, be it directly or indirectly, endorse corporal punishment or in effect, turn a blind eye to suspected child abuse is just insane. Who? Who? Who has done this?

      Please, just few names. That will suffice. Just what do think would happen if a mother, who returns home from work, and whose husband has taken care of the child in her absence, notices the child has deep red welts on his buttocks and is in considerable pain when trying to sit? What happens when she goes to the ER and one of your 70% corporal punishment approving physicians is called upon to treat this child. Is he just going to just brush it off as an example of acceptable discipline by a parent under the circumstances as it is presented to him that day?

      Is that just one of those, “shit happens” moments?

      Man oh man, is Geoffrey Fieger going to love you!

      Is that just one of those, “shit happens” moments?

Comments are closed.