# Putting P Values in Their Place

Although I am not a statistician, I find something very appealing about mathematics and statistics and am pleased when I find a source to help me understand some of the concepts involved. One of these sources intersects with my obsession with politics: Nate Silver’s website fivethirtyeight.com. Yesterday, during a scan of fivethirtyeight’s recent posts, this one by Christie Ashwanden caught my eye: “Statisticians Found One Thing They Can Agree On: It’s Time to Stop Misusing P-Values.”

P values and data in general are frequently on the minds of manuscript editors at the JAMA Network. Instead of just making sure that statistical significance is defined and P values provided, we always ask for odds ratios or 95% confidence intervals to go with them. P values are just not enough anymore, and Ashwanden’s article was really useful in helping me understand why these additional data are needed (as well as making me feel better about not fully understanding the definition of a P value—it turns out I’m not alone. According to another fivethirtyeight article, “Not Even Scientists Can Easily Explain P-Values”). One of the bad things about relying on P values alone is that they are used as a “litmus test” for publication. Findings with low P values but not contextual data are published, yet important studies with high P values are not—and this has real scientific and medical consequences. These articles explain why P values only can  be a cause for concern.

And then there was even more information about statistical significance to think about. A colleague shared a link to a story on vox.com by Julia Belluz: “An Unhealthy Obsession With P-Values Is Ruining Science.” This article a discussed a recent report in JAMA  by Chavalarias et al “that should make any nerd think twice about p-values.” The recent “epidemic” of statistical significance means that “as p-values have become more popular, they’ve also become more meaningless.” Belluz also provides a useful example of what a P value will and will not tell researchers in, say, a drug study, and wraps up with highlights of the American Statistical Association’s guide to using P values.—Karen Boyd

# Point of View: A Conversation With Cheryl Iverson

(Editor’s Note: Point of View is an occasional series that features an interview with someone in the world of publishing. You are already familiar with Cheryl Iverson from her thoughtful and well-reasoned answers to our “Questions From Users of the Manual” section—so it’s only fitting that she be the first victim participant.—Brenda Gregoline, ELS)

Cheryl Iverson does not peeve about her grammatical peccadilloes as one might imagine of a woman who has spent her career editing, overseeing editing, and serving as the AMA Manual of Style committee chair for the last 3 editions—she continues as co-chair of the 11th edition, a work in progress. Although one might say that she doesn’t have any major pet peeves when it comes to grammar, she does admit, “I still get aggravated at the incorrect use of apostrophes like I – t- apostrophe – s. Those are not things that I would be willing to treat lightly.”

But concerns about splitting an infinitive or ending a sentence with a proposition, “some of those rules that people learned in grammar classes in grade school 50 years ago,” would be better off forgotten. She speculates that people who no longer understand the reason for the rule will either avoid the use or argue adamantly about using, say, different from rather than different than when in the end it doesn’t matter if meaning is clear. “That’s what I think. It’s good that we’ve gotten away from these old rules without understanding where they came from, which makes it hard for people to know when to bend a rule or when to disregard a rule.”

However, such discretion at the start of her career was discouraged.

“When I was a new copyeditor, the stylebook was the Bible. It was all about ‘If you want to publish in our journals, this is what we do.’” The mandate was so deeply impressed on her that it wasn’t until her short stint at the University of Chicago Press in the books  department that she began to rethink the approach.

First, while editing a small book on ancient Egyptian irrigation systems, she assiduously looked up all the city names mentioned in the manuscript and changed them in accordance with her source, an effort that left the author less than pleased.  “He wanted them all reinstated.  He said that there is controversy about some of these names. ‘I don’t care if your source used X. I want Y.’”

The exchange made Iverson pause. “I stopped to think that consistency is far more important in a journal article because all of these articles that are edited by different people are grouped together, whereas a book stands alone, so if your author wants to use something quirky, that’s just that author and it’s just in that book.”

The vise grip of the conforming to the immutable-style-rules directive loosened again when an author challenged a style point, again while she was at the University of Chicago Press. So she brought her question to her supervisor and said, “I looked everywhere in the manual for this. I can’t find it. … This author wants to have such and such.”

“Cheryl,” the boss responded, “the manual is not a Bible. It’s just a guide.”

Iverson’s experience as a young copy editor is not uncommon. “When people are new, they want to know, What’s the rule? What’s the rule about? They want there to be a rule because rules help them”—especially when trying to explain their editing decisions to authors. “But then when you have been editing a while you’re more mellow. I don’t think it is a bad thing in other words. I don’t think it is a bad thing [to relax the rules] as long as you don’t get sloppy and let errors get through.”

Helping new copy editors navigate the rough terrain of medical articles that are filled with jargon, abbreviations, numbers, complex tables, and figures has been a singular pleasure. “I loved when a relatively new editor would learn the job,” Iverson said, describing how individuals would come to her and say, “‘I don’t know what to do with this table.’ Then later, they would come in and say, ‘This table is really difficult, but I thought I would do X.’ You could see that they were starting to develop their own judgment.” Although she remains in partial retirement, Iverson responds to questions from AMA Manual of Style users and posts answers on the AMA Style Insider blog. The questions range from as simple as how many spaces go after a period to as difficult as coaching a writer whose bosses disagree with a grammatical decision despite all kinds of evidence supporting her or his position. She responds to those questions with the same compassion and sense. And she receives the same pleasure in coaching them to resolution.

“I see [that] our job as editors is to facilitate reading so that you don’t have to stop and go back and say wait what did that say again?”

To illustrate her point she recalled a cartoon presented at a Council of Science Editors conference. In the cartoon, the characters debated, “Who is more important, the author or the editor?” Each character argued for the author or the editor until the final panel, which said “The reader is what is important. That’s who we are working for.”—Beverly Stewart

# Quiz Yourself

Which of the following sentences is correctly punctuated?

We conducted a randomized placebo-controlled trial.

OR

We conducted a randomized, placebo-controlled trial.

We conducted a randomized placebo-controlled trial.

When fewer than 3 modifiers are used, avoid adding a comma if the modifiers and the noun are read as one entity, such as randomized placebo-controlled trial.—Laura King, ELS

(Editor’s Note: This month our favorite proofreader takes a bus journey and speedily ruminates on the Self, commas, exclamations, and the astonishing loveliness of ambiguity in a world [a style-manual world, anyway] that prioritizes precision.—Brenda Gregoline, ELS)

When last I percolated here, my dregs, if you got down to them—and how could you not, 3 tall paragraphs easier to guzzle than a brick, let alone the world and all its oceans, interconnected though they may be?—could have been read as “When you get right down to it, we are all raccoons.” That’s the thing about dregs: a hard stop at the bottom of the carafe, with nothing beneath but a countertop stretched over whatever doodads populate the ceiling of hell, doesn’t invite interpretation. As with tea leaves, you can rearrange the dreglettes, but even if you’re a fool for random patterns they’re not going to inspire sleepless nights or a revised liberal arts curriculum. Granted, we are nothing if not raccoons, but that’s not exactly a fresh thought. Einstein, Calamity Jane, the Dionne Quintuplets, and virtually everyone whose name is droppable has, if only for one brief sliming moment, accessed their inner round table and considered that the omnivore rattling around in their garbage can was essentially him- or her- or themself/selves. John Donne said as much, King Arthur sat as much, and various enlightened beings have lived as much, perfecting a state of oneness with raccoons, buzzards, pancakes, whatever illusory forms rattled, circled, were flooded with syrup, or in any other way made a specious case for their independent existence.

That’s not exactly where I was going last month, but the dregs said “Stop here or you’ll penetrate plexiglass (alteration of “Plexiglas” per Merriam-Webster’s).” I’ve no desire to penetrate anyone’s plexiglass, least of all my own, so I stopped there and am only now refilling my pot with a few words about our peculiarly human masks and tails. First, let’s play Us and Them. Splitting our cozy Self into species and regarding raccoons as the “other,” it becomes clear that they are the most punctuated animals in the forest, each tail a rhapsody of serial commas, each face a question mark verging on exclamation, flexible paws cradling dependent and independent clauses as easily as crayfish and stones. As they lope, their arched backs become semicolons on the move, and if you approach one from behind you’ll find the elegantly apportioned tail initiating a furry sentence that terminates in the most fulsome period this side of Go, Dog. Go! The inky button is positioned squarely (by which I mean roundly) beneath Dick Turpin’s gaze and pulses in crinkledom over Dracula pearls designed to puncture and sift all your trashcan allows. Since humans crawled out of their wormholes and began to strive for literacy and other forms of one-upsmanship, there has not yet been a sentence so well-turned that it could possibly be mistaken for a raccoon. Similarly, squirrels aren’t allowed to compete in the Olympics.

If no athlete’s grace and dexterity can rival a rodent’s and no sentence yet composed can achieve the pixie perfection of a thief bent on depriving the city dump of your pie crusts, then why don’t we just give up? I suppose every refugee from Eden has struggled with or, being nonsquirrels, clumsily scuffled with that question when they weren’t clumsily scuffling with the question of whether or not the world was about to end. If an answer occasionally emerges from the fray, it might be as simple as this: there’s something to be said for ambiguity, for breaking an apple’s perfect skin with your sawed-off fangs and wondering whether it was the right thing to do, knowing it might not have been. Like Simple Simon, who as a connoisseur of pies fruity and not was far more complicated than the legendary folk hero we celebrate April 1 and every other day dedicated to human inadequacy (which would be all of them so far, unless world hunger was eliminated yesterday), “ambiguity” as an answer to anything, including what it means to be human, is a hard stop that keeps on going.

A visit to the local library (imagine that!) suggests that there’s actually everything to be said for ambiguity and nothing much to be said for anything else, though if your library contains scientific journals such as the ones our Manual of Style is designed to tease into perfection you’ll find a compelling counterargument. A less compelling but more visceral argument for the existence of precise diction, at least in its ability to describe persons as the sum of one of their parts, may be encountered on the Chicago 66 bus. Sadly in the first case and ho-hummily in the second, those are exceptions. In language we seek and prize precision, as if it could possibly turn us into porpoises, but language itself thrives on ambiguity. Flush with multiple meanings, the thrill of the hint, subsurface innuendo erupting into a geyser of accusation, then freezing midair into flurries of doubt, ambiguity is an ever-expanding state of expression gobbling into its definition the essences of uncertainty, hesitation, disgruntlement, past and looming failure, vague gnawing the likes of which beavers and termites would surely deem unprofessional. At its ragged outer edges ambiguity encompasses a sadness only caffeine and the promise of love can dent, and then only till 3 o’clock in the afternoon or the recriminating hour before dawn, respectively.

There it is, not so much the nowadays mild curse but the punctuation mark coloring the word and wrenching it into a refined state of ambiguity. As Kay Francis once said to William Powell in the context of 2 characters united in love and destined for separate gallows, “This is living, isn’t it, Dan?” We punctuate; therefore we are. Are what? Raccoons, sure, but beyond that certainty a rash of beautiful bustling hesitations—saintly clause, fiendish fragment, each reinterpreting murk in the light of a moment just passed, anticipating and dreading the future, then, all too soon, drinking it in.—David Antos

# Questions From Users of the Manual

Q: A colleague is adamant about citing page numbers even if the reference is used more than once in a paragraph. Is this necessary?

A: There is a little advice on citing page numbers in the manual, just above section 3.7.

If the author wishes to cite different page numbers from a single reference source at different places in the text, the page numbers are included in the superscript citation and the source appears only once in the list of references.  Note that the superscript may include more than 1 page number, citation of more than 1 reference, or both, and that all spaces are closed up.

Example: These pages showed no sign of proactive sphincteric adduction.3(p21),9

You’ll notice that it’s not mandatory and our style includes, “If the author wishes.”  Page number citation can be helpful to the diligent reader who wants to go to the source cited and find the exact mention of the quoted material.—Cheryl Iverson, MA

# Quiz Yourself

Do you know the difference between the terms multivariable and multivariate? One term refers to multiple predictors (independent variables) for a single outcome (dependent variable), and the other term refers to 1 or more independent variables for multiple outcomes? Which is which?

Multivariable refers to multiple predictors (independent variables) for a single outcome (dependent variable). Multivariate refers to 1 or more independent variables for multiple outcomes. Therefore, analyses can be described as multivariable, to indicate the number of predictors, or as multivariate, to indicate the type of outcome.—Laura King, ELS

(Editor’s Note: We are a little late posting this update from the bold, fresh-roasted David—but it’s still January and still legit to think about the new year. And it’s always legit to think about commas!—Brenda Gregoline, ELS)

I’ve been percolating through a sieve called December and am ready to be poured into a new year. When I imagine myself as dark liquid I also imagine I’m filling one of my favorite cups to the brim, usually a small tapered one with a pecking hen, though my foursquare Yosemite one sporting a cute raccoon and all-caps DAVID is a real winner too. But this is different: 2016 is no cup I’ve known, no mug, no chalice, no shattered past or graspable present. December 31 is intent on pouring me into a vat of pure future, of which the only knowledge I have is really a bag of assumptions, such as that the Chicago River will continue to flow backwards and the world will not end. A recent year had the distinction, until it was over, of being a candidate for the year it all ended, something no cool person took seriously but the less smug secretly pondered. Was that 2012? I should know; I was there. There was a movie about it, so let’s look it up: yes, 2012 is the movie, and other sources specify the date as December 21. Unless I’m oblivious, the world didn’t end then, but I do recall it beginning when, either that day or the day before, a person actually asked me if I thought it would end. She asked cautiously, sincerely, a human comma breaking the flow of unthinking acceptance that an unpunctuated worldview promotes. That was 3 years ago now, and I think not one of the 1000+ days since has passed without a moment of appreciation for that strategically placed comma.

Now what do you suppose I said when asked if I thought the world would end later that day or the next day, depending on which day the question was posed, which I’m irked not to be remembering, especially since I’ve already said the current world began at that moment, which makes that moment momentous, to say the least? I dug into myself and excavated an honest answer, which is that I thought it wouldn’t end because it hadn’t ended yet and I’m used to continuance, believing in it not as a matter of faith but as a matter of habit, expecting the sun to rise each morning because I don’t trust it to do anything else. As I spoke I realized that, in my own dim way, I was being as smoothly uncritical as the snarksters who, subtext curling lasciviously around the scientific method, dismissed the whole Mayan calendar thing as the kibble of bozos. I too was riding a wave of unpunctuated assumptions, though on a sea of dork instead of a sea of cool. The earnest question was dry land where there had been none, a bit of Moby-Dick calling into question the efficacy of sailors. Had it not been posed, the world would have continued without the slightest hesitation, but the human comma created a pause in which thought struggled to find itself and, failing somewhat, has been continuing the struggle through each of these past 1000+ days.

You might suggest that my human comma was more of a question mark, but you’d be literalizing my friend based on the type of sentence she uttered rather than the form her sentence imposed. A semicolon possibly; maybe even a period, given that she stopped the world cold and started it up again in a different water park. But a human question mark? I think not. She, or he, or whatever she or he really is–we’re all on a continuum, you know, but not an unpunctuated one, not (for me, anyway) since December 20 or 21, 2012– tossed a silken boomerang into the assumption stream and now, though I have no idea what lies ahead, or beneath, as I begin to tumble into whatever 2016 is or isn’t, I do find one sustaining idea limning the boomerang’s surface, that punctuation exists not to separate but to illuminate, to gently render intelligible the formless murk through which we pilot our masked faces and ringed tails.—David Antos

# Questions From Users of the Manual

Q: Is it correct to leave “post” as a separate word in the following sentence? “These activities must take place from prelaunch to post launch.”

A: In section 8.3.1 (“When Not to Use Hyphens”), you’ll see the following:

Note that when post is used as a combining adjectival form, as in postmortem examination, it is set closed up. When it is used as an adverb, as in post hoc testing, it is set as 2 separate words.

So in your example, you would not close up “post launch.” However, the meaning of this sentence is ambiguous to me.  I would suggest rephrasing it to avoid an awkward construction and to clarify exactly what interval you are talking about. How about “These activities must take place both before and after launch.”?—Cheryl Iverson, MA

# Quiz Yourself

Edit the following sentence for correct usage of anatomy terms:

The investigators examined catheter-induced lesions of the right heart.

The investigators examined catheter-induced lesions of the right side of the heart.

Editor’s Note: Authors often err in referring to anatomical regions or structures as the “right heart,” “left chest,” “left neck,” and “right brain.” Generally these terms can be corrected by inserting a phrase such as “part of the” or “side of the” (§11.6, Anatomy, p 410 in print).—Laura King, ELS

Editor’s Note: Here begins an occasional series by one of The JAMA Network’s favorite overcaffeinated proofreaders, David Antos. He has a deep affection for parts of speech, punctuation, and widows and orphans. (See what I did there?)

Twice in the past week I’ve done the unthinkable, which, however, I found my mind was eventually willing to grasp, as if the unthinkable could be thought through without losing its privileged status of noble ignorance. Less like a raccoon’s grasping a crayfish than a raccoon’s washing pastry till it dissolves, my mind was the beneficiary of an organic shampooing so dense and luxurious that I got a little smarter right there in the shower. No blinding insight in my trickle-down acuity, but a soft perception that there was meaning to be had in twice forgetting to close a parenthetical expression, which is the horror I’d committed.

Do you call parens open/close or opening/closing? Or left/right? I tend toward overexpression, so you can imagine I favor the suffix, but, to keep this clean and tidy, like my hair, let’s do open/close, even though “close” reminds me of “clothes,” which I was bereft of in the shower though in polite society am usually not. Which is the point my shampooed mind was shoring into, that society becomes impolite when punctuation falters. In 2 emails I’d sent that week I’d properly initiated a parenthetical expression that, when gotten to the end of, nakedly submitted itself to the rest of the sentence without so much as a stitch of close. Well, actually, the 2nd email contained a couple bracketed morsels at its conclusion, and the close paren was likely forgotten in the heat of bracketing. No excuse that, but a circumstance lending a ray of thought to the unthinkable. I’d also preemptively included in each of those emails the assertion that, in the interests of sending them in a timely fashion, I was not going back to proofread before hitting the button of no return. I had my out, yet when, after sending, I did proofread, I felt guilt, shame, inner filth, as if I were Michelangelo’s socks when he finally crept down the Sistine Chapel’s ladder. I’d poured my soul into words and left out the close parens, rendering that soul holier than the aforementioned socks. And not in a good way.

Later, I washed my hair and, as a person does when feeling hair and contemplating sin, reflected on the week that was and the week that will never be. I thought of the close parens that were not allowed to see the light of another’s eyes and of the violent omissions recounted in the week’s headlines, and it began not to seem out of place to think that faulty punctuation is at the heart of a world gone wrong. Not that a sloppy email is to blame for unthinkable atrocities, but the sense of purpose, order, and unblinking inclusion punctuation provides mirrors the sense of moral rectitude we only occasionally achieve in our dealings with each other. Punctuation is never evil, just misplaced or missing, and as I stepped out of the shower I cheerfully anticipated my next trip on public transportation, for I would no longer regard the other commuters as merely people but as pieces of punctuation, possibly shoveled into a seat of crumbs or wedged pielike into each other, but innately good, even angelic. —David Antos