RTFS-plus

It's time to play "Anatomy of an Error" again. There's a cure for this sort of hed inaccuracy, but it requires a little more work than usual. Here we go:

And here's the lede from the story:

SHARM EL SHEIK, Egypt --Middle East leaders looking to contain and weaken Hamas forces now controlling the Gaza Strip converged at this Red Sea resort Monday.

But Hamas upstaged them by releasing the first recording of an Israeli soldier they captured a year ago, which had touched off a major Israeli intervention in Gaza.

So far, so goodish (though the Team Is An It Department is pained to see "Hamas" getting a plural pronoun). Where's the error?

The audiotape of Israeli soldier Gilad Shalit, along with the release of a new video from the Gaza Strip showing abducted BBC correspondent Alan Johnston wearing an explosive vest, overshadowed the modest Egyptian summit, which was meant to demonstrate support for Palestinian Authority President Mahmoud Abbas.

A good way to put the RTFS rule into practice is to isolate any fact claim in the hed and ensure that you can put your finger on the corresponding fact in the story. The main hed is pretty easy: Hamas is the actor, and upstaging (OK, it's the reporter's opinion, but for purposes of making the hed correspond with the story, it's going to count as a fact) is what Hamas did.

The deck is more of a challenge, because the copy editor has to assemble a couple of disparate fact claims into clauses to test the assertion "Group releases recordings of prisoners." Graf 2 gives us one clause: Hamas released a recording of a prisoner. But the fact claim is about recordings of prisoners, so on to graf 3: The tape (released by Hamas) and a video of the captive BBC reporter are what did the overshadowing. In effect, we have an active clause and a passive one:

Hamas releases tape of soldier
Tape of reporter released

See where the error comes from? The hed assumes that Hamas is the subject of both, and that's farther than the first three grafs allow. That's a cue to hunt through the rest of the story for the missing subject of the second clause. It isn't there (it's in the full story at the McClatchy Washburo site but appears to have been end-cut from the versions appearing in Charlotte and Miami), but that's not a clue to guess. It's a clue to visit the originating desk and ask who did the releasing of the videotape.

Turns out it wasn't Hamas; it's a different group whose agenda appears to include yanking Hamas's chain. So the hed's claim -- "Group releases recordings" -- is false.

Now, the reaction from the glass offices is likely to be a big yawn. "If you see an error, write us," the paper's Web site implores, but it needs a footnote: "Unless it's about the Middle East or something somebody from the Washburo did." That's unfortunate, because how clearly a news organization reports a topic is a reflection of how clearly it understands that topic. Hamas has been playing up its efforts to win the reporter's release, which says something about how Hamas wants to be seen by the outside world. Whatever your views on this part of the world, if you want to know what a group is up to, the group's perception of its interests is a really good place to start.

OK, end commercial for the Virtues of Realism. If this paper is the place you get your news about the fractious Near East, you could end up misinformed. Or you might just roll your eyes and turn to some source that gets the structure right.

Why is this sort of error more common in news than in sports? Sports deskers are expected to have a pretty high level of subject knowledge and to pay attention to developing stories they're likely to handle. News desks need to catch up -- especially if "credibility" is one of those things we're going to make central to our claim on audience time and attention.

Several papers that used the MCT article took approaches that avoided the assumption that fed the error. Fort Worth went with "Tapes of captured soldier, journalist mar summit." Miami used "Tape of missing Israeli soldier aired" (I'd quibble with "aired" to mean "posted on a Web site," but at least the hed stays away from the initial goof).

Bring your own brush

Ah, the comma of direct address! Such a little thing, yet so useful in telling these sentences apart:

Let's eat, kids!

Let's eat kids!

As our tabloid cousins apparently forgot last night. Are you guys going to provide the brushes and soap?

There are, no doubt, other meanings to be wrung from this. You may talk about them in the comments, or you may go [ahem] wash your mouths out.

I can has hyperforeignizations?

Some days you pick up the Times and wonder what was going on in the deskly mind:

Until recently, the Jordanian side of the Dead Sea was not much more than a barren stretch of land, biblically sacred but inhabited mostly by Bedouin goatherd.

Must ... make ... noun ... plural. Oh noes! Secret Meaning already is plural! Must ... remove ... -s! Oops, amputated wrong plural marker. Or maybe Safire's malign influence is metastatizing outward from the Magazine and we're reading "Bedouin" as the noun and "goatherd" as the modifier: the "attorneys general" or "Whoppers Junior" effect.

In the confusion there, somebody missed the joke, which is that "goatherd" already has a plural. It is "goatherds." As in "Chapter 11: Of what befell Don Quixote with certain goatherds." So does "bedouin" ("bedouins"), wherein things get slightly more entertaining.

True, "badawin" is a plural form of "badawi," a desert-dweller (although the Wehr and Cowan dictionary gives a nice clear hint about the status of the English noun when it defines "badawi" as "a Bedouin"). Here's what the OED says about how we got it:

First known to Europeans in Crusading times. The plural, being of most frequent use, was adapted in med.L. as bedu ni, bedewni, It. beduini, baduini, whence a sing. L. bedunus, It. beduino, F. beduin, etc.

And forms of "Bedouins" have been showing up in English for six centuries now: Bedoynes, Bedwins and Bedonians (hail, hail Bedonia!) among them. So even if the Safire effect was in play, it should have produced "Bedouins goatherd," since even at the Times you probably couldn't get away with "Bedouins goatherds." Although it'd be fun to try.

Anyway, one can only hope the high-level debates over how to pluralize what in the lede didn't leave the editors' guard down when this one came along:

Although Westerners may harbor suspicions — or perhaps guilt — about lounging seaside in an Arab country wedged between the West Bank and Iraq, the spagoers are attracted by Jordan’s reputation as a moderate and safe country, whose people are imbued with a Bedouin sense of hospitality.

It might have been nice of the desk to remind the writer that not all "Westerners" are Americans (though maybe he meant "badawins who live on the West Bank of the Hudson," rather than "people from countries where the prestige press doesn't fan irrational paranoia"). But the "guilt" part -- I'm trying to convince myself that's merely stupid, rather than blitheringly ethnocentric to the point of being racist, and not having a lot of success. You?

East wind. Rain.

Brief, totally self-indulgent time-out from the usual litany of complaints. The Official Dissertation Committee has had its way with the Official Dissertation and checked all the appropriate boxes anyway (short answers: Yes, there's an effect; yes, it works; no, I haven't figured out why government orientation and media orientation go in opposite directions despite the overlap, but I'm working on it). Pending a few revisions, the degree is complete.

Thus ends an interesting period in which on any given day, the mailbox at the Manor was likely to hold a solicitation from the AARP and a TO THE PARENTS OF! note from some bottom-feeding student loan provider. Everybody ought to try it sometime.

What's ahead? OK, here's a trivia quiz: Name this intersection! (You're at the southwest corner looking north; neither road is named after one of the Research Kitties.)

Make up your mind

What's wrong with this picture? Or, more to the point, what's wrong with the lede, CP and downpage heds?

Here they are again, in slightly larger type:

Bush may close Guantanamo
Ozone advice may become order
Firefighters may have been trying to save colleagues

Exactly! Four stories on the page, three violations of the then-again rule: Any hed saying "X may happen" must by law have a deck or subhed saying "Then again, it may not." Heds are supposed to tell the reader what happened, not what might not happen.

The biggest problem with "may" heds is that they're almost never news. "Bush may close Guantanamo" is by definition no truer today than it was yesterday (which appears to be "not very"; we're talking about closing the prison, not the base itself, aren't we?). What's news is that some officials are now saying semi-publicly (or that "the AP has learned," whichever) that the said closure appears imminent.

And when you waffle your 1A heds as a matter of policy, there's always a risk that everyone writing for the page will waffle in the same way. As in the example above.

Stick to what's new. If all you can do is tell people what you're guessing, you don't have much claim on their attention.

Paint self out of corner

Remember "In the Walls of Eryx"? The Lovecraft story about the prospector who gets trapped in the invisible maze on Venus and dies a miserable symbolic death, all the while unaware that the exit is just a couple feet behind him? Such is the fate that awaits the unwary hed writer who starts stacking preposed noun modifier atop preposed noun modifier:

It's correct, or at least not incorrect. Because of its distance from the subject, you pretty much have to hyphenate "tax-cut," ugly as it is, just so it's clear that you're talking about something "tax-cut" modifies and not a cut in taxes themselves. But whether you're counting words or picas, the hed's three-quarters over before you get the subject out of the way.

The hed writer (or the slot; as Lovecraft implied, the solution is always clearer when you aren't the one in the maze) needed to think like a flight attendant: Remember, the nearest exit may be behind you. Stop counting the distance from the hyphenated compound to the subject noun and turn the whole thing around:

Final votes likely today
on tax cut, budget, DOT

It's simpler, it takes the hyphen question off the table, and it gets to the point at the outset.

Speaking of Lovecraft, this just in from The Ridger and too good not to spread more widely:

Ya! Ya! We await a good old-fashioned platform fight at the convention.

Planet Fox: Special framing issue

And this just in from the Fair 'n' Balanced Network:

Enjoy the wire story here.

RTFS

And this just in from the crosstown competition:

Man exposed himself to child, officials say
Sounds pretty serious. Since it's in the hed, do you suppose it's true?

An Odessa man is in custody for allegedly exposing himself over the Internet to a Boone County sheriff’s detective posing as a teenage girl.

Uh, guess not.

Odessa police arrested Alleged Cluelessgoober, 24, Friday at his home and seized a computer and a Web cam under a search warrant, according to a news release from the Mid-Missouri Internet Crimes Task Force.

The investigation started in January after Cluelessgoober contacted a person he believed to be a 13-year-old Boone County girl over the Internet. The "girl" was actually a task force investigator, Boone County sheriff’s detective Andy Anderson said.

Blink! Blink! Scare quotes! So if you "ignored" the "lede," you could still "catch on" that the "girl" wasn't "actually" a "girl" at "all."

Why complain? Well, it's sort of Brave New Week Of Inventing The Future Of News Week around here, and it's hard to see why we should even bother if we can't even read our own ledes.

Sunday morning hed roundup

The bad, the bad and the ugly from this morning's romp through the news:

A time to adore Ruth
Nearly festive service honors her devotion to family, Christ

OK. We're familiar with the source paper's fetishistic attitude toward the Graham family (used to be the only thing that would get a staffer overseas was a Billy Graham crusade). But did anybody think to look "adore" up before slapping it into the hed? Gosh, let's:
1. To worship as a deity, to pay divine honours to. (Now almost confined to poetry.)
2. (In the usage of R.C. Ch.) To reverence with relative or representative honours.
3. absol. and intr. To offer worship.
4. techn. To kiss the hand, to a sovereign, etc. (So explained by Selden, but perh. never so used.) Obs. Also, To elect (a pope) by adoration.
5. fig. To reverence or honour very highly; to regard with the utmost respect and affection. Now (also in trivial use), to like very much.

The second and third parts of 5 might work, but they seem pretty much like ongoing states -- not the sorts of thing brought on by the occasion of the funeral. Recommendation: Turn down the wailing and gnashing of teeth a couple notches. It's cloying and irritating, and you can be as respectful as you need without it.

5,000 COLLECTED IN 3 HOURS
Signs point to tossing more illegal ones
Crews sweep streets to find violators as city considers upping fines

Somebody got just a bit too clever -- or a bit too smitten with Elongated Yellow Fruit Syndrome. The trouble here is that I have no idea what "ones" means: Is it like "cold ones"? Is it like "illegal aliens"? There's no help from either of the decks, or from the lede:

Peter Harden has made it his mission to clean up a corner of Providence Road in southeast Charlotte. Every Friday night, he says, is out of control as new signs pop up for the weekend.

What is he trying to clean up -- is it something the signs are advertising? Escort services? Weekend parties? If you can hang in there until the third graf, it comes into focus at last:

Twenty pairs of city employees and volunteers fanned out across the city to clean up illegal signs.

Ah. It's the signs themselves. Dear deskers, if you want to be subtle, try to be a bit more blunt about it. I'm coming to these in the grim, coffee-deprived hours of the morning. As do many people who read heds, whether online or in print. (Did we check the math, by the way? Each of these pairs was collecting more than 80 signs an hour?)

Cops nab 3 after tortured boy, missing girl and mom's body found at Wisconsin house
Sometimes the sheer Foxness of it all simply overwhelms one, doesn't it?

Diagramming contest

Ruth Bell Graham, the wife of the world’s foremost evangelist yet shined outside his shadow in her own right, died at 5:05 p.m. today at home in Montreat in the N.C. mountains.

No further comment.

Daily editing quiz

Today's editing quiz comes from the Fair 'n' Balanced Network:

Here's the lede, with a link to the story:

NEW YORK — Shoppers at Wal-Mart stores across America are loading carts with merchandise — maybe a flat-screen TV, a few DVDs and a six pack of beer — and strolling out without paying. Employees also are helping themselves to goods they haven't paid for.

Your questions:

1) Is the hed true?

2) If not, what would be a better figure to use?

3) Could anybody at Fox get a clue if they walked through a clue bar at closing time with $100 bills hanging out of their pockets?

Post answers in replies, pls.

When zombies attack

This just in from the Grauniad, courtesy of Triangle buro chief Strayhorn:

Police shoot dead woman
waving gun at officers in car park

Hed split as reported on Guardian Unltd. Wouldn't be right to make that sort of thing up, would it?

Short attention span funnies

Here's the AP, spreading multiculturalism and contextitude all across the world:

MIAMI -- Defense attorneys in the Jose Padilla terrorism support trial are going to great lengths to suggest to jurors that jihad is not necessarily Muslim holy war and that mujahedeen could just as easily be freedom fighters as terrorists.

Seems a bit like card-stacking to suggest that one side is "going to great lengths to suggest" that, um, the sun rises in the East. The AP's own holy writ (in an update that dates to the 2002 edition) describes "jihad" thus:

Arabic noun used to refer to the Islamic concept of the struggle to do good. In particular situations, that can include holy war.

But it's a lot to ask of the AP to read its own damn stylebook, let alone remember stuff from its freshman class in world religions.

The next point is even more fun, because if anyone has "gone to great lengths to suggest" that "mujahedeen could just as easily be freedom fighters as terrorists," the AP needs to stand up and take some credit:

Reagan said failure of the Soviet occupation was "due to the spirit and will of the majority of the Afghan people, and to the mujahedin, the freedom fighters, who continue to resist the Soviet invaders." He said the freedom fighters' "forces and their will remain intact." (12/26/82)

Hamid Navid of the Washington-based Islamic Unity of Afghanistan Mujahedeen (Freedom Fighters) told the demonstrators over a loudspeaker that Afghans "will fight until the last Russian is expelled from our country." (12/27/84)

Eashan Jan Areef, a Washington-based spokesman for one of the guerrilla groups, called Shevardnadze's comments "another Soviet trick. They want to weaken the ranks of the mujahedin (Moslem freedom fighters)." (1/7/87)

The point, of course, is that "mujahedeen" doesn't mean either "terrorists" or "freedom fighters." It means "people who j-h-d" (the mu- prefix does such to lots of verbs; it's one of the cool things about Arabic nounage). Whether those folks are "terrorists" depends less on whether they're blowing stuff up than on whose stuff they're blowing up. The "freedom fighter" bit is primarily the doing of the mendacious Republicans who ran the executive branch in those days, but the AP was a willing co-conspirator.

Sermon topic? Seems it'd be nice of "objective" journalism to point out that languages aren't scary. It's people who do awful things that are scary. Arabic doesn't have a secret stash of words for "snow" or "camel" or "religious death sentence." Those are the inventions of credulous people who ought to have better stuff to be scared of. They should be enlightened, then hit gently on the head with a stick.

And if the AP's that hard up for editors and writers who can find words in a dictionary that goes backwards and is arranged by consonantal roots, maybe it can hire some of the Arabic-speakers Your Pentagon seems to consider superfluous in these peaceable times.

Are we there yet?

Just a quick breakfast treat from America's Newspapers.

Another day on Planet Fox

Check out the No. 3 tale on display at the Fair 'n' Balanced Network's Web site:

Why does a story like rate a mention on the same page as a missing teenager and another missing teenager? It's not even about American baby names -- and, if you were wondering, it really doesn't have a thing in the world to do with the Diana documentary in the reefer. (It's not a "poll" either, of course, but that's just Fox not bothering to read the text before writing the teasers.) It's there for one reason only, which you've probably figured out by now:

THEY are coming RIGHT AT our baby names!!!!!!!!!!

It must be nice to have a corporate stablemate like The Times o'London, though. You have a free source of international news -- some of it actually true -- from a paper that still has a sheen of respectability from its pre-Murdoch days. And you get that charming sense of numerical accuracy that plagues all newspapers, even the ones that don't have mythical creatures in their nameplates. Here's the Times itself:

Muhammad is now second only to Jack as the most popular name for baby boys in Britain and is likely to rise to No 1 by next year, a study by The Times has found.

That's a traditional shell game for doctoring ledes. The "study" counts up a dozen or so likely transliterations among names provided by the Office of National Statistics. Hence the first finding. The second ("and is likely to rise to No 1...") is more interesting, because it appears to have no factual basis at all. Here's the fourth graf:

Although the official names register places the spelling Mohammed at No 23, an analysis of the top 3,000 names provided by the Office for National Statistics (ONS) puts Muhammad at No 2 once the 14 spellings are taken into account. If its popularity continues – it rose by 12 per cent last year – the name will take the top spot by the end of this year.

Wow. Wonder where that comes from? Could it be the table at the end?

1 Jack 6,928, 2 Muhammad (all spellings) 5,991, 3 Thomas 5,921, 4 Joshua 5,808, 5 Oliver 5,208

So how's it going to do that? If Jack stands still and Muhammad increases at another 12% this year, Jack is still going to be 200+ brats ahead. Muhammad's going to have to rise at about 15% to catch up. Does that mean the rate of increase in Muhammad is going to grow 25% a year? So five years out, there will be twice as many Muhammads as Jacks each year? And eventually, British tinies will have to start naming their teddy bears Muhammad just to keep up with the overflow?

Good old Fox. Always there to remind you the world isn't as safe as you thought.

It's already overcrowded ...

Here's an early front-runner for Annoying, Stupid Hed of the (still-young) Month:

Two offenses here. First, question heds are never to be used to soften or attribute assertions. Under rare circumstances, they can be considered for actual questions. Here's how you can tell the difference: "Paper or plastic?" is a question. "X is too Y" is an assertion. Got it?

Now for the major offense. Journalism's claim to being "objective" rests on its ability to stay grounded in the empirical world. We talk about stuff we can measure -- hence the ban on heds that talk about miracles and on photos of deities manifesting him/her/it/themselves on foodstuffs -- and we're open about the scales we use for measuring it. We try to stick with valid and reliable measures, so skeptics or critics have a good chance of getting the same answers.

OK? Let's proceed to the text.

For years, nine flags have flapped in the breeze atop Steamer's Sports Pub in east Charlotte. That changed Sunday afternoon when owner Bill Nolan climbed onto the roof and took down six American flags.

Nolan was cited May 18 for violating a city ordinance that allows his business to fly only three flags. After learning of the $350 appeal fee, Nolan decided to give in for fear of losing the bar's certificate of occupancy, he said.

See the problem? "Flags on roof of bar" is not a valid or reliable index of patriotism. It's a measure of how many flags you have on the roof of your bar and nothing else. Asking whether the display is "too patriotic" assumes that increases in patriotism are somehow indexed by increases in the amount of multicolored cloth you wave around, and that's the sort of deliberately offensive, knuckle-dragging notion that ought to be squeezed out of people before they get to grade school.

Longtime readers of the fishwrap in question have always been a bit surprised when readers accuse it of being part of the Librul Media Conspiracy. This sort of race-to-the-bottom pandering makes pretty clear where its sympathies lie. Perhaps, next time somebody deems Flags On Bar Roof to be a 1A story, the desk can appoint a designated grownup to give it the sort of hed it deserves.

(apple+orange)/2 = grapes!

Preemptive warning, in case any readers out there are tempted down the Path of Sin by the comrades at Real Clear Politics. What's wrong with this picture?


Exactly. The top line, or "RCP average." Quick, what's the average of 34% (among 600 likely voters), 27% (among 500 registered voters), and 29% (among some unspecified number of registered voters)? Hint: You can throw in 41% (among who knows how many of what) and toss in 31% (among 801 likely voters) for good measure and you still won't get "30.0."

Well, that's not entirely fair. You can call it "30.0%" all you want. But you might as well call it the Loch Ness Monster, the wreck of the Edmund Fitzgerald or the piece of cod which passeth all understanding, all of which bear the exact same resemblance to the original data as the alleged result presented as the RCP average.

The average of apple and orange is not "grapes." You can't get an average of unlike things. If someone tries to smuggle this bit of nonsense past the vigilant checkpoint that is your copydesk, take it out and shoot it.

Shortest off-season yet

It seems as though the Silly World Series was just last week, and here we are in the peak of silly season again. Where does the time go?

In other words, even though it's way too early in the season to do so, it's already time to complain about inept reporting of public opinion polls. Step forward, The State of Columbia!

Clinton, Giuliani lead in S.C., poll says
Hillary Clinton and Rudy Giuliani lead the races for the 2008 S.C. Democratic and Republican primaries, according to a new poll from Winthrop University and ETV.

Problematic conclusions for a bunch of reasons, only one of which the story bothers to mention, and it gets that one wrong. So let's reiterate the basic rules of reporting on polls:
1) The mainbar must include the minimum details needed for a marginally smart reader to judge the poll's reliability: Population sampled, sample size, dates poll was conducted, confidence interval (aka "margin of sampling error") and confidence level. If you kick the story back to cityside for these details and the assigning editor says "What's a confidence level?" -- well, what would you do with a sports editor who couldn't tell you what an earned-run average was?
2) Never draw conclusions that go beyond your data. A survey of registered voters eight months before primary season starts tells you what registered voters say eight months before primary season starts. It doesn't, and can't, address questions of who has cause to worry or who's in a tougher fight than whom.
3) Don't use language that goes beyond your data.
4) Doublecheck all the arithmetic. Then doublecheck it again.

OK, how about some details?

Clinton, a U.S. senator from New York, has an almost 10-percentage-point lead over her closest Democratic competitor, U.S. Sen. Barack Obama, D-Ill. However, former New York Mayor Giuliani is in a much closer fight among Republicans.

While you're at it, particularly when the story holds to the front and reefers to details inside: Put all the relevant numbers in the mainbar too. Don't make me go hunting for the chart. According to which, by the way, Clinton is less than 8.5 points ahead of Obama.

Still, many Democratic and Republican voters remain uncommitted seven months before the first 2008 ballot is cast in South Carolina. The poll found that about 30 percent of voters in each party - and a similar percentage of independents - remain undecided.

I'm hunting for the damn chart again, and it says the poll ended May 27 -- eight months before the S.C. Democratic primary (assuming the paper got it right last Sunday). And the proportion of undecided voters is a hint of how risky it is to talk about who's leading anything at this point. (Yes, we're also still waiting to be told it's a poll of registered voters.)

The poll also showed Giuliani has more reason for concern. His lead over U.S. Sen. John McCaincq of Arizona - 18.6 percent to 14.4 percent - is within the poll’s margin of error.

One, the poll doesn't show anything about who has reason for concern. That isn't what it measures. Two, let's just go ahead and retire the phrase "within the poll's margin of error" (which, ahem, sends me hunting for the chart again). It's irrelevant here. Three -- gotta love that cq, huh?

...As for the front-runners, the poll shows Clinton dominates Edwards’ native state, a state where her campaign organization is dwarfed by those of candidates with a fraction of her polling percentage.

"Dominates" is an opinion, not a fact, and a poorly based one at that. Clinton is at about the level with "undecided" among registered Democrats eight months before the primary.

And Giuliani, whether he’s in first or behind McCain - uncertain given the margin of error - remains in shouting distance of victory in a state where many Republican voters would fall opposite him on social issues.

Everybody talks about this "margin of error," but nobody bothers to tell us what it is or why it might matter. Let's review for a moment, then. The margin of sampling error* describes the band in which nonchance cases can be expected to fall at a given confidence level. Using the general standard of 95% confidence, 19 samples out of 20 will be within plus or minus the margin (3.8 percentage points for the whole sample here). So there's one chance in 20 that your sample is outside this range and the real population figure -- which we're trying to estimate -- is something way different from what our poll predicts.

Why insist on the confidence level? Because the margin is meaningless without it. You can lower the margin of sampling error to 1.9 points with a flick of your wrist. All you need to do is accept a one-in-three chance that you're wrong.

Here's how it's explained (err...) in the methodology blurb:

The margin of error ranges from plus or minus 3.79 percent to 6.01 percent.

Why the range? The size of the group involved in answering each question.

Consider this example: Of all S.C. voters who responded, 53.4 percent said the war in Iraq is the most important issue facing the country. That result has a margin of error of only 3.79 percent because of the 670 people surveyed, meaning as few as 49.61 percent of the sample or as much as 57.19 percent could believe the war is the most important issue.

Uh, sort of. Those are differences of about 7%, even though it's 3.8 percentage points either way. And we know exactly how much of the sample (53.4%) believes the war is the most important problem; it's the figure for the population we don't know. And there's still a 5% chance that the population figure lies outside those bounds. (See what we meant about confidence level?)

That's why "within the margin of error" is meaningless for describing a candidate's lead. Sampling error applies to each candidate's result, meaning you have to double the margin -- so candidate A's lower bound is above candidate B's upper bound -- to be sure at your confidence level that the difference is real. (How likely a difference is to reflect a real lead, though not necessarily the one the poll indicates, is a different matter.)

However, the survey also found 61.9 percent of Democrats and 50.8 percent of Republicans named the war as most important. But each of those results came from groups smaller than the 670 surveyed, giving a higher margin of error — 6.01 percent, meaning as few as 55.89 percent or as many as 67.91 percent of Democrats believe the war is the most important issue.

Now we're screwing up the arithmetic. These samples are different sizes: The poll had about 25% Democrats and about 40% Republicans. The maximum margin of sampling error at 95% confidence for Democrats on this question is about 7.6 percentage points. It's slightly smaller for this question and for the presidential preference question (about 6.9 points, given the gap between Hillary and not-Hillary votes), but it gives you an idea of how quickly a shrinking sample size can change the band in which your nonchance results operate.

This is an interesting poll, but that doesn't mean it's exciting, which is a large part of the problem. The writer's going for drama where there isn't any. One effect of that (well, that and saying 8.4 points is "almost 10") is an impression that the paper is taking sides -- that it's plumping for Clinton on the news pages. That's not a good impression to give.

Let's pretend it's spring training and practice doing things right. Hit your cutoffs now and you're more likely to hit them in the pennant race.

* Please use its full name on first reference. Surveys are subject to a number of different types of error, and sampling is only one of them.

Hed grammar lesson

Ever scratch your head and wonder why you got a handout in J4400 that said: "If your hed's main verb comes from a relative clause, you wrote the wrong hed"? Here's what it looks like in real life.

This isn't the sort of "bad" grammar that violates silly prescriptive rules nobody believes in. Nor the sort of out-and-out miswiring that changes the meaning around. As heds go, this one is wired together perfectly, and unlike some heds it has the distinct advantage of being true in the bargain. What makes it bad grammar is that it uses the wrong wiring diagram:

5 from 82nd Airborne killed in crash
RALEIGH, N.C. --Five soldiers who died this week when a CH-47 Chinook helicopter crashed in Afghanistan were members of the 82nd Airborne Division, officials said Friday.

This is the diagram for a first-day hed. "(Are) killed" is the verb in the main clause. The subject is "5" and the "from 82nd Airborne" tells you which "5." It would have been a fine hed online on Wednesday, when the crash happened, or in print on Thursday.

But by now, "5 killed" is old news. The news is the victims were with the 82nd Airborne. That means "killed in crash" needs to become a relative clause (as in the lede, more or less) and move leftward to modify "5." We need to fill in the equational verb, to signal that we're back to the main clause and not in an endless loop of dog-who-chased-the-cat-who-killed-the-rat clauses. So we get:

5 killed in crash were from 82nd Airborne
There's still a lot of room for improvement. From the hed, there's no way to tell this wasn't a traffic accident. "Afghan crash victims" or "copter crash victims" would be a start. But at least it gets stuff in the right order.

Moral: Grammar makes meaning. If you get a meaning that doesn't correspond with your intent, you have the wrong grammar -- even if it's otherwise flawless.