Tag Archives: accuracy studies

Speed versus accuracy in journalism: towards a new debate

Today’s edition of my weekly column in Columbia Journalism Review looks at the issue of speed versus accuracy in journalism. I hope you’ll take a moment and read it, as it relates to this post. Think of the column and post as branches on the same tree.

My column looks at the issue in terms of the consequences of rushing out stories in today’s media environment, and why scoops aren’t what they used to be. This post examines the value of speed, and how it should be weighed against accuracy.

I confess that I started working on this several months ago, when New York Times public editor Clark Hoyt wrote a column that struck a familiar chord: the rush to publish news had apparently led to some inadequate reporting and sourcing in the paper.

It was a good column, the kind of inside look at how a controversial story unfolded that you’d expect a public editor to provide. I found myself agreeing with many of the points he made. But then Jon Landman, then the deputy managing editor in charge of the Times’ online operations, responded to Hoyt and expressed a different view of how the paper handled the story. Most important, Landman gave one of the best expressions I’ve read of the value of speed in journalism. He didn’t argue against accuracy; he simply said that the two need not always be seen as enemies.

Here’s how Landman began his third paragraph:

Of course working fast increases the chance of error and clearly that is a danger to acknowledge seriously and address carefully. But absence of error isn’t the only value. If it was, we’d long ago have scrapped daily and weekly newspapers and magazines in favor of refereed scholarly journals. Speed is a value too.

It’s rare to see a newspaper editor deal so frankly with the issue of accuracy, to not trot out the old “accuracy is one of our most important values” line. The truth is that, while accuracy is valued in journalism, it’s often subjugated in favor of other values. As Landman writes, speed can trump accuracy in the minds of editors. It happens all the time. We shouldn’t pretend that accuracy is always the most important value when it comes to the actual practices of a newsroom. (We also shouldn’t forget, as I note in my CJR column, that the difference between making an error and getting it right is often a matter of making one or two phone calls. Accuracy is often easier and faster to achieve than we think.)

Let me be clear that I’m not suggesting Landman doesn’t care about accuracy, or that he’s advocating ignoring it as a standard operating procedure. He’s simply stating the reality of how journalism works: accuracy isn’t always the number one concern. There is ample evidence to back this up.

Landman isn’t alone in pointing this out. Philip Meyer, one of the most important journalism thinkers/academics of the last couple of decades, made a similar case in his book, The Vanishing Newspaper. He wrote:

A newspaper with a zero level of factual errors is a newspaper that is missing deadlines, taking too few risks, or both. The public, despite the alarms raised in [American Society of Newspaper Editors] studies, does not expect newspapers to be perfect. Neither do most of the sources quoted in the paper. The problem is finding the right balance between speed and accuracy, between being comprehensive and being merely interesting.

There is a balancing act when it comes to certain elements of accuracy. It’s never okay to get someone’s name wrong, or to make a mistake about an easily verifiable factual error. I’m sure both Hoyt and Landman would agree with that. But, in some cases, editors have to make a call about whether they have all the facts, not just the right ones. That’s the kind of thing Meyer is referring to, and, I suspect, so was Landman. These calls have been made for decades, if not centuries, so they aren’t new to the online world. (What is new are the consequences of an incorrect report.)

When it comes to the online environment, Landman argues that it enables a media organization to improve the accuracy process, rather than degrade it:

When the reporting process plays out in public, that’s a good thing. Readers can and do participate. Their participation has a salutary effect on quality — millions of amateur editors catch a lot that a few professional ones miss. And the process of constant checks on the unfolding story produce incentives to keep pushing. In the Kennedy-Paterson story, the never-ending news cycle ultimately contributed to a good result — a story that got to the bottom of the strange back-and-forth between the Paterson and Kennedy camps, sorting facts from rumor and accusation.

This is akin to Jeff Jarvis’ mantra of “publish and correct.” Here’s what Jarvis wrote in a 2006 Guardian column:

We need to recognise that the internet alters how media operate. Blogs – whether written by professionals or amateurs – tend to publish first and edit later, which can work because the audience will edit you. In this medium, stories are never done; rather than turning into fish-wrap, they can grow and become more factual and gather new perspectives, thanks to the power of the link and, yes, the correction.

We all make mistakes. We’re human. And the internet makes our humanity more apparent than polished print and broadcast do.

I have to admit that even though he wrote the (excellent) foreword to the Regret the Error book, this philosophy always made me a bit uncomfortable. I suspect Hoyt may feel the same way. Why not wait 15 minutes or even an hour if it means getting the entire story right, rather than just most of it? When I have this internal argument with myself, I reply to that question by noting that readers can — and often do — spot things that journalists wouldn’t realize even if they waited all day before publishing. So there’s value in getting it out there. Accuracy is not always an absolute. I hate having to write that, but it’s true. Some things are non-negotiable, but others have shades of grey.

This is why journalists need to at least take a few moments and think about why they’re publishing something, and if their news values – speed, accuracy, and otherwise — are in proper alignment.

That’s one process we should never sacrifice for the sake of speed.

Regret the links

In case you hadn’t noticed, I recently added a “What I’m Reading” sidebar to the site. It’s over there to the right. I link to relevant articles that I find interesting or of note. And now, every once in a while, I’ll post a round-up of some of those links to make sure you don’t miss out on them. They’re all worth checking out. Enjoy.

Press Accuracy Rating Hits Two-Decade Low – Pew Research Center
pewresearch.org | September 14, 2009
More bad news re: trust and accuracy.

DISPUTATIONS: Spy Games | The New Republic
The New Republic | September 16, 2009
Victor Navasky demands satisfaction from The New Republic.

FACTS, ERRORS AND THE KINDLE | More Intelligent Life
moreintelligentlife.com | September 4, 2009
I’m interviewed in this Economist story about book errors and corrections.

Accidental headline of the year | Media Monkey | Media | guardian.co.uk
Guardian | September 1, 2009

A horror story involving the correction of a published scientific article – Statistical Modeling, Causal Inference, and Social Science
stat.columbia.edu | August 25, 2009


“In a way it is surprising that we do not make more mistakes.”

That’s a line from a blog post by Guardian subeditor (copy editor) David Marsh. It’s long been a common refrain from journalists, especially editors. In fact, Mitchell V. Charnley said basically the same thing in the introduction to his 1936 study of newspaper accuracy, the first of its kind.

“As common as the layman’s superficial generalization that ‘the newspaper is always wrong’ is the newspaper man’s defense that the wonder is that so few errors get into print,” he wrote in an academic article published in 1936. Charnley also noted the “appalling opportunities for error in the smallest story.”

Marsh marshaled that maxim while explaining that today’s copy editors have to juggle laying out stories, editing copy, and writing headlines and photo captions. Other tasks are often piled on top of those. Now add the fact that many newspapers have reduced the ranks of copy editors, and you have an environment that’s ripe for error.

“One of our best subs, taken to task this morning for what I described as the unforgivable crime of putting an acute accent on the artist Edgar Degas’ surname in last week’s paper, held his hand up to the offence but pointed out that he had been working on seven different pages under severe time pressure,” Marsh writes. “Doubtless he had corrected many mistakes but the one he missed was, of course, what everyone noticed.”

Marsh chose to tackle this topic because of an old chestnut offered up by readers — the idea that the quality of newspaper writing, editing and accuracy is far worse than it was decades ago. Call it the Golden Age of Newspapers Gripe. Marsh writes that “like most golden ages, this one was entirely mythical.” His evidence is this astoundingly bad paragraph from an edition of the Guardian published 45 years ago:

The Republican National Comittee decided in the spring that its chances of the White House in 1964 would be very slim indeed if it did not capture California, the second largest state, in 1962. Nobody less than its strongest possible vote-getter would do to defeat the incumbent Governor, Edmund (Pat) Brown. When it said this, Mr Nion was looking towards Washington, but the committee was liiking at Mr Nixon. He would have to oick the candidate, and if he oicked another man, eho lost, the party would be loth to nominate for the Preidency a national leader whose influence could not carry his own state in a state election. Yet, if Mr Noxon ran himself and won, he would practiclly forsweat the presidency; for, like allaspiring governors, he has been bocal and bitter about men who use the governor’s mansion as a springboard int the White House.”

The findings of newspaper accuracy studies have been relatively consistent since Charnley’s 1936 report. Most have found an error rate of between 40 and 60 percent, meaning roughly half of all newspaper news stories in U.S. newspapers have some kind of error, be it factual or one of a more subjective nature. (These studies gather data by asking sources to fill out questionnaires about an article in which they were featured.)

But it should be noted that the most recent — and comprehensive – study found one of the highest error rates on record. Also, the recent rash of cuts and buyouts in American newsrooms, coupled with the demands of producing an online edition, have already offered at least one example of reduced quality.

Charnley was right about the many opportunities for error within any single article. That hasn’t changed. What has changed is that, inside many newspaper newsrooms, you have less fewer people doing more work. Copy editors, in particular, have become editors, proofreaders and designers. (And, as Marsh illustrates, bloggers.)

The danger today is that the Golden Age of Newspapers Gripe may soon become reality, even if Marsh’s maxim remains true.

(With thanks to Andrew Phillips for spotting Marsh’s post.)