One of the long-standing accuracy debates in journalism centers around whether you should repeat the original error in a correction. I examined this issue two years ago, and I support repeating the mistake in order to ensure people understand the nature of the error.

That issue will likely never be settled, as different news organizations embrace different policies for different reasons. It would be a stretch to suggest that this is a raging debate, but it’s a precursor to what has emerged as the urgent accuracy debate of the moment.

This past weekend, several major media organizations wrongly reported and tweeted that Rep. Gabrielle Giffords had died in the Saturday shooting incident in Tucson. (For a full timeline of the incorrect tweets and the discussion around them, see this post on RegretTheError.com; I also collected notable after-the-fact commentary in this post.)

Put simply, the question is this: To delete or not to delete? Should news organizations and reporters delete tweets after it becomes clear they’re in error? In the wake of incorrect tweets on Saturday, different news organizations and reporters came down on different side of this debate. Below is a summary of the arguments offered for and against deletion, along with an outline of a proposed Twitter function that could help enable better corrections on the service.

Differing Opinions

According to some accounts, NPR was the first news organization to report that Rep. Giffords had died. Its reporting was also shared on Twitter. Andy Carvin, who oversees social media and community at NPR, sent out the errant tweets, and he later explained why he chose not to delete them:

With around 2 million people following @nprnews and @nprpolitics, deleting the original tweet wouldn’t have altered the fact that many of those followers had already seen the mistaken tweet and retweeted it. So based on that reasoning, I decided to be transparent about the mistake and not try to hide it.

Was that a good or a bad decision? At the time I felt it was a reasonable decision, given the circumstances, and still feel comfortable with the decision. I can imagine if I had deleted it, we’d be reading news stories and blog posts today about NPR trying to cover our tracks on Twitter.

WBUR, a public radio station in Boston, also explained why it chose to keep the incorrect tweet intact:

We have decided NOT to delete the erroneous tweet, because it serves as part of the narrative of this story. Facts can change fast when news is breaking, and that leads to errors. We need to own the error, not hide from it. But we also need to rectify the error and explain ourselves to people who trust us. Deleting the tweet would do more to harm trust than perserving [sic] it would do to harm truth.

On the other side of the coin are CNN, Reuters, and the PBS NewsHour, all of which deleted their Giffords-has-died tweets. Steve Safran at Lost Remote summarized the arguments in favor of deletion:

One argument in favor is to stop or slow the retweeting. But this is difficult, if not impossible. And it is tempting but impractical to call for a squad of people to monitor tweets. For hours after it was reported she was alive, people kept discovering the original tweet that she was dead, retweeting it to their friends without seeing the update. In several cases, the retweet of the incorrect report came three or more hours after the report first spread …

I also heard from Teresa Gorman, who does social media and online engagement for the PBS NewsHour. She explained why they deleted:

@craigsilverman in all transparency, we did tweet NPR’s news, but I chose to delete it because it kept getting retweeted hours later.less than a minute ago via HootSuite

I weighed in on this question via Twitter:

I’m not in favor of news orgs deleting incorrect tweets. Take time to push out correct info & contact RTers to alert to new info.less than a minute ago via TweetDeck

So you know where I stand. In addition to the reasons cited by Carvin and WBUR, I’d add that the standard for fixing errors online is to add a correction to the same piece of content. (Scrubbing is unethical.) Since tweets are self contained, you can’t go back and add a correction to a message, which is part of the challenge of correcting information on Twitter.

That said, there are some emerging best practices, and you can read my previous column that offered four guidelines for pushing out a correction on Twitter.

As explained by Safran, the primary reason cited for deleting an incorrect tweet is that it can prevent people from continuing to retweet it. This is valid. Have a look at the number of retweets for NPR’s initial, incorrect report, according to Topsy. Using the same measure, less than half the number of people then retweeted the update/correction. It’s not an exact measurement, but Topsy used the same protocols for both tweets, which suggests there is an issue.

The same dynamic seemed to be in play when, last year, the @BreakingNews account sent a mistaken tweet about an Icelandic volcano eruption. On Twitter, it seems, people don’t retweet corrections as often as they tweet the initial, incorrect news. As I wrote in my previous column about correcting tweets, “Remember that when something is retweeted, it takes on more authority among people and search engines—so your job in issuing a Twitter correction is to get it retweeted as much as possible.”

One good general practice that I’ll add to my previous Twitter correction tips is that a news organization should make an effort to reach out to people who retweeted the incorrect information in order to make sure they pass along the new, correct information. We have a responsibility to follow up on our correction tweets and help give them the push and distribution they require.

Aside from that guidance—and the suggestion that news organizations never delete incorrect tweets—there’s a role for Twitter itself.

A Twitter Correction Function

My suggestion is for Twitter to enable a correction feature, much in the same way it created a retweet function. What would a useful Twitter correction feature look like? Here’s my suggestion, and please share your feedback in the comments:

User Controlled: First things first: It’s neither feasible nor desirable to have Twitter play a role in determining which tweets do or don’t deserve a correction. This feature has to be controlled by users and only policed by Twitter when there are abuses. Which means the system has to have proper safeguards to prevent abuse.

Notification, Not Exactly Correction: In my vision, the Twitter correction function would let the owner of an account notify all retweeters that a corrected tweet has been issued. Note that I’m not suggesting a user have the ability to force a correction to be retweeted on other people’s accounts. That would too easily lend itself to spam-like uses. So, for me, the solution is to enable someone to automatically send @replies to everyone who retweeted the initial, incorrect tweet in order to inform them of the correction. The function itself takes the form of targeted reply tweets.

Workflow: Here’s how it would look in action:

• User tweets “Mrs. Smith has won the election.”

• That tweet is retweeted by 150 users, using Twitter’s official retweet button.

• Thirty minutes later, it becomes clear the original tweet was incorrect.

• User goes back to their original, incorrect tweet and selects the “correction” button on Twitter.com (or their preferred client), which prompts them to write a corrected tweet. “Correction: Votes still being counted in Mrs. Smith race. She has not been elected.” User hits send.

• Users who retweeted the original tweet receive this reply message: “@username: Correction: Votes still being counted in Mrs. Smith race. She has not been elected. Pls RT”

• The above reply tweet would also be highlighted with a special correction icon, much the same way retweets are highlighted with the square arrows icon from Twitter. (May I be so bold as to suggest using the Report an Error Alliance icon? We’ll give it to Twitter to use for free.)

• Users who click the retweet button on the correction message will also send the message out with the correction icon, thus helping draw attention to it among their followers.

Concerns: There are of course implementation issues, not to mention the fact that we have no idea if Twitter has any interest in this proposal. Aside from those issues, one obvious concern with my proposal is that the correction function could still be abused. By retweeting someone you are opening yourself up to receiving what is in effect a mass message. So I’m still concerned about nasty folks turning this correction feature into a way to spam users. There’s also something of an educational challenge in teaching people how it works. Though I suppose the same can be said for Twitter as a whole. I’ve tried to explain hashtags and retweets to newbies, and it’s not as if those are the most natural things in the world, either.

Omission: One possible way to integrate corrections into Twitter would be to use the annotations feature. I left this out of my proposal because annotations are not yet live, and I’m probably not the best person to figure out how to make this happen. My hope is that annotations might enable you to create a “correction” attribute for a tweet that helps it attract attention as a correction. I believe annotations can be useful in helping create an effective corrections framework for Twitter, and would love to learn more about how that might work.

So that’s what I’ve come up with, faults and all. Help me improve it by adding your thoughts below.

Correction of the Week

“In a Jan. 6 story, The Associated Press reported CNN anchor Piers Morgan’s claim that his interview with then-British Prime Minister Gordon Brown turned the tide of Gordon’s election campaign in his favor. While Brown got a boost in the polls after the interview, the story should have noted that he subsequently lost the 2010 election to a Conservative Party-led coalition.” - Associated Press

Craig Silverman is the editor of RegretTheError.com and the author of Regret The Error: How Media Mistakes Pollute the Press and Imperil Free Speech. He is also the editorial director of OpenFile.ca and a columnist for the Toronto Star.