Slate’s good strategy for correcting errors on Twitter, elsewhere

On Saturday night, Slate made a very funny, embarrassing error on Twitter:

Javier Bardem and Vladimir Putin aren’t exactly lookalikes. It’s a funny mistake, and thanks to Twitter’s recent changes the mistaken image loomed large in people’s timelines. Then came the correction:

Slate social media editor Jeremy Stahl employed a simple but effective strategy: he issued the correction as a reply to the original tweet. That’s why the correction begins “@Slate,” and it’s why it refers to the photo without having to show it again. The result is anyone viewing the original tweet can see the resulting correction in the stream of replies:

People viewing the correction tweet on its own can also see it’s part of a conversation linked to the original, offending tweet. (If you reply to yourself, then anyone who follows you will see the reply. So it’s just as good as sending it as a normal tweet, in terms of who can see it.)

Stahl figured a simple way to link the corrective tweet to the original. If you also look at the retweet accounts shown in the above image, it seems that the reply strategy doesn’t hurt distribution at all.

‘Almost never’ delete a tweet

Stahl says he’s been doing it this way for about a year, and it’s working well.

“In the past year, I’d say, I’ve started correcting mistakes within tweets as replies,” he told me by email. “Prior to that, I was running separate correction tweets after making errors, but then I realized that the original incorrect tweets were still out there and if people didn’t see both tweets they’d miss the correction. This seems to be an effective way of not only issuing a correction, but indicating the error on the mistaken copy.”

Exactly: it uses Twitter’s improved integration of conversations in the timeline to place the correction in context.

If Twitter isn’t going to offer a specific correction feature, it’s up to the people who need it to find better ways to use all the things you can do with the platform. Hats off to Stahl for doing that.

As for how he ended up confusing Javier Bardem and Vladimir Putin, Stahl told me:

We had run a story on Javier Bardem that I used a photo for the previous day, and I believe my photo of Bardem was next to my photo of Putin in my files list. I just accidentally uploaded the wrong one and sent it out without realizing it.

Due to a miscommunication with Slate’s deputy social media editor, Stahl said the mistaken tweet actually ended up being sent again a day later. Though he “almost never” deletes a tweet, Stahl did in that case.

“My logic was that if the same thing happened in an article, say if somebody accidentally copied and pasted the same error into a story twice, I would think it would only be necessary to correct the mistaken sentence once and leave one version of the corrected sentence up,” he said.

Corrections on Facebook, the Web

Slate follows a similar process for corrections on Facebook: It adds the correction as a reply. But Stahl also noted that he views the ability to “Hide” a post on Facebook as being different than deleting a tweet.

“If you delete a tweet, you are putting it down the memory hole, removing it from the visible web,” he said. “But if you hide a Facebook post from your wall after correcting, that Facebook post still exists, just not on your wall. To me it would be be like removing a mistake from the top of our homepage, but leaving the article and the correction up on the web.”

The “memory hole” concept is core to how Slate approaches corrections, according to editor David Plotz.

“Generally speaking, we don’t want to memory hole our mistakes,” he told me by email. “When we make an error, our readers should know that we made it, and know what it was, and they should know that the instant we make the correction, and 10 years later, too.”

Along with the social media corrections discussed by Stahl, there are two other notable aspects to how Slate handles corrections. The first is it publishes a post every Friday that lists the corrections for the week.

It’s a great way to highlight the publication’s errors/corrections, as the post shows up as a story just like any other, as opposed to being ghettoized somewhere else on the site. Reading the weekly roundups also gives a sense of the kind of errors Slate makes on a regular basis. (No. 1 cause, as far as I can tell: misspelled names.)

The other interesting approach comes on the offending articles. Slate places an asterisk at the end of the sentence where the original error occurred. (See this example in the story that featured the now-famous pic of Javier Bardem.)

Click on the asterisk and you’re brought to the correction at the end of the article. It’s a great to way to show the context of a mistake.

“We’ve found that readers really appreciate it and we want to be as upfront as possible about our mistakes,” said Miriam Krule, the Slate copy editor who oversees corrections.

We have made it easy to comment on posts, however we require civility and encourage full names to that end (first initial, last name is OK). Please read our guidelines here before commenting.

  • melitagnm105

    my Aunty Grace got a nearly new blue Kia by working part
    time from the internet. look at this now B­u­z­z­3­4­.­ℂ­o­m

  • http://www.CraigSilverman.ca CraigSilverman

    Thanks for that comment and tip, Tim! I’ve tested this with TweetDeck and it appears to work there as well.

  • Tim Farley

    A little hack I use when replying to yourself like this on Twitter: When you are using the web interface (i.e. twitter.com) to compose the reply, you can delete the reference to your own handle and it will still be a reply. I also find this useful for threading multi-part tweets together.

    So far I’ve only seen this work on the web interface, I don’t know of a Twitter client that behaves this way. But I’m sure someone will correct me on that.