Q & A steps up screening on tweets

Q & A has stepped up its moderation of tweets following last week’s #AbbottLovesAnal stuff-up.

Producer Peter McEvoy told Media Watch the show has a three stage electronic moderation and 2-3 human moderators, but will increase the level of electronic filtering and add a second senior producer at the point of final moderation.

But he rejected the idea of dropping the tweets altogether, telling the programme the number of mishap tweets was extremely low amongst 22,000 published tweets since it introduced the #qanda hashtag.

“You don’t cancel test cricket because a few streakers run on the field and you don’t shut down the live #qanda twitter conversation because an inappropriate twitter handle gets past the moderation controls,” he advised.

Here is the full response from Q & A to Media Watch:

1. Do you have any computer filtering systems set up to scan for obscenities in tweets?
Yes we have a three stage moderating process using electronic filtering of offensive language etc and then two stages of human moderation.

Does it also screen Twitter handles? If not, why not
In the past we have not used electronic filtering on handles because it would be ineffective (see your examples) and would unfairly exclude legitimate names e.g. “McManus”

2. Are enough staff involved in checking the content of tweets and Twitter handles?
We have two stages of human moderation with generally 2-3 moderators each week at the frontline selecting comments which meet the Q&A guidelines from the filtered tweets. Those selected tweets are then passed on to another senior producer who chooses the tweets that actually go on screen. This system has worked well over a number of years with only a very small number of errors amongst the 22,000 tweets that have gone on screen.

3. Are changes now being made to how you manage the broadcasting of tweets?
Because Q&A has been targeted for deliberate attack by people trying to break through the moderation system we have decided to increase the level of electronic filtering and add a second senior producer at the point of final moderation.

4. Inappropriate Twitter handles have made it to air on Q & A in past years (for example, Seymour Cox in 2014, Smellmikehunt in 2012) Does this week’s offensive twitter handle involving the Prime Minister demonstrate you’re unable to manage live tweets and they should be dropped?
Q&A has created a public space that invites Australian citizens to question political leaders and express and share their views on politics and other issues. That process inevitably involves risks. Q&A has had 75,000 audience members asking around 2,500 questions. Since the program introduced the #qanda twitter conversation it has generated over 5 million twitter comments from almost 400,000 users and published 22,000 twitter comments onscreen. In that context the number of disruptive incidents is very, very small. You don’t cancel test cricket because a few streakers run on the field and you don’t shut down the live #qanda twitter conversation because an inappropriate twitter handle gets past the moderation controls.