Archive — Bias

Groundbreaking report reveals racial bias in English football commentarySachin NakraniThe Guardian

Footballers' legs of different colours

A very interesting-sounding study has analysed English language football commentary from seven broadcasters and 80 live matches.

RunRepeat ratio-adjusted its numbers to account for the fact there were 1,361 comments about lighter-skinned players and 713 about darker-skinned players and found the former group more widely praised for intelligence (62.60%), hard work (60.40%) and quality (62.79%). Commentators are also 6.59 times more likely to talk about the power of a player if he has darker skin and 3.38 times more likely to reference his pace.

The study also found that 63.33% of criticism from commentators in regards to the intelligence of a player is aimed at those with darker skin, while the figure for quality is 67.57%.

Comment

Microsoft’s robot editor confuses mixed-race Little Mix singersJim WatersonThe Guardian

Jade Thirlwall and Leigh-Anne Pinnock from Little Mix

How about this for dystopia? MSN have replaced human news editors with a robot powered by Microsoft artificial intelligence technology. The problem is, it has already begun making racist decisions.

And then, in case you thought the story wasn’t already absurd enough, this:

In advance of the publication of this article, staff at MSN were told to expect a negative article in the Guardian about alleged racist bias in the artificial intelligence software that will soon take their jobs.

Because they are unable to stop the new robot editor selecting stories from external news sites such as the Guardian, the remaining human staff have been told to stay alert and delete a version of this article if the robot decides it is of interest and automatically publishes it on MSN.com. They have also been warned that even if they delete it, the robot editor may overrule them and attempt to publish it again.

Then the article ends on a delicious snippet — that Microsoft itself is concerned about the reputational damage this scheme will cause to its AI technology.

I’m immediately reminded of Microsoft’s disastrous Tay experiment.

Comment

Academia uneasy with lack of diversity on ‘dude walls’ of honourNell GreenfieldboyceNPR

A 'dude wall' featuring several portraits of white old men

On the campaign to remove — or make less prominent — walls of portraits of old white men from academia’s past.

“It just sends the message, every day when you walk by it, that science consists of old white men,” says [neuroscientist Leslie] Vosshall. “I think every institution needs to go out into the hallway and ask, ‘What kind of message are we sending with these oil portraits and dusty old photographs?'”

While defenders of dude walls warn of erasing history, the counterpoint is powerful:

…some argue that the old portraits themselves have erased history, by glorifying white men who hold power while ignoring the contributions to science and medicine made by women and people of colour.

Celebrations of individuals in this way always make me wary. It seems to be particularly common in higher education, where awards and buildings are routinely named after white males.

But very few breakthroughs are truly the work of a single individual. The people honoured in this way are likely the people most adept at taking all the credit for other people’s work.

Comment

Could Brexit break the BBC? The tensions, the bewildering question of ‘balance’ — and how to get it right — Mark Damazer, Prospect Magazine

Could Brexit break the BBC? The tensions, the bewildering question of ‘balance’ — and how to get it right — Mark Damazer, Prospect Magazine

An impressively thoughtful piece from the former Radio 4 controller, on why the BBC is struggling to remain unbiased amid Brexit.

One senior presenter put it like this: “We should encourage debate… while being more militant about our core approach—that we are fact-based, and question and test all sides of the debate. We should not be doing vanilla ‘on the one hand’ versus ‘on the other hand’ journalism. I am sympathetic to the arguments about the danger of ‘false equivalence,’ and think we should be clear about the weight of arguments. But if a substantial number of people believe, so to speak, that bananas are blue we have to treat that seriously. Seriously, but robustly.”

This article also briefly covers some of the limitations of TV news bulletins, and explains why in some aspects radio performs better. I do find it difficult to watch a bulletin like the 10 O’Clock News (I think I even watched the piece he mentions from Mansfield, with my head in my hands). In that format, it is impossible to cover anything in real depth — and that seems to be the true problem at the moment.

Comment

Why data is never raw — Nick Barrowman, the New Atlantis

A man checking dials

Why data is never raw — Nick Barrowman, the New Atlantis

Why it’s wrong and perhaps even dangerous to expect “raw data” to be neutral and strictly factual.

How data are construed, recorded, and collected is the result of human decisions — decisions about what exactly to measure, when and where to do so, and by what methods. Inevitably, what gets measured and recorded has an impact on the conclusions that are drawn.

1 comment