think about it
Your cart is empty
Visit The Shop

are your tweets sexist?

Image: Johannes Jansson

Image: Johannes Jansson

‘Wow, this user got the equality of retweeting almost down to an art,’ says Twee-Q, a website that analyses your retweets on Twitter based on the gender of the original tweeter.

Boy, am I relieved! It definitely ranked high on my list of concerns. In fact, it comes in third, just after the price of avocado and whether Benson is going to die on next week’s SVU, but before paying my phone bill and making deadline at work.

Okay, I may be taking the piss a little bit, but I find the whole idea of Twee-Q borderline ridiculous.

Twee-Q stands for ‘Twitter Equality Quotient’. Basically, you head over to the site, pop in any Twitter handle you like, and it analyses the retweets out of the last 100 tweets and sees how many were from male users and how many from female.

It’s the result of partnership between Joan Smith, a British writer and activist, and Crossing Borders, a Swedish organisation that promotes equal rights and opportunities. Apparently the site marks ‘the publication of her new book on contemporary gender roles and equality, The Public Woman’, by discovering how equal social media conversation is.

They say:

Gender equality begins in the conversation, when people who have something to say communicate with people who are interested to listen, who are curious and attentive, and who are not afraid of change. When discussion and reflection leads to action. That’s when the magic happens. But how equal is a conversation? What if the core of the conversation is unequal? What if we rather listen to, acknowledge and pass on opinions or thoughts from a particular gender? Simply put: what if we generally evaluate the arguments of a particular sex higher, perhaps without even knowing it ourselves? Well, in that case the conversations are broken.

I have a few gripes with this. To explain a couple, we’ll have a look at my Twee-Q score.

@melcomber scored 8.6 out of 10. The closer to 10 you are, the less sexist you are. 54% of my retweets were from men, and 46% were from women. Who do I mainly retweet? In the last 100 I’ve retweeted quite a few comments on marriage equality and the NDIS, but before that I’ve spent a hell of a lot more time retweeting a bookstore in England.

That’s one gripe. I have 1,810 tweets. Twee-Q is analysing a mere 18.1% of them. That’s not very many. It’s also not very many if I were to examine someone like Justin Bieber (@justinbieber) who has 22,000 tweets (who scored a 4.6, in case you care). What about the people who have hundreds of thousands of tweets?

Also, how you can assign gender to an organisation? It has been a very long time since I signed up to Twitter, so I’m not entirely sure of the information required at registration anymore, but gender is no longer displayed on profiles. So does Twee-Q scour an account’s profile information to determine gender? If a man initially created a company profile, does Twee-Q then assume that every tweet is from a male staff member? They don’t say.

Am I digging too far into this? Probably. That’s not really my main issue anyway.

What annoys me is the fact that Twee-Q is assuming that an opinion is assigned value based on the gender of the tweeter, rather than the content of the tweet.

Out of my last hundred, I didn’t retweet @TriciaMorosin, @LindaDrummond or @JuliaGillard because they have vaginas. I retweeted them because they said something that was worth sharing. Same goes for any man, company, organisation or fake banana (no, really, a fake banana, @dermottbanana).

Honestly, I could rant all day about this major bug-bear of mine, which I expect people to disagree with. Should a woman be given a job over man to improve the appearance of a company, or should the best person be chosen for the job? Should a woman win an award because a woman hasn’t won it for a while, or should the person who simply deserves it the most be chosen? Should a woman or a man be retweeted more for the appearance of a balanced conversation, regardless of the content of their tweet?

Also, where does it stop? While correcting my retweets for their gender imbalance, should I make sure I retweet more Asian people, more lesbians, more of the elderly?

Okay, okay, I’ve gone over the top here, but at least I can be certain I have illustrated my point, that the whole notion of Twee-Q is over the top.

Oh, and I hate to break it to them, but women do have equality in conversation on Twitter – they have just as much opportunity to be retweeted as men do. I think the word Twee-Q is after is actually equity.

Semantics aside, at least I will sleep tonight knowing that an analytical tool has congratulated me for my completely intentional gender balanced tweeting. Unfortunately Lip’s own Jo Mandarano won’t, saying to me yesterday, ‘Well, with a 4.4, I’m obviously a raging sexist.’

One thought on “are your tweets sexist?

  1. “What annoys me is the fact that Twee-Q is assuming that an opinion is assigned value based on the gender of the tweeter, rather than the content of the tweet.” <- I can see where you're coming from, but it strikes me that this is based largely on technological limitations: it would require incredibly deep AI to analyse the content of every retweet in order to assess bias, and Twee-Q is really just a quick hack. It's an interesting experiment (that really just functions as cheap marketing from Crossing Boarders) that probably shouldn't be taken too seriously. It reminds me a lot of the "Gender Guesser" tool, which attempts to analyse a user's gender based on a short sample of writing (http://www.hackerfactor.com/GenderGuesser.php). It usually works, but it's a gimmick.

    (Also worth noting: if you click "FAQ" down the bottom of the Twee-Q page, it explains how it assigns a gender to a Twitter account. I think this is where it's really flawed: the system will only recognise retweets from names that match its limited list.)

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>