How to mislead with numbers
I’ve always been a stickler for what I call the reverse-stat. Whenever I read that 40 per cent of people believe that John Key is the greatest human being on Earth my immediate response is to flip it and go “well that means that 60 per cent don’t”. It’s a comforting way to get past some of the hype that we see from media around use of data.
Data is supposed to elevate journalism above opinion and provide some rigidity and integrity. But this is not always the case as the two examples I will talk about show.
Firstly, there is this brilliant deconstruction of the New York Times’ use of graphs on how it retold the story of the Gaza invasion by Israel.
https://medium.com/aj-news/not-fit-to-print-when-good-design-goes-bad-cc52931a2ce0
It’s a great read and well worth reading the entire thing. The key notes from it is that data does not make the reporting any more or less clear, it is merely another way of reporting (Keith Ng aside of course, that man is 100 per cent integrity). So if the New York Times falls victim to biases with data then who on earth COULD we trust?
Not KiwiBlog, that’s for sure. David’s done a post called “Inaugural Media Opinion Statistics” where he attempts to dissect the media’s opinion of National and Labour and answer once and for all, just how biased against National is the media?
http://www.kiwiblog.co.nz/2015/04/inaugural_media_opinion_statistics.html
David tries to be even handed with it. As he is wont to do. He includes a number of caveats:
- It only covers “opinion” columns and editorials. It does not cover news stories. It is designed to shed light on what the newspaper or journalist/columnist thinks – rather than what the story is. Of course it is influenced by the stories of the moment.
- This is not an analysis of media bias. This is an analysis of opinion. It is quite legitimate for columnists and editorials to have views that are not split 50/50 between the parties. And it is fair to say one would generally expect an incumbent Government to be criticised more often than it is supported.
I think we can all admit that a post like this is entirely a dog whistle of media bias.
Anyway, to the data, well I don’t want to steal David’s work so I won’t present the pretty table David has made but unsurprisingly he finds that just about everyone hates National. Except the Press Editorial team (which is weird hey?). They are the only team/columnist who has a positive-towards-National ratio of higher than 40 per cent (it’s 67 per cent in case you care)(apparently). The next best is Fran O’Sullivan from the Herald who has written positively about National 36 per cent of the time.
Then he has a table that sets out the positive/negative columnists who have written about Labour. And oh would you look at that, everyone loves Labour apparently. Well not everyone, the Dominion Post editorial team and Vernon Small have both written more negative columns about Labour than positive but the remaining four have written overwhelmingly in favour of Labour.
This clearly shows that the media has a left-wing bias right? I mean it’s all there in the numbers.
Except let’s interrogate a few things here shall we.
David attempts to mitigate against future criticism by saying “one would generally expect an incumbent Government to be criticised more often than it is supported” which I guess is true, but I’m not sure why. And then we look at the dates covered: 1 November 2014 to 31 March 2015.
And in that time the following has occurred under National’s watch (as set out so angrily in my own post “Hubris“):
- the Snowden revelations were pored over with much more attention than prior to the election
- the Government decided to send troops to Iraq on spurious reasons and refused to put it to a vote in Parliament
- it was revealed that the Government had known about the 1080 threat for some months before alerting consumers
- the Sky City deal got shittier when Sky City said it needed more money
- the Northland By-Election happened after Mike Sabin resigned for mysterious reasons and National got its ass handed to it, and also acted like a bunch of self-entitled bell-ends offering bribe after bribe
- it was revealed that Key actually knew Dongha Liu a lot better than he portrayed.
Meanwhile Labour did fuck-all. Like, they’ve all but vanished really. There was a brief flurry when they gave a not-so-sly nod and wink to the voters of Northland to vote for Winston and Paddy Gower got all puppy-over-excited-he-runs-around-while-doing-wees about it but for the most part, Labour has been nowhere. Haven’t done anything to report on. Which is not a great thing, but hey it might explain why the media is SO GODDAMN BIASED AGAINST NATIONAL GODDAMN IT.
Because National have fucked up. They’ve spent the first part of this new parliament lurching from one fuck-up to the next. It would be altogether more surprising if there were more positive stories about National to be honest. They’ve been shit-house. There hasn’t been good news politically to cash in on.
So David, I admire your attempts to work your commenters up into a fury about how biased the media is, and I’m sure you know that a lot of the columnists you write about read your posts and that probably will give them pause for thought. But on the whole, it’s a big load of ass.
It’d be interesting to do a similar analysis of the six months prior to the last election.
“David attempts to mitigate against future criticism by saying ‘one would generally expect an incumbent Government to be criticised more often than it is supported’ which I guess is true, but I’m not sure why”
It’d be because the incumbent government can actually do stuff that has negative consequences, wouldn’t it? I mean, there’s not much point writing critically about the Opposition except to criticise their performance as Opposition (which still wouldn’t have real-world consequences in quite the same way that bad government decisions would).