Bad reporting from the Guardian

Yes, from the headline, I guess most of you will have thought that I was referring to this article yesterday on Ysgol Llangennech and Welsh-medium education generally.

However what I had in mind was this report on a poll commissioned by Chatham House on what position the EU should take in Brexit negotiations. The Guardian's headline reads:

Two-thirds of Europeans believe EU should take hard line on Brexit – poll

But look more closely at this graphic from the article:


Yes, it is technically true that two-thirds of those outside the UK said that the EU should not compromise its core principles ... but what about the additional 20% who said that the EU should not compromise at all?

Do the maths. The truth is that more than 80% of those questioned think that the EU should not compromise over its core principles. In fact, even a majority in the UK think that the EU should not compromise its core principles. If we needed proof that the UK government is not going to get the outcome it says it wants, this survey should add to it. Twenty-seven democratically elected governments are not going to ignore such overwhelming strength of opinion in their respective countries.

So what are we to make of such reporting? Is it sloppiness? Maybe. Or is it that the Guardian, like every other media outlet, is inclined to write stories that support its own agenda, or (being more charitable) has an inbuilt, unconscious bias that it simply isn't aware of?

I don't share the general anger that I can see in many of the pro-Welsh-language comments on the Guardian's Llangennech article. I think it would be better to accept that every news outlet operates this way, and to filter what any article says accordingly. I'm glad the Guardian wrote what it did, simply because exposing such bias (whether intended or unconscious) is the best way of dealing with it.

Bookmark and Share


Anonymous said...

Could you put this one on your blog list.

Michael Haggett said...

Happy to add your blog. It looks interesting.

Post a Comment