Sunday, March 09, 2008

March 6 Media Critique: One Year Analysis: "Reuters 2007 'Pictures of the Month' "

The rocket-scientists at HR have been busy looking at pictures and they are not happy.

Having previously challenged Reuters for its objectivity in a subjectively selected group of images.....

And what a fine basis for challenge that is!

Yes, having done so with such spectacular stupidity last year, HR decided once again to assess the "objectivity in a subjectively selected group of pictures”.

we were interested in seeing if that case represented a one-time aberration, or if Reuters has a serious bias problem in its coverage of the Arab-Israeli conflict.

Really, you couldn’t make this stuff up.

62% of images provoke sympathy for the Palestinians.

But HR do just make this stuff up.

So, let’s objectively (and subjectively) look at HRs subjective assessment of the objectivity in a subjectively selected group of photos.

62%. Oh. What does it mean though? HR don’t really say. Why is 62% not OK, and what would be OK? Again, HR is mute on the subject. Let’s assume, being complete idiots, HR think 50% would be “fair and balanced” (note: this is where I get all objective.) While HR get all heated up about a simple-minded notion of balance, the real issue is representation. Is the photo selection process representative of the pool of photos Reuters had to choose from, ie did Reuters only have 15 photos of Palestinian funerals to choose from or did they have 10,000? In case anyone from HR is reading, I’ll explain it simply as possible, with the classic example from statistics; green and red coloured marbles in a jar. We grab a handful of marbles and find we have 1 red and 4 green. In this explanation of probability, the more ‘trials’ of grabbing a handful of marbles we make, the greater certainty we have of knowing what is the actual proportion of red and green marbles in that jar. If we really want to know, we’d just pour the marbles out and count them and perhaps see that the ratio was indeed 1:4. For HR, the 62% figure indicates the selection is “unbalanced’. The only thing a figure like that can tell us is if the pictures chosen are representative, and HR could only do that by viewing all the photos in the pool, which they can’t because they only have the end result, ie. the handful of marbles and no access to the jar. Perhaps in the the pool of photographs, 80% of them "provoke sympathy". We can't know.

At a subjective level, we have HRs “unbalanced” claim. They don’t tell us what is ‘balance’ so it’s a little hard to comment on, but I’ll do my best. HR uses the photos of Palestinian funerals to try and make this argument. There were 15 photos of Palestinian funerals and none of Israeli funerals. Would ‘balance’ be 7 of each perhaps? But would this be “balanced”? Let’s return to a far more useful and objective criteria – representative. We can get some of idea of this by first looking at the number of funerals. For 2007 there were 373 Palestinians killed by Israel and, I’ll assume 373 funerals. 15 photos from Rueters, makes that a photo for every 24 Palestinian funerals. And there were 7 Israeli funerals, which is less than one-third of the ‘quota’ for a Palestinian funeral picture. So while we can’t see the pool of photos that were chosen from, we can see that, objectively, the funeral photos are representative of reality. HR, not being big fans of reality, don’t see this. If we followed their assumed notion of balance and included 7 of each, photos of Israeli funerals would be over-represented 24x. Is that “balance”? For HR, the answer would probably be yes.

Which leads us on to yet another subjective question, sympathy, and a quick review of HRs highly detailed and sophisticated methodology on the topic. How do they define sympathy? They don't. What are the criteria for determining that a photo “provokes” it? They don’t say. Who made the determinations? HR are silent . I can only assume it wasn’t any of the HR staff, because being such one-eyed pro-Israel partisans, a photo of dead or crying Palestinians would elicit exactly nil sympathy. The corollary of this is that we’ll have to assume that HR guessed as to whether or not a photo “provokes” sympathy, hence the 62% figure is 100% subjective.

Brilliant work.

I think it’ fair to say that subjectively……no wait, I‘ll go out on a limb here……. that objectively reviewing HonestReportings subjective assessment of the objectivity in a subjectively selected group of pictures -

It’s 100% horseshit.