: Job application #2…More on PLD and news coverage…
Kevin Munden sent an e-mail to Jay Manifold yesterday with more thoughts on power law distribution and news coverage. He copied me, and I am posting his letter here along with an Excel graph of the data.
Hi Jay:
Well, I’ve had a chance to do a bit more analysis on this topic.
I agree with your assessment that the NYT, WaPo, and Wire Services drive the Pareto distribution of news coverage. This is a structural requirement of the way the news is made. If the NYT and WaPo didn’t have the influence they do, some other media outlet would take their place.
Put yourself in the shoes of the editorial board of a B list media property. Say like my hometown paper, the Houston Chronicle. What sources are going to affect your call as to what’s going to appear on your paper, and what story-coverage assignments you’re going to hand out to your staff.
First, I’d look at local competitors, which in this case are all Television properties.
Second, I’d look at regional equivalents in San Antonio, Austin, and Dallas.
If I have a corporate parent, I’ll look there.
The wire services are always a way to cover a story where you don’t have coverage resources.
And you’d always look to the big boys (NYT, WaPo) to see what’s big in national and international coverage. How long do you make the list of “Big Boys”? If you’re dedicated to excellence and balance in coverage, the list would be long. If you’re lazy and have a gazillion other tasks at hand, you’ll be apt to keep the list short.
So the big boys may have influence well beyond their relative circulations, but not beyond their relative circulations *Among News Media Professionals*.
I’ve done some more work on google news. In search of artifacts specific to google, well their cutoff of 20 stories per topic skews things a good bit. And if their bot actually crawls 4500 news sources, it seems to turn up unusually low scoring “top twenties” in the business, sci/tech, entertainment, and health topics. By this, I mean a story with only 1 or two sources reporting on it. I have a hard time believing there are not stories out there with more coverage. The 20 limit must surely clip some more broadly covered stories in the US, World and Sports topics.
One artifact of my reporting methodology is that I had previously just counted the “and related” number. The way it should be done is to add that number to the number of citations displayed by google seperately. I did that for the Sci/Tech and Health Topics, but gave the other topics an adjusment of an “average” 5 citations.
Another Artifact is that some stories appeared both in the top stories and subtopics. This will show as a few “doubletaps” on the left hand of the aggregated graph.
All that being said, the aggregate graph looks to be pretty close to the Pareto norm graph.
I also looked at the individual topics. I went into this thinking that sports would be a very sharp distribution since there is a regional bias in sports reporting. Guess again. Was also suprised by the sharpness of the entertainment graph, which I thought nationally relatively homogenous. Less suprised by the Health result. I’d have to take any of the subtopic results with a large grain of salt, due to such small sample sizes.
Thanks for addressing all this on your site.
I’m CC:ing this to Andrew Cline, whose discussion I greatly enjoyed. Thank you Andrew.
Regards
Kevin Munden
