Posts by David Hood
Last ←Newer Page 1 2 3 4 5 Older→ First
-
Hard News: The Watching World, in reply to George Darroch,
Having worked a few years ago as a telephone interviewer and now spending a great deal of my working time designing, implementing and analyzing surveys, it sounds quite right.
The Rand approach in this U.S. election is being described as an interesting variant approach in surveying which is showing the same pattern as everyone else
-
Much of the pundit ranting kicked off a couple of weeks back, when Nate Silver was rating Obama at 75% likely, as nothing dramatic has happened to change the race (and even Bloomberg revealing the details of Romney's tax returns was not dramatic enough to have any affect) that percentage has climbed. the critical period seems to have been about a seek after the first debate, when the polls began to diverge from the pundits. I would also note that Nate Silver is among the most conservative of poll analysts, most others have been giving Romney an even smaller chance of victory.
That said, back when the probability was back at 72%, the Auckland Univeristy Department of Statistics' Stats Chat blog noted that the All Blacks have beaten Australia in 70% of their matches, as a way of putting it in context. At 85% (at the moment) we can make the analogy of the All Blacks winning pecentage in the 2000-2009 period (which was 82%)- Reliable favorites, but could still have a bad game or some bad refereeing decisions (voter suppression) go against them.
I was reading the other evening about the way sports commentary (and Nate Silver comes out of a sports analysis background) has learned to come to grips with statistics. Take home point, once you get all the coaches involved paying attention to the metrics and accepting them, you have a bit of a generational lag to the punditry. Article I was reading.
With New Zealand media, I am actually going to positively name check Jim Mora's The Panel, where it was clear (on those days I caught it and it was a topic for discussion) that people were aware of what sites like 538 were reporting and how it differed from the pundits. -
A scan of one of my father's slides
-
Hard News: Media3: Whistleblower Season, in reply to David Hood,
You turn it into a bouncy stonehenge
or, as my daughter just reminded me a bouncy marae
-
Hard News: Media3: Whistleblower Season, in reply to Ian Dalziel,
ok, how do you convert a Bouncy Castle?
You turn it into a bouncy stonehenge
-
Ditto
-
From today
-
There has been a lot of discussion in U.S. media commentary this week on insider narrative vs. big data narrative. It has been building from mid-October where the insider political narrative of a Romney surge and a neck and neck race stopped matching the results that the poll analysis people were getting. It has grown increasingly heated over the past week, but has also grown into a discussion of the importance of insider perspectives versus large scale analysis of public sources.
While insider sources is not the same thing as whistleblowers, I thought the parallels in discussion have been interesting. -
Capture: Spring is Like a Perhaps Hand, in reply to Nora Leggs,
here are couple more cups of gold calendula
I do like the depth of field you got on the second shot, well done.
-
From my mother's garden
