Stats Chat

There seems to be a view that the Roy Morgan political opinion poll is more variable than the others, even to the extent that newspapers are willing to say so, eg, Stuff on May 7

The National Party has taken a big hit in the latest Roy Morgan poll, shedding 6 points to 42.5 per cent in the volatile survey.

I was asked about this on Twitter this morning, so I went to get Peter Green’s data and aggregation model to see what it showed. In fact, there’s not much difference between the major polling companies in the variability of their estimates. Here, for example, are poll-to-poll changes in the support for National in successive polls for four companies

And here are their departures from the aggregated smooth trend

There really is not much to see here. So why do people feel that Roy Morgan comes out with strange results more often? Probably because Roy Morgan comes out with results more often.

For example, the proportion of poll-to-poll changes over 3 percentage points is 0.22 for One News/Colmar Brunton, 0.18 for Roy Morgan, and 0.23 for 3 News/Reid Research, all about the same, but the number of changes over 3 percentage points in this time frame is 5 for One News/Colmar Brunton, 14 for Roy Morgan, and 5 for 3 News/Reid Research.

There are more strange results from Roy Morgan than for the others, but it’s mostly for the same reason that there are more burglaries in Auckland than in the other New Zealand cities.

Automated location information can be very useful, but if the ‘location’ is an area and the automated result is a single point, it’s easy to get misled.

• I wrote about an interactive map of vaccine-preventable disease outbreaks, and the pertussis epidemic it maps as being in the  Ngaanyatjarra lands in Australia’s Western Desert.
• Mapping mentions of ‘Nigeria’ to the centre of the country was one of the problems of the 538 piece on kidnappings last week
• An analysis relating right-wing politics to porn viewing, among its other issues, had a lot of users mapped to Kansas, because that’s where the centre of the USA is.
• Wayne Dobson, who lives in Las Vegas, doesn’t have your cell phone. His house is where a lot of ‘uncertain location’ phone positions get mapped to.

From Cathy O’Neil: Ignore data, focus on power (and, well, most of the stuff on her blog)

From danah boyd and Kate Crawford: Critical Questions for Big Data

Two colour wheels, one using red-green-blue colour coordinates based on how monitors work, the other using a rescaled coordinate system based approximately on what the eye sees

The second set of colours isn’t as dramatic, but they are much better at being evenly spaced around the circle and equally bright.

Health and evidence edition

• Evidently Cochrane, a blog with non-technical explanations of Cochrane Collaboration review results
• Design process for a graphic illustrating the impact of motorbike helmet laws.  In contrast to bicycle helmet laws, laws for motorbikes do have a visible effect on death statistics
• Stuff has quite a good story on alcohol in New Zealand.
• The British Association of Dermatologists responds to ‘drinkable sunscreen’.
• 3News piece on Auckland research into extracts of the lingzhi mushroom. Nice to see local science, and the story was reasonably balanced, with Shaun Holt pointing out that this is not even approaching being anywhere near evidence that drinking the stuff would do more good than harm.

Team Ratings for Round 15

The basic method is described on my Department home page. I have made some changes to the methodology this year, including shrinking the ratings between seasons.

Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Performance So Far

So far there have been 87 matches played, 56 of which were correctly predicted, a success rate of 64.4%.

Here are the predictions for last week’s games.

Predictions for Round 15

Here are the predictions for Round 15. The prediction is my estimated expected points difference with a positive margin being a win to the home team, and a negative margin a win to the away team.

Team Ratings for Round 11

The basic method is described on my Department home page. I have made some changes to the methodology this year, including shrinking the ratings between seasons.

Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Performance So Far

So far there have been 80 matches played, 44 of which were correctly predicted, a success rate of 55%.

Here are the predictions for last week’s games.

Predictions for Round 11

Here are the predictions for Round 11. The prediction is my estimated expected points difference with a positive margin being a win to the home team, and a negative margin a win to the away team.

Following up on the “net tax” tangle, Keith Ng has a step by step explanation of how income tax and income distribution has changed over recent years in NZ.

You can also play with the visualisation yourself. Or, if you want to see the arguments about it, they’ll be on his Public Address post.

Maps are good because they take advantage of all the previous maps we’ve seen to provide background familiarity.  On the other hand, they use up both the available spatial dimensions before you’ve actually got any data, so you need to encode the information some other way. Colour is the obvious choice, but colour is much more limited than people appreciate.

Kieran Healy tweeted “It’s not like there’s a simple, tradeoff-free solution, but this is not a good map.”

And he’s right; it isn’t. There are too many categories, and some of them are ordered but not all of them, so colour isn’t enough. Even if you’re going to try, these aren’t the right colours: for example, orange should be between yellow and red.  About the most you could do clearly with a single map is the three-way split: Yes, same-sex couples can just roll up to the registry;  No, not this week; or It’s Complicated.

Jacob Harris pointed to an article at Source, describing the design of graphics for a story about marijuana legalisation. It does much better

They link to the classic piece “When maps shouldn’t be maps” by Matthew Ericson, which I’ve linked before. They also have a whole collection of articles on better maps, though it’s fairly programming-oriented.

A new map to let you see the impact of rises in sea levels on your area: this is Auckland with 13m sea rise

This doesn’t show the impact of storm surges, which are the big problem for a lot of eastern coastal Auckland (though not so much for Manukau Harbour).

(via everyone on twitter)