Posts from April 2021 (18)

April 30, 2021

Physical punishment of children – reporting

There’s a new research paper out from the Christchurch Health and Development Study, which recruited a group of people when they were born, in mid-1977, and has been following them ever since.  Those of the participants who were parents have been asked about physical discipline of their children on four occasions: when they were 25, 30, 35 and 40.  Obviously, over time, the number who are parents has increased (from about 150 to over 600), and the children have (on average) gotten older — when the parents were 25, most of the kids would have been pre-school; the group now includes a few very young children but many who are teenagers.

The good feature of birth cohorts like the Christchurch study is that you get to see the same people throughout the course of their lives; the bad feature is that at any given time everyone is exactly the same age.  In statistician jargon, age is completely confounded with period: you are completely unable to distinguish effects of ageing from time trends. When you see that the proportion of parents reporting hitting their kids has gone down from 77% to 42% over the 15 years, you can’t tell, at all, whether this is an effect of these specific parents getting older and more experienced or an effect of parents in general being less likely to hit their kids.  It’s hard (though not impossible) even to tell if it’s an effect of the kids being older.

As you’d hope, the research paper, in the NZ Medical Journal (paywalled) is very clear on this

…explanations include: increasing maturity of the parenting sample over time (less reactive, more experienced, older parents); a cultural shift towards the unacceptability of violence towards children over the period of the study; and the law change in 2007, which prohibited physical punishment and violence towards children. Given the nature of its design, it is not possible for the current study to distinguish between these explanations.However, it does not seem unreasonable to conjecture that all three processes are likely to have played a role.

And indeed it doesn’t seem unreasonable, as long as you recognise that the not seeming unreasonable isn’t a conclusion from the data and relies entirely on external plausibility.  The researchers do conclude that there’s still a lot of physical punishment going on, and that efforts are needed to stop it; the former is well-supported by the data and the latter is a policy response, not a scientific conclusion. That’s all good.

So let’s look at the reporting (some of this may have changed before or after I read it, of course)

  • Radio NZ: Number of parents smacking children drops by half in 15 years. No caveat about the study design meaning this conclusion is basically unsupported. Gets the journal name wrong.
  • 1 News: More than 40% of parents still use physical discipline years after law change, latest data shows. The story is better than the headline, and the Children’s Commissioner is quoted as saying “It’s representative of one cohort born in 1977, one group in one year in one generation, but there has been a discernible drop over the years.” I’d be happier if it was clearer from the beginning that this doesn’t claim to be representative of NZ in general over time.
  • NZ Herald. Parents’ physical punishment of children decreasing, but still common – report. Slightly better headline; much clearer in the story. “…the rate of physical punishment against children was higher when parents were younger, and then decreased with age… because of the way the study was designed, it couldn’t pinpoint how much the rates reduced because of the law change.”
  • Otago Daily Times. Parents still smacking, study finds. Good. “The authors warned that their method of studying a cohort of people over time meant they could not gauge what the attitude of new parents in 2021 might be to physical punishment. However, the research did suggest rates of smacking or hitting children were high enough to be a public health concern.”
  • Stuff. Physical punishment of children still ‘fairly common’, despite anti-smacking law change – study.  There’s no caveat about the study design, and the story says “New research, published in the New Zealand Medical Journal on Friday, examined how the prevalence of child physical punishment changed in the 15-year period between 2002 and 2017 – before and after the legislation came into force.”, which isn’t true. And that’s not a link to the research paper.
  • Newshub. Who’s most likely to use physical discipline against their kids revealed. The headline’s a bit dodgy given the non-representative group of parents, but the caveats are good “Because the study followed a cohort of parents who aged 15 years over the course of the study, “it is unclear what rates of physical punishment of children would be in studies of contemporary young parents”.

 

April 28, 2021

Top 14 Predictions for Round 23

Team Ratings for Round 23

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Current Rating Rating at Season Start Difference
Stade Toulousain 8.04 4.80 3.20
La Rochelle 7.04 2.32 4.70
Clermont Auvergne 6.64 3.22 3.40
Racing-Metro 92 5.07 6.21 -1.10
Lyon Rugby 4.71 5.61 -0.90
Bordeaux-Begles 3.07 2.83 0.20
Montpellier 2.31 2.30 0.00
RC Toulonnais 2.22 3.56 -1.30
Castres Olympique -0.82 -0.47 -0.30
Stade Francais Paris -0.83 -3.22 2.40
Brive -2.03 -3.26 1.20
Section Paloise -3.43 -4.48 1.00
Aviron Bayonnais -5.90 -4.13 -1.80
SU Agen -15.51 -4.72 -10.80

 

Performance So Far

So far there have been 146 matches played, 99 of which were correctly predicted, a success rate of 67.8%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Lyon Rugby vs. Clermont Auvergne Apr 25 41 – 30 2.90 TRUE
2 Stade Francais Paris vs. Section Paloise Apr 25 46 – 32 7.10 TRUE
3 Stade Toulousain vs. Racing-Metro 92 Apr 25 34 – 16 7.70 TRUE

 

Predictions for Round 23

Here are the predictions for Round 23. 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.

Game Date Winner Prediction
1 Aviron Bayonnais vs. Bordeaux-Begles May 08 Bordeaux-Begles -3.50
2 Brive vs. Stade Francais Paris May 08 Brive 4.30
3 Castres Olympique vs. Lyon Rugby May 08 Lyon Rugby -0.00
4 Montpellier vs. La Rochelle May 08 Montpellier 0.80
5 Racing-Metro 92 vs. Clermont Auvergne May 08 Racing-Metro 92 3.90
6 RC Toulonnais vs. Stade Toulousain May 08 Stade Toulousain -0.30
7 SU Agen vs. Section Paloise May 08 Section Paloise -6.60

 

Super Rugby Predictions for Week 11

Team Ratings for Week 11

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

 

Current Rating Rating at Season Start Difference
Crusaders 14.88 14.49 0.40
Blues 7.24 7.80 -0.60
Hurricanes 6.14 7.13 -1.00
Chiefs 4.71 4.38 0.30
Brumbies 3.53 1.47 2.10
Highlanders 3.52 2.70 0.80
Reds 2.75 1.59 1.20
Rebels -4.40 -3.51 -0.90
Waratahs -8.87 -5.02 -3.80
Western Force -11.52 -13.05 1.50

 

Performance So Far

So far there have been 38 matches played, 25 of which were correctly predicted, a success rate of 65.8%.
Here are the predictions for last week’s games.

 

Game Date Score Prediction Correct
1 Chiefs vs. Hurricanes Apr 23 26 – 24 4.40 TRUE
2 Western Force vs. Reds Apr 23 30 – 27 -9.70 FALSE
3 Waratahs vs. Rebels Apr 24 25 – 36 2.00 FALSE
4 Crusaders vs. Blues Apr 25 29 – 6 12.30 TRUE

 

Predictions for Week 11

Here are the predictions for Week 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.

 

Game Date Winner Prediction
1 Hurricanes vs. Highlanders Apr 30 Hurricanes 8.10
2 Blues vs. Chiefs May 01 Blues 8.00
3 Brumbies vs. Western Force May 01 Brumbies 20.50

 

Rugby Premiership Predictions for Round 18

Team Ratings for Round 18

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Current Rating Rating at Season Start Difference
Exeter Chiefs 7.03 7.35 -0.30
Harlequins 4.29 -1.08 5.40
Bristol 4.05 1.28 2.80
Sale Sharks 3.16 4.96 -1.80
Northampton Saints 1.34 -2.48 3.80
Wasps -1.28 5.66 -6.90
Bath -1.93 2.14 -4.10
Gloucester -2.56 -1.02 -1.50
Leicester Tigers -3.79 -6.14 2.30
London Irish -5.84 -8.05 2.20
Newcastle Falcons -8.02 -10.00 2.00
Worcester Warriors -9.55 -5.71 -3.80

 

Performance So Far

So far there have been 95 matches played, 60 of which were correctly predicted, a success rate of 63.2%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Bristol vs. Exeter Chiefs Apr 24 12 – 20 2.70 FALSE
2 London Irish vs. Harlequins Apr 24 21 – 25 -5.90 TRUE
3 Gloucester vs. Newcastle Falcons Apr 25 35 – 24 9.80 TRUE
4 Leicester Tigers vs. Northampton Saints Apr 25 18 – 23 0.00 FALSE
5 Worcester Warriors vs. Sale Sharks Apr 25 32 – 35 -8.90 TRUE
6 Wasps vs. Bath Apr 26 39 – 29 4.50 TRUE

 

Predictions for Round 18

Here are the predictions for Round 18. 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.

Game Date Winner Prediction
1 Bath vs. Bristol May 09 Bristol -1.50
2 Exeter Chiefs vs. Worcester Warriors May 09 Exeter Chiefs 21.10
3 Harlequins vs. Wasps May 09 Harlequins 10.10
4 Newcastle Falcons vs. London Irish May 09 Newcastle Falcons 2.30
5 Northampton Saints vs. Gloucester May 09 Northampton Saints 8.40
6 Sale Sharks vs. Leicester Tigers May 09 Sale Sharks 11.40

 

NRL Predictions for Round 8

Team Ratings for Round 8

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Current Rating Rating at Season Start Difference
Storm 15.87 14.53 1.30
Roosters 12.41 10.25 2.20
Panthers 12.19 8.88 3.30
Rabbitohs 9.67 7.73 1.90
Eels 7.46 1.68 5.80
Raiders 1.89 6.98 -5.10
Warriors -0.99 -1.84 0.90
Sharks -1.21 -0.76 -0.40
Dragons -1.96 -4.95 3.00
Sea Eagles -3.36 -4.77 1.40
Knights -4.68 -2.61 -2.10
Titans -5.98 -7.22 1.20
Wests Tigers -8.26 -3.07 -5.20
Cowboys -10.91 -8.05 -2.90
Broncos -11.44 -11.16 -0.30
Bulldogs -12.71 -7.62 -5.10

 

Performance So Far

So far there have been 56 matches played, 38 of which were correctly predicted, a success rate of 67.9%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Panthers vs. Knights Apr 22 24 – 6 20.30 TRUE
2 Titans vs. Rabbitohs Apr 23 30 – 40 -13.30 TRUE
3 Eels vs. Broncos Apr 23 46 – 6 18.20 TRUE
4 Sharks vs. Bulldogs Apr 24 12 – 18 18.70 FALSE
5 Cowboys vs. Raiders Apr 24 26 – 24 -12.30 FALSE
6 Wests Tigers vs. Sea Eagles Apr 25 6 – 40 4.50 FALSE
7 Roosters vs. Dragons Apr 25 34 – 10 15.90 TRUE
8 Storm vs. Warriors Apr 25 42 – 20 19.30 TRUE

 

Predictions for Round 8

Here are the predictions for Round 8. 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.

Game Date Winner Prediction
1 Raiders vs. Rabbitohs Apr 29 Rabbitohs -4.80
2 Storm vs. Sharks Apr 30 Storm 20.10
3 Broncos vs. Titans Apr 30 Titans -2.50
4 Panthers vs. Sea Eagles May 01 Panthers 18.60
5 Bulldogs vs. Eels May 01 Eels -17.20
6 Knights vs. Roosters May 01 Roosters -14.10
7 Warriors vs. Cowboys May 02 Warriors 9.90
8 Dragons vs. Wests Tigers May 02 Dragons 9.30

 

April 21, 2021

Knowing what to leave out

The epidemic modelling group at Te Pūnaha Matatini (who work a few floors above me) won the Prime Minister’s Science Prize for their work on modelling the Covid epidemic in New Zealand.   There have been some descriptions in the media of their models, but not so much of what it is that mathematical modelling involves.

A good mathematical model captures some aspect of the way the real process works, but leaves out enough of the detail that it’s feasible to study and learn about the model more easily.  The limits to detail might be data available or computer time or mathematical complexity or just not understanding part the way the process works.  Weather models, for example, have improved over the years by using more powerful computers and more detailed input data, enabling them to take into account more features of the real weather system and more details of Earth’s actual geography.

The simplest epidemic models are the SIR and SEIR families.  These generate the familiar epidemic curves that we’ve all seen so often: exponential on the way up, then falling away more slowly. They are also responsible for the reproduction number “R”, the average number of people each case infects.  The simple models have no randomness in them, and they know nothing about the New Zealand population except its size.  There’s a rate at which cases come into contact with new people, and a rate at which contacts lead to new infections, and that’s all the model knows.  These models are described by simple differential equations; they can be projected into the future very easily, and the unknown rates can be estimated from data.   If you want a quick estimate of how many people are likely to  be in hospital at the epidemic peak, and how soon, you can run this model and gaze in horror at the output.  In fact, many of the properties of the epidemic curve can be worked out just by straightforward maths, without requiring sophisticated computer simulation.  The SEIR models, however, are completely unable to model Covid elimination — they represent the epidemic by continuously varying quantities, not whole numbers with uncertainty.  If you put a lockdown on and then take it off, the SEIR model will always think there’s some tiny fraction of a case lurking somewhere to act as a seed for a new wave.  In fact, there’s a notorious example of a mathematical model for rabies elimination in the UK that predicted a new rabies wave from a modelled remnant of 10-18 infected foxes — a billion billionth of a fox, or one ‘attofox’.

The next step is models that treat people not precisely as individuals but at least as whole units, and acknowledge the randomness in the number of new infections for each existing case.  These models let you estimate how confident you are about elimination, since it’s not feasible to do enough community testing to prove elimination that way.   After elimination, these models also let you estimate how big a border incursion is likely to be by the time it’s detected, and how this depends on testing strategy, on vaccination, and on properties of new viral variants.  As a price, the models take more computer time and require more information — not just the average number of people infected by each case, but the way this number varies.

None of the models so far capture anything about how people in different parts of New Zealand are different.  In some areas, people travel further to work or school, or for leisure. In some areas people live in large households; in others, small households. In some areas a lot of people work at the border; in others, very few do.  Decisions about local vs regional lockdowns need a model that knows how many people travel outside their local area, and to where.  A model with this sort of information can also inform vaccination policy: vaccinating border works will prevent them getting sick, but what will it do to the range of plausible outbreaks in the future?  Models with this level of detail require a huge amount of data on the whole country, and serious computing resources; getting them designed and programmed correctly is also a major software effort.  The model has an entire imaginary New Zealand population wandering around inside the computer; you’re all individuals!

A mathematical modelling effort on this scale involves working from both ends on the problem: what is the simplest model that will inform the policy question, and what is the most detailed model you have the time and resources and expertise to implement?  Usually, it also involves a more organised approach to funding and job security and so on, but this was an emergency.  As the Education Act points out, one reason we have universities is as a repository of knowledge and expertise; when we need the expertise, we tend to need it right now.

April 20, 2021

Top 14 Predictions for Round 22

Team Ratings for Round 22

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Current Rating Rating at Season Start Difference
Stade Toulousain 7.63 4.80 2.80
La Rochelle 7.04 2.32 4.70
Clermont Auvergne 6.97 3.22 3.80
Racing-Metro 92 5.47 6.21 -0.70
Lyon Rugby 4.37 5.61 -1.20
Bordeaux-Begles 3.07 2.83 0.20
Montpellier 2.31 2.30 0.00
RC Toulonnais 2.22 3.56 -1.30
Castres Olympique -0.82 -0.47 -0.30
Stade Francais Paris -1.31 -3.22 1.90
Brive -2.03 -3.26 1.20
Section Paloise -2.95 -4.48 1.50
Aviron Bayonnais -5.90 -4.13 -1.80
SU Agen -15.51 -4.72 -10.80

 

Performance So Far

So far there have been 143 matches played, 96 of which were correctly predicted, a success rate of 67.1%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Section Paloise vs. Aviron Bayonnais Apr 17 43 – 33 8.20 TRUE
2 Castres Olympique vs. Stade Toulousain Apr 18 26 – 24 -3.80 FALSE
3 La Rochelle vs. Lyon Rugby Apr 18 38 – 23 7.10 TRUE

 

Predictions for Round 22

Here are the predictions for Round 22. 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.

Game Date Winner Prediction
1 Bordeaux-Begles vs. Montpellier Apr 24 Bordeaux-Begles 6.30
2 Aviron Bayonnais vs. Castres Olympique Apr 25 Aviron Bayonnais 0.40
3 Brive vs. La Rochelle Apr 25 La Rochelle -3.60
4 Lyon Rugby vs. Clermont Auvergne Apr 25 Lyon Rugby 2.90
5 RC Toulonnais vs. SU Agen Apr 25 RC Toulonnais 23.20
6 Stade Francais Paris vs. Section Paloise Apr 25 Stade Francais Paris 7.10
7 Stade Toulousain vs. Racing-Metro 92 Apr 25 Stade Toulousain 7.70

 

Super Rugby Predictions for Week 10

Team Ratings for Week 10

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Current Rating Rating at Season Start Difference
Crusaders 14.42 14.49 -0.10
Blues 8.12 7.80 0.30
Hurricanes 6.00 7.13 -1.10
Chiefs 4.92 4.38 0.50
Brumbies 3.41 1.47 1.90
Reds 3.20 1.59 1.60
Highlanders 3.03 2.70 0.30
Rebels -4.75 -3.51 -1.20
Waratahs -8.29 -5.02 -3.30
Western Force -12.10 -13.05 0.90

 

Performance So Far

So far there have been 34 matches played, 23 of which were correctly predicted, a success rate of 67.6%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Highlanders vs. Blues Apr 16 35 – 29 -0.10 FALSE
2 Rebels vs. Brumbies Apr 16 20 – 26 -2.20 TRUE
3 Chiefs vs. Crusaders Apr 17 26 – 25 -4.70 FALSE
4 Western Force vs. Waratahs Apr 17 31 – 30 1.80 TRUE

 

Predictions for Week 10

Here are the predictions for Week 10. 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.

Game Date Winner Prediction
1 Chiefs vs. Hurricanes Apr 23 Chiefs 4.40
2 Western Force vs. Reds Apr 23 Reds -9.80
3 Waratahs vs. Rebels Apr 24 Waratahs 2.00
4 Crusaders vs. Blues Apr 25 Crusaders 11.80

 

Rugby Premiership Predictions for Round 17

Team Ratings for Round 17

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Current Rating Rating at Season Start Difference
Exeter Chiefs 6.42 7.35 -0.90
Bristol 4.65 1.28 3.40
Harlequins 4.44 -1.08 5.50
Sale Sharks 3.52 4.96 -1.40
Northampton Saints 1.03 -2.48 3.50
Bath -1.59 2.14 -3.70
Wasps -1.62 5.66 -7.30
Gloucester -2.66 -1.02 -1.60
Leicester Tigers -3.47 -6.14 2.70
London Irish -5.99 -8.05 2.10
Newcastle Falcons -7.92 -10.00 2.10
Worcester Warriors -9.91 -5.71 -4.20

 

Performance So Far

So far there have been 89 matches played, 56 of which were correctly predicted, a success rate of 62.9%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Northampton Saints vs. London Irish Apr 17 44 – 26 10.60 TRUE
2 Sale Sharks vs. Gloucester Apr 17 25 – 22 11.70 TRUE
3 Exeter Chiefs vs. Wasps Apr 18 43 – 13 10.60 TRUE
4 Harlequins vs. Worcester Warriors Apr 18 50 – 26 18.10 TRUE
5 Newcastle Falcons vs. Bristol Apr 18 17 – 34 -6.90 TRUE
6 Bath vs. Leicester Tigers Apr 19 21 – 20 7.10 TRUE

 

Predictions for Round 17

Here are the predictions for Round 17. 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.

Game Date Winner Prediction
1 Bristol vs. Exeter Chiefs Apr 24 Bristol 2.70
2 London Irish vs. Harlequins Apr 24 Harlequins -5.90
3 Gloucester vs. Newcastle Falcons Apr 25 Gloucester 9.80
4 Leicester Tigers vs. Northampton Saints Apr 25 Leicester Tigers 0.00
5 Worcester Warriors vs. Sale Sharks Apr 25 Sale Sharks -8.90
6 Wasps vs. Bath Apr 26 Wasps 4.50

 

NRL Predictions for Round 7

Team Ratings for Round 7

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Current Rating Rating at Season Start Difference
Storm 15.60 14.53 1.10
Panthers 12.43 8.88 3.60
Roosters 11.68 10.25 1.40
Rabbitohs 9.98 7.73 2.20
Eels 5.59 1.68 3.90
Raiders 3.15 6.98 -3.80
Sharks 0.90 -0.76 1.70
Warriors -0.72 -1.84 1.10
Dragons -1.23 -4.95 3.70
Knights -4.91 -2.61 -2.30
Wests Tigers -5.05 -3.07 -2.00
Titans -6.28 -7.22 0.90
Sea Eagles -6.57 -4.77 -1.80
Broncos -9.57 -11.16 1.60
Cowboys -12.16 -8.05 -4.10
Bulldogs -14.82 -7.62 -7.20

 

Performance So Far

So far there have been 48 matches played, 33 of which were correctly predicted, a success rate of 68.8%.
Here are the predictions for last week’s games.

Game Date Score Prediction Correct
1 Broncos vs. Panthers Apr 15 12 – 20 -21.30 TRUE
2 Knights vs. Sharks Apr 16 26 – 22 -4.30 FALSE
3 Storm vs. Roosters Apr 16 20 – 4 5.00 TRUE
4 Sea Eagles vs. Titans Apr 17 36 – 0 -3.90 FALSE
5 Rabbitohs vs. Wests Tigers Apr 17 18 – 14 21.00 TRUE
6 Raiders vs. Eels Apr 17 10 – 35 5.70 FALSE
7 Dragons vs. Warriors Apr 18 14 – 20 4.30 FALSE
8 Cowboys vs. Bulldogs Apr 18 30 – 18 4.30 TRUE

 

Predictions for Round 7

Here are the predictions for Round 7. 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.

Game Date Winner Prediction
1 Panthers vs. Knights Apr 22 Panthers 20.30
2 Titans vs. Rabbitohs Apr 23 Rabbitohs -13.30
3 Eels vs. Broncos Apr 23 Eels 18.20
4 Sharks vs. Bulldogs Apr 24 Sharks 18.70
5 Cowboys vs. Raiders Apr 24 Raiders -12.30
6 Wests Tigers vs. Sea Eagles Apr 25 Wests Tigers 4.50
7 Roosters vs. Dragons Apr 25 Roosters 15.90
8 Storm vs. Warriors Apr 25 Storm 19.30