When your results should give you pause for thought

It's entirely possible to construct methods of measurement that will prove anything you want them to. But when you do so it's always worth just checking your results to see if they make some sort of sense. As with that delightful survey from nef a few years ago, where they tried to list the best places in the world to live coming up with Vanuatu as the answer. That they arrived at a place where a penis sheath is the major fashion accoutrement and they worship the Duke of Edinburgh as a Living God (which he is of course) leads to a certain questioning of the metrics they used to decide upon "best place to live". So it is with this report about how the children in England are horribly downtrodden, depressed and unhappy:

Children in England are less happy and satisfied with their lives than those in the majority of other European countries and North America, with only South Korean and Ugandan children worse off, a study by The Children's Society has found.

Although 90% of English children in the study rated themselves as having relatively good wellbeing levels, England still ranked ninth out of a sample of 11 countries around the world in the study, which involved 50,000 children – behind countries such as Romania, Spain, and Algeria and ahead of only South Korea and Uganda.

When we look at the details of the report we find that three of the four happiest places (in the larger sample) to be a child are Greenland (60,000 people stuck on an ice floe), Armenia (per capita GDP around $6,000) and Macedonia (per capita GDP $10,000 or so, under a third of the UK). Among the smaller sample of 9 countries the very best place to be a child is apparently Romania: and aren't we still sending teddy bears to the appalling orphanages there?

Perhaps being a child in England can be made better but it's not entirely obvious that this report is using the correct ways of measuring that "best place". Or even methods that are even remotely sensible.