A nice article appeared on-line in the latest Economist. Their ‘Buttonwood’ columnist simply pointed to the A-level results of his old school, Westminster:

• Grade As were 40% of all A-levels taken in 1988 but rose to 90% by 2009,
• Grades A & B together rose from 64% to 98% in the same period,
• passes, already 97%, became 100% in 1995 and in every year since.

I summarise; numbers are given on the school’s website for each year in the period.

As he points out, Westminster can’t have changed that much. This was always a private-sector academic hot-house – selective entry, driven parents, hard-working pupils and the best teachers money can buy – but in 1988 ‘only’ 40% of A-level entries resulted in an A grade. What can have changed since then to boost that 40% to 90%? Tighter selection, yet more driven parents, even harder working students, even better teachers? Or exams that, from 1988, began to be based on a national curriculum, to be modularised, marked to ‘absolute’ rather than comparative standards and with retakes allowed as a matter of course? On this evidence alone we can’t be certain, but the balance of probabilities is surely very much in favour of less demanding exams.

However we don’t have to be satisfied with the example of Westminster. Their results sit comfortably with the differently derived evidence from the Centre for Evaluation and Management (CEM) in the Education Department at Durham University. It samples 50,000 children annually and found between 1988 and 2005 an inflation of two grades in most A-level subjects, i.e. a 1988 grade C would get you a 2005 grade A, and three grades in Physics and Maths – D gets you A. So they have established not only a change in standards, but reasonably precise estimates, by subject, of the extent of that change.

The CEM is respected around the world. Its staff have serious analytical, experimental and statistical abilities; most have first degrees in Physics, not humanities subjects, and reading their reports is only for the statistically literate.

So the evidence for grade inflation is very solid indeed and we have pretty good estimates of its extent. Everyone in education knows about these results; they just don’t talk about them in public. There is a tendency for educationists to pick holes in the methodology of work that casts doubt on current educational ideology and these criticisms are often quite sophisticated. The problem is that the methods and approaches that are being defended have no experimental basis at all. Big, rigorous experiments are nitpicked to defend 50-year-old conjectures; narratives that have never been tested in the field but are plausible in the common rooms of Education Departments.

The question is, what are we trying to achieve by this change? Presumably it was intentional, though we were never told that it would happen, indeed the educational establishment still denies in public the reality that is staring us in the face. The alternative is that it was not intentional, an accident suffered by tens of millions of children, implying a level of incompetence and subsequent cover-up that would surely cause heads to roll in any organisation with pretensions to professionalism.

Forty years ago we worried about our best qualified young people emigrating to countries with more opportunities and lower tax rates – the ‘brain drain’. We’ve certainly solved that problem

At present in the UK there is a tendency to assume that attributing the structural deficit to the previous government is a political action designed to justify an ideological commitment to cuts in spending. But there is another structural deficit in the developed world, in the USA, and  whereas the UK’s structural deficit was created by a chancellor of a party of the left, in the USA it was caused by a president of the right, a big-spending republican – a contradiction in terms if ever there was one. This is not a political criticism it is an economic one; structural deficits are bad no matter who creates them.

And both men had the initials GB. Spooky, or what? Either way we will be paying for their delusions for a decade at least.

The paper described work done in the Chernobyl exclusion zone to measure the changes in biodiversity in the zone and to estimate the effect that contamination had on any change in diversity. The BBC’s headline was “Chernobyl zone shows decline in biodiversity”. A side header quoted “The truth is that these radiation contamination effects were so large as to be overwhelming.” Tim Mousseau, University of South Carolina. This seems a clear statement, but this is a subject riddled with controversy and checking its context might be a good idea.

There was also the fact that nowhere in an article of over 600 words that purported to be about a reduction in diversity was the extent of the reduction quantified.  Neither was there any quantification of what “overwhelming” might mean, nor of how representative is the exclusion zone of the conditions that might someday apply to you and me.

Starting at the beginning, it took 60 seconds to find from Wikipedia that the exclusion zone has a 30km (19 miles) diameter around the plant, though there have been some adjustments, both gives and takes, to make it a more accurate contour of a level of contamination. So, this isn’t about the tail, stretching hundreds of miles to Sweden, of minor contamination but about a rather limited area around the plant. That would be no comfort if you had lived there, but it is very, very different from the levels of contamination in the Baltic states and Scandinavia.

It took just moments more to get an abstract of the article from “Ecological Indicators”. This made it clear that the paper was trying to establish the validity and accuracy of possible methods of measuring changes in biodiversity – it was not primarily about determining the loss of diversity around Chernobyl, that was just a good example to work with. The abstract did not quantify the change in biodiversity in the Chernobyl exclusion zone, but it did say that contamination effects accounted for between 1.5% and 26.5% of the variance in biodiversity in the 9 animal taxa measured.

The conclusion of the abstract was “This suggests that standard breeding bird censuses can be used as an informative bio-indicator for the effects of radiation on abundance of animals.” Not quite the BBC’s headline.

I would have to spend $37.50 to get a copy of the full paper and I could get a couple of very decent bottles of wine for that, so I can’t be sure of what else I might find in the body of the paper. But it is clear that:

• The statement about the ‘overwhelming’ effect refers to the proportion of the (square of) the variations that is attributable to contamination, and that proportion seems to be 1.5% to 26.5%, not what most people would call ‘overwhelming’.
• The scientists did find a reduction in biodiversity, disproving assertions to the contrary by other “scientists” who had done no measurements (!), but I would like to be told the extent of the reduction to get some idea of how serious it is. The variations in biodiversity might be 90%, or 30%, or 10%, or even less.  I don’t know.

And the BBC didn’t tell me.

Surely the impression given by the BBC’s presentation is that there are big changes in biodiversity, overwhelmingly caused by contamination, yet nowhere are we given the measures that might justify this impression, or otherwise. Nor are we told over what area this applies.

How about “Scientists found a reduction of 8.3% in biodiversity, up to 26.5% of this change being accounted for by the change in contamination”. I have no idea from the BBC’s report what the right numbers are, but I am sure they (or something like them) were in the original paper, and even surer that this is what the scientists were trying to express, so why 600 mushy words where one sentence with two numbers would have given the essence of the story, and much more accurately?

I know that graduates in the humanities who make up a huge (overwhelming?) proportion of BBC staff are functionally innumerate and scientifically illiterate, but why do they imagine the rest of us are similarly afflicted? There is no reason to employ well-paid people to translate precise scientific statements into vague and misleading ‘narratives’.