Then you refill A to the brim. Does A now have more, less or the same amount of water compared with B? The correct answer is the same. Obvious, innit?

Not to me, it isn’t. My first answer was that A has more water. B, I reasoned, would now be empty because the quickest way of re-filling A, rather than going to the tap again, was to take the water back from B. I did wonder why anybody would want to keep pouring water back and forth but I knew it was supposed to be some kind of intelligence test and that psychologists have strange ideas about what we should do with our time.

This question was printed in the Sunday Times to illustrate new research, first reported in The TES, that purports to show Year 7 children are less able than in 1976. The alleged decline is not a small one. If the results are correct, the average child today is two years behind a child of his or her parents’ generation - on basic maths and science skills at least.

I don’t doubt the soundness of the research, which comes from King’s College, London, and involved 10,000 children. I merely wish to retain a corner of scepticism in my mind. Tests, no matter how well designed, test an ability to do tests. Many who struggle with written maths tests can work out their winnings from the bookmaker in seconds. Context is all. If you can solve crossword and Sudoku puzzles, it shows you are good at puzzles.

It doesn’t necessarily signify other skills, beyond the necessary grasp of words and figures. Likewise, as Gertrude Stein might have said, a beaker is a beaker is a beaker. As I get older and more remote from the testing and examining we all suffered in our schooldays, I find I more frequently fail those supposedly simple tests that papers print to illustrate the modern child’s stupidity.

All the same, I am inclined to believe the researchers’ explanation for their results. Messing around with sand and water has been all but banned from primary schools; children get out less and go on fewer traditional seaside holidays. The messing around was for developmental volume and heaviness tests, which require a familiarity with liquid and solid materials. Children may do more tests than in 1976; but they have less first-hand, physical experience. The second, I guess, has more than cancelled out the first.

The study is entirely consistent with an American book I quoted here last year, which argued that children are getting smarter. IQ levels have been rising since the 1940s but the increase has accelerated over the past decade or so. This, the book argued, is because video games and gadgets exercise our brains as never before and make us good at pattern recognition, spatial geometry and other skills on which conventional IQ tests put a high premium.

The implications are rather startling. In the mid-1970s, people argued that children didn’t go to school to play in sandpits; they could do that kind of thing at home. So began the movement from 1960s-style “learning by playing” - a road that led to the austere “basics” regime in today’s primary schools. But it seems that experience of playing around and handling physical things is exactly what modern children lack and that they already have ample opportunities for more cerebral stimulation. So those sixties’ trendies, if they weren’t right at the time, are right now.

As for my own difficulty with the beakers, that is easily explained. I never did like sand and water, even as a child.