An article in the Guardian (Teenagers' IQ scores can rise or fall sharply during adolescence) caught my eye recently, so I did a bit of digging. The research being quoted was originally published in Nature magazine and was funded by the Wellcome Trust.

However, I think a better response to the original research appears in Psychology Today -  Intelligence Is Still Not Fixed at Birth. In his article, Kaufman points out the following drawbacks of Ramsden et al's research:

It was a small sample, it's not clear that the tests they administered were age-appropriate for all of their participants, and some of the high-IQ children tested at the first time point may have found the test too easy (showing 'ceiling effects'). Future research should look at larger samples, using a number of different tests, on a wider range of ages. It remains to be seen if the same levels of plasticity are evident among older populations.

Fine. At any rate, it's a neat little study and contrary to an article in The Guardian, which stated that the study "contradicts a long-standing view of intelligence as fixed", the study is actually quite consistent with what IQ researchers have been finding for over 50 years! No sensible IQ researcher would say that intelligence is fixed. Anyone who does tell you that is a nutjob.

In fact, leading IQ test makers fully acknowledge that people don't have "an" IQ. That's why they are increasingly moving away from a focus on total IQ scores to observing how children behave on the different sections of the test.

As always (or almost always), the best thing to do is look at the orginal research paper. Unfortunately it'll set you back $32!