Sue Ramsden at University College London and her colleagues took a batch of 33 healthy adolescents (aged 12-16), with a wide mix of abilities. She tested their IQ in 2004 and then again 3-4 years later. During each test session, each child had their brains scanned using fMRI. Neither the participants nor their parents had a clue that they would be tested again. Consistent with earlier research, scores changed quite a bit from one testing session to the next. Thirty-three percent of the participants showed a clear change in their total IQ score. To highlight the range: One participant showed an increase in IQ score by 21 points, whereas another showed a decrease of 18 points. These are dramatic changes!
The researchers weren't just interested in test score changes, but were also interested in whether the changes were meaningfully reflected in brain structure. If they could link their changes to brain structure, then it would suggest these IQ fluctuations aren't merely the result of "measurement error." Indeed, they found that changes in the verbal portions of the IQ test (e.g., vocabulary, verbal comprehension) were related to changes in grey matter density and volume in the left motor cortex; a region that is known to be activated by the articulation of speech. Changes in the non-verbal portions of the IQ test (putting blocks and pictures together in their proper arrangement, completing pictures appropriately) were related to to grey matter density in the anterior cerebellum; a region that is known to be activated by motor movements of the hand. Interestingly, the grey matter changes that were associated with verbal and non-verbal IQ changes were not associated with regions normally found to be associated with IQ (frontal and parietal regions) Their results suggest that particular forms of intelligence (but not overall IQ scores) are reliant, at least in part, on sensorimotor skills and that such fluctuations aren't purely the result of 'measurement error'.