Are IQ tests what we really mean when we mean someone is intelligent? What about that kid who can't do calculus one year, then the next year they can? Haven't they gotten smarter?
Yes unnormalized IQ scores of kids increase when they get older. That's why only the normalized to their age IQ score is reported. I won't get seduced into a discussion about whether IQ is really intelligence. It's the best proxy we have that's all that matters for this discussion.
But it can make perfect sense that your rank doesn't change within your cohort. The better you are above your cohort, the more encouraged you are, and the more help you'll get from the system. So in that sense it can be the case that it's very hard to change your intelligence.
That's an assumption. It could be true or not. If true you'd expect steadily increasing IQ stores for people who are slightly above average, which is not what happens. The same effect should be observed in weaker students if it's true. I'm not interested in conjecture. I'm looking for peer-reviewed papers.