I think what you are seeing in the lower growth target is probably regression to the mean. Simply, those who score at the tails of the distribution are statistically expected to score closer to the mean on re-test. (In this case, lower.)

...pronounced like the long vowel and first letter of the alphabet...