... Individual differences in learning proficiency show increasingly higher correlations with IQ directly in relation to the following characteristics of the learning task.
1. Learning is more highly correlated with IQ when it is intentional and the task calls forth conscious mental effort and is paced in such a way as to permit the subject to "think". It is possible to learn passively without "thinking", by mere repetition of simple material; such learning is only slightly correlated with IQ. In fact, negative correlations between learning speed and IQ have been found in simple tasks that could only be learned by simple repetition or rote learning but were disguised to appear more complex so as to evoke "thinking" (Osler & Trautman, 1961). Persons with higher IQs engaged in more complex mental processes (reasoning, hypothesis testing, etc.), which in this specially contrived task only interfered with rote learning. Persons of lower IQ were not hindered by this interference of more complex mental processes and readily learned the material by simple rote association.
2. Learning is more highly correlated with IQ when the material to be learned is hierarchical, in the sense that the learning of later elements depends on mastery of earlier elements. A task of many elements, in which the order of learning the elements has no effect on learning rate or level of final performance, is less correlated with IQ than is a task in which there is some more or less optimal order in which the elements are learned and the acquisition of earlier elements in the sequence facilitates the acquisition of later elements.
3. Learning is more highly correlated with IQ when the material to be learned is meaningful, in the sense that it is in some way related to other knowledge or experience already possessed by the learner. Rote learning of the serial order of a list of meaningless three-letter nonsense syllables or colored forms, for example, shows little correlation with IQ. In contrast, learning the essential content of a meaningful prose passage is more highly correlated with IQ.
4. Learning is more highly correlated with IQ when the nature of the learning task permits transfer from somewhat different but related past learning. Outside the intentionally artificial learning tasks of the experimental psychology laboratory, little that we are called on to learn beyond infancy is entirely new and unrelated to anything we had previously learned. Making more and better use of elements of past learning in learning something "new" - in short, the transfer of learning - is positively correlated with IQ.
5. Learning is more highly correlated with IQ when it is insightful, that is, when the learning task involves "catching on" or "getting the idea". Learning to name the capital cities of the fifty states, for example, does not permit this aspect of learning to come into play and would therefore be less correlated with IQ than, say, learning to prove the Pythagorean theorem.
6. Learning is more highly correlated with IQ when the material to be learned is of moderate difficulty and complexity. If a learning task is too complex, everyone, regardless of his IQ, flounders and falls back on simpler processes such as trial and error and rote association. Complexity, in contrast to sheer difficulty due to the amount of material to be learned, refers to the number of elements that must be integrated simultaneously for the learning to progress.
7. Learning is more highly correlated with IQ when the amount of time for learning is fixed for all students. This condition becomes increasingly important to the extent that the other condition listed are enactive.
8. Learning is more highly correlated with IQ when the learning material is more age related. Some things can be learned almost as easily by a 9-year-old child as by a 18-year-old. Such learning shows relatively little correlation with IQ. Other forms of learning, on the other hand, are facilitated by maturation and show a substantial correlation with age. The concept of learning readiness is based on this fact. IQ and test of "readiness", which predict rate of progress in certain kinds of learning, particularly reading and mathematics, are highly correlated with IQ.
9. Learning is more highly correlated with IQ at an early stage of learning something "new" than is the performance or gains later in the course of practice. That is, IQ is related more to rate of acquisition of new skills or knowledge rather than rate of improvement or degree of proficiency at later stages of learning, assuming that new material and concepts have not been introduced at the intermediate stages. Practice makes a task less cognitively demanding and decreases its correlation with IQ. With practice the learner's performance becomes more or less automatic and hence less demanding of conscious effort and attention. For example, learning to read music is an intellectually demanding task for the beginner. But for an experienced musician it is an almost automatic process that makes little conscious demands on higher mental processes. Individual differences in proficiency at this stage are scarcely related to IQ. Much the same thing is true of other skills such as typing, stenography, and Morse code sending and receiving.
...
Arthur Jensen, 1980
Bias in Mental Testing (p 327-328)