Kail and Park (1990) put
this hypothesis to the test by giving 11-year-
old children and adults more than 3,000 tri-
als of practice at mental rotation. They found
that both groups sped up but that adults
started out faster. However, Kail and Park
showed that all their data could be fit by a
single power function that assumed that the
adults came into the experiment with what
amounted to an extra 1,800 trials of prac-
tice (Chapters 6 and 9 showed that learning
curves tended to be fit by power functions).
Figure 14.10 shows the resulting data, with
the children's learning function superim-
posed on the adult's learning function. The
practice curve for the children assumes that
they start with about 150 trials of prior prac-
tice, and the practice curve for the adults assumes that they start with 1,950 trials of prior prac-
tice. However, after 3,000 trials of practice, children
are a good bit faster than beginning adults. Thus,
although the rate of information processing increases
with development, this increase may have a practice-
related rather than a biological explanation.
Qualitative and quantitative developmental
changes take place in cognitive development
because of increases both in working-memory
capacity and in rate of information processing
Körkel, and Weinert (1988) looked at the effect of expertise at various age levels.
They asked German schoolchildren at grade levels 3, 5, and 7 to recall a story
about soccer, and they categorized the children at each
grade level as either experts or novices with respect to
soccer. The results in Table 14.1 show that the effect
of expertise was much greater than that of grade level.
Moreover, on a recognition test, there was no effect of
grade level, only an effect of expertise. Schneider et al.
also classified each group of participants into high-
ability and low-ability participants on the basis of their
performance on intelligence tests. Although such tests
generally predict memory for stories, Schneider et al. found no effect of general ability level, only of knowledge for soccer. They argue
that high-ability students are just those who know a lot about a lot of domains
and consequently generally do well on memory tests. However, when tested on
a story about a specific domain such as soccer, a high-ability student who knows
nothing about that domain will do worse than a low-ability student who knows
a lot about the domain.
In addition to lack of relevant knowledge, children have difficulty on mem-
ory tasks because they do not know the strategies that lead to improved memory.
The clearest case concerns rehearsal. If you were asked to dial a novel seven-digit
telephone number, I would hope that you would rehearse it until you were con-
fident that you had it memorized or until you had dialed the number. How-
ever, this strategy would not occur to young children.
Cattell (1963) proposed a distinction between
fluid and crystallized intelligence;
refers to acquired
refers to the ability to reason or to solve
problems in novel domains. In Figure 14.12, fluid intelligence, not crystallized
intelligence, shows the age-related decay. Horn (1968), elaborating on Cattell's
theory, argued that there is a spatial intelligence that can be separated from fluid
intelligence. Table 14.3 can be interpreted in terms of the Horn-Cattell theory,
where crystallized intelligence maps into the linguistic factor (tests 1 to 3), fluid
intelligence into the reasoning factor (tests 4, 5, and 7), and spatial intelligence
into the spatial factor (test 6). Fluid intelligence tends to be tapped strongly in
mathematical tests, but it is probably better referred to as a reasoning ability
rather than a mathematical ability. It is a bit difficult to separate the fluid and
spatial intelligences in factor analytical studies, but it appears possible (Horn &