INTELLECTUAL ABILITIES AND TYPICAL–MAXIMAL
PERFORMANCE
Although not a precise match, Hebb’s (1942) distinction of Intelligence A and
Intelligence B and Cattell’s (1943) distinction of fluid intelligence (Gf) and
crystallized intelligence (Gc) provided a reasonably close categorization of
abilities that are associated with maximal performance and typical perform-
ance, respectively. Intelligence A and Gf are most associated with relatively
decontextualized information processing, reasoning, and memory. The pro-
totypical measures of Gf, such as the Raven Progressive Matrices Test (Pen-
rose & Raven, 1936; Raven, Court, & Raven, 1977) or the so-called Culture
Fair Intelligence Test (CFIT; Cattell & Cattell, 1957) present the examinee
with test items that allow for minimal transfer of learning or knowledge in
solving the problems. Instead, they rely on the examinee’s ability to use mem-
ory resources and reasoning skills to derive the correct answers to the items.
The Raven test is unspeeded, but the CFIT has relatively strong time limits
on performance. In both cases, the examinee must devote a maximal level of
attentional effort to obtain the highest possible score.
Tests of general information (such as are found on the WAIS–III, Wechs-
ler, 1997 or the Stanford–Binet IV, Thorndike, Hagen, & Sattler, 1986) are
good prototypes for Intelligence B and Gc—that is, knowledge that the
examinee has acquired and maintained over a long period of time. What
makes these measures particularly appropriate for assessing typical intellec-
tual performance, is that the examinee must know the information prior to
testing—it cannot be derived in the testing situation. If an examinee is asked
to name his or her state’s elected politicians, a search of long-term memory is
needed, but this activity generally requires far less intellectual effort than
solving an abstract spatial reasoning test item.
Many scales of intellectual abilities assess a mixture of typical behaviors
and maximal performance. A test of reading comprehension, for example,
draws substantially on previously learned skills and knowledge (such as
reading skills, vocabulary knowledge, and even sometimes factual knowl-
edge), and new learning (which involves allocation of working memory re-
sources to understanding a new text passage). For this reason, it is not un-
usual to find that such tests correlate substantially with both Gf and Gc
factors. Although Gc type tests can provide the best single estimate of intel-
ligence (e.g., the information test on the WAIS is the most highly correlated
subscale with overall IQ, see Wechsler, 1944, and the Ebbinghaus comple-
tion test has the highest correlation with an estimate of Spearman’s general
ability factor (g), see Spearman [1927]), the traditional approach to omni-
bus intelligence assessment includes a wide sampling of both Gf and Gc-
associated items.
- TYPICAL AND MAXIMAL PERFORMANCE 121