PRETEND INTELLIGENCE 91whether RT correlates with IQ, thus supporting the conclusion that IQ
is really a mea sure of some physiological or neural effi ciency. For exam-
ple, in the 1990s, Arthur Jensen (a well- known supporter of Galton)
spoke of “individual diff erences in speed or effi ciency of the vari ous el-
ementary pro cesses,” and how “ those diff erences account for the diff er-
ences in per for mance on psychometric tests.”^20 Th e hope engendered was
that of discovering the mother lode in a psychological gold mine.
Initially, some excitement was created in two ways. First a small cor-
relation (0.2–0.3) was found between IQ and a modifi ed RT test: the ap-
propriate reaction had to be quickly chosen from up to four alternatives
(e.g., d i ff er ent buttons for diff er ent light signals). Variability of individual
per for mance in this so- called choice- reaction time was also weakly cor-
related with IQ.
Th e prob lem is what to make of it. As always, we must not accept such
correlations as causally meaningful without controlled experiments. Such
small correlations indicate, anyway, that there are a lot of other things
causing diff erences in per for mance. And they may not even be cognitive
in origin. Douglas Detterman showed how RT involves a lot other than
simple response effi ciency.^21 Individual diff erences can stem from mis-
understanding instructions, familiarity with the equipment, motivation to
do the task, sensory acuity, learned response strategies, time spent on sen-
sory pro cessing and motor action rather than decision time, attention,
arousal, task orientation, confi dence, and anxiety. Such research appears
to be up another cul de sac. But, like frustrated yet hopeful prospectors,
IQ devotees keep returning to RT.
Ability for Complex Cognition?
Of course, the favorite claim is that IQ somehow refl ects individual dif-
ferences in the ability for complex cognition (reasoning, thinking, prob-
lem solving, etc.). Linda Gottfredson, in an article in 2007, claimed that
“the active ingredient in tests of intelligence is the complexity of their
items... that makes some more diffi cult than others (more abstract, more
distracting information, require inferences, etc.).” She provides examples
of such complexity, as seen in test items. For example, completing a 3 × 3
matrix item like those in the Raven test (see fi gure 3.1) requires moreThis content downloaded from 139.184.14.159 on Tue, 17 Oct 2017 13:52:12 UTC