trying them out in advance. Th en those items that around 50 percent of
testees get right are kept in the test, along with smaller proportions of the
items that either very many or very few testees get right: “It is common
practise to carry out item analy sis in such a way that only items that
contribute to normality are selected.”^6
As it happens, few natu ral biological traits are distributed in that way.
Mea sures on almost all basic physiological processes— visual acuity, resting
heart rate, basal metabolic rate, and so on—do not display normal distribu-
tions. And of course, myriad human traits— language variation, tastes in
dress or food, use of tools and technologies, knowledge of the world, and
ways of thinking— vary at a completely diff er ent, cultural, level compared
with physical traits. In the 1980s, Th eodore Micceri amusingly titled a
review of such distributions “Th e Unicorn, the Normal Curve, and Other
Improbable Creatures.”
Ironically, those traits that are most impor tant to survival are the
very ones that do not have a normal distribution. Natu ral se lection
itself produces more and more phenotypes with values that would have
previously been above average. Th at is, of course, the aim of artifi cial
se lection in crops and animals.
We would surely expect most aspects of human potential to come into
that category of “impor tant to survival.” Indeed, some recent brain studies
by György Buzsáki and Kenji Mizuseki concluded that “at many physi-
ological and anatomical levels in the brain, the distribution of numerous
par ameters is in fact strongly skewed... suggesting that skewed...
distributions are fundamental to structural and functional brain organ-
ization. Th is insight... has implications for how we should collect and
analyse data.”^7
Buzsáki and Mizuseki go on to review the overwhelming evidence
against the bell- shaped distributions of psychological functions. From
sensory acuity and reaction times, to memory word usage and sentence
lengths, individuals simply do not vary that way.
I return to this question below. In the meantime, it needs to be
stressed that if the bell- shaped curve is the myth it seems to be for
these traits, then the model used in behavioral ge ne tics is already
suspect, the statistics are inappropriate, and estimates for causes of
variation may be seriously wrong.
This content downloaded from 139.184.14.159 on Tue, 17 Oct 2017 13:51:26 UTChttp://www.ebook3000.com