PRETEND GENES 45Th e trait value is then thought to be an expression of the corresponding
ge ne tic charge, more or less attenuated by the environment (of which I have
more to say later in the chapter). You do not need to be a statistical wizard to
appreciate that in a large population of such random gene combinations,
the distribution of trait will be like the normal curve: lots of individuals
having middling values and diminishing numbers of them toward the ex-
treme values. Th is is the famous bell- shaped curve. It is shown for human
height in fi gure 2.2. Behavioral ge ne ticists have to assume that intelligence
is made up and distributed in exactly that way. It is the only way their statis-
tical models can be applied to data. We know that height is normally dis-
tributed. But does that distribution apply to intelligence?
FIGURE 2.2
Th e normal curve of height measurements— shorter values to the left , taller values to
the right.Height intervalsFrequencyBell- Shaped Curves and Unicorns
Th e prob lem is that heritability estimates can be seriously fl awed if the
population distribution is not bell shaped— for example, if it is skewed
one way or the other. What the behavioral ge ne ticists fail to mention
is that IQ tests have been deliberately constructed so that the scores
will exhibit such a distribution. Th is is done by devising test items andThis content downloaded from 139.184.14.159 on Tue, 17 Oct 2017 13:51:26 UTC