Educability-and-Group-Differences-1973-by-Arthur-Robert-Jensen

68 Educability and Group Differences
NOTES
1. ‘Population’ is a statistical concept. A population is defined by a set
of operational criteria by which individuals are either included or
excluded as members of the population. A population is thus
defined also as all individuals who meet the criteria for inclusion.
Mean values of various phenotypes in the population and their
dispersion (variance or average difference among members of the
population) are estimated on random samples drawn from the
defined population. A sample is truly random when every member
of the population has an equal chance of being included in the
sample.
2. The term ‘genetic value’, (also called genic or genotype value), has
a technical quantitative meaning. It does not imply any ‘value
judgment’ about the trait in question. An individual’s ‘genetic
value’ is his hypothetical position on the trait’s scale of measurement
if he had developed in the ‘average’ environment of the
population of which he is a member and in which h2 has been
estimated. Conversely, the individual’s ‘environmental value’ is his
hypothetical position on the trait scale if his genetic endowment
were the average of the population. While ‘genetic value’ refers to
all the genetic components that constitute the individual’s genotype,
the term ‘breeding value’ refers only to the additive genetic effects,
or heritability in the narrow sense. An individual’s breeding value
is the hypothetical mean position on the scale of trait measurement
of all of the individual’s potential offsprings when his mate or mates
are genetically at the population mean and when the offspring have
developed in an average environment for the population.
3. It should be noted that a multiplicative model depends upon a scale
with a true zero point and no zero or negative value; the multiplicative
model cannot be applied to deviations from means, which is
what practically all psychological test scores consist of. In any case,
there is little difference in predictive precision between an additive
and a multiplicative model unless the variations in G and E are
very large in comparison to their means. There would be no
practical difference between the models in estimating heritability
when the coefficients of variation (i.e., the ratios or/Mean G and
txE/Mean E) are less than about 0-20.
4. A detailed treatment of the use of transformation of the scale of
measurement in quantitative genetics can be found in Chapter 17
of Falconer’s Introduction to Quantitative Genetics (1960).
5. Epistasis is the interactive (i.e., non-additive) effect among genes at
different loci in the chromosomes. For example, the phenotypic
effect of gene A plus gene B may be greater (or less) than the sum