"If something exists, it exists in some amount.
If it exists in some amount, then it is capable
of being measured."
René Descartes, Principles of Philosophy, 1644
The first intelligence test was the
Binet-Simon
test, developed by Alfred Binet, a French educator, in 1905. Binet was
asked by the education department to develop a way of identifying
children who were behind in their academic performance so that they
could receive remedial education.
In these pre-IQ days, the interest was in
comparing a child's score on the test with peers of the same age, to
determine who needed extra help.
The Binet-Simon test was revised in 1908 and
1911 and then received a major revision in 1916 and was renamed the
Stanford-Binet Intelligence Test. Part of the revision was the
invention of the "Intelligence Quotient", better well-known as IQ.
Lewis Terman
(an American professor from Stanford University) then worked with Binet
to produce a major revision in 1916. The Binet-Simon test was
renamed as the Stanford-Binet test. The Stanford-Binet can be used
with children from the age of about 2 and a half years.
More about the
Stanford-Binet Intelligence Test...
Part of Terman's major revision was to
re-conceptualize the relationship between Chronological Age and Mental
Age.
Terman realised that the ratio between MA and CA
was a better indicator of intelligence than MA alone e.g., for example
The formula for calculating IQ developed by
Terman was 100 x MA / CA.
There are some problems with calculating I.Q.
scores in this way.
For example, do Person 1 and Person 2 have the
same level of intellectual superiority?
A second problem with the IQ = 100 x MA/CA
formula is in measuring adults:
Mental Age does not steadily increase throughout
the lifespan, but Chronological Age does. So, an 85-year is not likely
to have gained any greater capacity than when he/she was an 60-year old.
Mental Age levels off around the end of adolescence, but Chronological
Age gets higher, resulting in lower IQ scores as adults get older!
A third problem was that IQs do not fall
strictly along a bell-shaped curve. There is a higher incidence of very
high (there something like 50 times are many people with very high IQs
than expected from a normal curve), with a similar though not as large
effect for those with very low scores.
Read more...
A fourth problem was that it was difficult to
convert to IQ scores from other achievement and ability tests.
Because of these problems, MA is no longer used
in calculating IQ scores - instead "deviation IQ" is used.
Deviation IQ uses a scale which is based on the
actual rarity with which ratio IQ scores actually occur. Thus, it
deflates the bulges in the curve, and makes a smooth bell-curve out of
the data. Basically, people are allocated to percentiles rather than
actual scores, so "deviation IQs" perfectly fit a normal curve. This is
pretty simple to do, by looking up a chart that shows the match between
someone's percentile rating and deviation IQ scores (based on a
distribution with a mean of 100 and a standard deviation of 15). See Generic IQ Chart; also see Murphy & Davidshofer (p.31).
For example, if one scores in the 96th
percentile on an IQ test, then this can be read off a generic IQ chart
as an IQ of 128.