Flynn effect - IQ increase 3 pts/decade

Hal (hfinney@shell.portal.com)
Tue, 17 Oct 1995 15:33:48 -0700


I found Flynn's papers at the library, and they are quite interesting.
"The Mean IQ of Americans: Massive Gains 1932 to 1978", Psychological
Bulletin 1984, Vol. 95, No. 1, 29-51. Abstract:

"This study shows that every Stanford-Binet and Wechsler standardization
sample from 1932 to 1978 established norms of a higher standard than its
predecessor. The obvious interpretation of this pattern is that
representative samples of Americans did better and better on IQ tests
over a period of 46 years, the total gain amounting to a rise in mean IQ
of 13.8 points. The implications of this finding are developed: The
combination of IQ gains and the decline in Scholastic Aptitude Test
scores seems almost inexplicable; obsolete norms have acted as an
unrecognized confounding variable in hundreds of studies; and IQ gains of
this magnitude pose a serious problem of causal explanation."

By the time of a later paper he has an explanation of a sort.
"Massive IQ Gains in 14 Nations: What IQ Tests Really Measure",
Psychological Bulletin 1987, Vol. 101, No. 2, 171-191. From page 187:

"The literature makes clear what real-world behavior we have a right to
expect from those at various high IQ levels: Above 130 they find
school easy and can succeed in virtually any occupation; above 140
their adult achievements are so clear that they fill the pages of
American Men of Science and Who's Who; above 150 they begin to
duplicate the life histories of the famous geniuses who have made
creative contributions to our civilization." Flynn then goes on to
estimate that the percentage of the Dutch population who would have
scored at 150 or above in 1952 was .04%, but if people today took those
same tests the percentage who would achieve that score would be 2.27%,
an increase by a factor of almost 60, "which means that the Netherlands
alone has over 300,000 people who qualify as potential geniuses. The
result should be a cultural renaissance too great to be overlooked."

Flynn then commisioned a study of a variety of European newspapers and
magazines to see if there had been any reports about a dramatic
increase in genius and mathematical and scientific discovery, but there
were none. (This surprises me, because comments about the
ever-increasing pace of science and human knowledge are something of a
cliche, but perhaps this is never expressed in terms of increased
quality of scientists, rather a matter of quantity.)

"The only hypothesis that can cover the facts is that the Ravens test
does not measure intelligence but rather a correlate with a weak causal
link to intelligence.... The Ravens test measures a correlate of
intelligence that ranks people sensibly for both 1952 and 1982, but whose
causal link is too weak to rank generations over time." Flynn proposes
therefore that such tests do not measure intelligence but what he calls
for lack of a better term "abstract problem-solving ability (APSA)".

He goes on to explain the fall in SAT scores in the face of increases in
abstract IQ test scores as an indication that SAT tests do not measure
APSA. Rather, they rely more heavily on concrete knowledge at a high
school level. APSA may be helpful in some parts of the test but
deterioration in the quality of education is a larger effect.

I have not yet read these papers in detail, but I don't find Flynn's
arguments completely convincing. It's not clear that he is justified in
taking his Netherlands data and concluding that 2% of people today would
score at 150 on a 1950's IQ test. The data from 1950 were not that
complete and he had to put a few different studies together to estimate
the distribution. Also, by 1982 enough people were getting close to a
perfect score that the distribution was becoming non-normal and that had
to be corrected for as well.

As for the claim that we should notice if we were living in a society
full of people who would have been considered geniuses in the past,
that's not so clear. Maybe it is partially a matter of competition -
success in life is not so much a matter of how smart you are, but of
how much smarter you are than everyone else. And as I mentioned above,
actually it does seem like progress in some areas has accelerated.
Maybe people have moved into fields where individual achievement is
less obvious. Thousands of anonymous programmers and chemists toil
away in industrial settings for good salaries rather than writing
papers and becoming famous. There are just so many cultural variables
that I would hesitate to rule out the possibility that people really
are more intelligent on average today. Checking the newspapers hardly
seems like the right way to answer the question.

Hal