INTELLIGENCE IN CHINA svN&~@l
RICHARD LYNN Dc
U$sf*
University of Ulster, Northern Ireland s~3"*
,3@
Studies of the intelligence of' Oriental peoples in Japan, Hong Kong, Taiwan, Singapore Y-v6xUc{F
and the United States have typically reported slightly higher mean IQs than those of [
&51m^
British and American Caucasoids. Recently results have become available for a 0`:0m/fsU
standardization of Raven's Standard Progressive Matrices in the People's Republic of A\LMmg
China. The results show that Chinese 6-15 year-olds obtain a mean IQ of 102.1 in relation k*J0K=U|
to an American Caucasoid standard of I00. r3'0{Nn+
A number of studies have shown that the Oriental or Mongoloid peoples EW]rD
have higher mean IQs than whites or Caucasoids of European origin in the cJMp`DQzc
United States and Britain. Much of the literature on this question is reviewed w~Aw?75t
in Lynn (1987). The IQ advantage of Mongoloid populations has generally ^tI
,eZ
been found to lie between 2 to 8 IQ points. A number of studies of 6YNd;,it>p
intelligence in Japan indicate that the mean IQ is approximately 105. Studies #mFIZMTRd
of the intelligence of children in Taiwan and Singapore have obtained mean p[>!;qI
IQs of approximately 103. A recent study of children in Hong Kong obtained `@RTfBBg
a mean IQ of 110 (Lynn, Pagliari & Chan, 1988). The extensive literature :wtK'ld
on the intelligence of ethnic Orientals in the United States was reviewed by tw,uV)xm
Vernon (1982) who concluded that the mean non-verbal IQ was about 110 FG/1!8F
and the verbal IQ about 97, which can be averaged to give a figure of 103.5. y2cYRHN[X}
Although data are now available on the intelligence levels of Mongoloids m9 1Gc?c
in a variety of locations, the jewel in the crown is missing. This is the 7*.nd
intelligence of the population in the People's Republic of China. The interest OQ&l/|{O0?
in obtaining data on intelligence in China is twofold. Firstly, this is the @P@?KZ..v!
homeland of by far the greatest number of Mongoloids. Secondly, the high ''tCtG"
Xi
intelligence levels of Mongoloids in several locations could have been ]V^ >aUlj
determined by selective emigration from China. This is suggested by Flynn c~SR@ZU
(1989) for the Chinese in the United States, and the same argument could be MaN6bM
applied to the Chinese populations of Taiwan, Hong Kong and Singapore. a[I
: ^S
Data on the intelligence level of the population of mainland China are crucial \8g'v@$wG
for the resolution of this problem. The required data have recently become VX0}x+LJ
available and are the subject of this report. "Doz~R\\
METHOD W^k95%zBM
A Chinese standardization of the Progressive Matrices was carried out by Q\rf J||
Hou Can Zhang of Beijing Normal University in the mid-nineteen eighties. {OIB/
The standardization sample consisted of 5,108 individuals drawn as a uVN2}3!)Y
stratified sample from the 6 principal administrative areas of China and from }]<0!q &xB
small, medium sized and large towns within these areas. The results are given 9(6f:D
in the form of norm tables in Raven and Court (1989). These norm tables >P@g].Q-
consist of raw scores for a number of age groups and the percentile ;Nk,bb K
equivalents of raw scores, set out in the same format as those given for the F
JCs$0
United States in Raven (1986) and for Britain in Raven (1981). ~bf4_5
RESULTS nvbzC tC
The American norm tables give the most precise percentile equivalents for -!q^/ux
raw scores and for this reason are the most satisfactory standard with which 5If.[j{
to compare results from China and elsewhere. For the Chinese data it is cDS\=Bf
possible to derive American percentile equivalents for 20 age groups over the CVE(N/&b
age range 6-15 years. These have been collapsed into 5 age groups, the %tB7 &%ut
percentiles transformed into IQs and the Chinese results given in relation to Y O
&@
American IQ means of 100 and SDs of 15 in Table 1. Mean IQs of British ]n}aePl}oU
children are also given in the Table, making the same calculations from the z;DNl#|!L
British standardisation sample. C cPOK2
TABLE 1: MEAN SCORES AND IQs OF AMERICAN, BRITISH s@zO`uBc
AND CHINESE CHILDREN ON RAVEN'S STANDARD (1
(~r"4I
PROGRESSIVE MATRICES. G>=Fdt7Oc
It will be seen that the Chinese children obtain consistently higher means z^vfha
than the American and a marginally higher overall mean than the British. ijP`fM8
The mean IQ of American children is depressed by the presence of .J<t]
substantial numbers of blacks in the population. The mean IQ of white ^hNl6)hR
Americans is 102.2 (Jensen and Reynolds, 1982). Hence in relation to an `R;i1/
American Caucasoid mean IQ of 100, British children obtain a mean of 101.2 =WT&unw}
and Chinese children of 102.1. The statistical significance of the higher mean SSxp!E'
obtained by the Chinese children in relation to the American can be tested by ,BUrZA2\U$
calculating the standard errors. The difference between the means is greater ,oykOda:|
than twice the standard errors and can therefore be considered statistically |M`'
significant. The Chinese-British difference is not statistically significant. tq}MzKI*
DISCUSSION /% g+|C
The results show that the intelligence level of children and adolescents in G 9|2
KUG
mainland China is slightly higher than that of Caucasoids in the United States 'i',M+0>jC
and in Britain, although the Chinese-British difference is not statistically S/"G=^~
significant. In evaluating the result, account needs to be taken of the very low By waD?
living standards in China. =^1jVaAL
Few dispute that intelligence is to some degree determined by envi- |sd G<+
ronmental conditions and standards of living. Thus children adopted by oqAO@<dL!
middle-class families enjoying good living conditions obtain higher IQs than & .0A%
their siblings reared in working-class families (Dumaret, 1985; Capron & da<