For some, the IQ, intelligence quotient, is a magic word. Knowing that one has a high IQ definitely boosts his/her self-esteem, even though he/she may be in a career trouble. A widely held view is that Jewish and Asians have the highest IQ, followed by whites, both of which are slightly above average (100), while blacks are stigmatized by having the lowest IQ. There is no doubt that something fishy is behind this conclusion; but ironically, empirical results repeatedly show the same order of IQ by race, although some inconsistencies exist.

The recent effort by (in)famous Philippe Rushton from the U of Western Ontario and Arthur Jensen from UC Berkeley demands some attentions, if not critiques. The 60-page review in the current Psychology, Public Policy, and Law (2005, 11:2 235-294) traced 30 years of research on racial differences in intelligence measurements. An interesting note is that the authors used “cognitive ability” instead of “intelligence” in the title, presumably to broad the paper’s scope and assume an objective attitude.

This lengthy review concluded that the racial difference in intelligence indeed had genetic component. It was the group biological characteristics, not the social or environmental factors that underlay the racial difference of intelligence. Although some may be offended and willing to jump onto them and beat them to death, cool-headed researchers would like to choose a more scientific approach. That is, we need to critically evaluate their evidence and interpretations.

Given the massive evidence, complicated study design, and possible misinterpretation even among trained researchers, several evaluation criteria should be established beforehand.

First, what kind of research we value the most? There are several types of studies. At the first line is the ecological study in which the population average differences of IQ are compared and sought to be explained by population level (group level) characteristics such as average education or income. However, the group level variables don’t necessary reflect the individual characteristics. For example, those having lower IQ may have higher income, even though at the group level, those affluent countries had average higher IQ than those poor African countries. Due to this ecological fallacy, this type of study would be discarded as evidence.

The second line study is cross-sectional study at individual level. Here everything is assessed at individual (person) level. However, in terms of explaining whether race, biological and social factors are causal factors of IQ differences, the cross-sectional studies are of little value, except for exploratory purpose. The reversal cause-effect pathway is always possible. Causes should occur before effects. Therefore, interpreting results from these studies should be cautious.

To solve the time problem, observational follow-up studies in which individuals are followed for a certain period to collect the outcome of interest, are of the most importance, given that experimental studies are not practical and rare in social science. In experimental studies, people are assigned to different groups (either randomly or quasi-randomly). Random experimental studies are “gold standard” for causal inference. Fortunately, there are “natural experiments” such as migrations, adoption studies, and twin studies which have similar power to assess causal effect as those experimental studies. I will emphasize these studies in later discussion.

Next, what is the intelligence? An uneducated person may be a skilled craftman. Nobody will say he/she is not intelligent. On the other hand, the popular IQ only measures reading, memory, mathematics and logic, and some motion abilities. It is not a complete and accurate representation of intelligence. It excludes aspects such as emotion control, leadership, and athletic abilities which are more important in future success. Furthermore, IQ is measured based on tests, mostly in verbal and written language, thus subjected to bias. Objective measures are more relevant. In Rushton’s review, the authors indeed focused on measures such as brain size and weight, reaction time, and even brain metabolic activities using functional magnet resonance imaging (fMRI). However, how these measures relate to general intelligence is largely unknown.

The third important point one must keep in mind that “the cause of individual differences within groups has no necessary implication for the cause of the average difference between groups” (p5). That is, although intelligence has a large part of genetic component (e.g., more than 50% heritability) which may explain IQ difference between Joe (black) and Jane (white), it does not necessary explain the difference between overall black and white difference. It is very possible that social factors such as education between blacks and whites may cause the group level difference, as it is well known that blacks have low average education.

Finally, the concept of heritability should be carefully defined. Heritability is computed as the genetic contribution to the individual differences in a particular population, statistically assessed by percent of total variance explained by genetic variation. Therefore, the heritability is a concept of proportion or probability. It is used to describe population, not individuals. Given the complicate gene-gene, gene-environment, and environment-environment interactions, there is no specific gene that fully explains the intelligence. Even if a phenotypic trait such as intelligence is largely genetically determined (which is not necessary true), environment such as education can change it.

Now we are ready to formally critique the review (please forgive me for the above lengthy preamble). The debate is “whether any significant part of the mean black-white IQ difference is due to genetic rather than purely cultural or environmental in origin”. Notice that the question is at group level, not at individual level. In addition, it is also possible that the contribution of environmental factors to IQ may be different within blacks and whites. For example, education may be more important in improving one’s IQ in one group than in the other group. The same is true for genetic contributions, that is, the effect of genetics may be different within blacks and whites. Therefore, the statement that the difference of mean IQ between blacks and whites is due to genetics actually says that given the same environmental conditions, blacks still have lower IQ than whites. The inference is made from subtraction, an indirect method.

The hereditarians assert that the persistent racial difference of IQ (and intelligence in general) is due to genetics, even if everything else is the same. On the contrary, culture-only sociologists (environmentalists) hold the view that it is social environment factors that explain the racial difference of IQ. Improving these factors will eliminate the IQ difference. One needs to remind him/herself that it is not uncommon in science that the same results may have different interpretations, and both could be true.

Now here come my critiques.

My first question is what the race is. With convenience, race is commonly defined by physical characteristics such as skin color and facial shape. It is more politically correct to define race based on geographic origin such African, East Asian, and European descendents. Ideally, one wish to define race based on genetic composition such as polymorphisms if we assume race exist as a biological term. Given the pervasive racial mixture in the US and Latin America, most blacks have some percent of white “blood”. Therefore, the genetically deterministic view between whites and blacks in the US is skeptical (discuss in detail later). Nevertheless, racial comparison worldwide still makes sense given most African never travel out of their hometown.

To be continued……