I’m about to do something dangerous.
I want to address something that, while unable to do so in another class, has been bothering me since I read about it, and that deserves more attention than it likely gets. I doubt most students are pleased to approach this subject, but approached it shall be, nonetheless. I want to discuss the disparate scores that follow racial lines in tests of intelligence, and how communication, not intelligence, is a potential explanation for the issue.
Communication is defined as “the exchange of information between individuals” (Penn State University, 2016, 1) in very basic terms. Communication, however, takes many forms. One form of communication that is often discounted is the administration and subsequent scoring of written tests. Consider that for a moment; the test writer has devised a series of questions whereby they attempt to glean understanding of the test-taker’s knowledge of a subject. By contrast, the test taker is attempting to communicate their proficiency by means of answering the questions as accurately as possible. It falls into the category of an exchange of information between individuals, and has a defined purpose, which most communication of a serious nature will.
So here’s the thing: decades of research has gone into the application of tests that people take for a variety of reasons; everything from the teller survey at the bank to the GMAT exam has been thoroughly researched and studied for effectiveness, creating a linkage of communication whereby answers might be gleaned about a particular paradigm. Intelligence tests are one such item; a carefully designed examination asking questions to which the answers can show a particular amount of cognitive ability; perhaps the most well known is the Wonderlic Personnel Test, a 50 question exam administered over the course of 12 minutes, made famous for its use as an intelligence measurement tool by the National Football League. There are probably as many intelligence tests out there as there are blades of grass, but the Wonderlic, like other accepted tests, is able to boast high quality results thanks to construct validity; most intelligence tests in common use agree with one another to a high statistical level.
So why is it that there are clearly delineated scores along racial boundary lines?
The question is posed in part by a study from the mid ’90’s that was published as a book called “The Bell Curve.” In it, authors Charles Murray and Richard Herrnstein discuss, among other things, what we refer to today as “adverse impact” as it relates to the ratio of scores between two particular races (Penn State University, 2016, 2). This concept, like many others in posts I tend to write, was difficult to research, as no one seemed willing to approach the subject directly. When one simply types, “the bell curve book” into Google’s search bar, the results of the entire first several pages are all refutations and arguments against the research and the positions taken by Herrnstein and Murray, a snap-to-attention cattle call of the anger that is so near the surface of anything related to the racial relations in America today. Actual research is harder to find.
But finding research, when there is research to be found, is never impossible, and what I found sobered me. A study carried out in 2005 found that the two races in question had a consistent disparity of 1.o-1.1 standard deviations of difference in terms of IQ test results (Ruston & Jensen, 2005). In a personnel selection class taught at Penn State, the class text discussed the concept of “adverse impact” subsequently needed to be controlled for, as disparate scores due to racial precedent could cause an Equal Opportunity issue (Gatewood, Field, Barrick, 2016). It would seem that the research is out there, but is difficult to find, much of it having been explicitly to refute “The Bell Curve” and the various studies that have found similar results. I find that troublesome, as the scientific method requires objectivity; if you set out to refute something in an experiment, you will. Unfortunately, that invalidates much of the research on the issue. As the Penn State lesson states, “In response to The Bell Curve, many of the highest profile psychologists studying intelligence endorsed the previously mentioned one-page summary of scientific findings on intelligence testing, titled “Mainstream Science on Intelligence”, in The Wall Street Journal” ( Penn State University, 2016, 2). Anytime one sees someone endorsing something (or someone) because ‘the alternative can’t be correct/right/good,’ you have a serious problem.
I’m going to be as clear as possible; I do not believe the issue to be one of racial superiority/inferiority. Rather, I’m looking at this from the vantage point of communication: if the results are skewed to one side, rather than run around in the testing room like a chicken with its head cut off, crying out “diversity, diversity!” we need to look at the problem objectively, as there is a miscommunication of some sort happening between certain test takers and the test, which is driving the results difference. If we can find where the communication goes wrong, we can probably find a way to close the gap between the races and improve everyone’s scores, imbuing confidence and pride in intelligence into an entire generation of Americans, race immaterial. Yes, the research may first yield values that people could lose their careers over; right now does not feel like a good time in America to start running experiments on racial differences on intelligence tests. But being willing to address the problem is the only way to fix it; endorsing previously unendorsed articles because they don’t agree with a controversial study isn’t going to get us anywhere.
Gatewood, R., Feild, H., Barrick., M. (2016). Human Resource Selection. 8th Ed. Boston, Cengage.
Penn State University. (2016). OLEAD 410, Lesson 4: Global Communication. Retrieved from https://psu.instructure.com/courses/1802572/modules/items/21179072
Penn State University. (2016). PSYCH 490, Lesson 4: Intelligence. Retrieved from https://psu.instructure.com/courses/1803843/modules/items/21158266
Rushton, J. Philippe; Jensen, Arthur R (2005). “Thirty Years of Research on Race Differences in Cognitive Ability” (PDF). Psychology, Public Policy and Law.
Thank you for posting such a thorough follow-up, Dr. Redmond. You manage to articulate exactly what my original post was going for; that the inflexibility of micro-level influences on the communication between a test and a test taker could be responsible for a part of the cultural/race gap in IQ tests, and should be looked into further to help encourage a potential fix on a global scale, ending the conversation with a ‘got it!’ rather than essentially shelving what is a very useful tool (intelligence tests) due to their potential adverse impact on certain groups. If I get a chance this semester, I’ll try to share a bit of the research I did yesterday at the expense of being even remotely productive at the office (oh, how aspiring grad students spend their idle hours!). I’d imagine all parties in the discussion would be interested to see some of it. It’s just not quite ready yet.
Also, if I actually managed to hurt anyone’s feelings… well, only you can decide whether or not I am a heathen and a bastard or a gentlemanly rogue; also, a bit of a nerd in either case. Let’s all agree to get along; learning is more fun that way, and we’ll probably be able to teach each other something.
Cheers,
-JR
This conversation got a bit flame-like rather fast, so I’m going to hop in here and try to clarify a few things and hopefully settle the conversation down. But as this topic is a bit beyond the scope of this class (it’s actually a psychology topic where a ton of research has been done on this idea, and I encourage you to read more about it) I’ll keep this much shorter than it could be.
But before getting into the details, I will ask that in the future rather than jumping to conclusions about others or making snarky remarks, that everyone asks clarifying questions of the other person, as that leads to greater learning opportunities for all. In particular, you may find that you are saying the same thing but using different cultural perspectives or understanding.
Intelligence testing has been around for a long time. They originated with Louis Terman around 1912 in France. In that time, they have primarily been developed by psychologists in Europe and the United states at academic institutions, which until very recently have been primarily white males of European descent both at the researcher and student level. The researchers, as with all humans, unintentionally incorporated their own cultural norms into the test when the wrote the tests. The students that the researchers then validated the tests on, also unintentionally contributed their culture to the test. So over time, white European culture became a subtle background for these tests. In particular the way they are worded, so they tend to show differences between races when in reality none truly exist.
An analogy that I like to use that we would think entirely absurd is giving an English intelligence test to someone who speaks another language, for example Japanese. The Japanese person would fail the intelligence test, but we know quite clearly that is because of poor communication between the tester and the test taker via the medium of communication of the test, not because the Japanese person is less intelligent. As a test is not flexible in the way it communicates ideas the way people can be in person, those tests accumulate data that shows differences between races.
Again, there has been a lot of research on this topic and I encourage everyone to learn more about it as this is an important time in American history to make change in this area. Here are some tips for searching in the library (I do not recommend using the public internet as it will be nearly impossible to sort real science from people spouting hate or rumor posing as “Science”):
First, narrow the search to “Scholarly and Peer-reviewed”
Second I would further narrow to the field of “psychology”
These first two steps reduce the number of articles from hundreds of thousands to roughly 14,000, which is a much more manageable to comb through.
I would then use variations in search terms like “Race” “IQ” “Intelligence Test” and “Cultural Bias”. Something else you might want to try to add to that list is “meta-analysis” which is a summary type of article (typically 10 years or more) that reviews all the literature on a topic.
There are a lot of good articles, but if you are looking for a quick summary Kwate (2001) does a decent job discussing the problems with intelligence testing across races that I summarized above.
Reference
Kwate, N.O.A. (2001). Intelligence or Misorientation? Eurocentrism in the WISC-III. The Journal of Black Psychology, 27(2). 221-238.
This actually is a pretty good conversation. The discussion of race is obviously a complicated and touchy one. I’m going to return to add some points to this conversation later this week as there is some empirical evidence and there is some theory as well behind this. It’s been a long time since I read up on this, so I have to go back and do some digging. But knowing both of you, I know that no one here is racist and that the intent of the words is to broach a tough topic and hopefully actually reduce racism, which is the whole point of psychologists and other social scientists collecting this data in the first place. The state of the world has to be known before any solution can be proposed.
I don’t think you understood my message correctly. I was putting forth the suggestion that, for whatever reason (since no other explanation has been validated), communication in the form of tests and their administration could be to blame for discrepancies that fall along racial lines. This is a broad look; it could be in the wording of the questions (cultural/language disconnect), the nature of how something like problem-solving is being measured (triggering fluid intelligence as opposed to memory recall), and is a possible source in the discrepancies on the tests that are found worldwide. Education, while a part of the discrepancy, does not effectively explain (see the sources I listed in the main comments section below the post). Education does not have a serious impact on fluid intelligence. IQ scores rarely vary from adolescence through adulthood. While memory recall plays a part, it is not the primary cognitive factor being measured.
My major is clearly bleeding through, here. Essentially, education plays a smaller part, the larger part of the problem is undefined. Given the nature of global communication in the 21st century, it is possible that improvements in standardized communication of cognitive capacity would help to close a gap that shouldn’t be there.
I hope (for all of the CogPsych stuff) that this makes more sense.
I’ve replied in the main comments section to my post. I really shouldn’t have bothered (since we both know what your opinion is) but I wanted there to be opportunity for discussion. Understand, I’m not alleging empirical evidence for a race divide, but looking for a valid, quantifiable explanation as to why the testing gap exists to begin with, and solutions as to what can be done. Given your apparent propensity for all things race-related, I would love your thoughts on the research in the articles cited and on the general topic.
-J
Yes, Marvin, clearly my addressing a difficult question that could perhaps be answered by research into the way communication (and a potential lack thereof) is carried out on intelligence tests is just a “thin veil of racism.” Veil is spelled with an ‘e,’ by the way.
As to the question of access, there are two points that I could have fleshed out better in the post, which I’ll do now (though I doubt you’ll take notice), using real science! Instead of, you know, whatever it was your rather condescending comment was going for.
First point: not all intelligence tests have anything to do with memory recall, but with fluid intelligence; while related to each other, you don’t have to have a post-doc to be able to recognize patterns or predict trends (Salthouse, Pink, 2008). Second, to that point, research has indicated that access to education is a part of the issue, but does not account for more than a part of the problem (Hunt, 2010). In fact, there is substantial evidence to the contrary (Mackintosh, 2011). Basically, research has determined very little, only that most suggestions of why the discrepancy is there are incorrect. There is, therefore, no answer known as to why there is a dichotomy between certain races (but hint: whites aren’t at the top!), and I suggested that communication methods between test takers and test administrators (on a global scale, see the citations) could be an avenue that has not yet been considered.
I’m not going to spell out the science for you; you can look up my sources yourself. But you did prove a part of my point in the post; it’s a dangerous subject to look into objectively, as the allegations of racism and rather nasty comments begin flying before the scientific method can ever work its way out of the closet to try and find a good answer to the problem (just look at your own behavior for reference). I hope for the sake of the protected groups who score lower on average than others that opinions like yours can be muted long enough to find the actual problem and fix it. Diversity is meant to be achieved smartly, not in the involuntary movements of a decapitated fowl. That gets us nowhere as a society but onto the dinner table, extra salty for the experience.
Works Cited (Science!):
Hunt, E. (2010). Human Intelligence. 1st Ed. London, Cambridge University Press.
Mackintosh, N. (2011). IQ And Human Intelligence. 2nd Ed. London, Oxford University Press.
Salthouse, T.A., Pink, J.E. (2008). Why is Working Memory Related to Fluid Intelligence? US National Library of Medicine. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2485208/
If I understand your message correctly you are saying that sometimes IQ test are set in the manner that certain protected status groups are less capable of succeeding at due to the design of the test. I believe that the results are less connected to the matter of race but more so the level of education people taking the test have. While the test is not set up by deliberate intent to discourage certain groups from taking the test and diminish their potential I think that it is unnecessary to think it is due to the race of individual and not educational level of the person taking the test.
So when you say the discrepancy should be addressed I believe that it should be done by providing good quality done, the steps to which have already be done.
I cant help but think this post blanketed by a thin vail of racism, as you seem to disregard the advantages and disadvantages that some races are exposed to, comparatively to others. Communications is not the sole problem, but rather, access. Access to better education and broader opportunities for growth would improve test scores. “Communication” is not as directly correlated to the improvement of test scores like your assertion suggests. I also don’t think that wanting diversity should be compared to the involuntary movements of a decapitated fowl. Privilege allows you to speak such away, and here I would agree that maybe if you had access to more effective communication, I would not confuse your analogy for an insult–but great post!