Friday, November 2, 2007

Is the MCAT an IQ test?

Terry Morris of Webster's Blogspot noticed my reference to an IQ test being required for medical school admissions, and asked whether the standards for this exam are different for "minorities." I have been thinking about intelligence and medical school admissions lately, so I thought the question deserved a dedicated post here as an answer.

First, anyone who is familiar with the world of medical school admissions will recognize that, when I linked the phrase "IQ test" to the MCAT homepage, there was an implicit wink there. This is because the MCAT is supposedly a knowledge test, not an apititude test. Perhaps some background on the MCAT would be useful.

The Medical College Admissions Test (MCAT) is administered by the American Association of Medical Colleges (AAMC). It is divided into four sections: physical sciences, biological sciences, verbal reasoning, and the writing sample. The writing sample is generally considered unimportant. The other three sections are multiple-choice. The verbal reasoning section consists solely of a series of passages, typically excerpts from articles or essays on non-scientific topics, each followed by a few questions designed to examine the subject's understanding of the passage. The physical sciences and biological sciences are arranged mostly the same way; while there are a few stand-alone questions interspersed throughout, most of the questions are based on passages containing a few paragraphs of text on some scientific topic and sometimes a diagram, graph, or table. Officially, the physical sciences section tests one's knowledge of college-level introductory physics and general chemistry, while the biological sciences section tests one's knowledge of introductory biology and organic chemistry. Those are the core courses required for admission into all medical schools in the USA.

Each of the three multiple-choice sections is scored out of 15. The score is scaled relative to other examinees' scores, so simply dividing one's section score by 15 doesn't yield one's percentage of correct answers on that section. The score report, which goes to the examinee and all of the medical schools to which he applies, shows the examinee's score on each section, and the percentile for each section and for the total score relative to all others who took the test on the same day. To give an idea of what the distribution of scores is like, a total score above 30 is generally considered sufficient for admission to medical school, and the average score among students accepted to Harvard is 35. A score above 39 would place one in the 99th percentile, and scores above 42 almost never occur.

Virtually everyone thinks that the MCAT is a knowledge-based test rather than an aptitude test; that is, that it tests how well the examinee actually knows biology, general chemistry, organic chemistry, and physics. Pre-medical students taking these classes are always concerned about ensuring that the topics appearing on the MCAT are being covered. MCAT preparation courses run by companies like Kaplan or Princeton Review, while they regularly involve taking practice tests, focus on drilling the material. The AAMC publishes a complete list of scientific topics that are fair game for the test, which many students refer to when preparing. And after the exam, before scores have been released, everyone seems to attempt to gauge how well they did by how well they feel they knew the answers to the questions that were asked.

Yet I suspect the MCAT is really a test of critical reasoning; that is, it is closer to an IQ test than it is to a content test. First, there is the verbal reasoning section, which one can prepare for only by practice. To answer the questions requires critical reasoning, because they do not ask for content directly from the passage, but rather are oblique to it, asking for things like inferences which could be drawn from it or approximations of the author's implied but never directly stated position. The physical and biological sciences sections, to be fair, do require a good bit of background knowledge, but again, most of the questions do not simply ask directly about that knowledge but instead require interpretation of a passage, which often, but not always, requires that background knowledge to be fully understood. I can speak only from my own experience here, but my impression after taking the test and receiving my scores is that logical reasoning and critical thinking are more important than mastery of material, or perhaps that if we stipulate a basic level of understanding of the background material, one's logical reasoning and critical thinking skills are what determine one's score. I say this because I emerged from the test demoralized, having felt that I may have bombed the biological sciences section, because most of the passages seemed to be based on topics from advanced biology that had not been covered in the introductory biology courses I had taken. Yet when I received my scores, it turned out I had done quite well. This indicated to me that the ability to reason through the passages was more important than having memorized a large volume of facts.

The other reason I think the MCAT is an IQ test is that the results along race and sex lines are consistent with what we see in IQ tests. Believe it or not, the AAMC actually makes some of this data available on its website. In 2005, according to the most recent summary document available, men on average scored higher than women on every section, though the difference was least on the verbal reasoning section and greatest on the physical sciences section (which relies heavily on math skills), and women scored one grade better on the writing sample. Meanwhile, Asians and whites were equal in biological sciences, Asians edged out whites in physical sciences, and whites surpassed Asians in verbal reasoning. Blacks' averages were significantly lower than both groups in every section.

No one really knows whether the MCAT is an IQ test. Searches of both the general web and scholarly indices turned up nothing. It would be fascinating to see the results of a study designed to look into MCAT-IQ correlations, but in our political climate, I can't see such a study being done. Imagine submitting that grant application. Who would agree to fund it, knowing that the results will probably not only add fuel to the fire of undeniable group differences in intelligence, but also show that the medical school admissions process is a discriminatory one?

Now, before discussing the standards applied to "minorities," I must make two things clear, because of assumptions people tend to make about tests in general. I encountered these misconceptions many times during the admissions process, so I know it can be confusing for people not familiar with the MCAT. First, one does not pass or fail the MCAT; one simply gets a numerical score. Second, schools may do with that score what they will; there is no hard and fast cutoff for admissions. Think of the SAT--whether it's Harvard or your local State U, no college simply accepts everyone with a score above, say, 1200 and rejects everyone below. Just about every medical school will tell you that they consider many factors when making admissions decisions: GPA, academic background, extracurricular activities, work experience, volunteer experience, exposure to the medical field, research experience, letters of recommendation, the interview, and the MCAT. My anecdotal view is that because most schools are dominated by liberalism and wish to appear open-minded and humanistic and not narrowly focused on numbers, they tend to downplay the significance of the applicant's MCAT score and exaggerate the value they place on "soft" qualities like volunteer experience, but there is no doubt that other factors besides the MCAT are important.

That said, are the standards for the MCAT different for "minorities?" Undoubtedly. One can see this by looking at the AAMC's own published data. Among 2006 matriculants, blacks had the second lowest score in each of the three numerically scored sections, second only to Puerto Ricans. There is nothing more to say. Blacks, as well as Hispanics, Native Americans, and Pacific Islanders are accepted to medical school with lower MCAT scores than whites.

Why have I been putting the word "minorities" in quotation marks? Because not all minorities are created equal. Every racial or ethnic group in America right now other than whites are a "minority." But as you see from the AAMC's data, there is one minority group which sticks out like a sore thumb, being admitted to medical school with higher scores than whites: Asians, of course. Because of their high average IQs, Asians are over-represented in the upper echelons of society out of proportion to their numbers in the population. This has led to the creation of a new term, "under-represented minority" or URM. In medical school admissions, one does not speak of minorities, only of under-represented minorities. Though as you can see from the table, this is really just a euphemism for "non-Asian minorities," since every racial or ethnic group other than whites are "under-represented" in the medical profession, if you are a liberal and believe that every subset of the American population must reflect the racial proportions of the entire population.

So, the medical schools are sitting pretty. They can use an IQ test to screen their applicants, yet because no one thinks it's an IQ test, they don't have to admit they are doing so. At the same time, they can apply different standards to different racial groups, without taking heat for it because they are looking at the "whole person" rather than focusing on one narrow numerical score. It's a liberal's dream come true.

8 comments:

Atlantin said...

re: "Is the MCAT an IQ test?"

I took the MCAT in 1962. At that time it was a devastating test. Individuals threw up before and after the test as they understood their entire future depended on the exam, more so then than now, as at that time there were far fewer student positions in Medical schools than now and admission to a medical school relied heavily on the MCAT score. One had to take the test on a given day held only once a year at a designated location that required most students to travel to the location. I had to travel about 50 miles from my College and many had even further distances to cover to take the test. I took it at a large University lecture hall which held if every seat was taken more than 1000 people, maybe 2000. It was gigantic and had two levels, was oval and could be ( if movable walls were used ) divided in half or quarters. All the walls were retracted and the entire lot of seats were used with an empty row between each used row and and two empty seats between each test taker. Plenty of proctors walked the isles watching for cheating.

I can remember two questions that were given which at the time I had no clue as to their answer. One was define ESTIVATION and the other was define HANGING VALLEY. At the time I was a Chemistry major and Math minor.

The exam in 1962 was long enough that the test takers were given a break for lunch. But who could eat!

There was no question that the exam was an IQ test as when I entered Medical School in 1963. The Dean of the Ivy League Medical School told our new class that no one was admitted with an IQ under 130 and that over 20 had IQs over 150. The only way he could have known that was from the MCAT results.

Alas, since the early 1960s the MCAT exam has been constantly watered down to permit more women and "minorities" to appear to qualify for medical school. My sister, who was an Associate Professor of Internal Medicine at a Medical School ( until she accepted an executive position at a gigantic Insurance company to make some money ) served for about 15 years on the school's admission committee, has told me that the MCAT test was not nearly as difficult as it had been in the past. She also related that the intelligence of medical students declined over the years she was a professor.

The SAT exam has been also watered down over the years. The only exam to my knowledge that has not been degraded is the Graduate Record Exam. That was a really hard exam in the early 1960s when i took it.

If one is interested in IQ correlations with various tests check out this web site:
[www.iqcomparisonsite.com/GREIQ.aspx] For GRE
[www.iqcomparisonsite.com/Default.aspx] for general information

Anonymous said...

All tests are IQ tests. Some are just more g-loaded than others. Even the knowledge and IQ inter-correlate at some point; you wouldn't expect a low IQ person to wrestle with O-Chem or college for that matter. The LSAT, SAT, and pretty much all standardized tests correlate with IQ and the distribution of scores are predicted by highly g-loaded tests, even ones with no knowledge component like Raven's Matrices.

Anonymous said...

The MCAT, as I took it in 1993, was definitely a test of reasoning, much more than knowledge. I would say the biological science portion had a significant portion weighted toward shear knowledge of the field, but the physical sciences and verbal reasoning sections required very little prior knowledge of the subject. Reasoning using the content provided could probably get you the correct answer for about 95% of the questions in these latter two categories.

Anonymous said...

I don't see the point in this post. I must be missing something.

Anonymous said...

For a bunch of apparent doctors, you guys aren't that smart. Let's give standard IQ tests to a group of college students who have majored in liberal arts and have very little training in the physical and biological sciences. Suppose one person scores over the 90th percentile, and another person scores over the 70th percentile (these people come from comparable backgrounds, maybe they are even twins). Then have the 70th percentile person get a second degree in the biological or physical sciences. In the meantime, the 90th percentile scorer will pursue a master's in children's literature. Then, have the 70th percentile scorer study for the MCAT, including taking a prep course. Then have both of them sit down and take the MCAT. Nine times out of ten, the 70th percentile scorer will do better.

This hypo is meant to show that it is a trainable test and that people who are less naturally intelligent can outstrip those who are more naturally intelligent. It is pretty scary that you are self-styled doctors and you don't know this.

It would surprise me if you were 1/1000 of the doctor that Ben Carson is, or that Charles Drew, and Daniel Williams were. In case you didn't know (which you more than likely don't since you didn't know the MCAT is a trainable test), these gentlemen are Black.

Anonymous said...

Considering the MCAT is graded on a CURVE, it doens't matter how hard or easy it is....a score in the 95% today (about a 35) is the EXACT same as a score in the 95% 50 years ago. Unless you claim that the human species has lost intelligence in the last 50 years.

Fortunately, evolution doesn't work that fast.

Anonymous said...

I scored 40 ~15 years ago and got waitlisted 4 years straight. Oh yes, I am also an an asian-american man.

I cynically agree with your views about liberalism. The MCAT is nothing more than a hurdle, built upon other regulatory hurdles designed to limit the the supply of physicians and keep wages and prestige high.

Hermes said...

This hypo is meant to show that it is a trainable test and that people who are less naturally intelligent can outstrip those who are more naturally intelligent. It is pretty scary that you are self-styled doctors and you don't know this.

I'm not just a "self-styled" doctor; I'm an actual doctor, with an MD from a fully accredited US medical school.

Anyway, of course the MCAT requires some background knowledge. That is not the point. The individuals in your hypothetical did not start at the same baseline. But, controlling for background knowledge--that is, with two individuals having the same training in the biological and physical sciences--the one with the higher IQ will do better. This is because the test does not merely ask straightfoward questions of factual knowledge, but rather, requires the examinee to solve problems and make inferences.