Does the MCAT Fall Short in Predicting Physician Success

The MCAT has been used as one of the critical pieces to assess the readiness of students for medical school since its inception in 1928. Evaluations subsequent to its introduction clearly demonstrated it could help predict success rates in medical school. The drop-out rates, being as high as 50% in some medical schools, fell to single digit numbers. A large part of this decrease was credited to the MCAT and its ability to predict which students could survive the rigorous medical school curriculum (Evolution of the MCAT).

The AAMC itself points to the predictive value of the MCAT in terms of medical school success (Validity of MCAT). But, when this issue is looked at closely, there appears to be an inverse relationship between the predictive power of the MCAT from medical school to a physician in practice (MCAT Scores Predicting USMLE I, MCAT Predicting First Two Years of Medical School, Ethnic Performances on Licensing Exam and MCAT) Additionally, there appears to be variability of the predictive success of the MCAT with different gender, ethnic and racial groups(MCAT Prediction and Diverse Groups, MCAT Prediction of Clerkship Performance)This has led several researchers to look for other factors to use for admission and prediction of success as a physician (Prediction With Personality Characteristics, Prediction Using Psychosocial Factors, Poor Preclinical Performers).

When the full spectrum of literature is reviewed, my interpretation is:

1.  The MCAT predicts success in the first two years of medical school very well;

2.  The MCAT predicts success on the USMLE I well;

3.  The MCAT has less predictive value for the medical school clinical years;

4.  The MCAT has very little to no predictive value for predictions past the medical school years and into actual medical practice.

Why is this?

Refer To: MCAT and Cognitive Domain, MCAT and Affective Domain, MCAT and Psychomotor Domain, MCAT Validity in Perspective

Trick Questions on The MCAT?

I was recently in a session where two of the brighter students from a major university (a Pac-Ten university) were discussing their experiences with taking a MCAT prep course and their experiences on the MCAT exam. Both students had taken the same one of the big TWO commercial MCAT preparation courses. 

When one of the students described that the instructor, and based on the course’s philosophy, had told him that the MCAT tries to trick you by how they write their questions…its putting it mildly to say that I was profoundly awakened from the momentary snooze that had beset me. So, I later Googled “trick questions on the MCAT” and found that there is a belief by many that the MCAT has trick questions. The Student Doctor Network has a forum discussing this question.

Categorically, my belief is there is no reason for the AAMC to trick anyone…the test is difficult enough without tricks. And, based on my experience of analyzing real MCAT’s in detail, I cannot honestly recall one question that I would seriously call a trick question. So, what might this course and instructor mean by making this statement? Whereas I cannot speak for them definitively, I will offer my thoughts:

1. There are no trick questions on the MCAT


2. If there were truly “trick questions” on the MCAT,




1. There would be no effective way to prepare (too many ways to trick you)


2. AAMC would have been sued already




3. Any perception of trick questions is




1. a reflection of the lack of understanding of the structure of the MCAT


2. lack of effective preparation for the MCAT


3. lack of skills appropriate for taking the MCAT




4. Relying on mock or simulated exams will increase this misperception of “trick” questions on the real MCAT.


5. So, don’t waste your time trying to determine if a question is a trick question…because there are not any.

My belief is simple. When there is appropriate preparation of content and skills using real MCAT’s, the specter of “trick questions” will magically disappear.