Here are correlations with the average MCAT score of 110 US medical schools (not osteopathic, not black schools, not Puerto Rican Schools, not Duluth or University of MN impacted by Duluth, not Mercer, not MCP/Hahneman/Drexel, not DC schools. You can certainly see where the nation is going by where the MCAT and education and medical school debt is driving us. Again MCAT is only a reflection of socioeconomic status and past experiences. It is not a bad tool in that it does predict USMLE 1 well for about 75% of students, it is just a tool that is used incorrectly, especially on the 25% of students who we most need admitted, the ones who go where most needed. You can also see what students of the most urban origin and the highest socioeconomic status prefer in these correlations. For each medical school I calculated a % of the graduates of that school going into research, FM, rural practice, etc. For education variables I used Education Weekly sources. Ethnicity sources were % of White and Asian ethnicity by school for 1994, 1997, and 2000 graduates from AAMC Minorities in Medicine

Correlations of MCAT, Medical School Characteristics, Physician Distribution, and Career Choice

The MCAT average is the average of all three scores (verbal, biosciences, physical sciences) for 5 different internet sources compiled from medical school web sites from 2000-2003. Research was self designation as a medical researcher in the Masterfile. The family medicine percentage is all FP and GP graduates of a school graduating 1987-2000. The internal medicine column is the proportion of office-based IM compared to all IM graduates for a medical school. All rural docs include all choosing rural practice by zip code RUCA determination. The High School to College Graduate ratio for 1986 is a measure of state investment in breadth of education (High School Graduation) compared to depth (bachelors or more).

MCAT and researcher variables are in significant agreement across all of the comparisons. If medical school admissions committees used the GPA to a higher degree than the MCAT, there was much less potential for impact on the distributional specialties. Schools that lean more heavily on GPA than MCAT will likely have better graduation of the physicians most needed.

The same factors impacting choice of family medicine also relate the same way to graduation of all rural doctors and to the proportion of medical school graduates who end up in office based general internal medicine compared to all who specify internal medicine. This was calculated with office based general IM compared to general IM totals (shown above) and office based general IM compared to all who initially took internal medicine graduate training and there was no difference in these relationships. Office-based primary care physicians in poverty areas have the same level of agreement with the exception of high school graduation rates. This has to do with the poorer states have the lowest high school graduation levels. The breadth vs depth HS to College ratio measurement moves completely to 0.229 on the positive side for office-based primary care poverty.

Basically the variables above all relate strongly to socioeconomics and densities of population, education, and income.  Depth measurements complement one another, such as MCAT, % researcher, % subspecialty, % bachelor's degree graduates, % urban, and higher income levels.  The correlations of family medicine and office-based IM relate with the same significance, but in the opposite direction. FP and rural physician choice connects with breadth in education and admissions. High school graduation, % older students, % instate students, % low income students or people,

FP match or % choice of family medicine is available by state, medical school, ethnicity, country of origin, county, gender, and various divisions of birth origin and all relate to the above and each other in a consistent socioeconomic level with crosstabs, tables, correlations, and multiple linear regression.

It is much more than large city, or perceived medical school quality

It is about socioeconomic status

Medical Schools By Choice of Family Medicine for 1987 – 2000 Graduating Classes

* FP Likely groups were born in urban influence codes 2 and 3, rural areas, and military bases.

All we are measuring is relationships to socioeconomic status.

can do studies by RUCA, by urban influence code 1993 1 - 9 level by county, or by metro-non metro

can do rural per cent, or older, or instate admissions or MCAT scores

since it is socioeconomic status that we are really measuring, it does not matter

The Jefferson studies clearly link SES and choice of FP, with 13% in highest and 22% in lowest. My birth origin studies provide a further link and move it from individual schools, to nationwide. I calculated a birth origin ratio per 100000 population and here are the comparisons with income of origin and percent of students who chose FP from each birth origin group from urban influence codes 1 to 9, which also mark income levels as well

This says more about the origins of the student getting in as compared to geography or quality or other measures

The MCAT is a standardized test and may actually be even more biased for different and diverse students because of its "speeded" nature.

The MCAT as compared to basic science performance does vary by different types of students http://www.aamc.org/students/mcat/research/bibliography/velos001.htm

Here is the 2002 data on ACT tests nationwide by income level and ethnicity.

Over 51.5% of allopathic medical students in the US were from parents with incomes of over $100,000 a year for the class admitted in 2004. This increased from 23.5% of the class in 1997. AAMC Minorities in Medicine Studies

2000 MCAT scores by race/ethnicity and sex compared to ACT 2002 also