I can only speak from my perspective and share what we do, so in the following comments, when I say "we", you should know that n=1 on that. I honestly don't know exactly what other PDs do. It may be that some PDs do use medical school pedigree more than I do--BUT, I will say that there is a sense from some selection committee members I know that they like "scrappy" candidates, and that some students from some "elite" schools are viewed as possibly entitled and thus less scrappy. I know of one selection committee member at another program that feels that Harvard medical school students have "attitude" and don't work as hard. Just saying...
We do like to get students from the best medical schools, but I don't think the boost is as great as you think for the mythical "top 3" schools. In other words, there's a big difference in how we might view an applicant from a top brand medical school compared to an unknown medical school, but not much of a difference in how we might see a "top 3" medical student compared to a "top 20" medical student. To be honest, I don't even know what the true "top 3" schools are.
In our scheme, we have varied the amount of intangible influence we have given an application at the "select to interview" stage based on medical school pedigree from year to year. In order to do this, we have used an arbitrary ranklist we found on the internet that was easy to use--
medical-schools.startclass.com/-- not because I believe the methodology has any value or accuracy, but because it was easy for me to download the list into my spreadsheet
Schools are then lumped into groups (i.e., top 20, next 20, etc), and we calculate a "subtraction number" based on their group. In our algorithm, this gets factored into the "clinical clerkship" score. Without going into the details, essentially the applicants core clinical clerkship scores are "converted" to a number using a formula (complicated--takes into account the % of honors, % of high pass, etc.). We then subtract the number that represents the pedigree assessment. The final result is the "modified clinical clerkship score".
In some years, we have tweaked the algorithm to give more importance to the pedigree than in other years. In the years we've given it more importance, the schools were organized into groups of 20 (i.e., top 20, next 20, etc) up to 100, so that schools 1-20 have a "subtraction" number of 0, schools 21-40 have a "subtraction" number of 20, etc---after 100, everything gets a subtraction number of 100. DO schools get a subtraction number of 120. IMGs get a subtraction number of 150. For any state's most highly ranked state school (or 2 top schools for the big states like Texas), we cut the subtraction number in half--so, for example, the University of Iowa is #39 in the arbitrary meaningless ranking we use. Because it is the major state school for Iowa, instead of giving students from that school a "subtraction" score of 20 (since it is in the 21-40 pile), we give it a subtraction score of 10 (half of 20). The reason for this? Because some students choose to go to the best in-state school for financial reasons--it is often cheaper than the private schools that they may have gotten into. We wanted to somehow reflect that.
In other years, we tweaked the algorithm to have pedigree have less importance. In those years, we only had a 10 point difference between the tiers (top 20 had subtraction score of 0, next 20 had subtraction score of 10, etc). In those years, the students from schools that were ranked 100 or higher had subtraction scores of 50, not 100. DO schools had a subtraction score of 75, IMGs also 75.
The "modified" core clinical clerkship score (which is the formula-driven core clerkship score minus the pedigree subtraction number) is adjusted with a fudge factor so that the
average score and the
median score for our applicants is targeted around 500 points (it turns out our average and median scores are not that different on this number the way we have calculated it, which pleases me). The "modified" USMLE combo score (which is the combined step 1 and step 2 scores, with dampening of the scores for high outlier performance on step 1--see previous post, but essentially it minimizes the boost you get from a super high step 1 score) happens to
average around 500 points for our applicants (yes--that means you add step 1 and step 2 and the average applicant at our program approaches 500--usually at 498 or so).
So you see the impact. In the years we let pedigree most highly influence the score, there was about a 10% impact (100 points subtracted from a score that averaged 1000) to your score if you were from a relatively unknown medical school--but only a 2% impact for those students from the schools ranked 20-40 compared to those students ranked 0-20. Yes, in those years, the student from medical school #100 would have had a 100 point lower score compared to the student from medical school #1 if everything else was the same--meaning the equivalent of 50 step 1 points and 50 step 2 points.
Huge--when comparing applicant from the #1 med school to the applicant from the #100 medical school. But the applicant from the #40 medical school would have had only a 20 point lower score compared to the applicant from the #1 medical school (equivalent of 10 step 1 points and 10 step 2 points), and there would be no difference between applicant from #1 medical school and applicant from #20 medical school.
In the years we had pedigree least influence the score, there was really only a 5% impact to your score if you were from a relatively unknown medical school compared to a top 20 medical school (50 points out of 1000 average). If everything was the same between two students except pedigree, applicant from #1 medical school would have a 50 point advantage over applicant from #100 medical school (equivalent of 25 step 1 points and 25 step 2 points)--again, pretty significant. However, comparing applicant from #1 med school to applicant from #40 medical school, only a 10 point difference (equivalent of 5 step 1 points and 5 step 2 points)--and again, no difference between applicant from #1 medical school and applicant from #20 medical school.
Putting it all in perspective, however, remember that I only give out about 50-60% of the slots based on this formulaic approach. After that, for my last 40-50% of the slots, I look at the intangibles I discussed earlier. This is where leadership, research, rock-star recommendation letters, etc factor in. Remember also that this is just to get you the interview. Once we interview you, the combined USMLE and clinical clerkship score only accounts for 50% of your evaluation, with the other 50% based on your interview assessment--and even then, we eyeball the results and aren't beholden to the "numbers".
This is why I said to nickelbackfan that "many doors will open, but not
all doors may be open". Selection committees like to see relative outperformance--high USMLE scores, good core clerkship performance--when they select for interview, and reserve only a subset of the interview spots for candidates with "intangibles". Doing a research year is looked upon more favorably for candidate A who didn't need to do that to get a top-tier residency as compared to candidate B who had sub-average USMLE and core clerkship performance. We know that the research year for candidate B was almost certainly taken in order to boost chances of getting a better residency--for candidate A, it may have been a true expression of interest in an academic career that could boost the residency's "legacy" if the candidate goes to that residency and then becomes an academic leader.