Quotes by famous Dr G

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

TRuSchool

Full Member
10+ Year Member
15+ Year Member
Joined
Jul 21, 2007
Messages
124
Reaction score
0
Points
0
  1. Medical Student
Add to the list if you are a Goljan listener. Paraphrasing the cockiness of this man (deservedly so)

"Let's get started. As you know I'm Dr. Goljan. My students call me papi."
"Dont use Kaplan Pathology notes, use mine. Thats why you're here."
"Those people studying First Aid, they wont know anything. But you! You have my high yield, you wont see any surprises out there!"
"There is a linear correlation with my notes. The more you read, the higher your score."
"I could not imagine life without the ability to pass gas."
 
"I talked to students that got a 92, 94, 97. Pretty good scores huh?"

... not really
 
"I talked to students that got a 92, 94, 97. Pretty good scores huh?"

... not really

Actually, prior to about 5 years ago, your %tile actually meant your %tile and not just above a 230 something. My friend got a 228 in 2006 which was a 89.
 
Actually, prior to about 5 years ago, your %tile actually meant your %tile and not just above a 230 something. My friend got a 228 in 2006 which was a 89.
The means have always been around 220 with an STD of 15-20. I don't think a 228 would have been an 89 percentile.
 
well if he was talking about the 2-digit score and not % correct...

I donno...2-digit scores vs. percentile vs. percent correct are 3 very different things on this test.

A 2-digit score of 97 is a very solid (but hardly elite) performance (probably in the high 220s/low 230s).

Scoring in the 97th percentile, on the other hand, certainly would be elite. I know someone who took the available data and extrapolated it out statistically and 99th percentile corresponds roughly to 265ish.

97 percent correct - now that's a face-melting, rip-roaring, Terminator-esque performance. That would likely be 280+. I'm not even sure Edward F. Goljan himself could pull this off if he sat for the test today.
 
Last edited:
I donno...2-digit scores vs. percentile vs. percent correct are 3 very different things on this test.

A 2-digit score of 97 is a very solid (but hardly elite) performance (probably in the high 220s/low 230s).

Scoring in the 97th percentile, on the other hand, certainly would be elite. I know someone who took the available data and extrapolated it out statistically and 99th percentile corresponds roughly to 265ish.

97 percent correct - now that's a face-melting, rip-roaring, Terminator-esque performance. That would likely be 280+. I'm not even sure if Edward F. Goljan himself could pull this off if he sat for the test today.


blasphemous. take it back
 
Okay youz guys. Quick question:

Who would score higher if they sat for the exam: Edward Goljan or God?
 
The means have always been around 220 with an STD of 15-20. I don't think a 228 would have been an 89 percentile.

the means have not always been 220...it was a 216 my friends year and 15 years ago it was a 207 (one of my professors year). So in goljans time it probably was around 210. I saw my friends score and it definately was a 228 and 89%. So somehow the scaling has changed.
 
The means have always been around 220 with an STD of 15-20. I don't think a 228 would have been an 89 percentile.

The mean has actually increased a lot in the last few years. I think it was something like 200 or so 20 years ago,

You're right about the SD though, and thus the 228 not being 89%. They in fact got rid of the percentiles in 1999.
 
also, my favorite goljan moment would have to be that bit about micro vs. macronodular. sounds like he's calling a horse race. that one never seems to get old for me.
 
The mean has actually increased a lot in the last few years. I think it was something like 200 or so 20 years ago,

You're right about the SD though, and thus the 228 not being 89%. They in fact got rid of the percentiles in 1999.

Weird. How is it a standardized exam, again?

In my opinion, percentiles is the only way to make an exam standardized.
 
Weird. How is it a standardized exam, again?

In my opinion, percentiles is the only way to make an exam standardized.


I'm not sure about the 2 digit score on step 1, all it says is "The 2-digit score is not a percentile.The 2-digit score is derived from the 3-digit score. It is used in score reporting to meet requirements of some medical licensing authorities that the passing score be reported as 75. The 2-digit score is derived in such a way that a score of 75 always corresponds to the minimum passing score." (http://www.usmle.org/General_Information/FAQs/FAQs-Scores.html)

However, I know statistically standardized tests are usually set to a bell curve. This would indicate that the mean score (220 or whatever it is now) is the center (highest point on the bell.) 68% of the scores will be in a range of + or – one standard deviation from the mean. So w/a SD of say 17 and mean of 220 on the step 1, 68% of people will score between 203 and 237. Approximately 95% of the scores will be in a range of + or – two standard deviations from the mean. So (using mean 220 & SD 17) 95% of people will score between 186 and 254. Getting above a 254 means beating out 97.5% of people who took the test.

If you’re confused about the 97.5% it’s because 95% of people get between 186 and 254 which means 2.5% of people get below 186 and 2.5% get above 254. Meaning those scoring above 2 SD’s are beating 97.5% of people

Obviously I don’t know the actual mean and standard deviation but I think that’s how it’s supposed to work.
 
I'm not sure about the 2 digit score on step 1, all it says is "The 2-digit score is not a percentile.The 2-digit score is derived from the 3-digit score. It is used in score reporting to meet requirements of some medical licensing authorities that the passing score be reported as 75. The 2-digit score is derived in such a way that a score of 75 always corresponds to the minimum passing score." (http://www.usmle.org/General_Information/FAQs/FAQs-Scores.html)

However, I know statistically standardized tests are usually set to a bell curve. This would indicate that the mean score (220 or whatever it is now) is the center (highest point on the bell.) 68% of the scores will be in a range of + or – one standard deviation from the mean. So w/a SD of say 17 and mean of 220 on the step 1, 68% of people will score between 203 and 237. Approximately 95% of the scores will be in a range of + or – two standard deviations from the mean. So (using mean 220 & SD 17) 95% of people will score between 186 and 254. Getting above a 254 means beating out 97.5% of people who took the test.

If you're confused about the 97.5% it's because 95% of people get between 186 and 254 which means 2.5% of people get below 186 and 2.5% get above 254. Meaning those scoring above 2 SD's are beating 97.5% of people

Obviously I don't know the actual mean and standard deviation but I think that's how it's supposed to work.


I'm not confused about the statistics. I'm confused about how they (the USMLE) can pull the wool over everyone's eyes and claim that an exam with 8 heart sounds, 4 epi questions, and 3 neuro questions is standardized to one with 10 neuro questions, 3 heart sounds, and 2 epi questions. It's because, "you're all supposed to learn that, anyway." What if I dominate at neuro and suck at cardio? If I get the neuro heavy exam, I'll score better than another exam weighted in another subject. That, to me, is not standardized. Standardized should be the same number of questions in the same subject matter delivered on the same day, at the same exact time, with all scores delivered at the same instant. If you fail, better luck next year. Why can these exams not be delivered electronically by the accredited medical school?

Sorry to hijack the thread.

Back to Goljan:

His rant about writing tests was one of the best that I had ever heard. It was my facebook status for about a week and a half. I made sure plenty of my undergrad professors read it as well.
 
The mean has actually increased a lot in the last few years. I think it was something like 200 or so 20 years ago,

This is definitely true...pg 8 of FA 2010 states "the mean step 1 score for US medical students rose from 200 in 1991 to 222 in 2009". I think the year-over-year increases have been even steeper more recently than they were before also.

And I agree with the other sentiments expressed here...why the hell can't the actual USMLE qs be more standardized and the scoring more transparent? Usually "standardized" exams place a heavy emphasis on both of the above. My only guess for why the NBME does it like this is that the USMLE was really intended to be a pass/fail exam, and they would still prefer to think of it that way...

Just be thankful you guys aren't taking the COMLEX - that scoring system is shrouded in even more secrecy than the USMLE's.
 
" Patient is not gonna have choice A, B,C,D for Dx you have to be thinking 😕 doctor lol"
 
"...and then I decided to put the stupid stethoscope in my ears. This time I actually heard the murmur (as opposed to keeping it around my neck)"

I'm paraphrasing here big time, hope that doesn't take anything away from how awesome he is.
 
Top Bottom