Yea, I hate those posts too. Someone comes in having never posted before saying, "I got 271, is that good?" and then the argument ensues over how everyone who posts their score on SDN is ridiculous. blah blah. But that's what got me curious what the frequency of really good scores actually are. Unfortunately, the statistics available are kind of lacking. So I made do the best I could.
The numbers I used were from Charting Outcomes of the Match 2009. Disclaimer from the beginning: Due to roughly 7% of scores unknown at the time of the publication as well as those who didn't participate in the match for whatever reason (prematching or no interviews or whatever), the numbers are incomplete. I'm going with the assumption that in general it would even out.
Using the bar graphs from each specialty that lists the number of Step 1 scores for people who matched and didn't match, I put them all into a spreadsheet, so I could add things up.
What I found...
The document gives you the average scores for people who matched and didn't match...
US Sen
Matched- 225
Unmatched- 216
Weighted Average- 224
Independents
Matched- 217
Unmatched- 207
Weighted Average- 211
Overall weighted average of all test-takers- 218
Slightly lower than the usual 220ish that's quoted, but ok.
But for what I was really interested in was the frequency of the astronomical scores we read about here versus how frequent they actually are.
According to Charting the Outcomes, of all those who participated in the match, there were 359 scores of greater than 260 (~1.3%) and 1432 scores between 251 and 260 (~5.3%). Even going a little further, there were 2571 who scored 241-250 (~9.7%). Graphically, it was actually pretty close to a normal distribution, which surprised me. There was a slight positive skew to it.
So I started this post with no clear direction and was hoping to find a real point by now, but nothing really came to me. Basically all I found is what we already knew in that there is a sampling bias when all you see are those super high scores here on SDN.
Another curious thing I've noticed before but thought was kinda weird is that there is a handful of people with failing Step 1 scores who matched. It rarely happens (less than half a percent), but I still found it surprising that people matched without a current passing Step 1 score. Also, I don't know how the NRMP interprets multiple attempts- I assume it takes the passing score, but I'm really not sure.
Moral of the story I guess is don't go crazy seeing all the 250's and 260's posted in the experiences thread, there are plenty of people on the other end.
The numbers I used were from Charting Outcomes of the Match 2009. Disclaimer from the beginning: Due to roughly 7% of scores unknown at the time of the publication as well as those who didn't participate in the match for whatever reason (prematching or no interviews or whatever), the numbers are incomplete. I'm going with the assumption that in general it would even out.
Using the bar graphs from each specialty that lists the number of Step 1 scores for people who matched and didn't match, I put them all into a spreadsheet, so I could add things up.
What I found...
The document gives you the average scores for people who matched and didn't match...
US Sen
Matched- 225
Unmatched- 216
Weighted Average- 224
Independents
Matched- 217
Unmatched- 207
Weighted Average- 211
Overall weighted average of all test-takers- 218
Slightly lower than the usual 220ish that's quoted, but ok.
But for what I was really interested in was the frequency of the astronomical scores we read about here versus how frequent they actually are.
According to Charting the Outcomes, of all those who participated in the match, there were 359 scores of greater than 260 (~1.3%) and 1432 scores between 251 and 260 (~5.3%). Even going a little further, there were 2571 who scored 241-250 (~9.7%). Graphically, it was actually pretty close to a normal distribution, which surprised me. There was a slight positive skew to it.
So I started this post with no clear direction and was hoping to find a real point by now, but nothing really came to me. Basically all I found is what we already knew in that there is a sampling bias when all you see are those super high scores here on SDN.
Another curious thing I've noticed before but thought was kinda weird is that there is a handful of people with failing Step 1 scores who matched. It rarely happens (less than half a percent), but I still found it surprising that people matched without a current passing Step 1 score. Also, I don't know how the NRMP interprets multiple attempts- I assume it takes the passing score, but I'm really not sure.
Moral of the story I guess is don't go crazy seeing all the 250's and 260's posted in the experiences thread, there are plenty of people on the other end.
