Historical CBSE exam data and Score Comparison Tool

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

Bereno

Smoking Monkey
10+ Year Member
Joined
Apr 15, 2011
Messages
2,031
Reaction score
631
Hey guys, I have put an excel document together with the data from historical CBSE exams. I will keep updating it as the tests come. It might prove useful for comparing your results or for trying to determine a score to shoot for. Feel free to use it or share it with those at your school who are taking or have taken it. Anyway, study break over, back to the grind.

Note: the data used is sourced directly from someone at the AAOMS. If my data is wrong, the source of the error would be with them. Just an FYI :)
Members don't see this ad.
 

Attachments

  • CBSE Results as of August 2015.xlsx
    108.8 KB · Views: 1,068
Last edited:
  • Like
Reactions: 2 users
Bereno strikes again! Thanks mate.
 
No problem. Hopefully this will be easier to find than the one in the other thread. :thumbup:
 
Members don't see this ad :)
The Anhedonia said:
curious if AAOMS ever got back to you about this :)

EDIT: just noticed the other thread. nicely done Bereno.
Click to expand...

I tried to get a histogram of the distribution, but seems like a no go at this point... Either way, this is still useful data.
 
Bereno said:
I tried to get a histogram of the distribution, but seems like a no go at this point... Either way, this is still useful data.
Click to expand...

Keep up the good work. However, I would put in a disclaimer about this being based on the assumption of a normal distribution. In reality, there might be a bimodal distribution with many scores in the 60-80's range and another peak in the 40-50's. Either way, based on the information that we have, this document should be a good way for people to gauge where they fall.
 
alxspider said:
Keep up the good work. However, I would put in a disclaimer about this being based on the assumption of a normal distribution. In reality, there might be a bimodal distribution with many scores in the 60-80's range and another peak in the 40-50's. Either way, based on the information that we have, this document should be a good way for people to gauge where they fall.
Click to expand...

Yeah, I hear ya. If you look, there is a disclaimer about the Gaussian curve assumption in a comment on the rank and percentile columns. Its not blatantly obvious because I like a clean and simple look. :)

About it being bimodal, yes it could very well be. Personally, I am thinking it is more likely to be a fairly flat distribution (platykurtic) with a slight negative skew to the left. This is merely speculation on my part, but at the end of the day, all we have are the min, max, stdev, and mean. Not even a median or mode! haha


EDIT: I just updated the comment on the bell curve to be a little more clear. Let me know if I should update it further.
 
Last edited:
So I was finally able to get the needed info for this file to be updated. Enjoy. :thumbup:
 
Thank you so much. I really appreciate that.
 
Looks like the February 15th CBSE stats have been released. I have updated my file accordingly. Enjoy :)
 
that May test tho....

Thanks for getting everything together!
 
Yeah, after seeing just how consistent the score distributions appear to be, May is becoming more and more of an outlier. Its a bummer really...
 
Members don't see this ad :)
Bereno said:
Yeah, after seeing just how consistent the score distributions appear to be, May is becoming more and more of an outlier. Its a bummer really...
Click to expand...
Why is it a bummer? Were you expecting scores to go up? Now that we know where score averages might be, it helps in analyzing performance on the CBSE
 
alxspider said:
Why is it a bummer? Were you expecting scores to go up? Now that we know where score averages might be, it helps in analyzing performance on the CBSE
Click to expand...

True, the averages have proven to be very consistent, which is great. However, you might have missed that I was commenting on the May exam specifically. The consistent averages make May appear an outlier that happens to correlate with the released questions on that exam make it more suspect. Its a bummer that is that it appears the released questions falsely inflated many people's CBSE score, making it harder for those that did not see the posted questions to compete with those who did see it.
 
  • Like
Reactions: 1 user
Bereno said:
True, the averages have proven to be very consistent, which is great. However, you might have missed that I was commenting on the May exam specifically. The consistent averages make May appear an outlier that happens to correlate with the released questions on that exam make it more suspect. Its a bummer that is that it appears the released questions falsely inflated many people's CBSE score, making it harder for those that did not see the posted questions to compete with those who did see it.
Click to expand...

Rabble rabble rabble! In all seriousness though it does look like that paranoid guy raising a ruckus on SDN about that May CBSE was on the money :/
 
  • Like
Reactions: 1 user
weren't there a bunch of students, like 10 or 20 that took it from harvard and another large group from columbia? Someone mentioned in a thread that their school had an average over 80 at columbia or harvard. I think May prob just had a larger group of students taking the test.
 
alxspider said:
weren't there a bunch of students, like 10 or 20 that took it from harvard and another large group from columbia? Someone mentioned in a thread that their school had an average over 80 at columbia or harvard. I think May prob just had a larger group of students taking the test.
Click to expand...
I know someone once CLAIMED that many(over 40 or 50) had scores >90 in May exam without any tangible evidence. However, It seems very suspicious to me that May exam is the outlier with its avg almost 5 points higher than the rest of them. Are you saying the higher avg on May exam is only because of an increased sample size( number of applicants that took the exam)? Can you please elaborate more on this view?
Thanks.
 
  • Like
Reactions: 1 user
alxspider said:
weren't there a bunch of students, like 10 or 20 that took it from harvard and another large group from columbia? Someone mentioned in a thread that their school had an average over 80 at columbia or harvard. I think May prob just had a larger group of students taking the test.
Click to expand...

I have not heard about Columbia or Harvard's averages, so can you show me where you got that info? I'd be genuinely curious to look into it. The only numbers I am familiar with is here at UConn we have a median right around 75 or so (from the numbers I have seen at least). Our average is likely a little higher due to some very bright few, so its not all that useful IMO. Also, more people taking the exam that day would actually make it harder to deviate from what seems to be the standard 54 or so. That said, there were 240 people who took the test, making it the 3rd smallest group out of 4.

Totori said:
I know someone once CLAIMED that many(over 40 or 50) had scores >90 in May exam without any tangible evidence. However, It seems very suspicious to me that May exam is the outlier with its avg almost 5 points higher than the rest of them. Are you saying the higher avg on May exam is only because of an increased sample size( number of applicants that took the exam)? Can you please elaborate more on this view?
Thanks.
Click to expand...

In reference to the bolded part: I dont believe that claim that there were so many in the 90's... What I do believe is that some people saw the questions posted and studied them as if they would have studied any other set of practice questions. Then, when they saw them on the exam they did a lot better, maybe 10/12 points better or so. I also think the benefit of seeing these questions would have a larger affect on those who would have been getting in the 50's than those in the 80's. The fact is the very small few in the 80+ range are likely smart enough to get the questions right anyway. The few questions they would have gotten wrong, but got right from the posted questions would be small. Also remember that the questions posted were from memory, not verbatim with verified answers. As I said, it looks to me that the posted questions did benefit those who saw them, but I dont think it boosted a lot of people's scores into the 90s. All said and done, I agree that this was just a claim lol.
 
Last edited:
  • Like
Reactions: 1 users
In reference to the bolded part: I dont believe that claim that there were so many in the 90's... What I do believe is that some people saw the questions posted and studied them as if they would have studied any other set of practice questions. Then, when they saw them on the exam they did a lot better, maybe 10/12 points better or so. I also think the benefit of seeing these questions would have a larger affect on those who would have been getting in the 50's than those in the 80's. The fact is the very small few in the 80+ range are likely smart enough to get the questions right anyway. The few questions they would have gotten wrong, but got right from the posted questions would be small. Also remember that the questions posted were from memory, not verbatim with verified answers. As I said, it looks to me that the posted questions did benefit those who saw them, but I dont think it boosted a lot of people's scores into the 90s. All said and done, I agree that this was just a claim lol.[/quote]

I too speculate that's what really happened. Some people got lucky. This also shows the test failed to do what it is supposed to do by giving some unfair advantage over others even if it was an accident.
I am just sorry for those who did not benefit from the released exams because obviously they were disadvantaged. I am also disappointed at the fact that AAOMS did nothing to rectify the situation. They didn't even explain what happened officially.
 
Last edited:
  • Like
Reactions: 1 user
Yeah... rocky start but hopefully it will be smoother from here on out.
 
Stats are up, just waiting to hear about the number of test takers for this exam. As of now, the N is set to 100 by default. I will update this when I hear the actual number. :)
 
Thanks! Not sure if this is a problem for anyone else but I can't seem to see/enter anything on the first worksheet (CBSE comparison tool) because your comments are blocking it, and I can't move them away because they are protected...
 
Phish said:
Thanks! Not sure if this is a problem for anyone else but I can't seem to see/enter anything on the first worksheet (CBSE comparison tool) because your comments are blocking it, and I can't move them away because they are protected...
Click to expand...

Hmm, try not hovering over the headings? Not sure why it would be doing that. Try it on a different machine and let me know. :)
 
Works fine for me, thanks Bereno! Ouch, my score isn't look as pretty as it did initially.
 
Heard the number of test takers this time was a crazy 505. Got it from a friend that contacted AAOMS today.
 
Hmm, if that holds as the new norm it's ~1000 exams per year for ~400 OMFS applicants. Comes out to 2.5 exam attempts per applicant. That seems fair since a lot of people are retaking for better scores and/or taking practice runs at this thing before seriously studying.
 
KillaCam said:
Heard the number of test takers this time was a crazy 505. Got it from a friend that contacted AAOMS today.
Click to expand...

Wow. I am still waiting to hear back an official number, but if that's the case, that is nuts!
 
CalBearr said:
Hmm, if that holds as the new norm it's ~1000 exams per year for ~400 OMFS applicants. Comes out to 2.5 exam attempts per applicant. That seems fair since a lot of people are retaking for better scores and/or taking practice runs at this thing before seriously studying.
Click to expand...

As long as the AAOMS states that residency programs will only get your highest score, and will not know how many times you took the exam, you better bet people will take practice runs.
If that's the case, the average high score would be higher than the mid-50's that we're getting for each individual exam. Might even be in the 60's.
 
  • Like
Reactions: 1 user
AlbinoPolarBear said:
As long as the AAOMS states that residency programs will only get your highest score, and will not know how many times you took the exam, you better bet people will take practice runs.
If that's the case, the average high score would be higher than the mid-50's that we're getting for each individual exam. Might even be in the 60's.
Click to expand...

I totally agree. The average test result will still be in the mid 50s, but the average score that people apply with will likely be higher. I doubt it will be much above 60 though (I'm not thinking people will improve by a full SD)... Unfortunately, we will never know lol.
 
Wonder if there is any chance the AAOMS can release the breakdown, on how that mean is calculated. How many scored in the 50's, 60s, 70s, etc. It's probably going to be a positively skewed curve. I know plenty of people who take it with minimal studying, just to see what the test is about, and that's why our averages really suck compared to the med students (mean = 230 = 82 cbse).
The med students can only take usmle step 1 once, and if they pass, they can't retake it (except in very specific circumstances). We can take it as many times as we want.

I'm willing to bet that there is a trimodal distribution. One large one with <50, for people who took it for ****s and giggles. A smaller one in the 65 range, who studied, but underestimated the exam and didn't finish UFAP. And another one 75+, for those who studied the right things, and dedicated enough time for this test (and those lucky mofos who took the first 2 years with the med students).
 
  • Like
Reactions: 1 user
Bereno said:
Hey guys, I have put an excel document together with the data from historical CBSE exams. I will keep updating it as the tests come. It might prove useful for comparing your results or for trying to determine a score to shoot for. Feel free to use it or share it with those at your school who are taking or have taken it. Anyway, study break over, back to the grind.

Note: the data used is sourced directly from someone at the AAOMS. If my data is wrong, the source of the error would be with them. Just an FYI :)
Click to expand...

Thank you Bereno!!
 
  • Like
Reactions: 1 user
AlbinoPolarBear said:
Wonder if there is any chance the AAOMS can release the breakdown, on how that mean is calculated. How many scored in the 50's, 60s, 70s, etc. It's probably going to be a positively skewed curve. I know plenty of people who take it with minimal studying, just to see what the test is about, and that's why our averages really suck compared to the med students (mean = 230 = 82 cbse).
The med students can only take usmle step 1 once, and if they pass, they can't retake it (except in very specific circumstances). We can take it as many times as we want.

I'm willing to bet that there is a trimodal distribution. One large one with <50, for people who took it for ****s and giggles. A smaller one in the 65 range, who studied, but underestimated the exam and didn't finish UFAP. And another one 75+, for those who studied the right things, and dedicated enough time for this test (and those lucky mofos who took the first 2 years with the med students).
Click to expand...

I have asked and they would not share. I even asked if they could provide an unlabeled histogram, and its a no go. I would love to see what the distribution looks like, but I think I wont ever know for sure haha.
 
  • Like
Reactions: 1 user
I think we should make a pact. Bereno gets the data if we have any future aaoms presidents lurking here
 
For anyone curious and couldn't wait for Bereno's data, I just got an email about the feb 7th test and they said the average was 53.6 with st dev of 11. Range was 1 (?!?!) to 91. Frankly I'm quite amazed at that 1 cause you'd probably get higher score than that if you guessed on every question. They either had to TRY to do that poorly or they just answered a question and fell asleep or something. Overall, I felt this test was significantly harder than last august's test and the averages show. Wish I was as prepared for Aug one as I was for this one though. Oh well...best of luck everyone and I hope you were happy with your score Bereno
 
Sorry for the delay. I just got back from a trip to Chile and will post what I can soon. :)
 
  • Like
Reactions: 2 users
OK, I have posted what I have. I have an email out to try and get the number of test takers for both this past Feb and Aug exams. :)
 
  • Like
Reactions: 3 users
Great job Bereno, appreciate you doing this, and hope this spreadsheet becomes your life-long hobby to help kids 20+ years from now.
 
  • Like
Reactions: 3 users
How do you get the percentiles...are they just estimates using the mean and SD and assuming a normal distribution, or do you actually have insider info as to the individual percentiles?
 
I increased from 69 to 73, but stayed in the same relative percentile for higher score. Ah oh well, still a higher numerical score regardless. Leave it up to the guy with the 99 to screw up the curve. I'm not sure if I would enjoy being his/her co-resident or not.
 
vellnueve said:
I wonder how they compare scores between different instances of the test, since the averages are different each time.
Click to expand...

They don't need to take into account which exam you took because it's all standardized so they can compare applicants taking different exams.
 
AlbinoPolarBear said:
They don't need to take into account which exam you took because it's all standardized so they can compare applicants taking different exams.
Click to expand...
I agree. The score report clearly states that that they are standardized with a standard error of measurement of +/- 3 points and predicts that's 2/3 of the time the applicant will score within that range, regardless of test version. Thus there is no significant point difference between people who score within 3 points of each other.

I think the most valuable information gleaned from the NBME score are
#1 - can you pass USMLE Step 1. If you can get greater than a 67 (correlates to passing score of 192), then you have evidence that you can.
#2 - some programs have significantly less time to prepare for Step 1 (e.g. Case Western), thus a wider margin of preparedness for Step 1 would be preferred.
#3 - for the lucky few at pure P/F schools, this is the only cold hard way to evaluate them.
#4 - standardized tests are always the best way to compare candidates from different schools
 
Who got 99? Is this real life. Apparently it's not the first time according to the Excel sheet tho.
 
Last edited:
agent2362 said:
I increased from 69 to 73, but stayed in the same relative percentile for higher score. Ah oh well, still a higher numerical score regardless. Leave it up to the guy with the 99 to screw up the curve. I'm not sure if I would enjoy being his/her co-resident or not.
Click to expand...
He's actually a pretty cool kid. Very laid back kid...but he pretty much just worked out, slept, ate, and studied for like a year before taking the test
 
  • Like
Reactions: 1 user
gn4 said:
He's actually a pretty cool kid. Very laid back kid...but he pretty much just worked out, slept, ate, and studied for like a year before taking the test
Click to expand...

What school did he go to? If he went to a dental school that's integrated with the medical school, I'd understand how such a score is possible.
 
Member 6236 said:
What school did he go to? If he went to a dental school that's integrated with the medical school, I'd understand how such a score is possible.
Click to expand...

Columbia. They also had guy in mid 90s and another in high 80s...
 
  • Like
Reactions: 1 user
You must log in or register to reply here.
Next unread thread

Similar threads

FosterWallace
Retaking good CBSE score for better score
Replies
8
Views
872
JustAWanderingStudent
J
R
old CBSE exams
Replies
4
Views
803
IndifferentOsteoblast
IndifferentOsteoblast
R
HELP WITH CBSE
Replies
6
Views
516
roadtodmd1
R
O
CBSE studying
Replies
2
Views
734
FosterWallace
FosterWallace
O
bootcamp for CBSE
Replies
1
Views
528
Gustavo dos Santos
Gustavo dos Santos
Top