Compilation of 2015 MCAT Preliminary Percentiles, and Practice Exam Scores

[Xenith]

Hi everyone,

Sorry for posting this everywhere. I think it deserves its own thread considering the work we have put into it and the interest people have shown.

mcatjelly and I compiled data on final scores, preliminary percentiles, the AAMC full length, and company exams in the following google doc. Please check it out and give us feedback! Also, if your information is missing or incorrect, please edit it!

PLEASE DO NOT QUOTE! The embedded sheet will make this thread very slow if it is quoted.

Also, click this >link< to open in a new window (especially useful for mobile devices)!

PS: Can this be stickied?

PSS: For scores after July Here is a link to the form, and here is a link to the new spreadsheet. Thanks to mcatjelly!

---

Members don't see this ad.

 

[CarmenCygni]

I assume we're not able to edit this and we just post our scores here?

 

[ikat]

Actual: 520 -- 98%
PS: 131 -- 99%
CARS: 130 -- 97%
Bio: 131 -- 99%
Psych: 128 -- 86%


May 22 Prelim
PS: 85-100
CARS: 85-100
Bio: 85-100
Psych: 81-96
Overall: 90-100

AAMC FL: PS: 78% CARS: 91% BS: 88% Psych: 83%

AAMC Official Guide: PS: 80% CARS: 90% BS: 76% Psych: 90%

Kaplan Full Lengths (not completed in this order)
FL 3: PS:125 CARS:128 BS:125 Psych:127 Total: 505
FL 4: PS:125 CARS:127 BS:125 Psych:127 Total: 504
FL 5: PS:124 CARS:126 BS:125 Psych:127 Total: 502
FL 8: PS:125 CARS:128 BS:126 Psych:124 Total: 503
FL 9: PS:125 CARS:128 BS:126 Psych:124 Total: 503

 

[Flashfan]

Actual:
PS: 130 97%
CARS 128 87%
Bio 129 93%
Psych 129 93%
516-95%

Prelim:
PS: 85-100
CARS: 82-97
Bio: 85-100
Psych: 85-100
90-100%

 

[panda16]

Prelims: 85-100 all, 90-100 total

Actual

PS: 130 (97%)
CARS: 129 (93%)
Bio: 132 (100%)
Psych: 131 (98%)
Overall: 522 (99%)

@Xenith Assuming you will enter this stuff into the spreadsheet? My row is now protected and I can't enter anything.

 

[Xenith]

Prelims: 85-100 all, 90-100 total

Actual

PS: 130 (97%)
CARS: 129 (93%)
Bio: 132 (100%)
Psych: 131 (98%)
Overall: 522 (99%)

@Xenith Assuming you will enter this stuff into the spreadsheet? My row is now protected and I can't enter anything.

Hey mcatjelly can remove the lock, but you can post them in the sheet labeled copy of sheet 1.

 

[VegasPreMed]

Prelim May percentiles:
Chem/Phys: 85% - 100%
CARS: 85% - 100%
Bio: 82% - 97%
Psych/Soc: 85% - 100%
Total: 90% - 100%

Actual:
Chem/Phys: 130 (97%)
CARS: 131 (99%)
Bio: 128 (87%)
Psych/Soc: 129 (93%)
Total: 518 (97%)

 

[carpediem22]

Chem/Phys: 129 (93%)
CARS: 130 (97%)
Bio: 130 (97%)
Psych/Soc: 131 (98%)
Total: 520 (98%)

 

[FutureBenCarson]

Chem/phys:132
CARS:130
BIO:131
Psych/Soc:132
overall:525

 

[Alkaidius]

We should be more specific about this compilation's purpose at this point for new people posting. I feel like some people are misunderstanding this as some sort of hall of fame, when it should be treated as a compendium of practice test to actual test correlation (since the preliminaries are going to be all but useless at this point).

 

[carpediem22]

We should be more specific about this compilation's purpose at this point for new people posting. I feel like some people are misunderstanding this as some sort of hall of fame, when it should be treated as a compendium of practice test to actual test correlation (since the preliminaries are going to be all but useless at this point).

Most of us who have posted already have our practice scores and preliminary percentiles on the doc, and are just updating with our actual scores.

 

[Xenith]

All ranges are now editable! Please add your score if you would like! Otherwise I or someone else will get to it soon.

 

[Xenith]

You can edit! Go ahead and update yours on the google sheet.

 

[EverStriving]

It'll be really interesting to see the stats on which FLs are most informative, once all the May testers fill in their scores.

 

[NeuResurgens]

Someone had a 528? Holy.

 

[Xenith]

Someone had a 528? Holy.

No proof!

 

[EverStriving]

No proof!

Also, if you look closely at the histogram the AAMC published along with the percentiles, there is no column in the 528 bin. So looks probable that nobody got a 528 out of all the april/may testers.

 

[Xenith]

I wonder what the sample size was for the 4 testing dates. What do you think?

 

[EverStriving]

On some random internet article I recall seeing a figure around 13000, but I can't seem to find the article again. And who knows where they got that figure anyway....

But if 13,000 is correct, there would be around 13 individuals in the top 0.1%, which corresponds to 526-528 IIRC. Given that 526 and 527 are each multiple times more likely than 528, I think it's plausible to suppose that a 528 just didnt happen.

 

[carpediem22]

I dunno I think it is definitely possible. They purposefully made both 1 and 15 equivalents (118 and 132) more attainable with this test (on the old MCAT 45's were essentially a myth). Given the high number of people on here with 132s in individual sections, I would not be surprised if someone hit the jackpot with a full sweep.

 

[jd989898]

I don't even want to get into medical school anymore I just want that #1 spot on the google doc

 

[NeuResurgens]

I dunno I think it is definitely possible. They purposefully made both 1 and 15 equivalents (118 and 132) more attainable with this test (on the old MCAT 45's were essentially a myth). Given the high number of people on here with 132s in individual sections, I would not be surprised if someone hit the jackpot with a full sweep.

Heck, I never even heard of anyone getting a 43, let along 45. I do remember hearing that they made getting a perfect score more realistic for this test, just can't remember where. It seems like this test has a higher number of high scores coming out, but that could just be the already skewed MCAT scores typical of SDN users.

 

[EverStriving]

So just in time for June testers to get their prelims, I went through the google doc and tabulated all the preliminary percentiles people reported for April/May and what final scores they corresponded to:

All 85-100% prelim ranges corresponded to a final score of 129 or above.

Below the 85-100% range, there were no verifiable instances of a single preliminary range leading to two different scores. So presumably if you find your range on this table, you should be able to know your score on the section.

There seems to be close agreement between sections, so I also made a composite table to give an impression of the boundaries between the different scores.

 

[mcatjelly]

So just in time for June testers to get their prelims, I went through the google doc and tabulated all the preliminary percentiles people reported for April/May and what final scores they corresponded to.

All 85-100% prelim ranges corresponded to a final score of 129 or above.

Below the 85-100% range, there were no verifiable instances of a single preliminary range leading to two different scores. So presumably if you find your range on this table, you should be able to know your score on the section.

There seems to be close agreement between sections, so I also made a composite table to give an impression of the boundaries between the different scores.

This is great, thanks!

 

[starstruck20]

This is really helpful but don't you think it would be a good idea to create a study habits thread additionally? Like there was for the old one. I know many people found the old thread very helpful in creating their own study plan.

 

[mcatjelly]

This is really helpful but don't you think it would be a good idea to create a study habits thread additionally? Like there was for the old one. I know many people found the old thread very helpful in creating their own study plan.

"Thanks for using your free time to put this together, now please make something else for me." is essentially what I just read. Ain't nothing stopping you from messaging the mods to start a new stickythread!

 

[jd989898]

So just in time for June testers to get their prelims, I went through the google doc and tabulated all the preliminary percentiles people reported for April/May and what final scores they corresponded to:

thanks for that!

 

[jd989898]

Something I've noticed on the google doc is that no one with an 85-100% in every section ever scored below 520, even though those prelims correspond to as low as 516. I suppose if a tester obtains 85-100 in every section, they likely possess confounding traits making it more likely that they'd get a 130,131, or 132 regardless of the fact that those higher section scores are statistically rarer i.e. smaller bins on the histogram.

 

[Ad2b]

To all of you who took the Kaplan course and then posted here on your Kaplan v Actual... My I just bow in sincere thanks to you?!?

There is hope. My 501 diagnostic does not = doom (unless, of course, I just stop studying!)

 

[littlebear92]

Were the same percentile range bins used for every group of test takers from April to May?

 

[EverStriving]

Something I've noticed on the google doc is that no one with an 85-100% in every section ever scored below 520, even though those prelims correspond to as low as 516. I suppose if a tester obtains 85-100 in every section, they likely possess confounding traits making it more likely that they'd get a 130,131, or 132 regardless of the fact that those higher section scores are statistically rarer i.e. smaller bins on the histogram.

It could also be a statistical effect. There is only one way to get 85-100s across the board and only get a 516 (129s across the board) which is a very specific score combination and therefore exceedingly rare. There are four ways to get a 517, ten ways to get a 518, and so on. So even if the individual higher scores are less likely, there are a whole lot more paths to get to 520-524 than to either 516-519 or 525-528.

Another way to think about it is, a criterion of all 85-100s is not going to sweep up a lot of 516s because there just aren't that many people with that specific score combination. It will pick up more 517s and 518s, but still not that many. The chance of qualifying for all 85-100s goes up dramatically as your overall score goes up.

Were the same percentile range bins used for every group of test takers from April to May?

Unfortunately I don't think we have a large enough sample size to test this. I just used everyone that had both prelims and final scores posted.

 

[jd989898]

It could also be a statistical effect. There is only one way to get 85-100s across the board and only get a 516 (129s across the board) which is a very specific score combination and therefore exceedingly rare. There are four ways to get a 517, ten ways to get a 518, and so on. So even if the individual higher scores are less likely, there are a whole lot more paths to get to 520-524 than to either 516-519 or 525-528.

Another way to think about it is, a criterion of all 85-100s is not going to sweep up a lot of 516s because there just aren't that many people with that specific score combination. It will pick up more 517s and 518s, but still not that many. The chance of qualifying for all 85-100s goes up dramatically as your overall score goes up.

Agreed. The combinatorial explanation is likely a factor in many cases. Though the magnitude of this effect is still obviously dependent upon the sum of likelihoods associated with each way to achieve X score, which could theoretically still amount to a smaller probability than achieving even a 516 (which can only be achieved in one way). Ultimately, the probabilities can never be known, though, because of the various confounding factors.

 

[EverStriving]

Agreed. The combinatorial explanation is likely a factor in many cases. Though the magnitude of this effect is still obviously dependent upon the sum of likelihoods associated with each way to achieve X score, which could theoretically still amount to a smaller probability than achieving even a 516 (which can only be achieved in one way). Ultimately, the probabilities can never be known, though, because of the various confounding factors.

Ya, people who get a high score on one section are more likely to get high scores on other sections as well, so that should move the mean of the real distribution upward from the likelihood sum prediction. Only AAMC would have the data to answer this question unfortunately.

I also ran regressions comparing AAMC FL percent correct to final scores. Looks like for individual sections, B/B was the best predicted by AAMC FL performance, followed by CARS and C/P. Psych/Soc was by far the most poorly predicted, with an r-squared of only around 0.2 and a barely significant correlation. Overall score was the best predicted of all (by overall % correct), with an r-squared of around 0.63.

 

[carpediem22]

Something I've noticed on the google doc is that no one with an 85-100% in every section ever scored below 520, even though those prelims correspond to as low as 516. I suppose if a tester obtains 85-100 in every section, they likely possess confounding traits making it more likely that they'd get a 130,131, or 132 regardless of the fact that those higher section scores are statistically rarer i.e. smaller bins on the histogram.

Not true, some people in May got 519s and I think 518s

Edit: Just looked and those people aren't on the doc. However your theory is true just go down to 518 or 519 hah

 

[NeuResurgens]

To all of you who took the Kaplan course and then posted here on your Kaplan v Actual... My I just bow in sincere thanks to you?!?

There is hope. My 501 diagnostic does not = doom (unless, of course, I just stop studying!)

I'm doing Kaplan as well and I've found their stuff to be very difficult. Super calculation heavy and very focused on minute distinguishing details. It's encouraging to see that a lot of Kaplan users have gone on, by and large, to get very good scores (majority >515 I believe).

 

[carpediem22]

I'm doing Kaplan as well and I've found their stuff to be very difficult. Super calculation heavy and very focused on minute distinguishing details. It's encouraging to see that a lot of Kaplan users have gone on, by and large, to get very good scores (majority >515 I believe).

Real thing is not calculation heavy though, and also wasn't super minutiae focused... just my .02

 

[jd989898]

Real thing is not calculation heavy though, and also wasn't super minutiae focused... just my .02

Agreed. I probably used <5% of the information I memorized

Edit- this post is not intended to devalue studying. More is always better. Just enphasizinf that the test calls on a very small random subset of actual rote memorization type material.

 

[littlebear92]

I can only remember 4 calculations off the top of my head. And some of them, you can even make an educated guess by looking at the units that the answer choices are in and then dividing/multiplying accordingly

 

[NeuResurgens]

Real thing is not calculation heavy though, and also wasn't super minutiae focused... just my .02

Exactly - this is what I've been told by several people. I suppose over-prepped is better than under-prepped, provided you can digest all the information while studying!

 

[soccergurl1987]

Can you make the file editable again for those of us with prelim percentiles from June? Or would you like us to post here and/or wait until our final scores are out later in July?

 

[Alkaidius]

Can you make the file editable again for those of us with prelim percentiles from June? Or would you like us to post here and/or wait until our final scores are out later in July?

I don't think there's much point in putting in preliminaries anymore. June should be the last one's that receive preliminaries.

EDIT: Although, I guess it could still be helpful if people at least want to see how certain preliminaries correlate with your practice tests. That could be helpful.

 

[Ad2b]

So memorizing the entire page of stage birth to 3 and every subset with every neurological change and action potential with subsequent hormonal changes in the brains of infants through toddlers along with calculating the % change differential between the synapses is UN necessary!?

okay. back to a=v/t, vf^2=vi^2+2a(delta)x, or get higher test scores

 

[Mr. Beefy Lion]

June test taker:

Chem/phys: 85-100%
CARS: 65-80%
Bio: 85-100%
Psych/soc: 85-100%

Practice exam taken 11 days before:
Chem/phys: 88%
Cars: 83%
Bio:81%
Psych/soc: 86%

Xenith you are entering these right? I think the table is locked

 

[Mr. Beefy Lion]

So memorizing the entire page of stage birth to 3 and every subset with every neurological change and action potential with subsequent hormonal changes in the brains of infants through toddlers along with calculating the % change differential between the synapses is UN necessary!?

okay. back to a=v/t, vf^2=vi^2+2a(delta)x, or get higher test scores

Hahah definition of unnecessary my friend

 

[premedprep22]

June test:

C/P: 85-100
CARS: 76-91
Bio: 85-100
P/S: 85-100
Overall: 88-98

AAMC Practice test taken 2 weeks before:
C/P: 60%
CARS: 85%
Bio: 75%
P/S: 85%

TPR tests:
Course test 1: 503
Course test 2: 502
Full Length 1: 505

Kaplan test:
Full length 1: 497

God is Good!

 

[mcatpremed]

For the CP and BB sections that have a AAMC prelim range of 82% - 97 %, is it safe to assume the actual score will be a 128 ?

 

[mcatpremed]

Example:
2 test, same totals

One has a high P/S score , the other has a high C/P score ( the other sections have the same scores ).

Does anyone feel that the total scores will be looked at differently ?

 

[a_borgia]

June test:

Prelims: (90-100)/(85-100 in all)

AAMC practice test: Fuzzy recall on the numbers, but 92ish cumulative

Kaplan sunday 1/3 length (pre-studying): 502
Kaplan diagnostic 1/2 length (post-studying): 512
Kaplan FL1: 510
Kaplan FL2:512
Kaplan FL3: 514
Kaplan FL4: 516
Kaplan FL5: 515
Kaplan FL7: 516
Kaplan FL8: 519
Kaplan FL9: 518
Kaplan FL10/11: Took as sections, don't remember the score.

Sorry that I don't have section breakdowns on my practice tests; my access expired the day after I took my test. My highest sections on a Kaplan test were 130 on CP, 129 on CARS, 131 on BB, and 129 on PS. Obviously not all on the same test, but I took so many that I don't remember details.

 

[Okazaki Frag Grenade]

June test:

Prelims: (90-100)/(85-100 in all)

AAMC practice test: Fuzzy recall on the numbers, but 92ish cumulative

Kaplan sunday 1/3 length (pre-studying): 502
Kaplan diagnostic 1/2 length (post-studying): 512
Kaplan FL1: 510
Kaplan FL2:512
Kaplan FL3: 514
Kaplan FL4: 516
Kaplan FL5: 515
Kaplan FL7: 516
Kaplan FL8: 519
Kaplan FL9: 518
Kaplan FL10/11: Took as sections, don't remember the score.

Sorry that I don't have section breakdowns on my practice tests; my access expired the day after I took my test. My highest sections on a Kaplan test were 130 on CP, 129 on CARS, 131 on BB, and 129 on PS. Obviously not all on the same test, but I took so many that I don't remember details.

519 on a Kaplan?

 

[jd989898]

519 on a Kaplan?

Yeah srsly. If this guy doesn't get a 528... lol

 

[a_borgia]

Yeah srsly. If this guy doesn't get a 528... lol

Gal, actually, and I was the one chatting with you in the other thread about probabilities. It's worth nothing that those 519s were with longer breaks in between sections than I'd given on the test because I was insanely fatigued by that point in my studying - 516 was way more representative of my Kaplan score.

In other news, starting study for a June MCAT in May? Don't do it.