If upon retest, the test taker scores higher, the new score will have its own confidence band/confidence interval. However, multiple tests scores cannot be considered in isolation. You have to look at all the data points. In your example, the test taker’s true score would be in the 500-503 range since that is the intersection of the confidence bands from both test sittings.
The new test score does not necessarily show a “better hand” since it cannot be considered in isolation. You have to look at all the data.
Ideally, if a person takes and retakes a test multiple times (like dozens or hundreds of sittings) and you plot all the results, you will see a bell curve emerge with the high point/central point being the best representation of the individual’s true score.
Med schools tend to publicly report only the highest score since it makes them look better for US News rankings; however, this doesn’t mean that the highest test score is the only one considered when the adcomm reviews an applicant. They have to look at testing trends and confidence interval overlaps if there are multiple test scores. And I highly doubt adcomms are so short-sigted to accept a student who has 7 MCAT attempts, 6 being significantly below the school norm and only the last one being very high just to fluff up their USNews rankings. They’re not that stupid–and 1 higher than average test score isn’t going to make that much difference in their ranking. (And remember that bell curves do have outliers so that individual’s one high test score could just be a fluke. It happens.)
BTW, AMCAS has now limited the number of lifetime attempts one individual can have for the MCAT to 7. That number includes tests completed, but canceled after completion. The 7 chance rule is an attempt to mitigate the “practice effect” on test scores. It’s tenet of test design that an individual will always score higher upon retake because they are now familiar with the test format and question types. Familiarity with testing format skews test results. So any decrease in test scores upon retake is a strong indicator that the person originally scored at or near the upper end of the confidence interval.
And yes, to show significant or meaningful improvement, any score increase has to be above the previous confidence band interval. So in the case of the MCAT >3 points.
I would be surprised if they consider anyone who takes it that many times despite AMCAS limitation. Medical schools depend on test taking abilities and they are making a Step1 a pass or fail with no chance of repeating a passed exam and limiting failures to 3 times.
http://www.boardvitals.com/blog/9-things-to-know-before-taking-usmle-step-1-exam/
…also considering that a Step 1 is quantum leap more challenging than MCAT and many take it in the middle of the medical school second, which is not that far ahead from taking the MCAT. They also got to sustain sitting thru Step 1 for about twice as long as MCAT, so they have to bring up their “brain stamina”…all along making sure that you are NOT under excessive stress while preparing as it may cause some stress related conditions, like shingles (few had shingles in D’s class, while preparing for the Step 1). So, you better nail that MCAT on the first try!!!
This isn’t true any more. The new MCAT is 6.5 hours long (plus breaks). STEP 1 is 7.25 hours long (plus breaks).
I’ll bet most who sit for Step 1 feel it’s much longer in time than MCAT
@dheldreth - Thanks your experience sharing for your D, and I hope my D has similar result - “got 3 early IIs and 3 early acceptances”, Congratulations to her and your family.
Hello, I have a question on a slightly different topic. My D (sophomore in CA) wants to take Psych in a community college over summer. Is this allowed or will be looked down upon when she goes through the application process?
Assuming she’s taking the rest of her pre-reqs at her home school and assuming she’s not applying to any med schools that will not accept CC credits to fulfill admission requirements, one summer psych class won’t hurt her chances.
For a list of med schools that do not accept CC credits, please consult MSAR.
Thank you so much for your reply WayOutWestMom. Do we have to have MSAR membership to be able to find this information?
MSAR = Medical School Admission Requirements.
MSAR is a subscription-only website run my AMCAS. A full-access subscription costs $30 (I think) and an updated one is published every April. (BTW, wait until the new version become available before buying since your subscription ends when the update comes out.)
You can see a preview here:
[Medical School Admission Requirements–Preview]( https://services.aamc.org/msar/home)
But a list of CC credit policies isn’t available without a paid subscription.
Your daughter may be able to access MSAR (for free) at her college’s library, career center or health professions advising office since those often buy an institutional subscription.
Alternatively, your daughter could email the admission offices of the med schools she’s most likely to apply to (all her in-state med schools, med school associated with her undergrad) and ask about their policies regarding CC credit.
Some medical schools have CC and internet class credit policies posted on their admissions webpages.
Just thinking it’s time I signed on to this thread considering I’ll be one with a bottle (or three) of wine this go around.
@texaspg Something’s not working correctly with their stats on that Lizzym Score. If I throw in my guy’s pretty darn high MCAT coupled with his pretty darn high GPA, it comes up with a pretty darn low percentage rate - one that wouldn’t be possible to achieve in my math classes with the starting points.
What is the GPA supposed to be out of (since that seems to be the problem)? It evidently isn’t a 4.0 scale (at least not right now).
Are you using this one?
http://schools.studentdoctor.net/schools/lizzym_score
Because I just tried it, plugging in both D1’s and D2’s stats (using both their old MCAT score and the 2015 equivalent percentile)–and got the right LizzyM score for both of them.
The GPA is on a 4.0 scale
@creekland - I plugged in 522 and 3.9 and got a 77 and 92.7% below that score.
@texaspg I think that’s my point, so I’m wondering how it works. A 522 is in the 99th percentile of the MCAT. Plugging in a 4.0 (what should be 99th percentile of GPA) says that it’s merely at the 95.17% of overall scores.
How does that happen?
How does one end up with 3.06% higher, but remain within that top 1% of both scores?
If I put a 4.2 in for GPA, it still has 98.28% as a final.
@creekland - Too many people have a 4.0 so it would never be 99 percentile.
Too few people have 522 or above which is 99%ile.
What the graph is projecting is the total number who are in that range of LizzyM score.
I did notice the new MCAT score does not work as well on this chart. At 528, it still gives a 79 score for a 4.0 but it should be equal to about 45 and 4.0 which is about 99.999%
http://■■■■■■■■■■■■■■■■■■■■■■■■/threads/efles-mcat-2015-to-old-mcat-percentile-comparison-conversion-tables.1143689/
Someone calculated a 523 to be 39+ (not sure why it is lower than a 5 point differential -528-523 = 5). I think you do a lot better if you can convert from new to old and just add the mcat + gpa x 10. At about 80 score, there will be about 700 people at the most who meet that number.
I agree with both your posts for the most part. The new score is close, but I don’t think it’s as accurate if the percentages listed with the scores are also accurate. (GPA has a larger percentage at the top, of course, but one would think it should still be capped at 4.0 and not vary if one puts in higher amounts.)
It’s a curiosity to play with anyway. My guy is totally content with his stats/score and it will just be time that tells with his applications later in the year/season. It’s just the math person in me that got playing around with it since it’s a new “toy” for me.
https://www.aamc.org/data/facts/applicantmatriculant/157998/factstablea24.html
The LizzyM difference for GPA between 3.8 and 4.0 is 2 points.
Based on the attached charts, there are 29% whites, 13% non Asian minorities and 25% Asians who fall within that 2 point range for the applicant pool.
On average, 17,000 applicants per year may have a GPA within this range.
OTOH, if we add up all scorers of 39-45 for the same GPA range, we have only 1400 for 2 years or 700 per year. I checked the distribution at one time for MCAT scores of 42 or above and figured out there may be 100-200 altogether (they don’t spell out the exact counts) each year. So I think this distribution graph for each score seems to be pretty accurate, at least for the old MCAT.