Okay with this thread dying down, I wanted to end with the final statistics. This is for future test takers of the old mcat and is intended to be informational. Please take this with a grain of salt though, as this is an observational study (possibly badly done too since I'm not a social scientist and not a person who usually does observational studies). This is not a true experiment. Nothing was manipulated and data was collected only from posters of this thread. Nobody privately messaged me, so all the data is right here for others to reproduce.
Quoted below are posts that talked about the statistics. Then after that are the final outcomes.
View attachment 185431
Grey background bars: Expected score distribution for a population of 107 people. Based on AAMC published data of all 2013 test takers.
Blue bars: Predicted Scores
Red bars: AAMC average scores (rounded, e.g. 29.5 = 30, 35.3 = 35)
Gold bars: Actual exam scores from August 7th, 2014 - 8am and 2pm exam times
8am:
Total People:
30
AAMC Average
33.0, st. dev 3.2 (21 responses, 70% response rate)
Predicted
31.3 st. dev 3.4 (20 responses, 67% response rate)
Actual
32.5 st. dev 3.9 (20 responses, 67% response rate)
2pm:
Total people:
77
AAMC Average
31.3, st. dev 3.8 (40 responses, 52% response rate)
Predicted
29.4 st. dev 4.0 (40 responses, 52% response rate)
Actual
31.9 st. dev 4.8 (47 responses, 61% response rate)
Both Times:
Total people:
107
AAMC Average
31.9 , st. dev 3.7 (61 responses, 57% response rate)
Predicted
30.0, st. dev 3.9 (60 responses, 56% response rate)
Actual
32.1, st. dev 4.5 (67 responses,
63% response rate)
I kept track of people who joined the conversation after scores were released and old posters who have been a part of this thread before scores were released. I found no significant difference between the groups. So, people who appear out of nowhere on this thread once scores are released do not necessarily have insane scores (35+). There is a fair share of new posters who have lower scores too to balance out. We just tend to remember those 41s and 42s who come out of nowhere and not the 25s or 26s who add their posts into the thread.
Also important, I noted nine individuals who either have been a part of this thread pretty diligently and all of a sudden disappeared once scores were posted or individuals who have noted in their posts that they scored pretty low and understandably don't feel comfortable posting their scores. So the actual score average might be a little lower, and this might come into play for the analysis of the next paragraph. If all those people scored 10 points below their average, then yeah, we might have a problem. But I've seen quite a few people (not necessarily in this thread) scoring averages of 25s, expect a 30+, but actually got a 25. Thankfully those people posted their scores and made me realize those individuals exist too. So keep in mind, a person might not say their score because THEY feel it's below their expectation, even if it might be a good score or their score is to be expected. Therefore, I believe the actual averages will not significantly change due to the nine individuals. Some might have relatively good scores (30+), but were below their expectations
I compared AAMC averages with Actual scores from people who gave BOTH scores, not just one or the other. Overall there was no difference. 2pm's actual score was 0.7 points above their AAMC average and 8am was exactly the same. So no difference. It can be inferred that you will score your AAMC average. Trust your average! In fact, 14 people out of 55 people (25%) who gave me both scores scored the EXACT same AAMC average (after rounding the aamc average). That's quite a significant number of people. Fortunately, I scored 3 points above my average
However, my last two AAMC practice exams I took were a 39 and 37 (AAMC 9 and 8 respectively), and the average of those two was my actual score. I am so lucky. Thank you all who supported me
Now in contrast, the mean
difference between the AAMC average (not rounded) and Actual score was +0.47 (st. dev 2.34). The mean
absolute difference was 1.88 (st. dev 1.45). So if I am interpreting this right, even though on average people pretty much score their exact AAMC average (as suggested by the previous paragraph), if you look at the
absolute difference, it suggests otherwise. People tend to score about +/- 1.88 from their average. However, we predicted that people tend to score +/- 2 from their AAMC average, which isn't bad, and the data supports that!
In one sentence: Most likely, you will score your AAMC average, BUT on average, you will score +/- 1.88 from your AAMC average. Hope that makes sense.
Think of a bell curve centered around your average. The peak of the curve is your mode, which is your average. You are most likely to score your mode/average and you are most likely to stay within 2 points of your average. However, don't fret if you score +/- 3 or 4. Keep in mind it is a possibility, but a little less likely (the standard deviation of the absolute difference is 1.45, 1.88+1.45 = +/- 3.33)
Also note 8am was not necessarily easier than 2pm. Look at the AAMC averages from both times and compare to their respective actual scores. They are about the same. More people might have done slightly worse in 2pm than 8am, but those people were slightly worse exam takers to begin with.
Furthermore, we had about a 60% response rate of posters who posted at least one message saying they took this exam. This is good news! We were expecting like around 30% and not even close to 60% at best.
Finally, SDN exam takers are indeed above average. We have a mean, median, and mode of 32! Great Job!!! Proud of y'all, even those who might not have gotten what they wished for
Hope this is very informative! Peace out. Back to being a lurker, that is if I am ever on here again.
