- Joined
- May 16, 2018
- Messages
- 1,840
- Reaction score
- 4,680
Hey all. I was inspired by @planeblue 's data pulled from 2015-2016 and wanted to share an updated one from 2019-2020 with some new additions. While @planeblue went through each thread and jotted down data points, I pulled all of mine from my spreadsheet. I borrowed a lot of their formatting so thank you to them.
Format of my spreadsheet:
Sheet 1: II tracker
Sheet 2: links to add data
Sheet 3: II raw data
Sheet 4: Outcomes Summary
Sheet 5: Outcomes raw data
Sheet 6: Difference to the mean
Sheet 7: Average MCAT by school
Sheet 8: Average GPA by school
Sheet 9: Interview type/info
Sheet 10: By Month Summary
Sheet 11: MCAT Distribution
SCOPE:
From those, I pulled data from the II raw data to find the following...
(II: n=2290) (MCAT: n=2260) (GPA: n=2237)
Overall, from MSAR there were approximately 70,395 interview invitations reported as given out in the 2018 cycle. My sample represents 3.3% of the total interview invites as reported by SDN and Reddit members. It includes 147 US allopathic schools for an average of 16 II's per school.
DATA & METHODS:
I used SPSS and excel to compare the reported GPA and MCATs to AAMC's matriculant data found here
The matriculant mean for GPA is 3.73 (SD: 0.24) and for MCAT it is 511.5 (SD: 6.5)
To start, here are the basic statistics from my collected data:
The MCAT is normally distributed so I made a histogram for it and found the mean is 514.91 (SD: 5.469)
because the MCAT is normally distributed, I ran a two-tailed T-Test to see if the reported MCAT score was different in a statistically significant way from the reported matriculant mean. The p value = <0.000 and is stastically significant which should come as no surprise based on self reporting.
For GPA, things got a bit trickier since it is not normally distributed. I used a One-Sample WIlcoxon Signed Rank Test:
Feel free to correct me if I'm wrong, but I used the same One Sample Wilcoxon Signed Rank Test to see if the GPA was higher than the matriculant median in a statistically significant way and believe it is...
BUTTTTT for the stuff people seem to care more about, see below:
The data:
Total interview invites across schools:
Percentage of interview invites remaining by month:
Obviously there are limitations with this data:
A) I am not a statistician and my methods could be flawed in part
B) This is self reported data from SDN and Reddit where people likely tend to have higher scores (as seen by the statistically significant difference btwn matriculant mean for GPA and MCAT and the one reported on my spreadsheet)
C)The cycle is not fully complete - as it draws to an end, I will update this
D) This should not be taken as the be all end all for anyone at any school. Things change cycle to cycle and are unpredictable.
Going forward:
Please feel free to message me if there are any glaring errors. There are more statistical analyses that can be run on the data as well.
APPLY EARLY - this data is biased but it seems to emphasize what many adcoms on here say, applying early is key
Thank you to everyone that has posted their IIs or acceptances on my tracker this cycle 🙂
@Lucca @efle you might find this interesting
PS: Also if anyone is looking for my post-II A spreadsheet, you can find it here
and there is an attachment on here and idk how to delete it...sorry
Format of my spreadsheet:
Sheet 1: II tracker
Sheet 2: links to add data
Sheet 3: II raw data
Sheet 4: Outcomes Summary
Sheet 5: Outcomes raw data
Sheet 6: Difference to the mean
Sheet 7: Average MCAT by school
Sheet 8: Average GPA by school
Sheet 9: Interview type/info
Sheet 10: By Month Summary
Sheet 11: MCAT Distribution
SCOPE:
From those, I pulled data from the II raw data to find the following...
(II: n=2290) (MCAT: n=2260) (GPA: n=2237)
Overall, from MSAR there were approximately 70,395 interview invitations reported as given out in the 2018 cycle. My sample represents 3.3% of the total interview invites as reported by SDN and Reddit members. It includes 147 US allopathic schools for an average of 16 II's per school.
DATA & METHODS:
I used SPSS and excel to compare the reported GPA and MCATs to AAMC's matriculant data found here
The matriculant mean for GPA is 3.73 (SD: 0.24) and for MCAT it is 511.5 (SD: 6.5)
To start, here are the basic statistics from my collected data:
The MCAT is normally distributed so I made a histogram for it and found the mean is 514.91 (SD: 5.469)
because the MCAT is normally distributed, I ran a two-tailed T-Test to see if the reported MCAT score was different in a statistically significant way from the reported matriculant mean. The p value = <0.000 and is stastically significant which should come as no surprise based on self reporting.
For GPA, things got a bit trickier since it is not normally distributed. I used a One-Sample WIlcoxon Signed Rank Test:
Feel free to correct me if I'm wrong, but I used the same One Sample Wilcoxon Signed Rank Test to see if the GPA was higher than the matriculant median in a statistically significant way and believe it is...
BUTTTTT for the stuff people seem to care more about, see below:
The data:
Total interview invites across schools:
Percentage of interview invites remaining by month:
Obviously there are limitations with this data:
A) I am not a statistician and my methods could be flawed in part
B) This is self reported data from SDN and Reddit where people likely tend to have higher scores (as seen by the statistically significant difference btwn matriculant mean for GPA and MCAT and the one reported on my spreadsheet)
C)The cycle is not fully complete - as it draws to an end, I will update this
D) This should not be taken as the be all end all for anyone at any school. Things change cycle to cycle and are unpredictable.
Going forward:
Please feel free to message me if there are any glaring errors. There are more statistical analyses that can be run on the data as well.
APPLY EARLY - this data is biased but it seems to emphasize what many adcoms on here say, applying early is key
Thank you to everyone that has posted their IIs or acceptances on my tracker this cycle 🙂
@Lucca @efle you might find this interesting
PS: Also if anyone is looking for my post-II A spreadsheet, you can find it here
and there is an attachment on here and idk how to delete it...sorry
Attachments
Last edited:
