Preliminary 2022 McMaster Intervew Invites Results

[bearded frog]

I have tabulated the data from the results thread here as well as the 2 results threads on r3ddit:

https://forums.premed101.com/topic/115693-mcmaster-interview-invitesregrets-2022/

https://www.**DELETED**.com/r/premedcanada/comments/s36fjx/mcmaster_interview_invitesregrets_2022/

https://www.**DELETED**.com/r/premedcanada/comments/s3ylk3/mcmaster_invitedregrets_2022_repost/

The following is only for in-province applicants.

Please not this will in no way be a representative data set of all applicants. Clearly, as invitees make up the largest group. People are more likely to post if invited. Most obviously only a subset and probably not representative group will post about their application on the internet. Also there will be a bias towards higher stat applicants posting due to embarrassment etc. So take this all with a grain of salt. Also I haven't compared the r3ddit responses to premed but maybe I should...

I have done my best to remove duplicates of people posting both here and on r3ddit but there are a few examples of people with the exact same stats with different email times (and even some with the exact same stats and different outcomes!). Overall, it's clear that where you are within the casper quartile makes a big difference so unfortunately you can't be confident about anything, except that if you're in 3rd quartile you better have a 130 or higher CARS and a 3.8 or higher GPA if you want an interview. There were no invites to 2nd or 1st quartiles. (Nobody posted that they applied with a 1st quartile...).

There were 99 responses, 84 in province. Of those 84, 44 were invited, 6 were waitlisted, and 34 were rejected. 6/44 invites had a 3rd quartile casper (13.6%), and 22/34 rejected had a 4th quartile casper (64.7%). All waitlisted had a 4th quartile casper. A pie chart of results (I = interview, W = waistlist, R = regrets) by quartile follows:

First, the averages for each group, with green being invites, yellow waitlisted, and red regrets. As you can see, individually the standard deviations overlap with this small sample size, so hard to draw conclusions for any variable on its own:

The first graph which may actually be helpful is the total results showing the array of CARS and GPA (in bounds of all accepted applicants - there were rejected applicants with GPA/Cars lower than visible), with bubble size relative to casper quartile (tiny dot is 2nd quartile, large bubble is 4th), and green for invites, yellow for wait-list, and red for regrets:

As you can see, its confusing, because it's difficult to visually display data with 3 axis, so I felt it made the most sense to graph GPA and CARS distribution for invites and regrets within Casper quartiles. The first graph is for 4th quartile casper, with green for invite, yellow for waitlist, and red for regrets:

As you can see, clearly where you are within the 4th quartile will play a signficiant role, as the red/yellow/green overlap. Although interesting to note that all with a cars of at least 129 in the 4th quartile at least was waitlisted or invited. There doesn't appear to be a similar threshold for GPA, although worth noting that 4.0 GPA had the lowest invited CARS at 126. Also, 3.7X GPA definitely seems to be in play with at least a 127/128 CARS and a good casper.

The data for the 3rd quartile casper is a little more defined:

For 3rd quartile is seems that 130 is an absolute cut off, and potentially relatively more signficant than GPA, as there were some high GPAs with regrets due to a CARs 129 or lower. No 2nd quartile applicant was invited.

Looking at GPA and CARS on their own, as I said above, is not as helpful, but we may be able to see ranges.

Here is a histogram of applicant GPAs. light green is invited with 3rd quartile casper, dark grey is invited with 4th quartile casper. Light red is regrets with 3rd quartile casper, and dark red is regrets with 4th quartile casper:

As you can see, lots of overlap, but 3rd quartile casper has a clear cut off at 3.8.

Below is the same graph but for CARS:

Here we can start to see some maybe-almost-somewhat-normally distributed results, with again a 3rd quartile casper cut off of 130, and no regrets to 131 and above.

That's all I have so far, open to suggestions for other analysis

[rmorelan]

Ha, thanks for going through that! I will have to go through it in more detail.