CaRMS School Match Rate

[Aetherus]

Every year, it seems people try and compare the match rates of every school in order to determine which school is the most successful in matching their students to their desired specialty. However, simply comparing the match rates does not paint the whole picture. People commonly state that a certain school has a lower match rate because their students went for a more competitive specialty or that a certain school has a high match rate because most of their applicants go into primary care.

 

This is my attempt to come up with a standardized approach to comparing school match rates.

 

I will post my results for the 2018 Match and then explain my methodology. If anyone has any constructive criticism, that would be awesome.

 

School	Standardized Match Rate
Queen’s University	1.12183974
Dalhousie University	1.11964593
University of Ottawa	1.06570349
McGill University	1.00310852
University of Alberta	0.99333422
Northern Ontario School of Medicine	0.97475584
Western University	0.97029931
University of Toronto	0.97021267
McMaster University	0.96812175
Memorial University of Newfoundland	0.96625383
University of Calgary	0.9641772
University of Manitoba	0.9625414
University of Saskatchewan	0.95720077
University of British Columbia	0.94367284
Université Laval	0.94277376
Université de Montréal	0.85887693
Université de Sherbrooke	0.82016583

 

Methodology

 

Calculating the Competitiveness Score

In order to compare each school, I created a competitiveness score using table 34 from the 2018 CaRMS Data. I used the inverse of the Supply>Demand ratio’s from the CaRMS talk to generate a score for each specialty. I then multiplied this by the amount of students in each school who ranked this specialty first. This was added up for a total score and then divided by the amount of students participating in the match. This created a Competitiveness Score for each school based on how difficult the match strategy was for the school.

 

Calculating the Standardized Discipline Score

Using Table 16 from the 2018 CaRMS Data, I created a standardized score for each school based on how many people match to their 1st choice discipline, 2nd choice discipline etc. Students who matched to their top discipline had a 1x multiplier, students who matched to their 2nd choice discipline had a 0.5x multiplier. Students who matched to their 3rd choice discipline had a 0.25x. Students who matched to their 4th choice discipline had a 0.125x multiplier. Finally students who matched to lower than their 4th choice discipline had a 0.0625x multiplier. I then added all these up and divided by the total amount of individuals participating in the match to give the Standardized Discipline Score.

 

Calculating the Standardized Match Rate.

Standardized Match Rate=SMR

Competitiveness Score= CS

Standardized Discipline Score=SDS

SMR=CS*SDS

 

Let me know what you guys think!

[a7x]

I don't think this captures the whole story of people's rank order lists. Specifically, by not including any factor relating to location preference (which we don't know from the Carms data) it makes the numbers you generate relatively meaningless, some home schools are more competitive than others for example, this is a fact borne out by the fact that some schools have virtually no unfilled spots even in less competitive disciplines year after year.

[Edict]

Its a nice attempt no question, but the preferences are so complex and there are so many factors that you just can't compare medical schools across the country. Quebec's match is very isolated compared to the rest of Canada due to language, people tend to stay local, people also do tend to move away more commonly for certain locations etc. 

[la marzocco]

Queen’s have been leading the pack for quite a few years. They have been pretty consistent no doubt. Don’t know how they do it.

[indefatigable]

3 hours ago, Edict said:

Its a nice attempt no question, but the preferences are so complex and there are so many factors that you just can't compare medical schools across the country. Quebec's match is very isolated compared to the rest of Canada due to language, people tend to stay local, people also do tend to move away more commonly for certain locations etc. 

I agree with this - a lot of people will prioritize location for example, and a given specialty in TO will be more "competitive" than in other locations.

However, it's an interesting way of looking at the data and trying to capture competitiveness.

I'd suggest that in the SDS, since the difference between 2nd and 3rd choice could be minimal in terms of competitiveness, it might be better to use table 17 to objectify the 1st, 2nd, 3rd choice, etc.. instead of the multipliers 1, 1/2, 1/4... (e.g. weighted average like the CS)

Moreover, in QC, this might not be the case at all, since people tend do broader electives so choice 1, 2 and 3 could actually still be competitive specialties (for instance plastic surgery as 2nd choice).    

But I'm not sure it really makes sense to compare QC to the rest of Canada, since the movement between the two language groups is almost zero (McGill excepted).  
 

[bearded frog]

28 minutes ago, carmscarmscarms said:

for anyone who matches to something that they didn't expect, there is now a new website to discuss swapping spots with someone else in the country
(just like residencyswap.org in the USA): 

https://www.residentransfer.com/

Transfers don't work that way, you can't just swap spots with someone.

[thestar10]

42 minutes ago, carmscarmscarms said:

for anyone who matches to something that they didn't expect, there is now a new website to discuss swapping spots with someone else in the country
(just like residencyswap.org in the USA): 

https://www.residentransfer.com/

This is a scam.

[Tullius]

As far as I can tell, this is the most convincing analysis of match rates to date.  CaRMS really should be providing this information, with the location preferences incorporated so that medical school applicants can have complete information, and schools can look for ways to improve their students' prospects.  But since they don't provide the location preference data (as far as I know), or granular enough data to compile it manually, perhaps the OP's chart is the best we can hope for.

 

Well done.

[1D7]

Match rates don't tell the whole story if you don't have data on specialties students applied to and rate of matching into backups.

I know at my school the previous years match rate was terrible but we had a fantastic one this year. Anecdotally this is purely because more people applied to less competitive specialties and more people backing up.

[Aetherus]

1 hour ago, Tullius said:

As far as I can tell, this is the most convincing analysis of match rates to date.  CaRMS really should be providing this information, with the location preferences incorporated so that medical school applicants can have complete information, and schools can look for ways to improve their students' prospects.  But since they don't provide the location preference data (as far as I know), or granular enough data to compile it manually, perhaps the OP's chart is the best we can hope for.

 

Well done.

Thank you!

It would also be nice if CaRMS could publish data that had Quebec and the rest of Canada separate. The reality of the Quebec match is very different than the rest of Canada and combining the data together obscurs the reality of the match for both Quebec Grads and the rest of Canada.

[Tullius]

Why do you think combining the data changes things? Is it not true that quebec grads fare less well at matching to competitive residencies?

[NeuroD]

5 minutes ago, Tullius said:

Why do you think combining the data changes things? Is it not true that quebec grads fare less well at matching to competitive residencies?

It's more that they apply to unique set of programs that most of the rest of the country can't apply to (french).

[PurpleS]

I feel like there are individual factors that cannot be captured in statistics no matter how we analyze the numbers. Home school advantage is hard to quantify. Some schools don't even have certain programs (no derm at Western, or Queen's for example). Individuals who gun hard for specific specialities will likely do so regardless of which medical school they go to. For smaller programs, personality and "fit" is also important and cannot be quantified easily. Match rates will also fluctuate year to year based on class interest - so unless you combine data from every year it's hard to know which schools are "best" at matching students. Competitiveness of specialities also fluctuate from year to year - for example, 2019 was a difficult year for IM.

The reality is, students should not be picking their medical schools based on the CaRMS match rate. Who knows how the match might change in 3 - 4 years by the time they're done.

[Snowmen]

On 3/29/2019 at 8:21 AM, Tullius said:

Why do you think combining the data changes things? Is it not true that quebec grads fare less well at matching to competitive residencies?

For instance, we had a presentation today by a CaRMS employee who mentioned that a lot of the data they provide is completely irrelevant to us. Regarding the elective in match discipline data that they've started to release, she was saying that the data was completely upside down with almost no one having completed 3+ electives in their matched discipline while almost all the cases where someone matched without an elective were in Quebec. She had a few more examples like this.

[orangeisthenewblack]

13 hours ago, Snowmen said:

For instance, we had a presentation today by a CaRMS employee who mentioned that a lot of the data they provide is completely irrelevant to us. Regarding the elective in match discipline data that they've started to release, she was saying that the data was completely upside down with almost no one having completed 3+ electives in their matched discipline while almost all the cases where someone matched without an elective were in Quebec. She had a few more examples like this.

Do you mean 3 or more, or greater than 3 electives in their matched discipline? I'm sure there's tons of people who completed four 2-week rotations since there's so many competitive programs out there and lots of applicants to those programs. Your statement would assume that a majority of people are rather splitting their elective weeks evenly in two (or more) specialties, presumably equally to be somewhat competitive in both. 

[Aetherus]

11 minutes ago, orangeisthenewblack said:

Do you mean 3 or more, or greater than 3 electives in their matched discipline? I'm sure there's tons of people who completed four 2-week rotations since there's so many competitive programs out there and lots of applicants to those programs. Your statement would assume that a majority of people are rather splitting their elective weeks evenly in two (or more) specialties, presumably equally to be somewhat competitive in both. 

I think what Snowmen is trying to say is that the Match Strategy in Quebec are quite different then the rest of Canada. It is not uncommon for an applicant to go for Plastics and Ophtho and match to one of them while having split their electives.

[orangeisthenewblack]

23 minutes ago, Aetherus said:

I think what Snowmen is trying to say is that the Match Strategy in Quebec are quite different then the rest of Canada. It is not uncommon for an applicant to go for Plastics and Ophtho and match to one of them while having split their electives.

Ah, that makes a lot more sense thanks!

[Snowmen]

12 hours ago, Aetherus said:

I think what Snowmen is trying to say is that the Match Strategy in Quebec are quite different then the rest of Canada. It is not uncommon for an applicant to go for Plastics and Ophtho and match to one of them while having split their electives.

Exactly. Most schools have had limits on the number of electives in one discipline for a while. For instance, it is one at my school unless you get approval from admin (which is always granted so in practice, it's two except for "non-clinical electives" like radiology or pathology which is a hard-limit at one). In practice, that means that the elective strategy is very different and I'm being called a PM&R "gunner" for doing two 2-week electives in the field.