The Mmi Interview Scoring And Its Impact On Acceptance

[bearpuppy]

Hi all,

 

I've come to the conclusion that UC is a school where the MMI is basically the deciding factor on admission. To many of you this may seem obvious, but I do think it's somewhat unfortunate. I'll explain what I mean:

 

In the 2013/2014 cycle, the pre interview score for +/-1sd was 104.7 - 119.1 for successful applicants. The final score for successful was >210. The MMI distribution was fairly tight (as in other years) -- 2013/2014 seemed to range from 8  to ~16.5 out of 20. The mean was 13.03 and the std dev. 1.61. That means that 68% of interviewees scored between 11.42 and 14.64. Thus, if you were to score 11.42 you would get a standardized score of 85 (vs mean of 100) and this would then go from a 15 pt difference to a 7.5 pt difference in your final score.. This is pretty significant, considering that on average, you got 1-2 less points on an interview station. This doesn't seem to be a good differentiating factor between applicants.

 

As an example, lets say a friend and yourself score the same pre-interview: 110. Thus you need to hit the average on the MMI (standardized score of 100) to get an acceptance. You scored 11.42 on average on the MMI and your friend 13.03.... Your friend gets accepted (210 final), and you don't (202.5 final). 

 

To me, it seems to me that for most people the MMI is what determines your admissions... and all those other things many of us had worked hard for -- good grades, volunteering, research/academics, reference letters, etc. -- ends up playing a minor role in the decision.

 

Personally, I think it would make more sense to either (1) increase the spread of the MMI scores (maybe change it to out of 30), and/or (2) reduce the weight of the interview in the final score. It seems somewhat unfair that a small difference can essentially prevent you from getting in..

 

Anyway - just my two cents.

 

 

[Tacrolimus]

People say "GPA is king"... nope. MMI is. Also, I completely agree with you. Whether you get accepted or not literally depends on 1 mark out of 20 on the interview. 14/20 means you're likely okay (15/20 = admission), and 13/20 means you're fighting an uphill battle. 12/20 means pretty much no chance. 

 

If the standard deviation is similar this year too, every one point on the interview is essentially ~9 points, which is easily the difference between acceptance and rejection. In the end though, it is a heavily subjective process, but we also don't have a better option than that, so... 

[amichel]

People say "GPA is king"... nope. MMI is. Also, I completely agree with you. Whether you get accepted or not literally depends on 1 mark out of 20 on the interview. 14/20 means you're likely okay (15/20 = admission), and 13/20 means you're fighting an uphill battle. 12/20 means pretty much no chance. 

 

1/20 on the average score, but that could also be 1-2 points per station.  I'd say that's significant. 

[LeBronto2019]

UofC is a hater!  

[JohnGrisham]

How else would they differentiate though? GPA is king in the sense that if you don't have it, you're fighting an extreme uphill battle.

But once everyone has the 4.0, they need to be able to differentiate.

From everything i've read though, I think UofC does a great job at being transparent and running a good process.

[Tacrolimus]

How else would they differentiate though? GPA is king in the sense that if you don't have it, you're fighting an extreme uphill battle.

 

But once everyone has the 4.0, they need to be able to differentiate.

 

From everything i've read though, I think UofC does a great job at being transparent and running a good process.

Oh, I completely agree. They are probably the most transparent school in Canada, and I really appreciate that they disclose so much information. It's also encouraging in a sense that I know it's just one part (albeit very important) of the application I need to fix for next year. 

[Consequence]

every single part of the admissions process is unfair in some small way. VR scores can deviate significantly with a few correct guesses, high GPAs are easier to obtain in some programs/school, reviews of an applicant community advocacy are very subjective, etc etc.

 

the hope is that all of these together combine to form a relatively fair view of the applicant. the system isnt perfect, but its the best option available

 

i find this forum often finds faults with U of C's process because they are so open about scores / techniques to judge applicants. in the end, all schools in Canada are using a somewhat similar admissions criteria, some more or less fair

[bearpuppy]

every single part of the admissions process is unfair in some small way. VR scores can deviate significantly with a few correct guesses, high GPAs are easier to obtain in some programs/school, reviews of an applicant community advocacy are very subjective, etc etc.

 

the hope is that all of these together combine to form a relatively fair view of the applicant. the system isnt perfect, but its the best option available

 

i find this forum often finds faults with U of C's process because they are so open about scores / techniques to judge applicants. in the end, all schools in Canada are using a somewhat similar admissions criteria, some more or less fair

 

I don't mean to hate on Calgary.. All I'm saying is that I think the interview is given disproportionate weight given its tight distribution of scores and on the low value that is then placed on other aspects of the application that have in my opinion taken much more effort to accomplish.. I think having the interview out of 30 and having it count 25% of the final score would be more reasonable.

[bearpuppy]

1/20 on the average score, but that could also be 1-2 points per station.  I'd say that's significant. 

 

I have to disagree here.. Since the stations are broken down into 10 pts for the rubric and 10 pts for the impression, if you consistently get one point more in impression (for example), that would have a huge effect overall. Whereas if I was an interviewer, i think it'd be very difficult, over many applicants, to definitively identify a 1 (or 2) point difference in their abilities. If there was a bigger spread of values, that would alleviate this.

 

to reword the issue is that I think the error in the measurement of the MMI (obviously impossible for me to know what that is, but could be estimated using the average inter-station standard deviation for all applicants) is likely just as big as the standard deviation (~1.6 pts), and that means that you could easily suddenly end up with 7.5 (1SD = 15 * 0.5) lower (or higher) in the final score.. that's what I have an issue with.

[amichel]

I have to disagree here.. Since the stations are broken down into 10 pts for the rubric and 10 pts for the impression, if you consistently get one point more in impression (for example), that would have a huge effect overall. Whereas if I was an interviewer, i think it'd be very difficult, over many applicants, to definitively identify a 1 (or 2) point difference in their abilities. If there was a bigger spread of values, that would alleviate this.

 

to reword the issue is that I think the error in the measurement of the MMI (obviously impossible for me to know what that is, but could be estimated using the average inter-station standard deviation for all applicants) is likely just as big as the standard deviation (~1.6 pts), and that means that you could easily suddenly end up with 7.5 (1SD = 15 * 0.5) lower (or higher) in the final score.. that's what I have an issue with.

Yah, I don't think any of us know enough about interview scoring to draw these conclusions. 

 

EDIT for elaboration.   For example, what if (and neither I nor anyone on this forum has any idea if this is true) there are criteria for each station.  If on each station, you are consistently missing one important piece of criteria (say 'considers consequences of both options'), I'd say that's a significant difference. 

 

The point is we don't know. 

[bearpuppy]

Yah, I don't think any of us know enough about interview scoring to draw these conclusions. 

 

EDIT for elaboration.   For example, what if (and neither I nor anyone on this forum has any idea if this is true) there are criteria for each station.  If on each station, you are consistently missing one important piece of criteria (say 'considers consequences of both options'), I'd say that's a significant difference. 

 

The point is we don't know. 

 

agreed...and it definitely "is what it is" at this point..

[kevojy]

Thanks for this post! I thought I had an absolutely AWFUL MMI score, but if your numbers are correct then I was just shy of that acceptance letter or waitlist. Of course, that doesn't change anything. But at least I can stop beating myself up for an awful interview. Gives me hope for the next time I interview with a lot more practice and experience (I wasn't expecting to get an interview this year, so I was unprepared to the point where my first MMI was the real thing!).

 

And while we're on the topic of fairness, I don't think it's really right to say that there is no difference between someone who scored an 11.42 and a 13. You have to draw the line somewhere, and the fact remains that the 13 score outperformed the 11.42 score. I don't think this would change if they increased the score to being out of thirty or even forty. The only way they could 'add fairness' would be to add more stations, and I don't think that's realistic (12 is already pretty generous). 

[bearpuppy]

Thanks for this post! I thought I had an absolutely AWFUL MMI score, but if your numbers are correct then I was just shy of that acceptance letter or waitlist. Of course, that doesn't change anything. But at least I can stop beating myself up for an awful interview. Gives me hope for the next time I interview with a lot more practice and experience (I wasn't expecting to get an interview this year, so I was unprepared to the point where my first MMI was the real thing!).

 

And while we're on the topic of fairness, I don't think it's really right to say that there is no difference between someone who scored an 11.42 and a 13. You have to draw the line somewhere, and the fact remains that the 13 score outperformed the 11.42 score. I don't think this would change if they increased the score to being out of thirty or even forty. The only way they could 'add fairness' would be to add more stations, and I don't think that's realistic (12 is already pretty generous). 

 

It'll be interesting to see what the stats look like for this year and just how much spread there was in the MMI scores.