How Will The Old And New Mcat Be Compared?

[scaredofthedark]

So I've been seeing some scattered discussion about this and I think it's an interesting topic that warrants a thread.

 

I'm very curious to see how adcoms will compare scores between the old MCAT and the new one.  

 

As you guys are probably aware, both tests follow a 15 point scoring system for each of their sections.  It is important to note, however, that the percentile ranks of a score on a particular section are not equal across exams. For example, an 11 on the verbal section of the old exam would be in the 95th percentile.  The numerical equivalent of this score on the MCAT2015, a 128, would be in the 87th percentile.

 

An easy solution that many have suggested would be for adcoms to look at the percentiles alone.  Unfortunately, a comparison with percentiles across exams would only be justifiable if the sections of both exams are very similar to one another. I guess my argument here is that with the changes to the materials being tested in MCAT2015, like in the Physical Science section, the two sections may not represent the same skillset being tested (?) across exams and therefore would not make much sense to compare the percentiles across them.  (I'd love to hear your opinions about this point in particular). 

 

I think we can all agree that comparing based on numbers alone would not be the ideal way for adcoms to handle this. An 11/11/11 =/= a 128/128/128. I think it would be strange for the adcoms, especially for Western to set the same #-based cut-off for both exams. Anyone who had written the old MCAT would be disadvantaged in this scenario.

 

Finally, the last point I have seen made is to compare WITHIN exams. In other words, for a school like McMaster, apply a z-score to verbal scores of the old mcat and new mcat individually, and grant invitations based on how those applicants faired against others that wrote the same exam. This is definitely a route that could be taken... but I think this is also biased against writers of the old MCAT. Those who score poorly on the old MCAT would be more inclined to re-write, and those who re-write would probably make up a certain population of MCAT2015 writers. This would serve to increase the competitiveness of the old MCAT pool and disadvantage them as a result.

 

Well, that's it for my late night musings. I thought that this was a particularly unique problem due to the MCAT switch and I just wanted to share my views on it! I hope I was coherent (too tired to re-read x_x). Let me know what you guys think!! I'd love to hear what you guys make of all this!

[sdj78we9jtf2o3mgfvj298j]

For schools with cutoffs like Western my opinion is that they should set the cutoff to be similar as in past years and then adjust the cutoffs of the new MCAT accordingly to fill out the interview spots.

 

Queen's has always been a bit more holistic in its evaluation so applicants MCATs aren't directly compared with one another.

 

For Mac they only look at VR/CARS, which are similar enough where comparing percentiles would be appropriate IMO.

[Drosophila]

I also think a direct comparison is not ideal and many factors need to be taken into consideration. 

 

For instance, since the new MCAT has only been offered for one summer, many people have not had a chance to write it multiple times to get an excellent score.  

 

This is in comparison to the old MCAT.  People have had more time to write it numerous times to achieve a competitive score. This would increase the proportion of high scores, likely causing schools to increase their cutoffs to only interview a certain number of applicants. 

 

So for example with the old MCAT, there is probably a greater proportion of people applying with a VR score of 10 and above. In contrast, since there were less opportunities to rewrite the new MCAT, a smaller proportion of applicants would have a CARS score of 127 and above. By this logic, it would not be fair to use the same cutoffs for both tests. Maybe this is one of the reasons UofT lowered their cutoffs from 9 to 125(~8) this application cycle? 

[indefatigable]

 

I also think a direct comparison is not ideal and many factors need to be taken into consideration. 

 

For instance, since the new MCAT has only been offered for one summer, many people have not had a chance to write it multiple times to get an excellent score.  

 

This is in comparison to the old MCAT.  People have had more time to write it numerous times to achieve a competitive score. This would increase the proportion of high scores, likely causing schools to increase their cutoffs to only interview a certain number of applicants. 

 

So for example with the old MCAT, there is probably a greater proportion of people applying with a VR score of 10 and above. In contrast, since there were less opportunities to rewrite the new MCAT, a smaller proportion of applicants would have a CARS score of 127 and above. By this logic, it would not be fair to use the same cutoffs for both tests. Maybe this is one of the reasons UofT lowered their cutoffs from 9 to 125(~8) this application cycle? 

 

Except that a number of places only look at the most recent rather than best score.  As such, on an individual test there is little reason to suspect a rewrite will automatically increase the overall score.  The AAMC has stats on rewrites - the higher the score, the smaller the chance of increasing a score (without intrinsic factors like more preparation).   If only UofT were so understanding with GPA!

[Drosophila]

Yes you are right, I was overly optimistic by saying a rewrite means an increase in the score! 

 

However, if you look at the published AAMC statistics on rewrites I think it supports my argument. 

 

This is a little long but I’d love to know what others think of this and perhaps I am missing something. (Also, I should probably get a life lol)

 

I’ll use McMaster as an example since it only looks at VR and will make things simpler. Lets just say the competitive pool of VR scores is 10 and above (though I know they accept VR >6)

 

Looking at the AAMC rewrite data, it is clear that a large percentage of students either decrease their scores or don’t change but a percentage does improve. The overall effect is that the # of people that improve their scores to become competitive outweighs the number of competitive scores that decrease. It is important to consider N -  many more people with “uncompetitive” scores rewrite compared to those with higher VR scores. 

 

For example-

 

MCAT first attempt: 

 

-X are the number of people who get 10 or above —> X is the size of competitive applicant pool 

-those with VR 9 or under—>not part of competitive applicant pool

 

MCAT rewrite possibilities and effects on competitive applicant size:

 

-some people with a competitive MCAT score from attempt 1 (10 or above) will decrease to under 10 and thus, the number of the competitive applicants decreases (X decreases)

-those who in attempt 1 got 10 and above and in attempt 2 either increased their score, didnt change, or dropped to a minimum of 10 (eg. from 12 to 10) will not have an effect on the competitive applicant pool size (since they are already part of it and continue to be)

-many people who in attempt 1 got 9 and below will stay the same or decrease, which will have no effect on the initial competitive applicant size, since they were not part of it to begin with

-this is important: the total number of people who scored below 10 on their first attempt, rewrote and scored better (eg. scored 7 initially and got enough points to get 10 (like +3, +4, +5), or an 8 and increased by +2,+3, etc.) OUTWEIGHS (about 5 times) the total number of people who scored competitively on their first attempt, rewrote, and dropped their score to become uncompetitive. This would cause an overall increase in the competitive applicant pool number (more people with 10s and above). 

 

I hope this makes sense. Of course this is not exact and there are many other variables. This data is only from one section when most schools look at all sections- this might change things a bit.  Also, its based on students who rewrote in one summer though I think if more time is given to rewrite this might make the effects more pronounced.

 

AAMC statistics on all section rewrites: https://www.aamc.org/students/applying/mcat/data/mcat_stats/

on VR only: https://www.aamc.org/students/download/271686/data/retestervrscorechange.pdf