Predicting UCAT Scores from Mock Exam Results

[KM7]

Tbh I'm not sure how this would help you, and it's best to focus on your own score because 'good' is very subjective and varies by individual. Work on improving your scores to reach your goal (or a score you think is 'good') and focus on your mistakes. I always believe that expecting a lower score is totally fine, but you should always aim for the best possible score (in this case, 44/44), otherwise, you're just limiting yourself. Your opinion of a 'good' score is also subject to change as you start improving, as will your confidence levels. I know I didn't really answer your question, but focus on how you can improve!

but if you really want to know that badly, i'd be happy with 38+/44. this is probably going to change over time and has changed from my previous expectations. as i've already said, best to focus on your scores!

Have you hit 38+ in any of the official exams? I have noticed it is markedly more difficult than most other resources and have scored on average about 5 marks lower...

 

[cocodreams]

Tbh I'm not sure how this would help you, and it's best to focus on your own score because 'good' is very subjective and varies by individual. Work on improving your scores to reach your goal (or a score you think is 'good') and focus on your mistakes. I always believe that expecting a lower score is totally fine, but you should always aim for the best possible score (in this case, 44/44), otherwise, you're just limiting yourself. Your opinion of a 'good' score is also subject to change as you start improving, as will your confidence levels. I know I didn't really answer your question, but focus on how you can improve!

but if you really want to know that badly, i'd be happy with 38+/44. this is probably going to change over time and has changed from my previous expectations. as i've already said, best to focus on your scores!

Yeah, no I agree about it being individual and not limiting oneself. I was just really curious because at this rate I feel like I am not improving and the exam is so close I guess it's about where should i really be at this point in time (like a month away from the exam)

Have you hit 38+ in any of the official exams? I have noticed it is markedly more difficult than most other resources and have scored on average about 5 marks lower...

Yeah it is a bit weird, considering the passages are not actually super long, but some do require inferencing?

 

[ucatboy]

Have you hit 38+ in any of the official exams? I have noticed it is markedly more difficult than most other resources and have scored on average about 5 marks lower...

Yeah I remember getting low 30s for the official exams when my average on other sites was ~36-38

 

[N182]

Have you hit 38+ in any of the official exams? I have noticed it is markedly more difficult than most other resources and have scored on average about 5 marks lower...

Yes I also agree with that. Official has more inferences and a lot of the answers seem correct, but only one answer is the most correct, causing a lot of time being wasted looking at each statement and evaulating them.

 

[nb]

I just did my first mock from the official website and damn was it hard...
I just want to let everyone know that you shouldn't be disheartened from your scores from the official mocks. From what I've heard (and now witnessed myself), these official mocks (ESPECIALLY QR) are poorly designed and not even merely indicative of the real QR. I just got 18/36 for QR after getting 33 and 34's / 36 on other prep company websites.
So please take these scores with a grain of salt ppl

 

[Fili]

Yeah don't worry too much with QR stuff from the official website.
However, the other sections are pretty similar to the actual test, AR especially.

 

[nb]

Yeah don't worry too much with QR stuff from the official website.
However, the other sections are pretty similar to the actual test, AR especially.

Yeah, the other sections were comparable to other mocks I have done, just QR that was insane!

 

[someday]

I just did my first mock from the official website and damn was it hard...
I just want to let everyone know that you shouldn't be disheartened from your scores from the official mocks. From what I've heard (and now witnessed myself), these official mocks (ESPECIALLY QR) are poorly designed and not even merely indicative of the real QR. I just got 18/36 for QR after getting 33 and 34's / 36 on other prep company websites.
So please take these scores with a grain of salt ppl

How'd you go for VR? My score dropped to a 34/44 on the official one - just hoping I'm not the only one who finds it quite harder and yeah QR was crazy lmao

 

[AJ13]

How'd you go for VR? My score dropped to a 34/44 on the official one - just hoping I'm not the only one who finds it quite harder and yeah QR was crazy lmao

Dude a 34/44 in the real thing would be a great score! Sounds like you been nailing other mock exams?

 

[someday]

Dude a 34/44 in the real thing would be a great score! Sounds like you been nailing other mock exams?

Thanks! I didn't do substantially better in other mocks - just a few marks higher, but thats just because I the majority of my prep on VR practice. Just trying to keep my other sections up as well now

 

[rainfall]

I have completed 6 practice exams now, and I'm finding that the overall percentile I'm given is higher than most of the individual subtest scores I'm getting. In one practice exam I was below 70th%ile for VR (other substests were low 90s except DM which was 99th), but ended up getting overall 95th%ile. So I thought maybe it was DM that was skewing the overall results.

But then today I did a mock exam and got 88th%ile for DM and AR and low 90s for VR and QR. I ended up with an overall percentile of 96.

Has anyone else noticed their percentiles doing this? I can't think of an explanation for it but I must be missing something.

 

[ucatboy]

I have completed 6 practice exams now, and I'm finding that the overall percentile I'm given is higher than most of the individual subtest scores I'm getting. In one practice exam I was below 70th%ile for VR (other substests were low 90s except DM which was 99th), but ended up getting overall 95th%ile. So I thought maybe it was DM that was skewing the overall results.

But then today I did a mock exam and got 88th%ile for DM and AR and low 90s for VR and QR. I ended up with an overall percentile of 96.

Has anyone else noticed their percentiles doing this? I can't think of an explanation for it but I must be missing something.

It's probably some maths/statistics-related phenomenon - the sum of the individual standard deviations for each section seems to always exceed the overall standard deviation. I noticed this when looking at last year's UCAT summary test statistics, where 77.7 + 83.2 + 81.9 + 91.1 = 333.9 > 269.7. This means that if you score exactly one standard deviation above the mean for every single section i.e. 84th percentile, your overall percentile will actually be 333.9 ÷ 269.7 = 1.238 percentiles above the mean i.e. 89th percentile.

 

[Ali Issmail]

I've have completed my first mock subtests for verbal reasoning and decision making but i am sitting the UCAT in 2021.

For VR I scored 630/900 and for DM I scored 600/900. I am just wondering whether this is a good score to start off with this or is this a concerning score. Also, by how much should I be improving every time I complete a mock subtest.

Thanks in advance!

 

[AJ13]

I've have completed my first mock subtests for verbal reasoning and decision making but i am sitting the UCAT in 2021.

For VR I scored 630/900 and for DM I scored 600/900. I am just wondering whether this is a good score to start off with this or is this a concerning score. Also, by how much should I be improving every time I complete a mock subtest.

Thanks in advance!

Hey base 600’s is a great foundational score!
As far as improvement, it’s honestly dependent on the work you put in between each mock exam. If you try to focus on the styles of questions you got wrong then you are only going to improve.
Next year is still so far away but I respect your attitude trying to get prepared as early as possible

For example I started a little lower than you in my first mock (550ish average) back in February and it has taken me until now to start hitting the scores I set myself (4 full mocks + all mini mocks).
Bare in mind I Knew I was going to need to put this much time into prep to compete with other people in this forum (yikes).

 

[nb]

I've have completed my first mock subtests for verbal reasoning and decision making but i am sitting the UCAT in 2021.

For VR I scored 630/900 and for DM I scored 600/900. I am just wondering whether this is a good score to start off with this or is this a concerning score. Also, by how much should I be improving every time I complete a mock subtest.

Thanks in advance!

imho, I don't think you need to be too concerned with your scores at this point in time (You have a whole year to prepare). If you do really want to start preparing now, just focus on improving your speed (in reading and calculating for example in maths). I wouldn't recommend doing practice mocks now as you will very likely run out well before your real test. As Average Joe13 said, those are definitely decent scores to get (I got much lower on my first couple mocks).

 

[whys]

Hi everyone, I've got a question regarding these scores:

As you can see, I did appallingly in QR , which is unfortunately not very surprising. For this mock, I got an overall of 95th percentile, but how on earth does this work if I got 26th for QR?! Shouldn't the overall percentile be the average of the percentiles for each mock? And how come a 35/44 is equivalent to 900 for VR, but a 32/39 for DM is also 100th percentile but yields only 890? I also thought you needed to get at least 40+/44 for VR to get a 900, since the scaled score is not supposed to be a percentile, and stays the same no matter what.

 

[KM7]

Hi everyone, I've got a question regarding these scores:
View attachment 3452
As you can see, I did appallingly in QR , which is unfortunately not very surprising. For this mock, I got an overall of 95th percentile, but how on earth does this work if I got 26th for QR?! Shouldn't the overall percentile be the average of the percentiles for each mock? And how come a 35/44 is equivalent to 900 for VR, but a 32/39 for DM is also 100th percentile but yields only 890? I also thought you needed to get at least 40+/44 for VR to get a 900, since the scaled score is not supposed to be a percentile, and stays the same no matter what.

This platform uses the total raw score for the 4 cognitive sections to calculate your percentile, so your very good scores in VR and DM (and also AR as well) put you well ahead (probably very high 90s) and offset the poorer QR score. When you think about it, the higher QR scores probably range from 22-25+ (depending on which exam - e.g. I got 98%ile for 22/36 in exam no. 4), so you have only dropped about 6 or so marks from the majority of other test-takers in that section. You have more than made up for that difference in VR and DM - again, compared to others in this test.

I would be interested to know which exam this was... I am surprised a 35/44 scored a 100%ile in VR: usually, the threshold for 100%ile is a bit higher.

In regards to the scaled scores, I noticed a few months ago that this platform bases the scaled score entirely on the percentile given for each subtest. It uses the normal distribution scale and 2019 UCAT statistics (mean and standard deviation) to calculate what score you would have received if you beat 100% or 80% of the population. That's why for VR, the highest score was 900 last year, so that is the score you receive. But if you score in the 99%ile, you will receive 750. Same for DM (890 was the highest) and the others being 900 also. I tend to use the total scaled scores as a VERY general indicator of my performance, mostly using the percentiles for each subtest and overall.

This system is probably not as accurate in calculating scaled scores as you would hope for, especially considering the massive difference between scoring in the 99%ile and 100%ile for all of the subtests - even for QR is a difference of 50, DM = 80, etc.

To sum up, I would take these scaled scores with a pinch of salt. Overall, they tend to even themselves out to provide a general indicator of your performance against other test-takers on that platform - but don't stress over them too much. And don't expect to get an overall score in the 99%ile (based on 2019 stats) every time - or any score people are aiming for - due to the way they are calculated. Aim to maintain and/or improve your overall percentile (as your section percentiles can vary markedly with a difference of a few marks up or down) as it is the most reliable indicator of your performance.

 

[ucatboy]

Hi everyone, I've got a question regarding these scores:
View attachment 3452
As you can see, I did appallingly in QR , which is unfortunately not very surprising. For this mock, I got an overall of 95th percentile, but how on earth does this work if I got 26th for QR?! Shouldn't the overall percentile be the average of the percentiles for each mock? And how come a 35/44 is equivalent to 900 for VR, but a 32/39 for DM is also 100th percentile but yields only 890? I also thought you needed to get at least 40+/44 for VR to get a 900, since the scaled score is not supposed to be a percentile, and stays the same no matter what.

I think you gotta understand that percentiles work very differently fundamentally to raw scores or even scaled scores for that matter. For example, it's possible to get 96th percentile by getting 300 in VR and 900 in the other three sections (300 + 900 + 900 + 900 = 3000).

I would also take the scaled scores provided by the preparation platform in question with a grain of salt because exactly like KM7 said, they are based on the percentiles you achieved in the mock superimposed on last year's test statistics. It's definitely different to how Pearson does it (through IRT or whatnot), where each question is assigned a score value depending on its difficulty but the only way to get 900 is still to get all the questions in a subtest correct.

 

[whys]

Thanks everyone for clarifying

I would be interested to know which exam this was... I am surprised a 35/44 scored a 100%ile in VR: usually, the threshold for 100%ile is a bit higher.

This was exam 6. I was also quite surprised - I've gotten higher than this but got a lower percentile.

 

[nb]

Hey, can somebody please explain how a 98th percentile yields a 790 subtest score and a 100th yields a 900 subtest score... That means a 99th percentile will yield a score between 800-900 which is far too big of a range...