• We need your support!

    We are currently struggling to cover the operational costs of Xtremepapers, as a result we might have to shut this website down. Please donate if we have helped you and help make a difference in other students' lives!
    Click here to Donate Now (View Announcement)

UHS MCAT (all resources compiled here)

Messages
639
Reaction score
842
Points
103
I was randomly going through some merit lists for Medicine this year and while going through the data the economist eyes realized that there seemed to be some correlation so I did a regression analysis by randomly sampling some data from various merit lists and this is what I found 'with a few exceptions at both ends, generally there is a clear negative correlation between FSc scores and the MCAT of those students who got selected in government medical colleges' hence as the FSc Score goes up MCAT scores on a general scale tend to go down. My explanation for this psychological behaviour would be 'lack of energy and/or over confidence' . . .Numbers are wonderful



Linear regression alpha = 1.16
Linear regression beta(gradient) = -0.22

29gk17c.jpg
 
Messages
639
Reaction score
842
Points
103
^That was astonishing! :p FSc wale jitna books ko ratta maar chorte hain, according to that, they should ace the MCAT! :p

Yea that should happen, but on a general level that didn't happen, although some (very few out of all those selected in public colleges) did manage to balance their FSc scores and MCAT scores, lack of energy and/or over confidence really are the reason. But the interesting point is, if you think you are an average individual who genuinely works hard and is not really a genius then by looking at the those numbers one can predict what kind of FSc/A level score must he/she get to get a reliable relevant MCAT performance (neuro-science).
 
Messages
160
Reaction score
143
Points
53
I was randomly going through some merit lists for Medicine this year and while going through the data the economist eyes realized that there seemed to be some correlation so I did a regression analysis by randomly sampling some data from various merit lists and this is what I found 'with a few exceptions at both ends, generally there is a clear negative correlation between FSc scores and the MCAT' hence as the FSc Score goes up MCAT scores on a general scale tend to go down. My explanation would be 'lack of energy and over confidence' . . .Numbers are wonderful
Wow. That is certainly very...intriguing. This should be sent to the University of Health Sciences.
 
Messages
1,217
Reaction score
3,841
Points
273
after just a month's prep, an A Leveler can score 80% aggregate........

so i guess 3 months would be a sweet minimum to score 95% above......
 
Messages
639
Reaction score
842
Points
103
after just a month's prep, an A Leveler can score 80% aggregate........

so i guess 3 months would be a sweet minimum to score 95% above......

Yes. Either way the numbers do prove that as high scoring A level people tend to get between 935-990 A level equivalence they should as a result work very hard and with high motivation to score on the MCAT above the regression line, as the majority of the scores are below the regression line.

It would be interesting to do another regression analysis if I had data for only the A level people and their corresponding MCAT scores but the prob is many people after O levels switch to FSc so I can't be sure whether the numbers are for A levels equivalence or not.

What about doing a regression analysis for O levels equivalence score and MCAT scores ? That should still be interesting
 
Messages
1,532
Reaction score
478
Points
93
The reason behind the negative coorelation is not lack of preparation or loss in concentration but the difference in the patterns of the test. FSC marks consist of several different kind of papers. While MCAT is totally MCQ based which actually requires you to have a sound grasp of the concepts. And then the marking pattern is also different.
Plus, the addition of A'levels students into this statistical analysis makes it a bit error prone. As even those people with high equivalency at times tend to get a lower MCAT score. So if you remove the A'levels scores from this analysis, I'm sure the linear regression beta will be much less.
But if we ignore these errors, this data totally shows that KIPS and other academies offering MCAT classes are pretty much useless. :p
 
Messages
639
Reaction score
842
Points
103
The reason behind the negative coorelation is not lack of preparation or loss in concentration but the difference in the patterns of the test. FSC marks consist of several different kind of papers. While MCAT is totally MCQ based which actually requires you to have a sound grasp of the concepts. And then the marking pattern is also different.
Plus, the addition of A'levels students into this statistical analysis makes it a bit error prone. As even those people with high equivalency at times tend to get a lower MCAT score. So if you remove the A'levels scores from this analysis, I'm sure the linear regression beta will be much less.
But if we ignore these errors, this data totally shows that KIPS and other academies offering MCAT classes are pretty much useless. :p

Since the data is from candidates who made it to government colleges so the large majority of the data is from Matric/FSc background people therefore nullifying any A level bias together with random sampling so that's not an error. Only a very few really focused people in each merit list managed to keep up the balance between high FSc and MCAT scores, there were in fact an even larger number of people who scored more than 1000 on FSc but their MCAT was in lowest quintile of 900's. There is a certain neurotic tendency associated with it which is absolutely clear from those selected candidates whose FSc score is between 870 to early quintiles of 900's who are almost all lying above the regression line.

The conclusion is simply that it is not statistically significant to assume that those who score lower on the FSc are always going to score higher on the MCAT but that a high FSc score is not a guarantee to a high MCAT score
 
Messages
1,532
Reaction score
478
Points
93
I also want to clear another confusion among people here. I believe that A'levels students have an almost equal chance of making it to Medical colleges as compared to FSC students.
I observed that almost 30 students out of the 303 are from Alevels in KE. And almost the same number is found in other colleges. So about 10% of the intake is of A'levels students. This seems very less but we don't realize that A'levels students who appear for the MCAT are also less than 10% of the total applicants. I don't have any statistical proof for it. But I observed the number of A'levels students in KIPS classesand usually there were 6-8 A'levels students in every 80 students' class. So it is possible that A'levels applicants were even less than 10% of the total applicants.
So comparitively we have been more successful than the FSC students. And I believe that next year even more A'levels students will make it to public universities as the A* equivalency system is stabilized. So all those O'levels students who are stuck between the choice to do FSC or A'levels should not be influenced by any rumours whereby A'levels students are undermined for admission into public universities.
 
Messages
1,532
Reaction score
478
Points
93
Since the data is from candidates who made it to government colleges so the large majority of the data is from Matric/FSc background people therefore nullifying any A level bias together with random sampling so that's not an error. Only a very few really focused people in each merit list managed to keep up the balance between high FSc and MCAT scores, there were in fact an even larger number of people who scored more than 1000 on FSc but their MCAT was in lowest quintile of 900's. There is a certain neurotic tendency associated with it which is absolutely clear from those selected candidates whose FSc score is between 870 to early quintiles of 900's who are almost all lying above the regression line.

The conclusion is simply that it is not statistically significant to assume that those who score lower on the FSc are always going to score higher on the MCAT but that a high FSc score is not a guarantee to a high MCAT score

Like I mentioned above, 10% of data comprises of A'levels students and I believe that is enough to make the data error prone. Though, I still agree with your conclusion. But I believe that the explanation behind this negative correlation is not 'lack of energy and over confidence' as you put it. FSC exams are given over two years so you have more time for preparation and the stress level is much less. Whereas you have to prepare the whole 2 year syllabus with 3 months, plus there is negative marking in MCAT. So that makes it difficult for some people to adjust.
 
Messages
639
Reaction score
842
Points
103
Like I mentioned above, 10% of data comprises of A'levels students and I believe that is enough to make the data error prone. Though, I still agree with your conclusion. But I believe that the explanation behind this negative correlation is not 'lack of energy and over confidence' as you put it. FSC exams are given over two years so you have more time for preparation and the stress level is much less. Whereas you have to prepare the whole 2 year syllabus with 3 months, plus there is negative marking in MCAT. So that makes it difficult for some people to adjust.

Even those who scored lower on the FSc and higher on the MCAT had 2 years to prepare for FSc, if those who always had a higher FSc score because they had spent 2 years of concentration but then went on to get a lower MCAT score then there would be no ground for comparison. MCAT is the dependent variable in principle since FSc is given before the MCAT. The trend says that those who are scoring higher on the FSc on a general scale have a tendency to score lower on the MCAT and that's the principle. And again as far as 'error' is concerned it is negligible because of random sampling, all the 10 % A level or the 90% FSc results are not included but instead sampled in a representative proportion using Econometrics.
While, a lack of energy might not be the sole reason the principle stands that those who did make it to public colleges and had a higher FSc score had a tendency to do less well on the MCAT and there should be a reason other than how the test is marked because that is the standardized variable as those with lower FSc scores also did score higher in general to make it into the merit list which would not be explained by the fact that people find it difficult to adjust, since only the generalized population with higher FSc scores was finding it difficult to adjust
 
Messages
1,532
Reaction score
478
Points
93
It totally depends on what you define as low FSC marks or high MCAT marks. You are only considering those who made it to the medical universities and not the whole population who appeared in FSC and MCAT. So those who you define as the ones with low FSC marks, must have higher marks in MCAT to make it to the medical college. But your sample totally ignores those who had both low FSC marks and low MCAT marks because those people didn't make it to the medical colleges and so you didn't consider them. This is why there is a considerably high negative correlation.
With this data you can't judge whether the ones with low FSC marks were able to adjust to MCAT or not because only a part of this group is included in your data.
 
Messages
639
Reaction score
842
Points
103
Regression Analysis for O level scores and MCAT scores in the future of candidates selected in public sector medical colleges

Conclusion: (As expected from intuition) a higher O level equivalence (being indifferent to the fact whether a candidate goes on to do A level or FSc) generally will yield a higher MCAT score in the future.

Linear Regression Beta(gradient)= +0.41

c6pp4.jpg
 
Messages
639
Reaction score
842
Points
103
It totally depends on what you define as low FSC marks or high MCAT marks. You are only considering those who made it to the medical universities and not the whole population who appeared in FSC and MCAT. So those who you define as the ones with low FSC marks, must have higher marks in MCAT to make it to the medical college. But your sample totally ignores those who had both low FSC marks and low MCAT marks because those people didn't make it to the medical colleges and so you didn't consider them. This is why there is a considerably high negative correlation.
With this data you can't judge whether the ones with low FSC marks were able to adjust to MCAT or not because only a part of this group is included in your data.

Considering those who didn't make it to public sector medical colleges will fool the process of rational thinking since those who didn't make it their cognitive behaviour is not relevant to the equation. The statistical analysis will be helpful to those looking to get admission in the future by considering how those who did get into medical colleges did on the MCAT. So lower FSc score here refers to those who are between 880-920 on the x-axis and are certainly above the regression line for MCAT score in general hence proving the point that even with a lower FSc score if you stay focused, energetic you can make it to the public sector colleges. Busting the myths of a a ridiculously high FSc score and also that A level people can't get admissions if they don't have that. ( The principle of rationality therefore ignores those who had a low FSc mark and a low MCAT mark and didn't make it to medical colleges and also those who had a high FSc mark and a low MCAT score and didn't make it to medical colleges both of these are insignificant in studying the behaviour of that population that made it into public medical colleges which is what people need)
 
Messages
1,532
Reaction score
478
Points
93
Considering those who didn't make it to public sector medical colleges will fool the process of rational thinking since those who didn't make it their cognitive behaviour is not relevant to the equation. The statistical analysis will be helpful to those looking to get admission in the future by considering how those who did get into medical colleges did on the MCAT. So lower FSc score here refers to those who are between 880-920 on the x-axis and are certainly above the regression line for MCAT score in general hence proving the point that even with a lower FSc score if you stay focused, energetic you can make it to the public sector colleges. Busting the myths of a a ridiculously high FSc score and also that A level people can't get admissions if they don't have that. ( The principle of rationality therefore ignores those who had a low FSc mark and a low MCAT mark and didn't make it to medical colleges and also those who had a high FSc mark and a low MCAT score and didn't make it to medical colleges both of these are insignificant in studying the behaviour of that population that made it into public medical colleges which is what people need)

This is where the problem arise. The negative correlation occurs because you didn't both sides of the group with low fsc scores. It might not be relative to the equation but this makes your conclusion biased for the ones with high marks because a large proportion of this group has been included in your data. So to make a fair comparison here. You must include the whole data, and then you'll realize that the date will have much less negative correlation. Making your conclusion pretty much void.

That is what the advantage of MCAT is for us. It provides us with a platform to improve on marks that may be categorized as low in fsc.
 
Messages
639
Reaction score
842
Points
103
This is where the problem arise. The negative correlation occurs because you didn't both sides of the group with low fsc scores. It might not be relative to the equation but this makes your conclusion biased for the ones with high marks because a large proportion of this group has been included in your data. So to make a fair comparison here. You must include the whole data, and then you'll realize that the date will have much less negative correlation.

That is what the advantage of MCAT is for us. It provides us with a platform to improve on marks that may be categorized as low in fsc.

Including the whole set of data will also only 'reduce the gradient' but not convert the negative correlation into a positive one. And that is irrelevant, those who didn't make it into the merit lists are not a part of the equation. The human body runs on a specific level of energy, those who expend more of it suffer later, what these statistics do prove is that you need to balance your schedule once you have got high FSc marks to ensure you get even more on the MCAT. Just like here at McGill the full course load is 5 courses per term, the people who take 5 courses get a lower gpa than people, like myself who do 4 courses per term and score a way higher gpa not because of intelligence level difference but how the body spends its energy. Those who are responsible enough will from this statistic keep in mind that to be the next Kanza Noor they don't need to top FSc but to ensure that they keep a balance between the energy spent on FSc and the energy spent doing the MCAT.
 
Messages
1,532
Reaction score
478
Points
93
Including the whole set of data will also only 'reduce the gradient' but not convert the negative correlation into a positive one. And that is irrelevant, those who didn't make it into the merit lists are not a part of the equation.

You can't be sure. A significant portion of students with FSC marks 880-920 have not made it to medical colleges because of their low MCAT marks. The reduction in the gradient will be significant enough to neutralize the disparity. So to prove your initial conclusion, you have to include the whole data. The conclusion now is that it is POSSIBLE to get a high MCAT score despite getting low FSC marks.
 
Messages
2,884
Reaction score
415
Points
93
Including the whole set of data will also only 'reduce the gradient' but not convert the negative correlation into a positive one. And that is irrelevant, those who didn't make it into the merit lists are not a part of the equation. The human body runs on a specific level of energy, those who expend more of it suffer later, what these statistics do prove is that you need to balance your schedule once you have got high FSc marks to ensure you get even more on the MCAT. Just like here at McGill the full course load is 5 courses per term, the people who take 5 courses get a lower gpa than people, like myself who do 4 courses per term and score a way higher gpa not because of intelligence level difference but how the body spends its energy. Those who are responsible enough will from this statistic keep in mind that to be the next Kanza Noor they don't need to top FSc but to ensure that they keep a balance between the energy spent on FSc and the energy spent doing the MCAT.
You are right!
 
Messages
639
Reaction score
842
Points
103

There is an even higher number of people who despite having high FSc marks didn't score high on the MCAT and didn't get into public medical colleges and are going to be next years repeaters. Sanity suggests only the cognitive behaviour of accepted applicants should be studied to get accepted in the future, no one studies the behaviour of rejects since that is obvious. The conclusion is and stands: Out of all those representatively sampled applicants accepted into public medical colleges the 'overall' correlation from regression analysis (numbers are not opinions they are facts) those who had a higher FSc score tended to score less well on the MCAT for the MAJORITY.' So if someone wants to be that 'minority' who score higher on the FSC 950-1050 and also score above the regression line on the MCAT the regular behaviour needs to be altered, balancing the schedule is an art and is not found in general and that is the important to be consistent in the MCAT score and FSc score. The standard neurotic behaviour is therefore that those who already score high on the FSc in the majority come out with a relatively lower MCAT score (whatever was happening in their brain is up to them).
 
Messages
1,532
Reaction score
478
Points
93
You are not getting my point. I agree that only the relative data should be studied. But if the data you left out could directly effect your conclusion, then not including it makes the whole statistical analysis a futile excercise.
 
Top