Deleted member 66701
Deleted member 66701
Last edited by a moderator:
University exams, fairness, and dealing with it.
- Thread starter Illgresi
- Start date
No, you are completely wrong, Biological intelligence is normally distributed
http://www.ncurproceedings.org/ojs/index.php/NCUR2012/article/view/159
just like many of traits and natural phenomena, which is why the Gaussian distribution used so frequently throughout science because it reflects natural underlying mechanics. Intelligence, like height, hand size, gas molecule velocities, flower petal dimensions etc follow an underlying Gaussian distribution. A nice visualization of why Gaussian distributions are so common in nature is the galton board:
yes, I don't disagree that the distribution wont be purely Normal, there is typically skewness to the right. When universities adjust to a distribution they rarely adjust to a Normal Bell curve but a skewed Gaussian to account for that.
I know some of my exams were adjusted, and when I have marked exams for students in Switzerland making adjustments was part of the process before entering them into the database.
And that is what I said.
As I said above, most corrections are done against a skewed Gaussian if that is the approach taken, but more commonly they will look for things like particular extreme questions and remove them form the results. This highlights why professors prefer to set exams that verge on being too hard, because they have more options for up-marking and adjustments for outlier questions than if everyone gets 90%.
That explains it, no offense, i'm sure you rally are bright and top in the class. Different universities have different ranking system, in the US for example loads of student will be getting 90% or more.
I say this because I graduated top in my class and got the medal of outstanding merit for the performance. I achieved average grades in the low 70s, the highest grades were very low 80s for which I received additional prizes and some of my work went on to be published.. The year before no one graduated with a first, which is not uncommon in a smaller graduation class.
90% would be about 4 standard deviations above the mean, which is extremely rare.
Only if the exam is easy, which is the whole point. I get your point, some subject and exam styles can make it easier to achieve a higher grade. The thing is good exams at university will require knowledge, skills and experience above and beyond what is presented in lectures and course material. Presented material just acts as an introduction to the subject and you are expected ro do a lot of additional reading and research beyond the lecture notes- going both broader and deeper.
This becomes especially true in essay based exams. I've done exams that were open book, had all my lecture notes and course text books with me, but getting anything above about 70% required a lot of additional knowledge but also ability in formulating a good response to the question.
Some questions can be very open ended. Typical examples I had were "Design a system to do X under Y constraints". No set answer, you had to pool knowledge form different areas of the course but form outside the course as well, do some hard problem solving and analytics and ultimately derive a novel solution you wont have come across before, and then present this all in a coherent response within a tight time limit.
Deleted member 66701
Deleted member 66701
Yeah, my marks caused a bit of a stir last year, cross marked multiple times, called in to meetings to ask me about the subject matter (basically to find out if I was cheating) - they even handed my assignments over to an external body to have them validated. Just so far outside the norm, the results raised alarm bells at the uni - pretty impressed tbh how far they went to ensure the marks were warranted.
In the end, I only had one mark changed - and that was to the highest mark as the external assessor thought they too harsh when marking my assignments towards the end of the year.
Last edited by a moderator:
I'm absolutely sick of the institutional snobbery on here, too. It's almost as bad as TSR at times.
Well done on your achievement so far. I hope you have a smooth third year.
Out of curiosity, what's the course?
Deleted member 66701
Deleted member 66701
I'm absolutely sick of the institutional snobbery on here, too. It's almost as bad as TSR at times.
Well done on your achievement so far. I hope you have a smooth third year.
Out of curiosity, what's the course?
Yeah, I like the way D.P. has even tried to make Lancaster sound sub standard in his post above. I think people fail to realise that all the Universities cross mark and validate each others courses, so in reality courses are the same difficulty no matter the institution (with very few exceptions). However, what does change is the support the institutions offer to students - something that ex-polys are actually better at that "traditional" Unis. Ex-ploy are also often better at delivery (see NSS results).
Thanks for the best wishes - course is Interactive Media Development - basically Web Design at it's core with elements of HCI, App design/build, 2D Graphics and Video Production. Can't find the prospectus on Lancasters website - this is almost the same but we have more programming modules on our course (PHP and C# mainly):-
https://www.northumbria.ac.uk/study-at-northumbria/courses/interactive-media-design-ft-uufimd1/
Last edited by a moderator:
Congrats! If I remember right you are mature student and that goes a long way in achieving top results. Some of the people I graduate with who got very high marks had a lot of industry experience and were doing a masters simply as a means of career progression, need the certificate to move to C level roles etc. One already had a BSc in Math and had been a computer scientists as Xerox and then intel for 10 years with extension experience in machine learning etc., so of course he got over 90% in the very hard machine learning classes etc.
Lancaster is fine, but it isn't even Russell group.
You are right that teaching quality and resources are often better at ex-polys and lower ranked universities, but are wrong that the exams are all the same - they aren't Universities of the similar perceived standard will exchange papers for verification and sue external examiners form similarly ranked schools.
Studying maths at Ox-bridge is a very different experience to an ex-poly, they are not at all the same level of difficulty.
Deleted member 66701
Deleted member 66701
Being a Russell group Uni isn't an objective measure of quality. It's a marketing gimmick that gullible people have taken hook line an sinker.
Lancaster out performs 21 out of the 24 Russell group universities:-
http://www.thecompleteuniversityguide.co.uk/league-tables/rankings
Last edited by a moderator:
On the converse, studying at a lower ranked university, you have less opportunities and worse facilities than the higher ranked ones. What's more, your peers at higher ranked universities will be more focused and driven. Less so at the lower ranked universities. Remember too, you're primarily dealing with people at a formative and impressionable early part of life. Your peer group's behaviour can rub off on you -- if your peers are all studying hard, you will probably too. If no-one is taking exams seriously, you may be inclined to not panic as much.
Bath isn't a Russell university but it's still a very good university and at present has the number one student satisfaction in the country. Your assumption that any university that isn't a Russell group university must be worthless is simply unfounded and based on elitism.
Finally, in case you didn't know, Lancaster has soared up the tables in recent years. It's currently ranked ninth. Southampton and Bristol are both ranked lower.
Time to change the record.
Last edited:
I don't think that's correct at all. Having entire questions relating to things that haven't been taught is a terrible idea and I can't believe anyone really sets exams like that. Rather, all questions should be based on taught content but with the highest grades (~75% plus) reserved for students who can demonstrate through their answers that they have studied beyond the core taught content. Obviously this varies a bit according to the subject matter and exam style.
Lancaster doesn't make it in the tip 100
http://www.topuniversities.com/university-rankings/world-university-rankings/2015#
I don't think you understand what the purpose of the Russell group os
nope and you're basically talking about IQ again...
well that was the point - it isn't, if you're going to say that a skewed distribution is used as a check then that is a different argument
except it isn't *that* rare... and it certainly isn't 4 standard deviations which is meaningless as I think we've accepted that a skewed distribution would have to be used for this anyway
that isn't really true, they have to meet certain standards but the idea that courses will be of the same level of difficulty is ridiculous - some courses at more competitive universities will be tougher than those at lower tier ones
it would be naive to assume someone studying mathematics at say London Met is takling courses at the same level of difficulty as say someone studying at Cambridge would be
Deleted member 66701
Deleted member 66701
I did say with certain exceptions, but even then, there isn't that big a gap between say a first at Cambridge and a first at London Met that many (elitist snobs) would assume there to be.
If you exclude the top 5 ranked unis and the bottom 5, I think you'll find all courses are broadly similar in difficulty and graduate output quality.
Deleted member 66701
Deleted member 66701
If you're using a global measure, then neither do 21 out of 24 Russell group Uni's......................
I do - it's a marketing gimmick. Did you read the article I linked?
So what did this mean then?