Student Fury Over 'impossible' Economics Exam

[StainlessSteelCat]

Yes, it doesn't look that much of a stretch from my GCSE/A level maths days. Once I had reassurance that the symbols didn't have any special mathematical meaning, I'd have probably had a stab at it.

I can imagine a BA Economics/joint honours student struggling with it though, and perhaps they are the complainers? Lastly, I think the students have nothing to fear grades wise. From memory, final year grading generally works on a bell curve rather than absolutes. If everyone has cocked up this question and the absolute results are all down, they'll simply shift the cut off for the getting a 1st, 2:1 etc.

It doesn't always work in your favour though. In my year there was a ton of us within a whisker of getting a 2:1 and some quandary about whether to nudge over the edge or not. Trouble is if they had done about 3/4 of the class would have got a 2:1 - and perhaps considered a career in research (and some were clearly not suited to it). So in the end, and to keep up academic standards, we all got 2:2s. Annoying at the time, but it hasn't substantially affected my career for the worse- most of those I knew who went into research were highly disappointed by the experience.

[Bloo Loo]

But if you read my rather brisk example above, its basically a graph of y = ax^0.5-bx² or otherwise written y = a(sqrt(x)) - bx²

where in general a is large compared to b, although this doesn't impact the general shape, it does of course impact the scale and sense of the curve.

But an equation of that form will start by rising as it is dominated by the ax^0.5 , but eventually the -bx² starts to dominate and so the curve then fall - ultimately there is an optimal number of people per City, which is this high point of the curve where dy/dx = 0 or in the case above d(output)/dN = 0

Eugenisism explained.

[justthisbloke]

To be fair to them, I'd be a bit miffed if a Q came up at the end of a £27k course for which I'd had no preparation.

But not that miffed as a slight bit of applied thinking will give you most of the answer. Hell, I can manage a reasonable stab at all those Qs and I've never studied economics and always struggled at maths (I'm an engineer by trade - maths was a necessary evil).

[Bloo Loo]

To be fair to them, I'd be a bit miffed if a Q came up at the end of a £27k course for which I'd had no preparation.

But not that miffed as a slight bit of applied thinking will give you most of the answer. Hell, I can manage a reasonable stab at all those Qs and I've never studied economics and always struggled at maths (I'm an engineer by trade - maths was a necessary evil).

Id be miffed if I hadnt even touched the subject...I suspect they had but the symbols were different,

[MrPin]

This thread has gone silly, so I need a lady with big hands to massage my buttocks!

[Bloo Loo]

Err this is a final year paper.

The answer needs A level/1st year UG maths.

thats why its bound to be wrong.

[rxe]

I saw that and like several others think I could have a reasonable stab at it. I've got engineering maths and zero economics training, I had to look up coordination cost, but that seems a pretty simple concept, the idea that collecting people into large groups has some sort of overhead.

Exponent is > 1 because traditionally the coordination cost gets higher as the group gets bigger. You need to work harder to get the food and energy in, you need people to do admin and provide services. I would argue that coordination costs are falling: cities are actually very efficient places to live, providing services to a high density population is much easier than a rural one.

Do I get a 2.2?

[SarahBell]

I did O level maths (1959) and could not have even approached this complexity.

However, A level Maths was full of calculus.

Unfortunately for me, that was in 1961 and now, I can't remember a thing!

I did A level maths - including maths mechanics. When I got to uni most of us needed a crash course in stats maths - me cos I'd not done it and everyone else cos they couldn't remember what they'd been taught.

I spose careers advisor should have told me to do maths stats at A level rather than the mechanics. I'd been told I didn't need a O or A level biology but taught myself the O level over the summer holidays and sat in in the November. I only got a B though.

[MrPin]

I saw that and like several others think I could have a reasonable stab at it. I've got engineering maths and zero economics training, I had to look up coordination cost, but that seems a pretty simple concept, the idea that collecting people into large groups has some sort of overhead.

Exponent is > 1 because traditionally the coordination cost gets higher as the group gets bigger. You need to work harder to get the food and energy in, you need people to do admin and provide services. I would argue that coordination costs are falling: cities are actually very efficient places to live, providing services to a high density population is much easier than a rural one.

Do I get a 2.2?

Pinversity says it's !

"in the post"

[onlyme2]

I saw that and like several others think I could have a reasonable stab at it. I've got engineering maths and zero economics training, I had to look up coordination cost, but that seems a pretty simple concept, the idea that collecting people into large groups has some sort of overhead.

Exponent is > 1 because traditionally the coordination cost gets higher as the group gets bigger. You need to work harder to get the food and energy in, you need people to do admin and provide services. I would argue that coordination costs are falling: cities are actually very efficient places to live, providing services to a high density population is much easier than a rural one.

Do I get a 2.2?

Yes, on the basis of experience that is pretty much the case, unless there are other factors overwhelming the model - such as the largesse of politicians only knowing how to splash the cash on vanity projects/cities, or their planning systems preventing more even growth.

[Riedquat]

The bit that's doing my head in, is in part b: "Provide intuition for your answers".

What?

Glad someone's picked up on that part. I'd be stumped there and found that part impossible since I'd be trying to provide logic. Still, remember this is supposed to be an economics exam...

[Byron]

Probably the intuition part is how you eventually explain why it all went wrong?

[Gigantic Purple Slug]

The answers were predictable.

[Marshmellow]

It's basically because since the age of 15 every single exam is just doing past papers, memorising all the buzzwords and answers and then repeating it in the exam

I distinctly remember a physics exam at A-level I sat where I was certain I had answered everything right. Basically I did, but because I hadn't done explanations using the keywords in the mark scheme I didnt get any marks in quite a few questions. I think it's kind of at a stage where a Professor in Physics would probably get a C or something if they sat a random A-level exam one day, although I do think that whoever puts the most effort in do get the best marks, but if you dont know the past papers and the answers/keywords they are looking for you are at a massive disadvantage.

Now when any question that comes up that is against the grain nobody can think. Usually the time is so tight that there isnt time to think, you just have to regurgitate the set pattern, although this did significantly improve at university, where as the lecturer was marking the exam you weren't bound to including key words or set methods in your answers to get the marks. But here the problem is the students are paying for this exam, and are arent going to be happy if they dont get their bit of paper at the end which they spent 30k on - of course they will kick up a fuss.

[SarahBell]

Now when any question that comes up that is against the grain nobody can think. Usually the time is so tight that there isnt time to think, you just have to regurgitate the set pattern,

Isn't that just poor exam technique to blame lack of time.

[Gigantic Purple Slug]

It's basically because since the age of 15 every single exam is just doing past papers, memorising all the buzzwords and answers and then repeating it in the exam

I distinctly remember a physics exam at A-level I sat where I was certain I had answered everything right. Basically I did, but because I hadn't done explanations using the keywords in the mark scheme I didnt get any marks in quite a few questions. I think it's kind of at a stage where a Professor in Physics would probably get a C or something if they sat a random A-level exam one day, although I do think that whoever puts the most effort in do get the best marks, but if you dont know the past papers and the answers/keywords they are looking for you are at a massive disadvantage.

Now when any question that comes up that is against the grain nobody can think. Usually the time is so tight that there isnt time to think, you just have to regurgitate the set pattern, although this did significantly improve at university, where as the lecturer was marking the exam you weren't bound to including key words or set methods in your answers to get the marks. But here the problem is the students are paying for this exam, and are arent going to be happy if they dont get their bit of paper at the end which they spent 30k on - of course they will kick up a fuss.

"I distinctly remember a physics exam at A-level I sat where I was certain I had answered everything right. Basically I did, but because I hadn't done explanations using the keywords in the mark scheme I didnt get any marks in quite a few questions. I think it's kind of at a stage where a Professor in Physics would probably get a C or something if they sat a random A-level exam one day, although I do think that whoever puts the most effort in do get the best marks, but if you dont know the past papers and the answers/keywords they are looking for you are at a massive disadvantage."

This makes sense to me, namely that in marking papers examiners want to see a full understanding of the process that led to the answer. This way they can award partial marks for demonstrating some understanding of the process. But it's completely contrary to :

"It's basically because since the age of 15 every single exam is just doing past papers, memorising all the buzzwords and answers and then repeating it in the exam"

The reason you are constantly being asked to show your working is precisely because the examiner wants to see your thinking and be able to reward that process, not the process of simply remembering a number or a fact.

In most lower level science/maths exams people are taught a process and then in the exam asked to demonstrate knowledge of that process by applying it in a new but similar situation (different values for example). The fact that you can demonstrate that basic knowledge normally means you are ready to learn more, as without being able to master the foundation skills it is really pointless moving forwards.

[Bossybabe]

The failure of some students to answer an exam question explains the tripling of house prices that happened when they were at primary school?

It would take too long to explain that a country that can't recognise and solve a basic maths problem at student level in economics is clearly unfit to run its economy. A fact that is all to evident when you look at the UK's. financial situation.

[Goat]

The bit that's doing my head in, is in part b: "Provide intuition for your answers".

What?

Using mikes simplified example y = ax^0.5-bx² they're just looking for you to suggest values for a and b. Actually the values aren't that important since we're not expressing what currency were using (£ = intergalactic hyper-pounds), it's the relationship between the 2 that is.

Hypothetically a = £100 and b = £1, which implies we start off spending 1% of our time co-ordinating with ourself and 4% when we have to co-ordinate with one other person, which seems to be in the right ballpark in my experience (actually I'd suggest that we change the equation to ....-b(x-1)² so that we don't have a cost of co-ordinating with ourself and then lose 1% for the second person).

Using the equation for values of x, y=

1 = £99

2 = £137

3 = £164

4 = £184

5 = £199

6 = £209

7 = £216

8 = £218.85

9 = £219.00

10 = £216.22.

Actually the biggest problem I have here is the choice of 0.5 and 2 as the powers in the equation, which means that the co-ordination cost catches up with the output much faster than in real life. Even if we assume that x represents a thousand people it doesn't seem very likely that 2x would only produce 37% more than 1x, I'd suggest values much closer to 1 for both powers to give a more real world outcome.

One final factor not included in the model is the benefit of division of labour, which should mean that the growth in output increases faster than the co-ordination cost until x is very large, something like y = ax^1.5 - bx^1.6 might produce a more realistic model.

Edit: actually my suggested model gives an optimum city size of 52,000 quadrillion people so perhaps something ITRO 1.1 and 1.2 instead.

Edit 2: not much better, 41,000 quadrillion, if you try 1.5 and 1.8 then that gives us a population of 2.5m and an output of 67 billion, which seems to be more in line with real world results in the 21st century.

[Goat]

Err this is a final year paper.

The answer needs A level/1st year UG maths.

+1, scary that 3rd years are even being set this question.

Economics at degree level is a very mathematical subject (try proving the rational expectations hypothesis from first principles if you doubt me), if they're being set this let alone struggling with it then the course must be junk.

[stormymonday_2011]

I have just realised this is my old Alma Mater

From the comments attached to the Independent article it looks as though the story is a bit more complicated than the BBC suggested.

http://www.independent.co.uk/student/news/university-of-sheffield-students-demand-investigation-over-highly-unfair-exam-10014553.html

The Sheffield economics degree is apparently divided into Quantative and Qualative modules. Students taking a single honours Economics degree are taught both sets of modules. Dual honours students doing degrees in subjects such as Politics and Economics appear to only attend the Qualative course and take the exam paper on the subjects covered in it. The issue appears to be that mathematical modelling techniques taught on the Quantative parts of the course ended up in the exam paper on the other parts of the course. Moreover, it appears that the issue was not just with Question 3 but also Question 1 which involved an understanding of Calculus which in my day was very much an extended O level or A level Maths topic and normally not taught to standard O level students. I dont know what the entry requirement for an Economics degree are at Sheffield so I dont know what base knowledge the students would have been expected to possess on joining a Dual Degree course. I dont suppose any mathematically minded Economics students doing a dual History and Economics degree would particularly like to turn up for the History paper and be presented with an original extract from some Anglo Saxon Charters in Old English or a piece of Latin medieval book hand as part of a question if they had not been taught palaeography or how to translate them on their course.

On edit - It appears that students are required to have Level B Maths at GCSE to take a BA Dual Degree in Economics and a subject such as Politics. They also have to take, and presumably pass, a Mathematical Methods for Economics 1 course in Year 1 of their degree. Year 3 appears to involve quite a choice of modules for students and I am not sure how they are structured into the exam.

https://www.shef.ac.uk/economics/undergraduate/degrees/ll12

Presumably the Question 3 quoted was from the Economics of Cities module.

https://www.shef.ac.uk/economics/undergraduate/degrees/modules/ecn303

One suspects the students on Dual Honours may have thought they had got the Mathematical element of the course out of the way in Year 1 so were not expecting to get stiffed with it in Year 3

Apparently the basic reading for the Economics of Cities module was

Brueckner, JK (2011) Lectures in Urban Economics, MIT Press

According to the Amazon blurb

Lectures on Urban Economics offers a rigorous but nontechnical treatment of major topics in urban economics. To make the book accessible to a broad range of readers, the analysis is diagrammatic rather than mathematical

[sPinwheel]

The bit that's doing my head in, is in part b: "Provide intuition for your answers".

What?

I assume they provided a bucket for that part.

[200p]

"Provide intuition for your answers".

If all else fails, grab your tinfoil hat, a shot gun, ammo, and food.

[ChumpusRex]

+1, scary that 3rd years are even being set this question.

Economics at degree level is a very mathematical subject (try proving the rational expectations hypothesis from first principles if you doubt me), if they're being set this let alone struggling with it then the course must be junk.

There's been a rash of comments on social media about this paper. Basically, there is one group of students saying that the paper has hysterically easy, and that a near identical question to this one had been walked through in the tutorials. There is a second group basically saying that it was impossible, and that much of the stuff wasn't taught.

That said, the question is trivial. It took me a couple of minutes to work out the terminology, but I reckon I could have got a decent mark for it in 15 minutes or so.

[Tankus]

42

[stormymonday_2011]

There's been a rash of comments on social media about this paper. Basically, there is one group of students saying that the paper has hysterically easy, and that a near identical question to this one had been walked through in the tutorials. There is a second group basically saying that it was impossible, and that much of the stuff wasn't taught.

That said, the question is trivial. It took me a couple of minutes to work out the terminology, but I reckon I could have got a decent mark for it in 15 minutes or so.

I am sure that the answers would be easy for anyone with A level Maths but probably not for those Dual Degree Arts students who had to take the extra tuition on top of their GCSE Maths

It would be interesting to know whether there was any timetabling clash between some of the subjects in the Dual Degree syllabuses.

They are not supposed to happen on well planned courses but I am aware of a fair few cases where students have been required to be at two places at once because of poorly constructed lecture and tutorial schedules. There are a lot of modules listed in Year 3 of that Economics course which to me suggest plenty of opportunity for timetable clashes with other departments. If that has happened leading to students missing vital elements of a course then they would have a very reasonable grounds for complaint. I dont suppose this sort of publicity is really going to sell Economics at Sheffield University as a subject for a Dual Degree students which means those highly paid Economics lecturers at Sheffield may be scratching around for students to fill some of those modules in years to come.