Cheat on your exams

Had a heated discussion with a friend the other day. I went to a school, where ‘exam techniques’ were part of the standard toolkit given to students to get them the best possible grades at GCSE, A-levels, and beyond. She didn’t, and so hadn’t ever heard of a special ‘technique’ for exams until uni. She felt robbed – why should one group of students get an advantage over the others, because their school taught them how to cheat the system?

Well, it’s a fair point; but my reply was that exam techniques really aren’t that complicated. In fact, you can boil most of them down to three simple rules: Answer the QuestionPlan Your Time and Plan Your Answers. Look, I can even explain each one in 100 words or fewer 😉

Answer The Question

Every stack of exam scripts that’s ever been marked from Socrates* onwards contains at least one howling stink-bomb of a perfect answer. The student has a deep and broad understanding of their subject. The answer is comprehensive, incisive, and backed up by copious references. Unfortunately, because they’ve misread the question and gone off on a tangent, you sigh, marvelling at the intelligence that manages to completely grasp a difficult concept like mitochondrial introgression, but utterly failed to comprehend the text of the exam paper. Nil marks. Don’t let it be you: read the question carefully and give the examiner only what they’ve asked for.

*Socrates may not have actually sat a single GCSE, but you get the point.

Plan Your Time

This one is too obvious for words, but you must practice and be self-disciplined in the exam for it to work. Basically, most students subconsciously assume that the relationship between time spent on a question and marks collected is linear, something like this: 

Linear plot. Uses xkcd lib for matplotlib/numpy on python

(Edit: err, ignore the axes’ values… oops!)

Wrong. It’s minimally true that the longer you spend on a question, the more marks you’ll probably* get. But most exams have more than one question, so you must balance trying to get top marks on, say, Question 1, with getting at least some marks on Questions 2, 3 & 4! This is why it’s important to get a grip on the real relationship between effort and marks. It looks a lot more like this:

Sigmoid plot. Uses xkcd lib for matplotlib/numpy on python

Can you see what’s going on here? Most exam questions are deliberately structured so that getting a third-class (‘D’) grade is relatively easy, then marks are awarded more-or-less linearly up to the top of an upper-second (‘B). Finally, first-class (‘A’) grade answers usually require substantially more insight – and deliberately and rightly so, since these grades are supposed to mark out the very best exam scripts, usually around 5-10% of the class at most.

In other words, if you spend half of a 100-minute, four-question exam on Question 1, you might get an A for maybe 25% of the marks… but at the cost of achieving a C, maximum, on the other three questions. You’ll be lucky to average a C. Plan your time.

*’probably’ because after you’ve spent too long on a question there’s a temptation to start chucking in the kitchen sink as well, and after a while you’re at risk of saying something stupid or wrong which might actually cost you marks.

Plan Your Answer

When you turn over the exam paper and pick up your pen, your head is likely buzzing – with caffeine, formulae and references you’ve crammed in at the last minute, and possibly with the dull ache of worry that your Mum will kill you if you don’t pass this exam. You want to calmly and methodically, putting the first two exam techniques to devastating effect and dazzling your examiners. You’ll actually probably grab the first question you like the look of and start writing immediately.

This is almost guaranteed to leave you sweating in a heap of confusion half-way through the exam, when you look up from polishing your first answer, realise time’s running out, and start flailing through the others. Without an answer plan you’ll find it hard to stick to time, and in the case of longer questions you’ll be more likely to stray off-piste as well.

Instead, invest some of the exam time into making a plan – with bullet points – for each answer. I always spend at least 5% of the time for each answer writing a plan. Do this on your answer script too, so that in the nightmare scenario that you run out of time the examiner can at least give you some marks. For instance:

Q2: Explain what is meant by Muller’s Ratchet, in the context of natural selection (30% of marks)

A: Outline answer:

  • Muller’s Ratchet (Muller, 1964) is accumulation of deleterious mutations in asexual populations
  • ‘Deleterious’ means lowering fitness of an organism
  • Asexual populations (or chromosomes) do not exchange genetic material in meiosis by recombination, unlike sexually-reproducing populations/chromosomes.
  • Mutations in genetic material occur over time at random
  • Most changes have little effect, or slightly deleterious
  • Some highly deleterious or advantageous
  • Natural selection filters randomly-occurring mutations. Under ‘neutral’ selection the effect of all mutations is negligible. Under ‘positive selection’ advantageous mutations’ benefit leads to their higher relative occurrence through evolutionary time. Under ‘negative selection’ deleterious mutations penalised highly
  • Muller’s Ratchet therefore implies that asexually-reproducing populations subject to negative selection will be disadvantaged compared to sexually-reproducing populations, as they cannot filter out deleterious mutations through recombination.

It probably took about 5 minutes to think about and write that plan – but it would likely get at least half-marks, on its own…

In many cases you will be able to get an idea of how many questions, and which topics, the exam will contain (using nefarious tricks like simply asking the course convenor). Armed with this, you can plan your overall exam strategy, with times. Something like this:

Plan for a 2 hour exam with 3 questions (choice of 7) starting at 10:30am:

  • 10:30 start exam.
  • 10:30-10:45 read exam and pick questions
  • 10:45 start first question
  • 11:15 start second question
  • 11:45 start second questions
  • 12:15-12:30 final proofreading

You can take your exam plan in with you (at least in your head) so you don’t waste valuable time trying to work out your timings in the real thing. You’ll also feel more confident and in control of your performance.

Bonus technique: Practice

The above three techniques will help you make the most of your hard-won knowledge (you did revise the content too, right?) but – trust me on this one, cos I’m an exam machine – you’ll be utterly unable to put them to effect without exam practice. This is probably the biggest difference between schools which actually spend precious teaching time on exam practice, and those which simply point students to blogs like this one as part of their revision.

Get a group of mates on the same course together, get a stack of past exam scripts, and practice in exam conditions. Compare, and mark each others’ scripts. Then repeat, again and again.

Last of all, don’t overdo the coffee. Good luck!

An appalling, terrible, borderline-criminal proposal from Westminster Council

I’ve just seen the proposal from Westminster Council on a section of the Quietway Grid in their patch from Covent Garden to Waterloo Bridge. A patch, lest we forget that’s used by a lot of bikes:

Cyclists on Waterloo Bridge waiting to cross the Strand (CTC/Roger Geffen)

I’m really pushed for time this week so I won’t write a detailed post on this. In short it is terrible. Virtually nothing proposed at all, and what little there is is for pedestrians, not cyclists! Loads of car parking retained, no modal filters.

Worst of all, on Waterloo Bridge and the Strand junction itself there is nothing, absolutely nothing to protect cyclists. A line of paint on the road is all that’s going to protect you from busses, lorries, coaches and cars. This is 2015, not 1995. The council are mad.

This is the absolute centre of civic life in London and if there isn’t a case for pedestrianising large swathes of it (allowing black cabs to some bits, and perhaps loading/deliveries outside peak hours), then I can’t think where in the UK it would be appropriate to pedestrianise.

As I say, I’m in a mega hurry, so here’s my pasted consultation response:

What on earth is the Council doing? These are appalling proposals. The Council seem to have utterly disregarded the LCDS2 and LCC recommendations for Quietway provision and instead decided to do virtually nothing. Where money *is* spent it is on pavement widening to benefit pedestrians, not cycling. There is no reason whatsoever why these streets, right in the heart of London, should be accessed by private cars. Taxis, yes, deliveries/loading yes (at prescribed times) but all private cars, all the time? Madness.

Unless additional permeability measures benefit bikes, and modal filters discourage cars, they will struggle to create a cycle-friendly network that encourages cycling as an everyday activity.

Furthermore lots of car parking and loading bays *are* retained which (contrary to stated on the plan) create pinch points and encourage car use. Instead the spaces should be inset to the pavement if they are to be retained at all.

Section 1 (Bow St etc) – there is hardly anything here proposed to comment on. No modal filters, nothing. Strongly oppose.

Section 2 (Covent Garden etc) – again nothing for cycling to comment on at all. The gain of raised tables will be more than offset by retained car parking and to include footway resurfacing as part of a ‘cycling scheme’ is laughable.

Section 3 (Strand) – the only element of this proposal that is genuinely welcome is dropping the junction across the Strand to carriageway level. The rest of the section propsal is nonsense (again, why are private cars to be allowed to drive and park outside the Lyceum, exactly?!) and the unprotected central feeder lane on Waterloo Bridge n/bound is utter madness. It belongs in the 1990s and will be incredibly dangerous – encouraging some novices to the centre of the traffic whilst providing zero protection. If a traffic lane can be lost then why not provide a fully segregated facility on this incredibly busy bridge?

Section 4 (Waterloo bridge northern end) – this proposal is so close to criminally liable I’m amazed the Council even let it out into the public domain. Given the documented incredibly high cycle mode share on this bridge, to protect cycle traffic with a **single white line of paint on the road** eg a mandatory on-road cycle lane, when there is a central reservation present, is beyond belief. This won’t do anything, *at* *all* to improve safety.

TO call this a cycling scheme is wholly disingenuous and the Council should consider seriously whether they have opened themselves to a judicial review by doing so.