OCR MEI Decision 1 (23-Jan-12) - Post-Exam discussion

I thought on the whole it was an OK paper. I was glad to see the linear programming question as just an 8 mark question, so no losing tonnes of marks on silly constraint mistakes.

I'm needing 90%+ to achieve a Cambridge offer for Natural Sciences, (preferably closer to 100 UMS) and in a mad rush of checking at the end, I'm quite sure I actually made my answer to the last question wrong. I originally had 70 seconds, and then quickly arranged the final activity into the empty block to make 65 seconds, without remembering that everything was its immediate predecessor.

I was wondering what you got for the last part of the last question, and if you know anything about where the boundary for 100 UMS usually is with the decision papers?

Reply 2

OP

Original post by Gould1610

I thought on the whole it was an OK paper. I was glad to see the linear programming question as just an 8 mark question, so no losing tonnes of marks on silly constraint mistakes.

I'm needing 90%+ to achieve a Cambridge offer for Natural Sciences, (preferably closer to 100 UMS) and in a mad rush of checking at the end, I'm quite sure I actually made my answer to the last question wrong. I originally had 70 seconds, and then quickly arranged the final activity into the empty block to make 65 seconds, without remembering that everything was its immediate predecessor.

I was wondering what you got for the last part of the last question, and if you know anything about where the boundary for 100 UMS usually is with the decision papers?

I was so glad that the Linear programming was just a solving question, rather than a formulating question.

I got 70 seconds for the final part-question. A friend said they got 60, but as the sum of the event lengths was 125, the minimum it could possibly be was 65 seconds. It might be possible to get 65, but I couldn't do it.

I'm not sure about the 100 UMS boundaries, I am afraid, so I would hope you haven't dropped anything more than a mark, in order to be sure.

My Answers (from memory, as complete as I can make it):

Section A:
Graphs:
- 1 cube: (see cube.png attached) - 6 vertices (degree 4), 12 edges
- 3 cubes: 14 vertices, 28 edges

Algorithms:
- Should get result of 1089 from given value and own value (provided you follow the instruction to give no repeated digits)
- Repeated digits: 1089 if only 2 consecutive digits are the same. 2 non-consecutive digits the same or all 3 digits the same returns "0"

Linear Programming:
- All lines should be shaded out ABOVE, all lines have negative gradient.
- Feasible Region has 4 points (plus origin)
- A - intersection of y-axis and (3); B - intersection of (3) and (2); C - Intersection of (2) and (1); D - intersection of (1) and x-axis. (numbering inequalities from top to bottom as (1), (2), (3))
- Solution: x=8, y=3, 2x+2y=25

Section B:
Networks:
- Dijkstra: Total Weight: 9
- Minimum Connector: Total Weight: 8
- Dijkstra finds least-weight connector for paths starting at a specific point, rather than least-weight connector for network

CPA:
- Original: 60 seconds, 3 people. 2 dummy arcs - after collecting equipment and utensils; after chopping celery and tomatoes.
- 1 person: 125 seconds (sum of times)
- 2 people: 70 seconds

Simulation:
- Using 0-1 --> A; 2-3 --> B [...] 8-9 --> E (other distributions are possible)
- P(0) - 1/25; P(1) - 14/25; P(2) - 6/25 ; P(3) - 4/25 !!!!!Not certain on these values - from memory!!!!!

(edited 12 years ago)

Cube.png12.0KB

Reply 3

you only got 8 edges for the three cuboid one? :s-smilie:

i got 28...

Reply 4

OP

Original post by gannalise

you only got 8 edges for the three cuboid one? :s-smilie:

i got 28...

Typo -.- You're right, 28 edges.

Reply 5

Ahh, yes I agree with all of those answers, except for the 3 cubes thing, I got 14 vertices and 28 edges, I'm not sure how it could possibly be 8, the one with one cube was 12 edges, and the one with 2 cubes was 20 edges.
I've been rearranging the activities here for quite a while now, I can't see the answer being below 70 seconds, because of D needing to have everything else finished, to be started at all.
Also, i only investigated one number for the final part of the algorithm question, too eager to get the linear programming question out of the way of course :P
I just investigated 255 (random) and got 1089.

I've just checked previous years' UMS/raw conversions and only Jun 2010 and Jan/Jun 2011 are available, and they have 70 UMS as about 50, 80 UMS as 55/56, so I'd guess 90 UMS to be about 61/62 and 100 UMS to be around 67-69. Then again though, I have to say that those three papers were harder than today's one, I thought, so I guess I'd better just hope they don't move all the UMS scale up too much.

I think may have dropped all 3 marks on that final bit, though possibly may retain one just for having done it in the right format or something, and I also may have dropped 1 for not investigating more of the repeated digit numbers in the algorithm question. Overall though I'd have to say perhaps 65-68 as a raw mark guess for myself?
I went in to it hoping for more, but if by some crazy luck i get 96-100 UMS then I won't consider retaking. Otherwise I will and it'll be a shame because that was an easy paper and a silly mistake!

Reply 6

i got 65 seconds for the histogram question with 2 people..but i potentially did it wrong O.o it would be so useful to have the paper and check :P

Reply 7

OP

Original post by Gould1610

Ahh, yes I agree with all of those answers, except for the 3 cubes thing, I got 14 vertices and 28 edges, I'm not sure how it could possibly be 8, the one with one cube was 12 edges, and the one with 2 cubes was 20 edges.
I've been rearranging the activities here for quite a while now, I can't see the answer being below 70 seconds, because of D needing to have everything else finished, to be started at all.
Also, i only investigated one number for the final part of the algorithm question, too eager to get the linear programming question out of the way of course :P
I just investigated 255 (random) and got 1089.

I've just checked previous years' UMS/raw conversions and only Jun 2010 and Jan/Jun 2011 are available, and they have 70 UMS as about 50, 80 UMS as 55/56, so I'd guess 90 UMS to be about 61/62 and 100 UMS to be around 67-69. Then again though, I have to say that those three papers were harder than today's one, I thought, so I guess I'd better just hope they don't move all the UMS scale up too much.

I think may have dropped all 3 marks on that final bit, though possibly may retain one just for having done it in the right format or something, and I also may have dropped 1 for not investigating more of the repeated digit numbers in the algorithm question. Overall though I'd have to say perhaps 65-68 as a raw mark guess for myself?
I went in to it hoping for more, but if by some crazy luck i get 96-100 UMS then I won't consider retaking. Otherwise I will and it'll be a shame because that was an easy paper and a silly mistake!

Sorry, the 8 edges was a typo, fixed now

I think you may lose something for not fully investigating the repeated numbers question, but get a mark for using the algorithm once with 255.
I'd guess you'd get a mark for attempting to do a histogram... That mark may be for using the correct scale? (1 sq. = 5 seconds, I believe?).

It did seem an easy paper, mostly because of the "perfect" structure to it (almost). Time management seemed too easy for a D1: I left the exam at 75 minutes, which is odd, as I've normally taken exactly 88 minutes in practice...

Reply 8

Original post by ATMarsden

Sorry, the 8 edges was a typo, fixed now

I think you may lose something for not fully investigating the repeated numbers question, but get a mark for using the algorithm once with 255.
I'd guess you'd get a mark for attempting to do a histogram... That mark may be for using the correct scale? (1 sq. = 5 seconds, I believe?).

It did seem an easy paper, mostly because of the "perfect" structure to it (almost). Time management seemed too easy for a D1: I left the exam at 75 minutes, which is odd, as I've normally taken exactly 88 minutes in practice...

Yes, I agree I've probably lost the final algorithm mark, and I did use a scale of one block for 5 seconds and it was undeniably a histogram, so I'd hope I've kept 1 mark there. [With the other two probably being 1) completely correct histogram, and 2) 70 seconds.]

I'll just hope those are the only 3 marks I've lost though, then. 69/72 sounds alright, and seems like it absolutely must be above 90 UMS, hopefully encroaching upon 100 :smile:

I'm quite relieved that I agree with your simulation question as well, I'd been a little too lax with time, and was forced to rush through that bit, and I found the question where it asked you to sum up the probabilities a little bit confusing to get my head around exactly what it wanted. But I took the method you seem to have taken and I got those answers so I'll keep positive about that.

No time to be negative about anything though, is there, with other exams all over the place too? :P

Reply 9

NA95

so what you do reckon is the grade boundary for an A? It was probably the easiest paper for D1 MEI so i'd say 65/72 for an A? any ideas?

Reply 10

Original post by Gould1610

I'm quite relieved that I agree with your simulation question as well

Actually, I got them different ways around, and I did use that rule.
I got 1 as the most likely, and 3 as the least likely of the ones that got any. 4 and 5 were poth 0.00, and I did 1, 4, 6 and 14 all over 25, just not in the order you've written them, but I think that's more likely to be a blip in memory, as opposed to either of us actually having got it wrong :smile:

Reply 11

Original post by NA95

so what you do reckon is the grade boundary for an A? It was probably the easiest paper for D1 MEI so i'd say 65/72 for an A? any ideas?

It will definitely be lower! that's over 90% for an A!!
It's normally around 80% and even though it was an easy paper, they can't make the grade boundaries that high. The percentages are normally lower than the UMS..I'd say around 58 :smile:

Reply 12

Original post by NA95

so what you do reckon is the grade boundary for an A? It was probably the easiest paper for D1 MEI so i'd say 65/72 for an A? any ideas?

I'd possibly not go as high as 65, I agree it was an easy paper, but it seems unusual to move all the grades up by 10 raw marks, that seems harsh.

I would find it more likely to be in the region of 60, maybe 61, but I couldn't say for certain

Reply 13

OP

Original post by Gould1610

Yes, I agree I've probably lost the final algorithm mark, and I did use a scale of one block for 5 seconds and it was undeniably a histogram, so I'd hope I've kept 1 mark there. [With the other two probably being 1) completely correct histogram, and 2) 70 seconds.]

I'm quite relieved that I agree with your simulation question as well, I'd been a little too lax with time, and was forced to rush through that bit, and I found the question where it asked you to sum up the probabilities a little bit confusing to get my head around exactly what it wanted. But I took the method you seem to have taken and I got those answers so I'll keep positive about that.

No time to be negative about anything though, is there, with other exams all over the place too? :P

The Simulation Question is really difficult to report on, purely because it requires having the numbers in front of me, but I believe those numbers to be the ones I got, and if you got similar, then I'll accept them as writ.

Original post by Gould1610

I'll just hope those are the only 3 marks I've lost though, then. 69/72 sounds alright, and seems like it absolutely must be above 90 UMS, hopefully encroaching upon 100 :smile:

Original post by NA95

so what you do reckon is the grade boundary for an A? It was probably the easiest paper for D1 MEI so i'd say 65/72 for an A? any ideas?

I'd say the A-boundary would be quite high... the 90% c.p. will probably be around 66, and the 80% c.p. around 62, I think. But this is by no means guaranteed.

The 100% c.p., unfortunately, will probably be 71 or even the full 72, from what I've seen elsewhere.

(edited 12 years ago)

Reply 14

NA95

haha thanks guys! I'm reassured now as I found the first question the most difficult :P everything else was quite simple! :biggrin:

Reply 15

Original post by ATMarsden

The 100% c.p., unfortunately, will probably be 71 or even the full 72, from what I've seen elsewhere.

Oh no, where have you seen something that suggests that? :frown:

Reply 16

eurof

Simulation 0.56 = 1 0.28 = 0 0.12 = 2 4 and 5 = 0 i probably lost a couple of marks on explain questions though all others were simple ... Activity arc everone get 2 dummy nodes?

Reply 17

OP

Original post by eurof

Simulation 0.56 = 1 0.28 = 0 0.12 = 2 4 and 5 = 0 i probably lost a couple of marks on explain questions though all others were simple ... Activity arc everone get 2 dummy nodes?

2 dummies, yes. One at the beginning: after grabbing the utensils and ingredients. The other after chopping celery and tomato.

Reply 18