The Student Room Group

Scroll to see replies

Reply 460
Original post by Mathsz0r
It really wasn't, if you'd done all the past papers you'd have seen how straight forward it was. There wasn't anything different to previous papers.


I agree , there were barely any 'trick' questions as is sometimes the case with stats. The section b questions on cumulative frequency and tree diagrams were also pretty straightforward and to be expected as they had not been any questions on these topics in the last two exam series.
As far as I'm aware the lowest A boundary for S1 has been 53 and the highest 60 so I'd predict this to be around 58/72. However, the boundaries are not decided on the difficulty on the paper but rather how everyone did so its impossible to say exactly what it might be. For a2 modules MEI makes the cut off at the top 25% of candidates , not sure about as.


Posted from TSR Mobile
Reply 461
Also I'm just wondering how people approached the median and inter quartile range question. The UQ was technically the 18.75th value, but apparently statisticians are divided on whether you should use the 19th value or exactly calculate the 18.75th and 6.25th. I think either should be fine but it would be interesting to see the common method .


Posted from TSR Mobile
Reply 462
Original post by tempz
Also I'm just wondering how people approached the median and inter quartile range question. The UQ was technically the 18.75th value, but apparently statisticians are divided on whether you should use the 19th value or exactly calculate the 18.75th and 6.25th. I think either should be fine but it would be interesting to see the common method .


Posted from TSR Mobile


Since it was from 25 pieces of data, the median was technically the 13th value (25+1 / 2), meaning the lower quartile must have been the 6.5th, and the upper quartile the 19.5th.
Reply 463
Sorry my bad , I don't remember the exact figures , what did you get for the IQR?, I remember calculating an outlier on the lower tail and none on the upper but that's about it


Posted from TSR Mobile
Reply 464
Also out of interest, on the final part of the last question, I got confused and thought they said she only stopped if she got all 3 darts in bull's eye. However, I did calculate and show the probabilities for both (HHH) and (MMMHHH) occuring, which were the two probabilities you had to calculate in the question before adding them together. However, seeing as I did not add these two values together and get the right answer, I'm relying on some method marks? Would I get any seeing as I clearly calculated and showed those two values? This question was out of four, so hopefully I'll recoup maybe 1, or possibly 2 marks this way.
(edited 10 years ago)
Reply 465
Original post by tempz
Sorry my bad , I don't remember the exact figures , what did you get for the IQR?, I remember calculating an outlier on the lower tail and none on the upper but that's about it


Posted from TSR Mobile

I have a feeling it was 0.88? Can't remember exactly, but I think it was that. And yer, I got that result for the outliers too :biggrin:
Reply 466
Original post by tempz
Sorry my bad , I don't remember the exact figures , what did you get for the IQR?, I remember calculating an outlier on the lower tail and none on the upper but that's about it


Posted from TSR Mobile


yeh the first value was the outlier
Reply 467
Original post by jhitman
Also out of interest, on the final part of the last question, I got confused and thought they said she only stopped if she got all 3 darts in bull's eye. However, I did right calculate and show the probabilities for both (HHH) and (MMMHHH) occuring, which were the two probabilities you had to calculate in the question before adding them together. However, seeing as I did not add these two values together and get the right answer, I'm relying on some method marks? Would I get any seeing as I clearly calculated and showed those two values? This question was out of four, so hopefully I'll recoup maybe 1, or possibly 2 marks this way.


If you clearly stated the probabilities of each occurring (HHH) and (MMMHHH) then you should get two method marks , only forfeiting the final method mark (for adding the 2) and the answer mark


Posted from TSR Mobile
Reply 468
Original post by tempz
Also I'm just wondering how people approached the median and inter quartile range question. The UQ was technically the 18.75th value, but apparently statisticians are divided on whether you should use the 19th value or exactly calculate the 18.75th and 6.25th. I think either should be fine but it would be interesting to see the common method .


Posted from TSR Mobile

i know, i think this is where i could have lost marks
Reply 469
Original post by tempz
If you clearly stated the probabilities of each occurring (HHH) and (MMMHHH) then you should get two method marks , only forfeiting the final method mark (for adding the 2) and the answer mark


Posted from TSR Mobile


i might have lost an answer mark by stupidly putting 5.82 instead 5.62 x10^-3. Actually i really can't remember if it was a 6 or 8 now :frown:
Reply 470
is it absolutely neccessary to add 1 to 25 before calculating the quartiles? If you didn't do that, will you lose marks because i checked both ways but got the same iqr?
Reply 471
One more thing to add Is that mei has recently become very strict with accuracy in stats modules , if your final answers aren't specified to 4sf max on long answers (eg. The var(x) calculation in the DRV question) then they now take away the final answer mark


Posted from TSR Mobile
Reply 472
Original post by juniorx
is it absolutely neccessary to add 1 to 25 before calculating the quartiles? If you didn't do that, will you lose marks because i checked both ways but got the same iqr?


No I don't think you will lose a mark , I recall also getting 0.88 for the IQR but can't be certain


Posted from TSR Mobile
Reply 473
Original post by tempz
I agree , there were barely any 'trick' questions as is sometimes the case with stats. The section b questions on cumulative frequency and tree diagrams were also pretty straightforward and to be expected as they had not been any questions on these topics in the last two exam series.
As far as I'm aware the lowest A boundary for S1 has been 53 and the highest 60 so I'd predict this to be around 58/72. However, the boundaries are not decided on the difficulty on the paper but rather how everyone did so its impossible to say exactly what it might be. For a2 modules MEI makes the cut off at the top 25% of candidates , not sure about as.


Posted from TSR Mobile


Although the questions themselves were nothing new, the test paper was at the same difficulty as recent past papers so I doubt there will an increase from 57 marks for an A.

It's obvious why everyone wants low grade boundaries, but realistically, everyone on this forum is such a small sample (and a biased one too) that predicting grade boundaries based on how difficult we found the paper is a bit pointless.

Could you explain what you mean by "For a2 modules MEI makes the cut off at the top 25% of candidates "?
Does it mean that MEI only considers top 25% when deciding grade boundaries?

Also how does MEI decide how many raw marks 100 UMS is? As raw marks seem to have different weighting depending on which range they are. I remember reading it was 2x raw mark difference between A and B + mark required for an A?
Reply 474
Original post by dada55
Although the questions themselves were nothing new, the test paper was at the same difficulty as recent past papers so I doubt there will an increase from 57 marks for an A.

It's obvious why everyone wants low grade boundaries, but realistically, everyone on this forum is such a small sample (and a biased one too) that predicting grade boundaries based on how difficult we found the paper is a bit pointless.

Could you explain what you mean by "For a2 modules MEI makes the cut off at the top 25% of candidates "?
Does it mean that MEI only considers top 25% when deciding grade boundaries?

Also how does MEI decide how many raw marks 100 UMS is? As raw marks seem to have different weighting depending on which range they are. I remember reading it was 2x raw mark difference between A and B + mark required for an A?


I meant they make the cut off for the A grade at the top 25% of candidates. This was stated by the chief examiner in his latest report , but only in regard to a2 units. The 90ums marks are published separately for a2 units by ocr , they otherwise don't have a conversion system like aqa for raw marks to ums but if the boundary for an A is 57 it is usually around 63/64 for 90ums and 69/70 for a hundred. In some units 100 is quite literally full marks (72/72). It all depends on relative performance


Posted from TSR Mobile
Reply 475
If the paper is exceptionally hard/ received quite poorly (not enough time to complete or poor style of questions) then the boundaries can fall quite significantly. For example with the FP1 exam in jan 2013 the A boundary was 45/72 one of the lowest mei has ever had, but some still found the paper easy and got a hundred. The ultimate grade boundaries are not set in regard to who found the paper difficult as this is subjective, rather it is calculated by averaging all the raw results and finding the top x%, in this case 25. That is the point I was illustrating


Posted from TSR Mobile
Reply 476
Original post by Law-Hopeful
For question 2ii I forgot to multiply by 4!... How many marks do you think I will lose (out of 4) for that as the rest of my answer was fine?

i thought it was just 4c3 which is 4?
(edited 10 years ago)
Reply 477
Original post by tempz
I meant they make the cut off for the A grade at the top 25% of candidates. This was stated by the chief examiner in his latest report , but only in regard to a2 units. The 90ums marks are published separately for a2 units by ocr , they otherwise don't have a conversion system like aqa for raw marks to ums but if the boundary for an A is 57 it is usually around 63/64 for 90ums and 69/70 for a hundred. In some units 100 is quite literally full marks (72/72). It all depends on relative performance


Posted from TSR Mobile


i think the c1 grade boundaries were like that
Reply 478
Original post by sarjan100
For the last question did people get 0.0069?

no sorry, it was 5.62x10^-3
4ii) e(x)= 1507.5 var(x) = 445511.25
That's the wrong answer, in mark schemes of past question you round up E(X), in this case .you round E(X) up to 1508 and get Var(X) of 440004.

Latest

Trending

Trending