Competition - post a photo you've taken of nature >>

OCR S2 (not MEI) 21 Jun 2013

bbchen

Hope Mr M's upload his answer soon.......

Scroll to see replies

Reply 1

TaiGunner

How long do they normally take :/

How did u find it ?

Reply 2

bbchen

My answer:

1. i) 89,90,91,91,92
ii) 91 repeat therefore biased + improvement

2. 0.240

3. i)E(y)=170
S.D= 89.4
ii)a)b)c) expain

4. Reject H0

5. i) sketch
ii) a/4
iii) T bigger variance?

6. E(X)=36.68
Var(X)=56.25
Do not reject H0

7.N(42 27.3)
Reject H0

8.i) 0.693
ii) a)b) increase

9.i) expain
ii) explain
iii) B0=18 0.0053
iv) 8.5

Reply 3

bbchen

Original post by TaiGunner

How long do they normally take :/

How did u find it ?

I think alright

Reply 4

Aamath

Anyone else find it hard?

Reply 5

TaiGunner

I got T2 error as fall for both
Then Bo=13
And lambda between 8.5 and 8.6 :/

Reply 6

Infinity1

Original post by TaiGunner

I got T2 error as fall for both
Then Bo=13
And lambda between 8.5 and 8.6 :/

Agreed

Reply 7

Infinity1

Question 1

(i) 89,90,91,91,92
(ii) Method is biased since 91 occurs twice, and is therefore more likely than the other 3 numbers
Improvement: Multiply by 100 instead of 834 and select accordingly. If number is outside range or already selected, select a new number and repeat.

Question 2

0.2399

Question 3

(i) Mean = 170, Standard Deviation = 89.4
(ii)(a) When do you use CLT - To approximate the distribution of the mean of Y as a normal distribution
(b) Why do you need it - The distribution of Y is unknown
(c) Why can you use it - because n is large

Question 4

0.9554 - Do not reject ...

Question 5

(i) Graph
(ii) a/4
(iii) Variance of T will be greater because T is a uniform distribution... (unsure)

Question 6

0.0524 - Hence do not reject..

Question 7

0.9481 - Hence reject ..

Question 8

(i) 0.6925
(ii) Decreases
(iii) Decreases - P(Type I) increases, so P (Type II) decreases

Question 9

(i) Breakdowns must be independent. (why is this necessary) - ....
(ii) Unlikely to be the case as if one car drives into another, the two breakdowns are clearly not indepdent of one another
(iii) Bo = 13
P(...) = 0.0739
(iv) Root is between 8.5 and 8.6 (sign change)

Please add to answers or change if incorrect

(edited 10 years ago)

Reply 8

hi1345

I got E(y) = 166.2 or sonething for q3:/ wasn't sure how to do it and used no cc :/ also for 5b got like a/4 + 3/4 or something of that nature. Also for q7 did people use the 1/2n cc? So 1/240? This exam was pretty hard overall :/

Reply 9

bbchen

Original post by TaiGunner

I got T2 error as fall for both
Then Bo=13
And lambda between 8.5 and 8.6 :/

i knew is fall, but no idea why i wort increase at the exam~~~ :confused:

Reply 10

MDG1

I used 1/2n so got 0.937 so rejected H0
Could someone also explain 8i) the Type II Error. Thanks
Also why was Bo 13, I put 12

(edited 10 years ago)

Reply 11

Infinity1

Is there another thread for this exam or why are there so few comments haha

Maybe if the title were June, instead of Jun, it'd be easier to find..

Reply 12

JAS274

Original post by Infinity1

I agree with most of these... I can't remember the probabilities for the hypotheses though cos I used the phi instead of converting it. And I got a different answer for Q2 but I had no idea what CC to use!

Reply 13

TaiGunner

Was there a 1/2n correction :/
I remember doing a 0.5 corection but nothing about 1/2n

Reply 14

crazymanny00

i thought the poisson statement improvement was that the breakdowns had to occur at a constant AVERAGE rate, not just constant, probably wrong though. however it did say improve his statement, not add to it

Reply 15

TaiGunner

Original post by crazymanny00

Yeah I did average too

Where's Mr m?

How many marks do you think I'll lose for missing a 1/2n correction ? Was there even a 1/2n corection ?
And for the improvement on he first q did anyone else say ignore repeated numbers :/

(edited 10 years ago)

Reply 18

_JC95

Original post by TaiGunner

How many marks do you think I'll lose for missing a 1/2n correction ? Was there even a 1/2n corection ?
And for the improvement on he first q did anyone else say ignore repeated numbers :/