[Ree69]

Couldn't find a thread for this - anyone else sitting this? The 'practice' papers in the textbook are really putting me off..

[CJAW]

Yep I'm sitting it. I've only done odd questions from the textbook practice exams, but I've done all of the OCR past papers from 2006 onwards and they've all been ok bar a few.
I'm ok on most topics apart from Chapter 2 - combining different distributions, mainly the bit where it's not clear whether it's a multiple of a random variable or a different observation...
How's it going for you?

[Ree69]

Yeah, I know exactly what you're talking about. So confusing! I think it's usually a different observation unless blatantly otherwise - like when they ask Var(X-4Y) for example.

Pretty well, this is my third time sitting this so hopefully I can ace it this time. I'm still a bit iffy on writing the null and alternate hypothesis (for z and t-tests) in terms of population parameters. I like to write them in words but I think this loses a mark.

[CJAW]

(Original post by Ree69)
Yeah, I know exactly what you're talking about. So confusing! I think it's usually a different observation unless blatantly otherwise - like when they ask Var(X-4Y) for example.

Pretty well, this is my third time sitting this so hopefully I can ace it this time. I'm still a bit iffy on writing the null and alternate hypothesis (for z and t-tests) in terms of population parameters. I like to write them in words but I think this loses a mark.

That's pretty much the conclusion I've come to. Often it's just Var(X-Y) anyway so it doesn't matter..

What's your best score so far? And isn't it usually just μx=μy for Ho and the opposite for H1, depending on whether or not it's a two tail or a one tail test? Except testing for the proportion where it would be px=py.
Just make sure you define what Px and Py are, the markscheme often uses p1=p2 instead.

[Ree69]

(Original post by CJAW)
That's pretty much the conclusion I've come to. Often it's just Var(X-Y) anyway so it doesn't matter..

What's your best score so far? And isn't it usually just μx=μy for Ho and the opposite for H1, depending on whether or not it's a two tail or a one tail test? Except testing for the proportion where it would be px=py.
Just make sure you define what Px and Py are, the markscheme often uses p1=p2 instead.

83. Need 88+ for the A*.

Yeah, but sometimes they'd ask you to define the notation in terms of its parameters. And sometimes (a lot of the time, actually) it's unnecessary (as in, there's no mention of it in the mark scheme).

I think I just need to get in to the habit of defining everything I write for the null and alternate hypothesis.

[NightBear]

Anyone done the June 2010 paper?

[CJAW]

(Original post by Ree69)
83. Need 88+ for the A*.

Yeah, but sometimes they'd ask you to define the notation in terms of its parameters. And sometimes (a lot of the time, actually) it's unnecessary (as in, there's no mention of it in the mark scheme).

I think I just need to get in to the habit of defining everything I write for the null and alternate hypothesis.

Ah not far off then!
I'm still a bit unhappy with some of the 'assumptions.'
Like when they ask, what assumptions must you make in using this test or similar. I've found that there seem to be an extra one or two in the mark scheme that I wouldn't of/hadn't thought of, which have no mention in the textbook!

(Original post by NightBear)
Anyone done the June 2010 paper?

Yes I have.

[Ree69]

(Original post by CJAW)
Ah not far off then!
I'm still a bit unhappy with some of the 'assumptions.'
Like when they ask, what assumptions must you make in using this test or similar. I've found that there seem to be an extra one or two in the mark scheme that I wouldn't of/hadn't thought of, which have no mention in the textbook!

Yeah, I know exactly what you're talking about! If I don't get the A*, I'll think back to this very conversation.

[Sgt.Incontro]

Bump to the top. The exam is yomorrow , how is everyone going?

This was posted from The Student Room's Android App on my HTC Sensation Z710e

[Sass93]

hey so has anyone done jan 2011?

If so could someone put working up for q6i. I get exactly the same as the ms up until 1-F(1/y) then i have no idea where the 1- part goes.

Thanks

Here is the link as well http://pdf.ocr.org.uk/download/pp_11...ce_473401.pdf?

[NightBear]

(Original post by CJAW)
Yes I have.

Oh, I had a question but I figured it out.

[NightBear]

(Original post by Sgt.Incontro)
Bump to the top. The exam is yomorrow , how is everyone going?

This was posted from The Student Room's Android App on my HTC Sensation Z710e

Not too bad, except for those sophisticated PDFs/CDFs and the extremely vague linear combination questions

[Ree69]

(Original post by NightBear)
Oh, I had a question but I figured it out.

Hey I just did Jan11 and for 8iic I got [0.0336, 0.3428] but the mark scheme says the answer is [0.0336, 0.3423].

Did you get something similar?

[CJAW]

(Original post by Ree69)
Hey I just did Jan11 and for 8iic I got [0.0334, 0.3428] but the mark scheme says the answer is [0.0336, 0.4323].

Did you get something similar?

Wasn't directed to me I know but no, I got their answer.

My working: [ (x̄ 1 - x̄ 2) +/-tsp *Root((1/22) + (1/20))]
t=2.704, s^2p=0.0342
Did you forget to square root the second s^2p? - that's the mistake I often make.
I wouldn't worry about .the 0.0334 (rounding error?), they accept that in the markscheme anyway.

[NightBear]

(Original post by Ree69)
Hey I just did Jan11 and for 8iic I got [0.0334, 0.3428] but the mark scheme says the answer is [0.0336, 0.4323].

Did you get something similar?

Can't recall what I got, but I remember not getting what the mark scheme had. I'm redoing that Q atm

No, got what the mark scheme has.
Did you use the correct value for t? That's where I went wrong at first, I think, but the answer should be way off in that case
Might be just rounding error?

[Ree69]

(Original post by CJAW)
Wasn't directed to me I know but no, I got their answer.

My working: [ (x̄ 1 - x̄ 2) +/-tsp *Root((1/22) + (1/20))]
t=2.704, s^2p=0.0342
Did you forget to square root the second s^2p? - that's the mistake I often make.
I wouldn't worry about .the 0.0334 (rounding error?), they accept that in the markscheme anyway.

(Original post by NightBear)
Can't recall what I got, but I remember not getting what the mark scheme had. I'm redoing that Q atm

No, got what the mark scheme has.
Did you use the correct value for t? That's where I went wrong at first, I think, but the answer should be way off in that case
Might be just rounding error?

No, I'm about 90% confident () that I've done tihs correctly.

CI =

=

=

= (4dp).

. Someone tell me where I've gone wrong .

[NightBear]

(Original post by Ree69)
x

I left out the upper limit just checked, you're right. I got 0.342775... which should round up to 0.3428 I double checked with their values and got the same answer.

OCR even though that answer will be accepted.

[Ree69]

(Original post by NightBear)
I left out the upper limit just checked, you're right. I got 0.342775... which should round up to 0.3428 I double checked with their values and got the same answer.

OCR even though that answer will be accepted.

Will it? In the mark scheme it says explicity (0.033-0.036, 0.0342-0.0346).

[NightBear]

(Original post by Ree69)
Will it? In the mark scheme it says explicity (0.033-0.036, 0.0342-0.0346).

Yeah, our answer is within 0.0342-0.0346, no?

[Ree69]

(Original post by NightBear)
Yeah, our answer is within 0.0342-0.0346, no?

Oh..ye. That was embarrassing.