stat271 (1 Viewer)

[Mini]

our current lecturer for stat271 - apart from not nearly having finished the course yet - is giving us no past exams or valuable info on our exam.

wondering if any of u wisened third years or fourth years would be able to tell me what the exam was like: theory based? tests? hard or easy in relation to stat272 exam (which i liked, yep)...

information would be v. v. helpful, thanks in advance

 

[ND]

I second this post. Any insight flyin'?

 

[ND]

Thanks for the info mate.

 

[Mini]

thanks flyin', much obliged

 

[CrashOveride]

flyin i will buy this 272 exam off you after exams?

 

[Grizzly]

If you get the same lecturer, he'll give it to you too (in the lecturers). Won't help you guys that much, as those who did the exam this year will tell.

flyin' i didn't know you were doing a B Sc as well..did you pick it up ?
one will never fail to amaze

 

[ND]

flyin' do you remember whether we had to know all the matrix crap he used to derive some of the tests? I haven't done anything to do with matrices for a year and can barely recall anything.

 

[ND]

I suspect they have been things to test in your exam.

I don't quite understand what you mean. Are you saying that you think it'll be in our exam? If so, what makes you say that?

 

[ND]

Ah ok, i hope that you're right.

 

[bong]

You guys had a look at the 2 way anova tute that was put up?? Is the significance level in Q2 part h correct?

Solution says: "The Bonferroni critical t for a = 0.05, df = 18 , k=2 tests is ~2.45"

So an individual significance of a/4 for each contrast (a/2k)? I thought it would've been a/2 ??

 

[RuledBy_Secrecy]

I think k is defined as the number of contrasts, which is 2 in this case.

On a different issue, I don't totally agree with the degrees of freedom in the anova table in q1. I think it should be 1,3,3,24,31. Do you think there's a mistake?

 

[RIZAL]

Hey, what are you guys focusing on for the exam? I can now do all of the 'tests' but the notes are basically ~50% proofs.... are any of you actually learning these?

and random question, what does lemma stand for in Neaman Pearson lemma (i didn't go to lectures).

 

[bong]

Testing... I should be ok, proofs and detailed math work - not so ok...

Any idea if we'll need to remember that change of variable stuff or order statistics from 272?

 

[RIZAL]

i'm in your situation bong. Testing is fine, rigorous proofs of every test not so OK.

The 'blue' formula sheet doesn't include table 8 (tabulated lower tail values for the Mann - Whitney U statistic).... Did he say in lectures that there would be no Mann Whitney U questions?

I'd say that we could get asked a change of variable question. Seriously, just have a quick flick back to ur 272 notes and memorise the generic change of variable formula. None of the change of variable questions in 271 are anywhere near as bad as they were in 272.

As for the order statistics, he only used them for the MLE of the uniform dist and as part of an assignment question...maybe just memorise the pdf's of the first and last ordered statistic?

I'm liking this subject more now...I'm upgrading my projection to a high pass.

 

[bong]

Didn't say anything about U test... suppose it'll be attached to the exam if needed

He did hint that there was one question that required some heavy math work... want to take a guess at what it might be?

but anyway, I have some room left on my crib sheets to fit the 3 little formlulas on change of variable and order stats

 

[ND]

Do any of you guys know the easiest way of calculating Sxy directly from the data? The formula i'd use is SUM(xi*Yi) - n.xbar.Ybar, is there an easier one?

 

[RIZAL]

Fuck i don't like the sound of that. I fucking hope it's not those gay fucking matrices.

yeah...any proof that involved finding the trace of a matrix I didn't bother with. Leave those proofs to Jacky Poon etc.

ND: not up to linear regression yet...damn.

 

[ND]

flyin' not sure if you're still reading this thread, but if so, do you remember any question from your exam with heavy math work? (might be similar since they're not releasing the past paper)

 

[Mini]

Do any of you guys know the easiest way of calculating Sxy directly from the data? The formula i'd use is SUM(xi*Yi) - n.xbar.Ybar, is there an easier one?

i'd say that's the easiest one, since they'd usually give u sum(xi*yi) and they'd also give you sum(xi) and sum(yi) so easy to calc the bars...

i reckon that one crazy math question will be for the HD aimers - much like that horrendous Poisson process one in stat272... not for me.

i'm just hoping being able to do the tests (and derive likelihoods and such) should do the trick.

 

[toknblackguy]

how are you guys finding the tutes?
is that the best thing we can base the exam on, cause so far i've only been looking at the examples in the notes