Please add, discuss, explain and correct in the comments
Things for copy/pasting: Chi2 H0: H1: X2 μ ρ Σ σ ≤ ≥ x̄ ȳ
1) a) r=0.549 b) n=8, 5% CV: 0.6215, reject H1 - no evidence of positive correlation. c) rs=0.714 d) n=8, 5% CV: 0.6429, acceptH1- evidence of positive correlation. e) Positive correlation between ranks but not linear.
2) a) Explain stratified sampling, just in terms of letters/general examples not numbers. b) H0: μp=μs+6, H1: μp>μs+6 5%CV = 1.6449 z=1.1888 < 1.6449, reject H1 + context statement c) No need to assume since CLT guarantees the sample means are approximately normally distributed with a large enough sample size.
3) a) x̄=1.55, s2=0.05667 b) P( | x̄ - μ | < 0.1) ≥ 0.99 z value for 0.5% (either tail) ≥ 2.5758 ((μ+0.1) - μ) / (0.5/sqrt(n)) ≥ 2.5758 0.1 / (0.5/sqrt(n)) ≥ 2.5758 sqrt(n) ≥ 12.879 n ≥ 165.87 Minimum n=166
4) a) k=0.47 b) 10% significance H0: μ=6, H1: μ≠6 μ in confidence interval, not significant, acceptH0, evidence for μ=6 c) σ=3.3741
5) Using a calculator a) P(L - 3C > 0) P=0.42015 b) P(C1+...+C8 + L1+L2+L3 > 20,000) P=0.0033861 c) Assume they are independent (possibly also random and taken from the assumed distribution)
6) a) H0: C~B(4, 0.5) is a good model for these data, H1: C~B(4, 0.5) is a bad model for these data 5 cells - 1 constraint = 4 degrees of freedom CV 1% Chi2 = 13.2767, X2 = 17.52, greater than CV, Reject H0 + context statement b) p=0.51 c) d=56.21 e=10.14 d) H0: The Binomial distribution is a good model for these data, H1: The Binomial distribution is a bad model for these data e) 5 cells - 2 constraints = 3 degrees of freedom CV 1% Chi2 = 11.3449 John should conclude his claim is wrong, as 16.9 was greater than the CV.
Hows everyone feeling, i find s3 pretty easy, just grade boundaries are annoying
I feel the same way, I can reliably get about 70 (usually fall down on definitions, all the actual stats is fine) but that's only just, or not quite, an A* most of the time - it would be great to get 95+UMS 🤞🤞
I feel the same way, I can reliably get about 70 (usually fall down on definitions, all the actual stats is fine) but that's only just, or not quite, an A* most of the time - it would be great to get 95+UMS 🤞🤞
Same tbh i need about 97-100 UMS because my fp2 and m2 is looking shocking right now
Same tbh i need about 97-100 UMS because my fp2 and m2 is looking shocking right now
Me too! I figure if my S3 and S2 resit are ~95 each, I only need low 80s in M2 or FP2 for the A* (270 UMS in 3 A2 modules, I'm pretty sure), and I don't think I'll be getting any higher than 85 in either :/
Beautiful thread. Thank you for setting it all up in a nice manner.
Does anyone want to form a quick revision group right before the exam?
Nobody in the group yet, but please do share with anyone who is sitting the S3 exam. Feel free to send screenshots of any hard questions you come across
https://i.imgur.com/eOGGzWr.jpg quick question do I have to add the sum oi^2/Ei row to gain marks or can I just leave it and just calculate directly. will I still get the marks?
https://i.imgur.com/eOGGzWr.jpg quick question do I have to add the sum oi^2/Ei row to gain marks or can I just leave it and just calculate directly. will I still get the marks?
Show yourself adding it together, it does gain you a couple marks i think.