STA261H1S Term Test 2 (Wed Evening)
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
STA261H1S Term Test 2 (Wed Evening)
2022
Question 1
(a) [5 Marks] A two sided 95% confidence interval for µ in the location Normal model N(µ, σ0(2)), calculated from a an observed sample is (54, 66). ( Note: sample size is not given to you)
Calculate the lower limit of the 90% confidence interval for µ, using the data from the same sample.
Solution
= = = 60
z(1+0.90)/2 = 1.64
ME = U L = 66 54 = 6
σ0 ME = 3.0612245
Lower limit of the 90% confidence interval
N(=)ot(x¯)do(.90)t(2)ed(=) n(6)0 fo(4)r(×)th(3.)ti(5)
(b) [8 Marks] Using R, I generated a sample of 49 observations from a Poisson(λ) distribution and calculated an approximate two sided 95% confidence interval for λ using the approximation based on the central limit theorem. The lower limit of this confidence interval is L = 1.64. Calculate the upper limit of this confidence interval.
Solution
L = − z √ =⇒ − √ − L = 0
Question 2
Suppose a sample of n = 10 employees were collected from a company and their income (in $1000) was reported. The incomes were s = 4.00, 5.26, 4.84, 6.77, 5.23, 5.64, 3.84, 6.43, 3.35, 4.28 ∼ N(µ, σ2 ), with both µ and σ unknown.
(a) [5 Marks] Estimate the population average income µ and the 80% confidence interval.
Solution:
> gam = 0.8
> barx <- mean(x); sx <- sd(x)
> barx; sx
[1] 4.964086
[1] 1.122237
> SE <- sx/sqrt(length(x))
> barx - qt((1+gam)/2, df = length(x) - 1) * SE
[1] 4.473273
>
> barx + qt((1+gam)/2, df = length(x) - 1) * SE
[1] 5.454898
The mean, = 4.96; the SD, s = 1.12, Lower limit = 4.47 and upper limit = 5.45 (Please check if this solution is right before marking)
(b) [4 Marks] Perform a two sided test to test H0 : µ = 6 at α = 0.1.
Solution:
> alph <- 0.1
> t.cal <- sqrt(n) * (barx - mu0)/ sx
> p.val <- 2 * (1 - pt(abs(t.cal), df = length(x) - 1))
> t.cal; p.val
[1] -2.919034
[1] 0.01705861
> ## Critical Value ##
> qt(alph, df = length(x) - 1)
[1] -1.383029
> t.test(x, alternative = "two.sided", mu = mu0)
One Sample t-test
data: x
t = -2.919, df = 9, p-value = 0.01706
alternative hypothesis: true mean is not equal to 6
95 percent confidence interval:
4.161286 5.766886
sample estimates:
mean of x
4.964086
We reject the H0 at α = 0.1. Students can make the decision based on p-value or the critical t. Note: To receive full marks you need to show full calculation for both (a) and (b)
Question 3
Suppose that x = (1, 0, 1, 0, 1) is an observed sample from Bernoulli(θ) distribution and the prior
distribution of θ has p.d.f.:
π(θ) = {0(4)θ3 ,
(a) [5 Marks] Calculate the posterior expectation of θ .
Solution: Note that the prior distribution of θ is beta(4,1) and thus the posterior distribution is beta(4+3 = 7, 5 - 3 + 1 = 3)
π(θ | s) = θ6 (1 − θ)2
Thus, the posterior expectation of θ 2 ,
E(θ | s) = θ 7 (1 − θ)2 dθ = = 0.7
Note for TA: Please check for any potential error in the solution.
(b) [5 Marks] Calculate the posterior variance of .
Solution: Thus, the posterior variance of 1/θ ,
E(θ−1 | s) = θ 5 (1 − θ)2 dθ = = 1.5
E(θ−2 | s) = θ4 (1 − θ)2 dθ = = 2.4 Thus, the variance, V ( ) = E(θ−2 | s) − E(θ−1 | s)2 = 0.15
Question 4
(a) [5 Marks] I generated a sample from a location Normal model with σ0(2) = 400. The two sided
95% confidence interval for the mean (µ) calculated using this sample I generated is (60, 70). This interval has all the information necessary to test the null hypothesis H0 : µ = 62 against the alternative H1 : µ > 62. Calculate the value of the observed z-statistic and the p-value for the z-test of the null hypothesis H0 : µ = 62 against the alternative H1 : µ > 62. Assume α = 0.05
Solution: The 95% confidence interval for the mean (µ) calculated using this sample I gener-
ated is (60, 70) =⇒ SE = = 2 = 2.55102
= = 65
z = 0 = 265(5)102(62) = 1 . 176 (4 points)
p-value = P (Z > 1.59) = 0 . 1197974 (2 point) ■
(b) [3 Marks] Calculate the number of observations in this study.
Solution: Let the number of observations = n. We know that
SE = σ0 / √n = 2.55102 ⇒ n = (σ0 /2.55102)2 = 61 ≈ 62
Question 5
A Biologists is interested in the proportion, θ of badgers in a particular area which carry the infection responsible for bovine tuberculosis. The prior distribution of θ is Beta(1, 19). The biologists captures a random sample of 20 badgers and tests them for infections. The observed number of
infections were 2. Assume that the
(a) [4 Marks] Find the prior mean and prior mode of θ .
Solution: This is a Bernoulli-Beta combination. The prior distribution of θ is a Beta distri-
bution with α = 1 and β = 19. Thus the mean 1/20 and mode 12 = 0. (b) [6 Marks] Find the posterior mean and posterior mode for the distribution of θ .
Solution: The posterior distribution is also a Beta distribution with parameters α = 1 +
n, β = 19 + n(1 − ) ⇒ α = 1 + 2, β = 19 + (20 − 2) ⇒ α = 3, β = 37. Thus, the posterior
2022-04-12