This document provides a statistical inference tutorial with assessed and additional questions. It begins by instructing students to submit their solutions as a PDF by a certain deadline. The assessed questions include: (1) finding an unbiased estimator for a probability density function, (2) determining whether statements about estimators are true or false, (3) identifying which statement about two estimators is false, and (4) calculating a confidence interval given information about a statistic's distribution. The additional questions involve additional confidence interval and estimator problems.
This document provides a statistical inference tutorial with assessed and additional questions. It begins by instructing students to submit their solutions as a PDF by a certain deadline. The assessed questions include: (1) finding an unbiased estimator for a probability density function, (2) determining whether statements about estimators are true or false, (3) identifying which statement about two estimators is false, and (4) calculating a confidence interval given information about a statistic's distribution. The additional questions involve additional confidence interval and estimator problems.
This document provides a statistical inference tutorial with assessed and additional questions. It begins by instructing students to submit their solutions as a PDF by a certain deadline. The assessed questions include: (1) finding an unbiased estimator for a probability density function, (2) determining whether statements about estimators are true or false, (3) identifying which statement about two estimators is false, and (4) calculating a confidence interval given information about a statistic's distribution. The additional questions involve additional confidence interval and estimator problems.
This document provides a statistical inference tutorial with assessed and additional questions. It begins by instructing students to submit their solutions as a PDF by a certain deadline. The assessed questions include: (1) finding an unbiased estimator for a probability density function, (2) determining whether statements about estimators are true or false, (3) identifying which statement about two estimators is false, and (4) calculating a confidence interval given information about a statistic's distribution. The additional questions involve additional confidence interval and estimator problems.
Please submit solutions to this tutorial as a single(!) PDF file. If you have any problems with getting your work into this format please get in touch. Please upload your solutions to the assessed questions (and any additional questions) to MMS under Coursework by 2pm Monday 8th February. Please clearly write your name, tutorial group number and tutor’s name on your solutions. Please submit the questions in the order given in the the tutorial sheet. The number of marks available for each question is given in [ ] brackets.
Assessed Questions
1. Consider independent random variables X1 , . . . , Xn each with probability density
function, 3x2 f (x) = 3 for 0 ≤ x ≤ θ, θ and zero otherwise. Find c such that cX̄ is an unbiased estimator of θ. [3] 2. Let X1 , . . . , Xn be independent and identically distributed random variables each with mean µ and variance σ 2 . Determine whether each of the following statements is true or false. (a) n1 ni=1 (Xi − X̄)2 is a biased estimator of σ 2 . P [1] 1 P n−1 2 (b) n−1 i=1 Xi is an unbiased estimator of σ . [1] (c) n12 ni=1 Xi is not a consistent estimator of µ. P [1] 3. Let X = {X1 , . . . , Xn }, with n > 2, be independent and identically distributed random variables each with P mean µ and variance σ 2 . Consider the following two estimators for µ, f (X) = n ni=1 Xi and g(X) = X1 +X 1 2 n . Which of the following statements is false? [1] (a) The variance of f (X) is greater than the variance of g(X). (b) f (X) is a consistent estimator of µ. (c) g(X) is not a consistent estimator of µ. (d) g(X) is an unbiased estimator of µ. 4. Consider random variables X = {X1 , . . . , Xn } and an estimator T (X) that will be used to find a confidence interval for a parameter φ ∈ R. Considering the continuous distribution of T (X), a statistician calculates that
T (X) − φ T (X) − φ P > 10 = 0.05 and P > −3 = 0.95. 2 2 Data are subsequently collected and an observed value of t(x) = 6 obtained. Given the observed data, calculate a 90% confidence interval for φ. [3] 1 Additional Questions
1. Consider an estimator T (X), X = {X1 , . . . , Xn }, for a parameter µ. From consid-
ering the continuous distribution of T (X), it is given that,
Calculate a 95% confidence interval for µ, given that you observe the estimate t(x) = 2.
2. An estimator T is used to calculate a confidence interval for a parameter θ ∈
R. Considering the continuous distribution of T , the statistician calculates that P(θ − T < 15) = 0.05 and P(θ − T ≤ 20) = 0.95. Which of the following is true, given that the statistician observes t = 1?
(a) P(θ − 1 ≥ 20) = 0.05.
(b) P(θ − 1 ≤ 15) = 0.05. (c) A 95% confidence interval is (16, 21). (d) A 90% confidence interval is (16, 21).
3. Consider a 95% confidence interval (3, 6) for a parameter θ. Which of the following interpretations is always true?
(a) If we were to repeat our experiment 100 times, we would expect the interval (3,6) to contain the true value of θ 95 times. (b) The true value of θ lies within (3, 6) with probability 95%. (c) If we were to repeat our experiment 100 times, we would expect the derived interval (not necessarily (3, 6)) to contain the true value of θ 95 times. (d) We are 95% certain that the true value of θ lies within (3, 6).
4. Let X1 , . . . , Xn be independent and identically distributed N (µ, σ 2 ) random vari-
ables, where µ is unknown and σ 2 is known. n 1X (a) State the distribution of X̄ = Xi . n i=1 (b) Hence, or otherwise, show that a 100(1−α)% confidence interval for µ is given by, z1−(α/2) σ X̄ ± √ , n α where z1−(α/2) satisfies P(Z ≤ z1−(α/2) ) = 1 − for Z ∼ N (0, 1). 2 (c) Suppose that we observe the following data values,
9.2 10.3 11.1 8.5 10.5 10.4.
Given these observed data, calculate a 99% confidence interval for µ, assuming that σ 2 = 4. 2 5. A teacher believes that the marks in a class test are normally distributed with mean µ = 80 (out of 100). Ten pupils take the test and achieve the following marks;
68, 82, 24, 71, 85, 87, 53, 91, 81 and 86.
In the following questions assume that the standard deviation is known to be
σ = 11.
(a) Find a 95% confidence interval for the average mark in the class test. Interpret your interval. (b) Comment on the validity of any assumptions that have been made.
6. Let X1 , . . . , Xn denote independent and identically distributed random variables.
Let T = f (X1 , . . . , Xn ) be a consistent estimator of some parameter θ. Which of the following statements is always true?
(a) T is an unbiased estimator of θ.
(b) T is a biased estimator of θ. (c) E(T ) → θ, as n → ∞. (d) E(T ) → θ and Var(T ) → 0 as n → 0.
7. Let X1 , . . . , Xn and Y1 , . . . , Y2n denote independent random variables each with
mean µ and variance σ 2 . n 2n ! 1 1 X X (a) Show that W1 = (X̄+Ȳ ) and W2 = Xi + Yj are both unbiased 2 3n i=1 j=1 estimators of µ. (b) Find expressions for the variance of W1 and W2 . (c) State, and justify, which of the estimators you would prefer as an estimator of µ.
