ECOM40006/ECOM90013 Econometrics 3 Department of Economics University of Melbourne Assignment 3 Solutions Semester 1, 2021 Let Y1, Y2, . . . , Yn denote a simple random sample from a population with probability density function f(y) = { θyθ−1, 0 < y < 1, θ > 0, 0, otherwise. 1. Show that the sample mean Y is a consistent estimator of θ/(θ + 1). [7 marks] Hint: First derive the mean of the population and then remember that laws of large numbers are your friends. 2. Derive a consistent method of moments estimator, θ˜ say, for θ. [1 mark] 3. Specify the log-likelihood function for this sample. [1 mark] 4. Derive the maximum likelihood estimator, θˆ say, for θ and prove that it is, indeed a maximum likelihood estimator. [3 marks] 5. Derive the Fisher information for the sample. [2 marks] 6. Suppose that someone wishes to test the null hypothesis null hypothesis H0 : θ = 1 against the alternative that H1 : θ 6= 0. State the true population density function and describe in words the implication for the population when this null hypothesis is true. [2 marks] 7. Derive likelihood ratio, Lagrange multiplier and Wald tests for the hypotheses of Question 6. In each case provide the decision rule that you would use in practice to apply the test, including any critical value(s) you may need. [12 marks] 8. Without appeal to the generic properties of maximum likelihood estimators, prove that θˆ is consistent for θ. [6 marks] Your answers to the Assignment should be submitted via the LMS no later than 4:30pm, Monday 10 May. Your mark for this assignment may contribute up to 10% towards your final mark in the subject. No late assignments will be accepted and an incomplete exercise is better than nothing. 1
欢迎咨询51作业君