HASTS 111 STATISTICAL INFERENCE
Assignment 1
2𝑦
0<𝑦<𝛽 𝛽>0
A1. Suppose 𝑦 has a distribution given by 𝑓(𝑦) = {𝛽2
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
a) Find the coefficient of kurtosis [5]
b) Calculate the coefficient of skewness [5]
c) Comment on the results for a and b above. [2]
𝑥
−1
A2. Given 𝑓(𝑥) = { 𝜆 𝑒 𝑥>0
𝜆
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
a) Find the maximum likelihood estimator for 𝜆. [4]
b) Suppose (2,3,5,,6,7,3,5) is a realization from the distribution, find the estimate for 𝜆. [1]
A3. Suppose a distribution for 𝑥 = 0,1,2,3, … … . . 𝑛 is Binomial with parameter 𝜃.
a) Find MLE for 𝜃. [4]
b) Show that the estimator for 𝜃 is unbiased and consistent. [4]
c) Given that 𝑓(𝑥, 𝜃) = 𝜃 2 𝑥𝑒 −𝜃𝑥 , 𝑥 > 0, 𝜃 > 0, find the method of moments estimator of
𝜃 based on a random sample 𝑋1 , 𝑋2 , ⋯ , 𝑋𝑛 . [4]
A4. A street vendor recorded his profit for ten days as follows
Day 1 2 3 4 5 6 7 8 9 10
Profit 3 5 7 10 11 14 14 7 8 19
a) Fit a linear model of the form 𝑦 = 𝛽0 + 𝛽1 𝑥 [4]
b) Use the model to estimate the profit made on the 15th day. [3]
A5. Show that 𝑆 2 is an unbiased and consistent estimator of 𝜎 2 . [5]
A6. a) Let 𝑋1 , 𝑋2 , ⋯ , 𝑋𝑛 be a random sample from the normal distribution given
1 1 𝑥−𝜇 2
by 𝑓(𝑥, 𝜇, 𝜎 2 ) = 𝜎ξ2𝜋 𝑒𝑥𝑝 ൜− 2 ቀ ቁ ൠ , − ∞ < 𝑥 < ∞. Find the MLE for 𝜇 𝑎𝑛𝑑 𝜎 2 [6]
𝜎
b) Give three advantages and three disadvantages of non-parametric tests over parametric
tests. [3]
END OF ASSIGNMENT (50 marks)