{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

STAT1302 (2008May)

# STAT1302 (2008May) - THE UNIVERSITY OF HONG KONG DEPARTMENT...

This preview shows pages 1–5. Sign up to view the full content.

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: THE UNIVERSITY OF HONG KONG DEPARTMENT OF STATISTICS AND ACTUARIAL SCIENCE STAT1302 PROBABILITY AND STATISTICS II May 23, 2008 Time: 6:30 p.m. — 8:30 p.m. Candidates taking ezaminations that permit the use of calculators may use any cal— culator which fulﬁls the following criteria: (a) it should be self-contained, silent, battery-operated and pocket-sized and (b) it should have numeral-display facilities only and should be used only for the purposes of calculation. It is the candidate’s responsibility to ensure that the calculator operates satisfactorily and the candidate must record the name and type of the calculator on the front page of the examination scripts. Lists ofpermitted/prohibited calculators will not be made available to candidates for reference, and the onus will be on the candidate to ensure that the calculator used will not be in violation of the criteria listed above. Answer ALL FIVE questions. Marks are shown in square brackets. 1. Let X1,X2, . . . ,X71 be independent and identically distributed random vari— ables. The cumulative distribution and density functions of X1 are, respec- tively, 6‘1e*(x“9)/9, :c > 6, 0, 5656, 1— e’lz‘ew, CE > 6, O, a: S 6, F(:t[6) ={ and f(a:[6) :{ for an unknown parameter 6 > O. Denote by X (1) and X the sample minimum and sample mean respectively. (a) Show that the statistic T = (Xu) , X) is sufﬁcient for 6. [6 marks] (b) By noting that X Z Xm always, show that the maximum likelihood estimator of 6 is X“). [4 marks] (0) Show that the density function of Xm is n6'1e“”(3‘9)/9. :1: > 6 :1: 6 = ‘ 7 g( l J 0, x39. [4 marks] (d) Show that as an estimator of 6, X“) has a bias equal to 6/n. [4 marks] (e) Using (d) or otherwise, ﬁnd a constant factor c such that cXm is an unbiased estimator of 6. [4 marks] S&AS: STAT1302 Probability and Statistics II 2 2 Let X], . . . ,Xn, Y1, . . , ,Yn be independent random variables such that X1~N(a,02) and K~N(0,a2), 2:1,. ,n, for some unknown a 6 (~00, 00) and a > 0. It is known that the N(a,, b2) distribution has the density function f(z[a,b) = \/217r_b2exp{—-(E?_b2a—)}, :r E (—oo,oo). (a) Show that the loglikelihood function of (a, a) is 1 n S(,LL,0) = —2nlna— ﬁgﬁXl—MQ—tyf}. [4 marks] (b) Find the maXimum likelihood estimator ([i, d) of (an). [4 marks] (C) Show that the Fisher information matrix of (a) a) is 0‘2 0 z ,0 2 n ' . (M ) [ 0 404 ] [6 marks] (d) Quoting the large-sample properties of the maXirnuIn likelihood estimator (i1, (3), find the approximate distributions of [L and (7 separately. [6 marks] (e) Suggest a method to calculate the standard errors of the maximum like— lihood estimates [i and [7 [2 marks] 3 Let X1, . ,Xn be independent and identically distributed random variables. The density function of X1 is O, :5 g (0, 1), for some unknown parameter 6 > 0 (a) Show that the above statistical model for X1, . , ,Xn possesses the mono— tone likelihood ratio property [6 marks] (b) Let 60 be a ﬁxed positive constant Consider the hypotheses HQIQSQO and H116>90. Show that if X1 is observed to be 2:1, for z z 1, i . . ,n, then the uniformly most powerful size a test of HO against H1 18 to reject H0 whenever 1— Fn(:c1:z:2 - - - Inldo) < a, where Fn(-|9) is the cumulative distribution function of I]; Xz [4 marks] S&AS: 100 095 025 020 015 ’ 010 005 0.00 STAT1302 Probability and Statistics II 3 (c) Show that X1 has the cumulative distribution function F1(\$|6’) : 339(1+ 6 — 6:5), for a: E (0,1). [2 marks] ((1) It is known that F1(0.9|6) = (0.9)6(1 + 0.16) is a decreasing function in 6 > O. The ﬁgure below plots it against 9. Plots of 0.99(1+0.1e) Suppose that n = 1 and X1 = 0.9 is observed. Deduce from (b), (c) and the above ﬁgure (i) a 95% lower conﬁdence bound for 6, (ii) a 95% upper conﬁdence bound for 6, (iii) a 95% equal—tailed conﬁdence interval for 6. [8 marks] 4. A FIFA World Cup tournament is an international football competition con— tested by national teams of the world. A total of 18 such tournaments had been held since 1930. The following contingency table shows how the win- ners, the runners—up and the third-places were distributed between countries in Europe and outside Europe. S&AS: STAT1302 Probability and Statistics II 4 Countries \ Title Winner Runner—up Third—place European non—European Total 18 1 8 18 We wish to test the null hypothesis H0 that the three titles (Winner, runner—up and third-place) all have the same probability to go to Europe, against the general hypothesis H1 that the three probabilities have no special relationships. (a) Set up an appropriate statistical model under the general hypothesis H1. [4 marks] (b) Calculate the estimated counts in the contingency table under H0. [5 marks] (c) Based on the estimated counts obtained in (b) conduct a 5% Pearson chi-squared test of HO against H1. [7 marks] f 1 2 3 4 5 6 upperquantile 3.8415 5.9915 7.8147 9.4877 11.0705 12.5916 5 The data below compare the person-hours lost in a month due to accidents at each of eight different industrial plants before and after a safety program was established. For 2' = 1, 2, . . . ,8, deﬁne D22B1_Aza where B1 and A. denote the numbers of person-hours lost? before and after the program was instituted, respectively. It is found that D 2 2le D1/8 = 59.1/8 : 7.3875. Plant Before program After program Difference i B. AZ D2 (D1 — DP 1 51.2 45.8 5.4 3.9502 2 46.5 41.3 5.2 4.7852 3 24.1 15.8 8.3 0.8327 4 10.2 11.1 -0.9 68.6827 5 65.3 58.5 6.8 0.3452 6 92.1 70.3 21.8 207.7202 7 30.3 31.6 —1.3 75.4727 8 49.2 35.4 13.8 41.1202 Total 59.1 402.9088 8&AS: STAT1302 Probability and Statistics II 5 Assume that (D1, . . . ,DS) constitutes a normal random sample with unknown mean u and unknown variance 02. We wish to test the null hypothesis H0: 11 = 0 against the alternative hypothesis H1: u > 0 using a one—sided t test. (a) Calculate an unbiased estimate of 02. [4 marks] (b) Calculate the t test statistic. [6 marks] (c) Determine whether you would reject H0 at the 5% signiﬁcance level. [4 marks] I 5% and 2.5% upper quantzles of tf 6.314 2.920 2.353 2.132 2.015 1.943 1.895 1.860 12.706 4.303 3.182 2.776 2.571 2.447 2.365 2.306 (d) It is suggested that a two—sample t test can be used to test HO: #3 2 MA against H1: 113 > uA, taking (B1, . . . , Ba) and (A1, . . . , A8) as two normal random samples drawn from N(/,LB,0§) and N(MA,01?) respectively, for unknown MB, MA and of. Do you agree with this suggestion? Explain. [6 marks] ...
View Full Document

{[ snackBarMessage ]}