Mathematical Statistics With Applications, 7th Edition Solution Manual

Preview (16 of 334 Pages)

100%

Purchase to unlock

Loading page image...

1Chapter 1: What is Statistics?1.1a.Population: all tires manufactured by the company for the specific year. Objective: toestimate the proportion of tires with unsafe tread.b.Population: all adult residents of the particular state. Objective: to estimate theproportion who favor a unicameral legislature.c.Population: times until recurrence for all people who have had a particular disease.Objective: to estimate the true average time until recurrence.d.Population: lifetime measurements for all resistors of this type. Objective: to estimatethe true mean lifetime (in hours).e.Population: all generation X age US citizens (specifically, assign a ‘1’ to those whowant to start their own business and a ‘0’ to those who do not, so that the population isthe set of 1’s and 0’s). Objective: to estimate the proportion of generation X age UScitizens who want to start their own business.f.Population: all healthy adults in the US. Objective: to estimate the true mean bodytemperatureg.Population: single family dwelling units in the city. Objective: to estimate the truemean water consumption1.2a.This histogram is above.Histogram of windwindDensity51015202530350.000.050.100.150.200.250.30b.Yes, it is quite windy there.c.11/45, or approx. 24.4%d.it is not especially windy in the overall sample.

Loading page image...

2Chapter 1:What is Statistics?’s Solutions Manual1.3The histogram is above.Histogram of U235U235Density0246810120.000.050.100.150.200.251.4a.The histogram is above.Histogram of stocksstocksDensity246810120.000.050.100.150.200.250.30b.18/40 = 45%c.29/40 = 72.5%1.5a.The categories with the largest grouping of students are 2.45 to 2.65 and 2.65 to 2.85.(both have 7 students).b.7/30c.7/30 + 3/30 + 3/30 + 3/30 = 16/301.6a.The modal category is 2 (quarts of milk). About 36% (9 people) of the 25 are in thiscategory.b..2 + .12 + .04 = .36c.Note that 8% purchased 0 while 4% purchased 5. Thus, 1 – .08 – .04 = .88 purchasedbetween 1 and 4 quarts.

Loading page image...

Chapter 1:What is Statistics?3’s Solutions Manual1.7a.There is a possibility of bimodality in the distribution.b.There is a dip in heights at 68 inches.c.If all of the students are roughly the same age, the bimodality could be a result of themen/women distributions.1.8a.The histogram is above.Histogram of AlOAlODensity1012141618200.000.050.100.150.20b.The data appears to be bimodal. Llanederyn and Caldicot have lower sample valuesthan the other two.1.9a.Note that 9.7 = 12 – 2.3 and 14.3 = 12 + 2.3. So, (9.7, 14.3) should containapproximately 68% of the values.b.Note that 7.4 = 12 – 2(2.3) and 16.6 = 12 + 2(2.3). So, (7.4, 16.6) should containapproximately 95% of the values.c.From parts (a) and (b) above, 95% - 68% = 27% lie in both (14.3. 16.6) and (7.4, 9.7).By symmetry, 13.5% should lie in (14.3, 16.6) so that 68% + 13.5% = 81.5% are in (9.7,16.6)d.Since 5.1 and 18.9 represent three standard deviations away from the mean, theproportion outside of these limits is approximately 0.1.10a.14 – 17 = -3.b.Since 68% lie within one standard deviation of the mean, 32% should lie outside. Bysymmetry, 16% should lie below one standard deviation from the mean.c.If normally distributed, approximately 16% of people would spend less than –3 hourson the internet. Since this doesn’t make sense, the population is not normal.1.11a.∑=nic1=c+c+ … +c=nc.b.iniyc∑=1=c(y1+ … +yn) =∑=niiyc1c.()∑=+niiiyx1=x1+y1+x2+y2+ … +xn+yn= (x1+x2+ … +xn) + (y1+y2+ … +yn)

Loading page image...

4Chapter 1:What is Statistics?’s Solutions ManualUsing the above, the numerator ofs2is∑=−niiyy12)(=∑=+−niiiyyyy122)2(=∑=−niiy12∑=+niiynyy122Since∑==niiyyn1, we have∑=−niiyy12)(=∑=−niiyny122. Let∑==niiyny11to get the result.1.12Using the data,∑=61iiy= 14 and∑=612iiy= 40. So,s2= (40 - 142/6)/5 = 1.47. So,s= 1.21.1.13a.With∑=451iiy= 440.6 and∑=4512iiy= 5067.38, we have thaty= 9.79 ands= 4.14.b.kintervalfrequencyExp. frequency15.65, 13.934430.621.51, 18.074442.753-2.63, 22.2144451.14a.With∑=251iiy= 80.63 and∑=2512iiy= 500.7459, we have thaty= 3.23 ands= 3.17.b.1.15a.With∑=401iiy= 175.48 and∑=4012iiy= 906.4118, we have thaty= 4.39 ands= 1.87.b.1.16a.Without the extreme value,y= 4.19 ands= 1.44.b.These counts compare more favorably:kintervalfrequencyExp. frequency10.063, 6.39721172-3.104, 9.5642323.753-6.271, 12.7312525kintervalfrequencyExp. frequency12.52, 6.263527.220.65, 8.1339383-1.22, 103940kintervalfrequencyExp. frequency12.75, 5.632526.5221.31, 7.073637.053-0.13, 8.513939

Loading page image...

Chapter 1:What is Statistics?5’s Solutions Manual1.17For Ex. 1.2, range/4 = 7.35, whiles= 4.14. For Ex. 1.3, range/4 = 3.04, while = s = 3.17.For Ex. 1.4, range/4 = 2.32, while s = 1.87.1.18The approximation is (800–200)/4 = 150.1.19One standard deviation below the mean is 34 – 53 = –19. The empirical rule suggeststhat 16% of all measurements should lie one standard deviation below the mean. Sincechloroform measurements cannot be negative, this population cannot be normallydistributed.1.20Since approximately 68% will fall between $390 ($420 – $30) to $450 ($420 + $30), theproportion above $450 is approximately 16%.1.21(Similar to exercise 1.20) Having a gain of more than 20 pounds represents allmeasurements greater than one standard deviation below the mean. By the empiricalrule, the proportion above this value is approximately 84%, so the manufacturer isprobably correct.1.22(See exercise 1.11)∑=−niiyy1)(=∑=niiy1–011=−=∑∑==niiniiyyyn.1.23a.(Similar to exercise 1.20) 95 sec = 1 standard deviation above 75 sec, so thispercentage is 16% by the empirical rule.b.(35 sec., 115 sec) represents an interval of 2 standard deviations about the mean, soapproximately 95%c.2 minutes = 120 sec = 2.5 standard deviations above the mean. This is unlikely.1.24a.(112-78)/4 = 8.5b.The histogram is above.Histogram of hrhrFrequency8090100110012345c.With∑=201iiy= 1874.0 and∑=2012iiy= 117,328.0, we have thaty= 93.7 ands= 9.55.

Loading page image...

6Chapter 1:What is Statistics?’s Solutions Manuald.1.25a.(716-8)/4 = 177b.The figure is omitted.c.With∑=881iiy= 18,550 and∑=8812iiy= 6,198,356, we have thaty= 210.8 ands= 162.17.d.1.26For Ex. 1.12, 3/1.21 = 2.48. For Ex. 1.24, 34/9.55 = 3.56. For Ex. 1.25, 708/162.17 =4.37. The ratio increases as the sample size increases.1.27(64, 80) is one standard deviation about the mean, so 68% of 340 or approx. 231 scores.(56, 88) is two standard deviations about the mean, so 95% of 340 or 323 scores.1.28(Similar to 1.23) 13 mg/L is one standard deviation below the mean, so 16%.1.29If the empirical rule is assumed, approximately 95% of all bearing should lie in (2.98,3.02) – this interval represents two standard deviations about the mean. So,approximately 5% will lie outside of this interval.1.30Ifμ= 0 andσ= 1.2, we expect 34% to be between 0 and 0 + 1.2 = 1.2. Also,approximately 95%/2 = 47.5% will lie between 0 and 2.4. So, 47.5% – 34% = 13.5%should lie between 1.2 and 2.4.1.31Assuming normality, approximately 95% will lie between 40 and 80 (the standarddeviation is 10). The percent below 40 is approximately 2.5% which is relativelyunlikely.1.32For a sample of sizen, letn′denote the number of measurements that fall outside theintervaly±ks, so that (n–n′)/nis the fraction that falls inside the interval. To show thisfraction is greater than or equal to 1 – 1/k2, note that(n– 1)s2=∑∈−Aiiyy2)(+∑∈−biiyy2)(, (both sums must be positive)whereA= {i: |yi-y|≥ks} andB= {i: |yi–y| <ks}. We have that∑∈−Aiiyy2)(≥∑∈Aisk22=n′k2s2, since ifiis inA, |yi–y|≥ksand there aren′elements inA. Thus, we have thats2≥k2s2n′/(n-1), or 1≥k2n′/(n–1)≥k2n′/n. Thus, 1/k2≥n′/nor(n–n′)/n≥1 – 1/k2.kintervalfrequencyExp. frequency184.1, 103.21313.6274.6, 112.82019365.0, 122.42020kintervalfrequencyExp. frequency148.6, 3736359.842-113.5, 535.18283.63-275.7, 697.38788

Loading page image...

Chapter 1:What is Statistics?7’s Solutions Manual1.33Withk=2, at least 1 – 1/4 = 75% should lie within 2 standard deviations of the mean.The interval is (0.5, 10.5).1.34The point 13 is 13 – 5.5 = 7.5 units above the mean, or 7.5/2.5 = 3 standard deviationsabove the mean. By Tchebysheff’s theorem, at least 1 – 1/32= 8/9 will lie within 3standard deviations of the mean. Thus, at most 1/9 of the values will exceed 13.1.35a.(172 – 108)/4 =16b.With∑=151iiy= 2041 and∑=1512iiy= 281,807 we have thaty= 136.1 ands= 17.1c.a= 136.1 – 2(17.1) = 101.9,b= 136.1 + 2(17.1) = 170.3.d.There are 14 observations contained in this interval, and 14/15 = 93.3%. 75% is alower bound.1.36a.The histogram is above.010203040506070ex1.3601234568b.With∑=1001iiy= 66 and∑=10012iiy= 234 we have thaty= 0.66 ands= 1.39.c.Within two standard deviations: 95, within three standard deviations: 96. Thecalculations agree with Tchebysheff’s theorem.1.37Since the lead readings must be non negative, 0 (the smallest possible value) is only 0.33standard deviations from the mean. This indicates that the distribution is skewed.1.38By Tchebysheff’s theorem, at least 3/4 = 75% lie between (0, 140), at least 8/9 liebetween (0, 193), and at least 15/16 lie between (0, 246). The lower bounds are alltruncated a 0 since the measurement cannot be negative.

Loading page image...

121Chapter 6: Functions of Random Variables6.1The distribution function ofYis∫−=−=yYyydttyF022)1(2)(, 0≤y≤1.a.2212121211)()(2)()()12()()(1++++−==≤=≤−=≤=uuuYuUFYPuYPuUPuF. Thus,11,)()(2111≤≤−=′=−uuFufuUU.b.22121211212)()(21)()()21()()(2++−−=−==≤=≤−=≤=uuuYuUFYPuYPuUPuF. Thus,11,)()(2122≤≤−=′=+uuFufuUU.c.uuuFuYPuYPuUPuFYU−==≤=≤=≤=2)()()()()(233Thus,10,1)()(133≤≤−=′=uuFufuUU.d..6/1)(,3/1)(,3/1)(321==−=UEUEUEe..6/1)(,3/1)21(,3/1)12(2==−−=−YEYEYE6.2The distribution function ofYis∫−−==yYydttyF132)1)(2/1()2/3()(, –1≤y≤1.a.)118/()3/()3/()3()()(32111−==≤=≤=≤=uuFuYPuYPuUPuFYU. Thus,33,18/)()(211≤≤−=′=uuuFufUU.b.])3(1[)3(1)3()3()()(32122uuFuYPuYPuUPuFYU−−=−−=−≥=≤−=≤=.Thus,42,)3()()(22322≤≤−=′=uuuFufUU.c.2/323)()()()()()(3uuFuFuYuPuYPuUPuFYYU=−−=≤≤−=≤=≤=.Thus,10,)()(2333≤≤=′=uuuFufUU.6.3The distribution function forYis⎪⎩⎪⎨⎧>≤<−≤≤=5.115.112/1102/)(2yyyyyyFY.a.)()()410()()(104104++=≤=≤−=≤=uYuUFYPuYPuUPuF. So,⎪⎩⎪⎨⎧>≤<≤≤−=−+11111664)(101200)4(2uuuuFuuU, and⎪⎩⎪⎨⎧≤<≤≤−=′=+elsewhere011664)()(1011004uuuFufuUU.b.E(U) = 5.583.c.E(10Y– 4) = 10(23/24) – 4 = 5.583.6.4The distribution function ofYis4/1)(yYeyF−−=, 0≤y.a.12/)1(31311)()()13()()(−−−−−==≤=≤+=≤=uuYuUeFYPuYPuUPuF. Thus,1,)()(12/)1(121≥=′=−−ueuFufuUU.b.E(U) = 13.

Loading page image...

122Chapter 6: Functions of Random Variables’s Solutions Manual6.5The distribution function ofYis4/)(yyFY=, 1≤y≤5.234123232)()()32()()(−−−==≤=≤+=≤=uuYuUFYPuYPuUPuF. Differentiating,()535,)()(2/123161≤≤=′=−−uuFufuUU.6.6Refer to Ex. 5.10 ad 5.78. Define)()()()(2121uYYPuYYPuUPuFU+≤=≤−=≤=.a.Foru≤0,0)()()(21=≤−=≤=uYYPuUPuFU.For 0≤u< 1,∫ ∫==≤−=≤=+uuyyUudydyuYYPuUPuF02221212/1)()()(22.For 1≤u≤2,∫ ∫−+−−=−=≤−=≤=uuyUudydyuYYPuUPuF202221212/)2(111)()()(2.Thus,⎪⎩⎪⎨⎧≤≤−<≤=′=elsewhere021210)()(yuuuuFufUU.b.E(U) = 1.6.7LetFZ(z) andfZ(z) denote the standard normal distribution and density functionsrespectively.a.).()()()()()(2uFuFuZuPuZPuUPuFZZU−−=≤≤−=≤=≤=Thedensity function forUis then0),()()()()(12121≥=−+=′=uufufufuFufZuZuZuUU.Evaluating, we find0)(2/2/121≥=−−πueuufuU.b.Uhas a gamma distribution withα= 1/2 andβ= 2 (recall thatΓ(1/2) =π).c.This is the chi–square distribution with one degree of freedom.6.8LetFY(y) andfY(y) denote the beta distribution and density functions respectively.a.).1(1)1()1()()(uFuYPuYPuUPuFYU−−=−≥=≤−=≤=The density functionforUis then10,)1()1()()(11)()()(≤≤−=−=′=−α−ββΓαΓβ+αΓuuuufuFufYUU.b.E(U) = 1 –E(Y) =β+αβ.c.V(U) =V(Y).6.9Note that this is the same density from Ex. 5.12:2),(21=yyf, 0≤y1≤1, 0≤y2≤1,0≤y1+y2≤1.a.20021212122)()()()(udydyYuYPuYYPuUPuFuyuU==−≤=≤+=≤=∫ ∫−. Thus,uuFufUU2)()(=′=, 0≤u≤1.b.E(U) = 2/3.c.(found in an earlier exercise in Chapter 5)E(Y1+Y2) = 2/3.

Loading page image...

Chapter 6: Functions of Random Variables123’s Solutions Manual6.10Refer to Ex. 5.15 and Ex. 5.108.a.uyuyyUedydyeYuYPuYYPuUPuF−∞+−−==+≤=≤−=≤=∫ ∫1)()()()(0212121221, so thatuUUeuFuf−=′=)()(,u≥0, so thatUhas an exponential distribution withβ= 1.b.From part a above,E(U) = 1.6.11It is given thatfi(yi) =iye−,yi≥0 fori= 1, 2. LetU= (Y1+Y2)/2.a.,21)2()()()(2220221212222121uuuyuyyyYYUueedydyeYuYPuPuUPuF−−−−−+−−==−≤=≤=≤=∫∫so thatuUUueuFuf24)()(−=′=,u≥0, a gamma density withα= 2 andβ= 1/2.b.From part (a),E(U) = 1,V(U) = 1/2.6.12LetFY(y) andfY(y) denote the gamma distribution and density functions respectively.a.)/()()()(cuYPucYPuUPuFU≤=≤=≤=. The density function forUis then0,)/()()(/1))((11≥==′=β−−αβαΓαueucufuFufcucYcUU. Note that this is anothergamma distribution.b.The shape parameter is the same (α), but the scale parameter iscβ.6.13Refer to Ex. 5.8;∫ ∫−−−=−≤=≤+=≤=uyuyyUdydyeYuYPuYYPuUPuF00212121221)()()()(=uuuee−−−−1.Thus,uUUueuFuf−=′=)()(,u≥0.6.14SinceY1andY2are independent, so2221121)(18),(yyyyyf−=, for 0≤y1≤1, 0≤y2≤1.LetU=Y1Y2. Then,∫ ∫−−=>=≤=≤=≤=11/21222112121212)(181)/()/()()()(uyuUdydyyyyYuYPYuYPuYYPuUPuF= 9u2– 8u3+ 6u3lnu.)ln1(18)()(uuuuuFufUU+−=′=, 0≤u≤1.6.15LetUhave a uniform distribution on (0, 1). The distribution function forUisuuUPuFU=≤=)()(, 0≤u≤1. For a functionG, we requireG(U) =YwhereYhasdistribution functionFY(y) =21ye−−,y≥0. Note thatFY(y) =P(Y≤y) =)]([)]([))((11yGFyGUPyUGPU−−=≤=≤=u.So it must be true that=−)(1yG21ye−−=uso thatG(u) = [–ln(1–u)]–1/2. Therefore, therandom variableY= [–ln(U– 1)]–1/2has distribution functionFY(y).

Loading page image...

124Chapter 6: Functions of Random Variables’s Solutions Manual6.16Similar to Ex. 6.15. The distribution function forYis.,1)(2bydttbyFybybY≥−==∫−FY(y) =P(Y≤y) =)]([)]([))((11yGFyGUPyUGPU−−=≤=≤=u.So it must be true that=−)(1yGyb−1=uso thatG(u) =ub−1. Therefore, the randomvariableY=b/(1 –U) has distribution functionFY(y).6.17a.Taking the derivative ofF(y),α−αθα=1)(yyf, 0≤y≤θ.b.Following Ex. 6.15 and 6.16, letu=()αθyso thaty=θu1/α. Thus, the random variableY=θU1/αhas distribution functionFY(y).c.From part (b), the transformation isy= 4u. The values are 2.0785, 3.229, 1.5036,1.5610, 2.403.6.18a.Taking the derivative of the distribution function yields1)(−α−ααβ=yyf,y≥β.b.Following Ex. 6.15, let()αβ−=yu1so thatα−β=/1)1(uy. Thus,α−−β=/1)1(UY.c.From part (b),y= 3u−1/. The values are 3.0087, 3.3642, 6.2446, 3.4583, 4.7904.6.19The distribution function forXis:FX(x) =P(X≤x) =P(1/Y≤x) =P(Y≥1/x) = 1 –FY(1/x)= 1 –()[]()ααβ=β−xx1, 0 < x <β–1, which is a power distribution withθ=β–1.6.20a.,)()()()()(2wwFwYPwYPwWPwFYW==≤=≤+≤=0≤w≤1.b.,)()()()()(222wwFwYPwYPwWPwFYW==≤=≤+≤=0≤w≤1.6.21By definition,P(X=i) =P[F(i– 1) <U≤F(i)] =F(i) –F(i– 1), fori= 1, 2, …, since forany 0≤a≤1,P(U≤a) =afor any 0≤a≤1. From Ex. 4.5,P(Y=i) =F(i) –F(i– 1), fori= 1, 2, … . Thus,XandYhave the same distribution.6.22LetUhave a uniform distribution on the interval (0, 1). For a geometric distribution withparameterpand distribution functionF, define the random variableXas:X=kif and only ifF(k– 1) <U≤F(k),k= 1, 2, … .Or sinceF(k) = 1 –qk, we have that:X=kif and only if 1 –qk–1<U≤1 –qk, ORX=kif and only ifqk, < 1–U≤qk–1, ORX=kif and only ifklnq≤ln(1–U)≤(k–1)lnq, ORX=kif and only ifk–1 < [ln(1–U)]/lnq≤k.6.23a.IfU= 2Y– 1, thenY=21+U. Thus,21=dudyand212121)1(2)(uuUuf−+=−=, –1≤u≤1.b.IfU= 1– 2Y, thenY=21U−. Thus,21=dudyand212121)1(2)(uuUuf+−=−=, –1≤u≤1.c.IfU=Y2, thenY=U. Thus,ududy21=anduuuUuuf−=−=121)1(2)(, 0≤u≤1.

Loading page image...

Chapter 6: Functions of Random Variables125’s Solutions Manual6.24IfU= 3Y+ 1, thenY=31−U. Thus,31=dudy. With4/41)(yYeyf−=, we have that[]12/)1(12112/)1(4131)(−−−−==uuUeeuf, 1≤u.6.25Refer to Ex. 6.11. The variable of interest isU=221YY+. FixY2=y2. Then,Y1= 2u–y2and21=dudy. The joint density ofUandY2isg(u,y2) = 2e–2u,u≥0,y2≥0, andy2< 2u.Thus,uuuUuedyeuf2202242)(−−==∫foru≥0.6.26a.Using the transformation approach,Y=U1/mso thatmmmdudyu/)1(1−−=so that the densityfunction forUisα−α=/1)(uUeuf,u≥0. Note that this is the exponential distributionwith meanα.b.∫∞α−α==0/1//)()(dueuUEYEumkmkk=()mkmk/1α+Γ, using the result from Ex. 4.111.6.27a.LetW=Y. The random variableYis exponential soβ−β=/1)(yYeyf. Then,Y=W2andwdwdy2=. Then,β−β=/22)(wYweyf,w≥0, which is Weibull withm= 2.b.It follows from Ex. 6.26 thatE(Yk/2) =()2/21kkβ+Γ6.28IfYis uniform on the interval (0, 1),1)(=ufU. Then,2/UeY−=and2/21ududye−−=.Then,2/212/21||1)(uuYeeyf−−=−=,u≥0 which is exponential with mean 2.6.29a.With22mVW=,mWV2=andmwdwdv21||=. Then,kTwmambwmwmwaWewewf/2/12/22)/2(2/3)(−−==,w≥0.The above expression is in the form of a gamma density, so the constantamust bechosen so that the density integrate to 1, or simply2/3232/3))((12kTmaΓ=.So, the density function forWiskTwkTWewwf/2/1))((12/323)(−Γ=.b.For a gamma random variable,E(W) =kT23.6.30The density function forIis2/1)(=ifI, 9≤i≤11. ForP= 2I2,I=2/Pand2/12/3)2/1(−=pdpdi. Then,pppf241)(=, 162≤p≤242.

Loading page image...

126Chapter 6: Functions of Random Variables’s Solutions Manual6.31Similar to Ex. 6.25. FixY1=y1. Then,U=Y2/y1,Y2=y1Uand1||2ydudy=. The jointdensity ofY1andUis2/)1(218111),(uyeyuyf+−=,y1≥0,u≥0. So, the marginaldensity forUis31)1(2012/)1(2181)(uuyUdyeyuf+∞+−==∫,u≥0.6.32NowfY(y) = 1/4, 1≤y≤5. IfU= 2Y2+ 3, thenY=()2/123−Uand()3241||−=ududy. Thus,)3(281)(−=uUuf, 5≤u≤53.6.33IfU= 5 – (Y/2),Y= 2(5 –U). Thus,2||=dudyand)33180(4)(2uuufU+−=, 4.5≤u≤5.6.34a.IfU=Y2,Y=U. Thus,ududy21||=andθ−θ=/1)(uUeuf,u≥0. This is theexponential density with meanθ.b.From part a,E(Y) =E(U1/2) =2πθ. Also,E(Y2) =E(U) =θ, soV(Y) =]1[4π−θ.6.35By independence,1),(21=yyf, 0≤y1≤0, 0≤y2≤1. LetU=Y1Y2. For a fixed valueofY1aty1, theny2=u/y1. So that121ydudy=. So, the joint density ofY1andUis11/1),(yuyg=, 0≤y1≤0, 0≤u≤y1.Thus,)ln()/1()(111udyyufuU−==∫, 0≤u≤1.6.36By independence,)(4212221221),(yyyyeyyf+−θ=,y1> 0,y2> 0. LetU=2221YY+. For a fixedvalue ofY1aty1, thenU=2221Yy+so we can write212yuy−=. Then,21221yududy−=sothat the joint density ofY1andUis21221121/41),(yuuyuyeuyg−θ−θ−==θ−θ/122uey, for 0 <y1<u.Then,θ−θθ−θ==∫/110/1222)(uuuUuedyeyuf. Thus,Uhas a gamma distribution withα= 2.6.37The mass function for the Bernoulli distribution isyyppyp−−=1)1()(,y= 0, 1.a.txtytYYpepypeeEtm+−===∑=1)()()(1011.b.ntniYtWWpeptmeEtmi]1[)()()(1+−===∏=c.Since the mgf forWis in the form of a binomial mgf withntrials and successprobabilityp, this is the distribution forW.

Loading page image...

Chapter 6: Functions of Random Variables127’s Solutions Manual6.38LetY1andY2have mgfs as given, and letU=a1Y1+a2Y2. The mdf forUis)()()()()()()(21)()()(2122112211tamtameEeEeEeEtmYYYtaYtatYaYaUtU====+.6.39The mgf for the exponential distribution withβ= 1 is1)1()(−−=ttm,t< 1. Thus, withY1andY2each having this distribution andU= (Y1+Y2)/2. Using the result from Ex.6.38, leta1=a2= 1/2 so the mgf forUis.)2/1()2/()2/()(2−−==ttmtmtmUNote thatthis is the mgf for a gamma random variable withα= 2,β= 1/2, so the density functionforUis0,4)(2≥=−uueufuU.6.40It has been shown that the distribution of both21Yand22Yis chi–square withν= 1. Thus,both have mgf2/1)21()(−−=ttm,t< 1/2. WithU=21Y+22Y, use the result from Ex.6.38 witha1=a2= 1 so that.)21()()()(1−−==ttmtmtmUNote that this is the mgf for aexponential random variable withβ= 2, so the density function forUis0,)(2/21≥=−ueufuU(this is also the chi–square distribution withν= 2.)6.41(Special case of Theorem 6.3) The mgf for the normal distribution with parametersμandσis2/22)(ttetmσ+μ=. Since theYi’s are independent, the mgf forUis given by[]∑∑∏∏σ+μ======iiiniinitYaUtUiiiatattameEeEtm22211)2/(exp)()()()(.This is the mgf for a normal variable with mean∑μiiaand variance∑σiia22.6.42The probability of interest isP(Y2>Y1) =P(Y2–Y1> 0). By Theorem 6.3, thedistribution ofY2–Y1is normal withμ= 4000 – 5000 = –1000 andσ2= 4002+ 3002=250,000. Thus,P(Y2–Y1> 0) =P(Z>000,250)1000(0−−) =P(Z> 2) = .0228.6.43a.From Ex. 6.41,Yhas a normal distribution with meanμand varianceσ2/n.b.For the given values,Yhas a normal distribution with varianceσ2/n= 16/25. Thus,the standard deviation is 4/5 so thatP(|Y–μ|≤1) =P(–1≤Y–μ≤1) =P(–1.25≤Z≤1.25) = .7888.c.Similar to the above, the probabilities are .8664, .9544, .9756. So, as the sample sizeincreases, so does the probability thatP(|Y–μ|≤1).6.44The total weight of the watermelons in the packing container is given by∑==niiYU1, soby Theorem 6.3Uhas a normal distribution with mean 15nand variance 4n. We requirethat)()140(05.415140nnZPUP−>=>=. Thus,nn415140−=z.05= 1.645. Solving thisnonlinear expression forn, we see thatn≈8.687. Therefore, the maximum number ofwatermelons that should be put in the container is 8 (note that with this valuen, we haveP(U> 140) = .0002).

Loading page image...

128Chapter 6: Functions of Random Variables’s Solutions Manual6.45By Theorem 6.3 we have thatU= 100 +7Y1+ 3Y2is a normal random variable with meanμ= 100 + 7(10) + 3(4) = 182 and varianceσ2= 49(.5)2+ 9(.2)2= 12.61. We require avaluecsuch thatP(U>c) =P(61.12182−>cZ). So,61.12182−c= 2.33 andc= $190.27.6.46The mgf for W is2/)/2()21()/2()()()(nYtYWtWttmeEeEtm−β−=β===. This is the mgffor a chi–square variable withndegrees of freedom.6.47By Ex. 6.46,U= 2Y/4.2 has a chi–square distribution withν= 7. So, by Table III,P(Y> 33.627) =P(U> 2(33.627)/4.2) =P(U> 16.0128) = .025.6.48From Ex. 6.40, we know thatV=21Y+22Yhas a chi–square distribution withν= 2. Thedensity function forVis2/21)(vVevf−=,v≥0. The distribution function ofU=Vis)()()()(22uFuVPuUPuFVU=≤=≤=, so that2/2)()(uUUueuFuf−=′=,u≥0. A sharpobserver would note that this is a Weibull density with shape parameter 2 and scale 2.6.49The mgfs forY1andY2are, respectively,11]1[)(ntYpeptm+−=,22]1[)(ntYpeptm+−=.SinceY1andY2are independent, the mgf forY1+Y2is2121]1[)()(nntYYpeptmtm++−=×.This is the mgf of a binomial withn1+n2trials and success probabilityp.6.50The mgf forYisntYpeptm]1[)(+−=. Now, defineX=n–Y. The mgf forXisntYtnYnttXXepptmeeEeEtm])1([)()()()()(−+=−===−.This is an mgf for a binomial withntrials and “success” probability (1 –p). Note that therandom variableX= # offailuresobserved in the experiment.6.51From Ex. 6.50, the distribution ofn2–Y2is binomial withn2trials and “success”probability 1 – .8 = .2. Thus, by Ex. 6.49, the distribution ofY1+ (n2–Y2) is binomialwithn1+n2trials and success probabilityp= .2.6.52The mgfs forY1andY2are, respectively,)1(11)(−λ=teYetm,)1(22)(−λ=teYetm.a.SinceY1andY2are independent, the mgf forY1+Y2is)1)((2121)()(−λ+λ=×teYYetmtm.This is the mgf of a Poisson with meanλ1+λ2.b.From Ex. 5.39, the distribution is binomial withmtrials andp=211λ+λλ.6.53The mgf for a binomial variableYiwithnitrials and success probabilitypiis given byiintiiYepptm]1[)(+−=. Thus, the mgf for∑==niiYU1∏+−=intiiUiepptm]1[)(is.a.Letpi=pandni=mfor alli. Here,Uis binomial withm(n) trials and successprobabilityp.b.Letpi=p. Here,Uis binomial with∑=niin1trials and success probabilityp.c.(Similar to Ex. 5.40) The cond. distribution is hypergeometric w/r=ni,N=∑in.d.By definition,

Preview Mode

This document has 334 pages. Sign in to access the full document!

Report

Study Now!

Document Details

Related Documents

The Statistics of Inheritance

Estimation and Hypothesis Testing

Lab Activity 5 Sampling Distributions

Probability and Statistics Assignment

STAT 250-004 Data Analysis Assignment 4

Normal Distribution - Amount of Sleep

Correlation and Confidence Intervals

Two-Sample Hypothesis Tests

Hypothesis Testing � Comparing Two Groups

Inferential Statistics Week 2 Solution

Company

Explore

Study Tools