Probability and Random Variables: Solutions To Problems 2021
Probability and Random Variables: Solutions To Problems 2021
Probability and Random Variables: Solutions To Problems 2021
Solutions to Problems
2021
Rahul Mukherjee
Indian Institute of Management Calcutta
1
BRING THESE NOTES TO EVERY CLASS RIGHT FROM THE BEGINNING
PROBABILITY AND RANDOM VARIABLES
Solutions to problems
3. The required probability depends on the composition of the two valves that are drawn at random
without replacement. There are three possibilities for these two valves: (i) both A, (ii) one A, one B,
(iii) both B, with respective probabilities
( 42 ) /( 92 ) = 3/18, (14 )(15 ) /( 92 ) = 10/18, ( 52 ) /( 92 ) = 5/18.
Moreover, the conditional probabilities, under (i), (ii), (iii), of the event E that both valves are non-
defectives are given respectively by
2 2
(1 – 0.1) = 0.81, (1 – 0.1)(1 – 0.2) = 0.72, (1 – 0.2) = 0.64.
1800 Both A E 243 No need to explicitly record the entries marked *
300 Not E *
One A, one B E 720 P(E) = (243+720+320)/1800 = 0.7128
1000 Not E * Hence the answer is obtained as 1 – P(E) = 0.2872
Both B E 320
500 Not E *
6. [F means fail, DA means directly acceptable (i.e., pass), MA means marginally acceptable, PF
means permanent fail]
1000 A 500 F 50 MA 20 (a) P(DA) = (450+240+140)/1000 = 0.83
PF 30 (b) P(MA) = (20+30+36)/1000 = 0.086
DA 450 (c) P(MA|F) = (20+30+36)/(50+60+60) = 0.506
B 300 F 60 MA 30 (d) P(A|MA) = 20/(20+30+36) = 0.2326
PF 30 (e) P(MA| not B) = (20+36)/(1000 – 300) = 0.08
DA 240
C 200 F 60 MA 36
PF 24
DA 140
3
8. (a) The answer is computed as
(0.6)(0.8)(0.5) + (1 – 0.6)(0.8)(0.5) + (0.6)(1 – 0.8)(0.5) + (0.6)(0.8)(1– 0.5) = 0.7
Here the first term corresponds to success in all components, while the next three terms correspond to
success in exactly two components, e.g., the second term corresponds to failure in the first component
and success in the second and third components.
(b) Analogously to (a), the answer is computed as
(0.3)(0.5)(0.2) + (1 – 0.3)(0.5)(0.2) + (0.3)(1 – 0.5)(0.2) +( 0.3)(0.5)(1 – 0.2) = 0.25
(c) J S 300*(0.25) =75 Here J means jam, NJ means no jam, S means success in at
1000 300 NS * least two components, NS means complement of S. No need
NJ S 700*(0.7) = 490 to explicitly record the entries marked *
700 NS * P(NJ|S) = 490/(75+490) = 0.8673
4
11. Upbeat X=2 Y=2 60 Here F means fine and NF means no fine
300 60
1000 X=3 F, Y = 2 27 (a) P(Fine ) = (27+90+60+12)/1000 = 0.189
90 NF, Y = 3 63 (b) P(X = 3 | Y = 2)
X=4 F, Y = 3 90 = (27+12) /(60+27+250+60+12) = 0.095
150 NF, Y = 4 60 (c) Given that the market condition was moderate,
Moderate X = 1 Y=1 150 the possible values of Y are 1, 2, 3 and 4, with
500 150 respective probabilities 150/500, 250/500, 60/500
X=2 Y=2 250 and 40/500. Hence the answer is
250 1(150/500) + 2(250/500) + 3(60/500) + 4(40/500)
X=4 F, Y = 3 60 = 1.98
100 NF, Y = 4 40 (d) Among the 63+90+60+28 = 241 cases with
Poor X=1 Y=1 100 Y=3, there are 63+28 =91 and 90+60 = 150 where
200 100 X equals 3 and 4, respectively. Hence by the two-
point variance formula, the answer is
X=2 Y=2 60 2
(4 – 3) (91/241)(150/241) = 0.235.
60
X=3 F, Y = 2 12
40 NF, Y = 3 28
12. X=2 Y=1, X+Y=3, W=0, T=3 (a) P(T > 6) = (150+80+8+32)/1000 = 0.27
300 Z=3 120 120 (b) Among the 150+80+32 = 262 cases with T=7,
1000 Y=2, X+Y=4, W=0, T=4 there are 150, 80 and 32 where Z equals 7, 8 and 9,
Z=4 180 180 respectively. Hence E(Z | T = 7)
X=3 Y=2, X+Y=5, W=0, T=5 = 7(150/262)+8(80/262)+9(32/262) = 7.5496
500 Z=5 350 350 (c) Among the 80+32 = 112 cases with W=2, there
Y=3, X+Y=6, W=0, T=7 are 80 and 32 where Z equals 8 and 9, respectively.
Z=7 150 150 So, by the two-point variance formula, V(Z | W = 2)
2
X=4 Y=3, X+Y=7, W=1, T=7 = (9 – 8) (80/112)(32/112) = 0.2041
200 Z=8 160 80 (d) Among the 80+80+8+32 = 200 cases with W > 0,
W=2, T=6 there are 80, 80, 8 and 32 where YZW equals 24, 48,
80 36 and 72, respectively. Hence E(YZW| W > 0)
Y=4, X+Y=8, W=1, T=8 = 24(80/200)+48(80/200)+36(8/200)+72(32/200)
Z=9 40 8 = 41.76
W=2, T=7 (e) P(Y is odd | T is even) = 80/(180+80+8) = 0.2985
32
13. This problem is solved by complete enumeration of all the sixteen cases
Case IIII IIIF IIFI IFII FIII IIFF IFIF IFFI FIIF FIFI FFII IFFF FIFF FFIF FFFI FFFF
X 0 1 1 1 1 1 2 1 2 2 1 1 2 2 1 1
Y 1 1 2 2 1 1 2 2 1 2 1 1 1 1 1 0
There are ten cases with Y = 1 (marked in bold). Of these, one, six and three correspond to X = 0, 1
and 2, respectively. Hence the answers to the three parts are 1/10, 6/10 and 3/10.
14. (a) Let X be the number of defectives among the two items drawn. The possible values of X are 0,
1 and 2, with respective probabilities ( 22 ) /( 42 ) = 1/6, (12 )(12 ) /( 42 ) = 4/6 and ( 22 ) /( 42 ) = 1/6. Hence E(X)
2 2 2 2 2 2
= 0(1/6)+1(4/6)+2(1/6) = 1, E(X ) = 0 (1/6)+1 (4/6)+2 (1/6) = 4/3, V(X) = E(X ) –{E(X)} = 1/3.
(b) Write D for defective and N for nondefective. The possible values of X are 2, 3, 4, 5, with
2 2
P(X = 2) = P(DD) = (0.2) = 0.04, P(X = 3) = P(NDD) = (0.8)(0.2) = 0.032,
5
2 2 2
P(X = 4) = P(DNDD) + P(NNDD) = (0.2)(0.8)(0.2) + (0.8) (0.2) = 0.032,
P(X = 5) = 1 – {P(X = 2) +P(X = 3) + P(X = 4)} = 0. 896.
(c) Let X be the number of items, among the five drawn, that survive for less than 10 hours. Then X is
–10/10 –1
binomial with n= 5 and p = 1 – e = 1 – e = 0.6321. Hence, writing q = 1 – p = 0.3679,
5 4
P(X ≥ 2) = 1 – {P(X = 0) +P(X = 1)} = 1 – (q + 5pq ) = 0.9354.
15. (a) Let X1,…, X30 denote the waiting times on the 30 occasions. These are independent, each
uniform with mean m = (0+10)/2 = 5 and variance, say, 2 . By the central limit theorem with n = 30,
then S = X1+…+ X30 is approximately normal with = nm = 150 and n . Hence P(S > 150) =
1 – ((150 ) / ) = 1 – Φ(0) = 0.5. Observe that explicit calculation of δ or was not needed.
(b) Let X1,…, X48 be the values of the quality characteristic for the 48 randomly chosen items. These
are independent, each uniform over [– 8, 4], i.e., each with mean m = (–8+ 4)/2 = – 2 and variance
2
2 ={4 – (–8)} /12 = 12. By the central limit theorem with n = 48, then S = X1+…+ X48 is
approximately normal with = nm = – 96 and n (12)(48) = 24. As M = S/48, we obtain
(108) (96)
P(4M < –9) = P{4(S/48) < –9} = P(S < –108) = = Φ(– 0.5) = 1 – Φ(0.5)
24
= 1 – 0.6915 = 0.3085.
(c) (i) Let X denote the score. From the given information,
0.2 = P(X > 80) = 1 – ((80 ) / ) , i.e., ((80 ) / ) = 0.8.
0.3 = P(X < 40) = ((40 ) / ) , i.e., 0.7 = 1 – ((40 ) / ) = (( 40) / ) .
Because Φ(0.84) = 0.8 and Φ(0.525) = 0.7, these two equations yield
(80 ) / = 0.84 and ( 40) / = 0.525,
i.e., {(80 ) / } + 2{( 40) / } = 0.84 + 2(0.525) = 1.89, i.e., / =1.89.
(ii) Let Y be the number of candidates, among the 100 randomly chosen ones, who score below 29.
Then Y is binomial with n =100 and p = P(X < 29) = ((29 ) / ) = ((29 44) / 7) = Φ(– 2.14) =
1 – Φ( 2.14) = 1 – 0.9838 = 0.0162.
Because n is large and p is small, Y may be supposed to have the Poisson distribution with λ = np =
–λ –λ –λ
1.62. Hence P(Y ≤ 1) = P(Y = 0) + P(Y = 1) = e + e λ = (1+λ) e = 0.52.
16. We first find k using the fact that the total probability is one, that is,
1 = k2+k+7k2 +k+2k+2k2+3k+2k = 10k2 +9k.
Hence k = 0.1 (the other root is negative, which is impossible as probability cannot be negative). Thus:
Value 4 5 6 7 8 9 10
Probability 0.01 0.10 0.17 0.20 0.02 0.30 0.20
(a) P(any particular year is unusual) = P(|X7| > 2) = P(X equals 4 or 10) = 0.01+0.20 = 0.21.
(b) Let Y be the number of unusual years among the four randomly chosen years. Then Y is binomial
4 3
with n = 4 and p = 0.21 [i.e., q = 0.79]. Hence P(Y ≤ 1) = P(Y = 0) +P(Y =1) = q +4pq = 0.8037.
(c) E(X) = 4(0.01) + 5(0.10) + 6(0.17) + 7(0.20) + 8(0.02) + 9(0.30) + 10(0.20) = 7.82.
17. Let X be the number of insects collected by the entomologist. The possible values of X are 1, 2, 3
and 4. Write N for not A, i.e., B or C. Then
P(X = 1) = P(A) = 1/3, P(X =2) = P(NA) = (2/3)(1/3) = 2/9,
P(X = 3) = P(NNA) = (2/3)(2/3)(1/3) = 4/27, P(X = 4) = P(NNN) = (2/3)(2/3)(2/3) = 8/27.
Hence E(X) =1(1/3) + 2(2/9) +3(4/27) +4(8/27) = 2.407. Note that X equals 4 if and only if the first
three trials yield N, irrespective of the outcome of the fourth trial.
6
18. (a) P(exactly two defective items) = P(X + Y = 2) = P(X=0, Y=2) + P(X=1, Y=1) + P(X=2, Y=0)
= P(X=0)P(Y=2) + P(X=1)P(Y=1) + P(X=2)P(Y=0) [because X and Y are independent]
2 2 2 2
= e– (1/4) + (2e– )(1/4) + (2e– )(1/4) = (5e– )(1/4).
The middle term in the above corresponds to “one of these defective items came from machine A
2 2
and the other from machine B”. Hence the answer is obtained as {(2e– )(1/4)}/{(5e– )(1/4)} = 0.4
(b) Let E and F denote, respectively, the events “at least one of the machines produced two or more
defective items” and “one of the two machines did not produce any defective item”. We need to
find P(F | E) = P(F E)/P(E). Now,
P(E) = P(X ≥ 2 or Y ≥ 2) = 1 – P(X < 2 and Y < 2) = 1 – P(X < 2)P(Y < 2) [using independence]
2 2 2
= 1 – {P(X=0) + P(X=1)}{P(Y=0) + P(Y=1)} = 1 – (e– + 2e– )(1/4 +1/4) = 1 – (3/2)e– .
Next, P(F E) = P{(X=0 or Y=0) and (X ≥ 2 or Y ≥ 2)}
= P(X=0 and Y ≥ 2) + P(Y=0 and X ≥ 2)
= P(X=0)P(Y ≥ 2) + P(Y=0)P(X ≥ 2) [using independence]
2
Note that P(X=0) = e– , P(Y ≥ 2) = 1 – {P(Y=0) + P(Y=1)} = 1 – (1/4 +1/4) = 1/2,
2 2 2
P(Y=0) = 1/4, P(X ≥ 2) = 1 – {P(X=0) + P(X=1)} = 1 – (e– + 2e– ) = 1 – 3e– .
2 2 2
Hence, P(F E) = (e– )(1/2) + (1/4)(1 – 3e– ) = (1/4)(1 – e– ).
2 2
So, the answer is given by {(1/4)(1 – e– )}/{1 – (3/2)e– } = 0.2712.
7
= P(Y ≤ 3) [write Y for the number of B in first five trials; then Y is binomial with n= 5, p =2/3]
4 5
= 1 – {P(Y=4) + P(Y=5)} = 1 – {5p (1 – p) + p } = 0.5391.
(c) Let T denote the number of I collected before the third B. Then T is negative binomial (here B is
treated as success) with r = 3 and p = 2/3, i.e., q = 1/3. Hence E(T) = rq/p = 1.5.
(d) Note that X = 3 + X*, where X* is the number of B collected before the third I, i.e., X* is
negative binomial (here I is success) with r = 3, p =1/3 and hence q = 2/3. So,
E(X) = 3 + E(X*) = 3 + (rq/p) = 3 + 6 = 9.
Similarly, Z = 5 + Z*, where Z* is the number of B collected before the fifth I, i.e., Z* is negative
binomial with r = 5, p =1/3 and hence q = 2/3. So,
E(Z) = 5 + E(Z*) = 5 + (rq/p) = 5 + 10 = 15.
Finally, E(Z – X) = E(Z) – E(X) = 15 – 9 = 6.
2
22. (a) From the given information 3 = {2 – (– h)} /12, i.e., h + 2 = ± 6, i.e., h = 4, because –h must
be less than 2. Thus X is uniform over [– 4, 2]. Hence
P(– 1 < X < 1 | – 5 < X < 1) = P(– 1 < X < 1 | – 4 < X < 1) = {1 – (–1)}/{1 – (–4)} = 0.4.
(b) Let Y be the number of customers, among the 100 chosen, having a negative X-value. Then Y is
binomial with n = 100 and p = P(X < 0) = Φ((0 – 6)/3) = Φ(– 2) = 1 – Φ(2) = 1 – 0.9772 = 0.0228.
Because n is large and p is small, Y may be supposed to have the Poisson distribution with λ = np =
–λ –λ –λ
2.28. Hence P(Y ≤ 1) = P(Y = 0) + P(Y = 1) = e + e λ = (1+λ) e = 0.3355
(c) If the distribution of X is exponential with mean m, then from the given information,
–30/m –20/m –30/m 2/3 2/3
0.343 = P(X > 30) = e , so that P(X > 20) = e = (e ) = (0.343) = 0.49, i.e.,
P(X ≤ 20) = 0.51. Hence the answer is obtained as
P(maximum of the three X-values exceeds 20)
= 1 – P(maximum of the three X-values does not exceed 20)
3
= 1 – P(all three X-values are less than or equal to 20) = 1 – (0.51) = 0.8673
23. Let the number of defects on any sheet have the Poisson distribution with parameter λ. Then the
probabilities for any sheet to have no defect, exactly one defect and exactly two defects are given,
λ λ λ 2
respectively, by p0 = e– , p1 = e– λ and p2 =(1/2)e– λ . Thus,
X is binomial with n =4, p = p0, Y is binomial with n =4, p = p1, Z is binomial with n =4, p = p2,
λ 2 λ
i.e., E(X) = 4p0, E(Z) = 4p2. Because E(Z) = 2E(X) (given), we get p2 = 2p0, i.e., (1/2)e– λ = 2e– ,
i.e., λ = 2, as λ must be positive.
(a) The number of defects on any sheet has Poisson distribution with parameter λ = 2, and hence
has variance 2. The same must hold, in particular, for the second of the four randomly chosen sheets.
So, the answer to this part is 2.
(b) Let T be the number of sheets, among the chosen four, with two or more defects. Then T is bino-
λ
mial with n = 4 and p =1 – (p0+p1) = 1 – (1+λ)e– = 0.594, as λ = 2. So, V(T) = np(1 – p) = 0.9647.
(c) Note that P(Y=3) = 4 p13 (1 p1 ) and P(X=1 and Y=3) = 4 p0 p13 . Hence
λ λ
P(X=1 | Y=3) = P(X=1 and Y=3)/P(Y=3) = p0/(1 – p1) = e– /(1 – e– λ) = 0.1856, as λ = 2.
24. Suppose the distribution of X is Poisson with parameter λ. Because P(X = 2) = P(X = 1), we get
λ 2 λ
(1/2)e– λ = e– λ, i.e., λ = 2. Also, E(Y) = 1(1/n)+…+n(1/n) = (1/n)(1+2+…+n) = (n+1)/2. As E(Y)
= 3, we get n =5, i.e., Y has possible values 1, 2,…, 5, each with probability 1/5.
(a) P(X + Y ≥ 3) =1 – P(X + Y ≤ 2) = 1 – {P(X=0, Y=1 or 2) + P(X=1, Y=1)}
= 1 – {P(X=0)P(Y=1 or 2) + P(X=1)P(Y=1)} [as X and Y are independent]
λ λ 2
= 1 – {( e– )(2/5) + (e– λ)(1/5)} = 1 – (4/5)e– [as λ = 2] = 0.8917.
8
(b) P(X + Y = 5) = P(X=0, Y=5) + P(X=1, Y=4) + P(X=2, Y=3) + P(X=3, Y=2) + P(X=4, Y=1)
= P(X=0)P(Y=5) + P(X=1)P(Y=4) + P(X=2)P(Y=3) + P(X=3)P(Y=2) + P(X=4)P(Y=1)
[as X and Y are independent]
–λ –λ –λ 2 –λ 3 λ 4
= ( e )(1/5) + (e λ)(1/5) + {(1/2)e λ }(1/5) + {(1/6)e λ }(1/5) +{(1/24)e– λ }(1/5)
2 2
= (1/5)e– {1 + 2 + 2 + (4/3) + (2/3)} [as λ = 2] = (7/5)e– .
The third and fourth terms in the above correspond to “none of them has more than three defective
2 2
spots”, and these two terms sum to (1/5)e– {2 + (4/3)} = (2/3)e– .
2 2
Hence the answer is obtained as {(2/3)e– }/{(7/5)e– }=10/21 = 0.4762.
(c) Note that Z = {X + g(Y)}/4, where g(Y) equals Y if Y ≤ 3 and 0 if Y > 3. Now, E(X) = λ = 2.
Since Y has possible values 1, 2,…, 5, each with probability 1/5, we also get
E{g(Y)} = g(1)(1/5) + g(2)(1/5) + g(3)(1/5) + g(4)(1/5) + g(5)(1/5) = (1/5)(1 +2 +3 + 0 +0) = 1.2
Hence E(Z) = [E(X) + E{g(Y)}]/4 = (2 + 1.2)/4 = 0.8.
25. Let Y be the number of persons, among the six with reservations, who report for the trip. Then
Y is binomial with n = 6 and p = 0.5, i.e., for y = 0, 1,…,6,
P(Y = y) = ( 6y )(0.5) y (1 0.5) 6 y = ( 6y ) / 64 .
Also, write X for the number of empty seats when the helicopter departs. The possible values of X
are 0, 1, 2, 3, 4, and
P(X=4) = P(Y=0) = 1/64, P(X=3) = P(Y=1) = 6/64, P(X=2) = P(Y=2) = 15/64,
P(X=1) = P(Y=3) = 20/64 P(X=0) = P(Y=4) + P(Y=5) + P(Y=6) = (15+6+1)/64 = 22/64.
Hence E(X) = 4(1/64) + 3(6/64) +2(15/64) + 1(20/64) + 0(22/64) = 1.125.
26. Let X or Y denote the lives (in hours) for any bulb of type A or B, respectively. It is given that
X and Y are both exponential. Let a and b denote the parameters of the exponential distributions of
X and Y, respectively. From the given information,
– 40/a – 20/a – 40/a 1/2 1/2
0.64 = P(X ≥ 40) = e . Hence P(X ≥ 20) = e = (e ) = (0.64) = 0.8, and
– 60/a – 40/a 3/2 3/2
P(X ≥ 60) = e = (e ) = (0.64) = 0.512, i.e., P(X < 60) = 1 – 0.512 = 0.488.
1/2 3/2
Similarly, P(Y ≥ 20) = (0.36) = 0.6, and P(Y ≥ 60) = (0.36) = 0.216, i.e., P(Y < 60) = 0.784.
Assume independence across bulbs.
(i) P(both survive for 20 hours or more) = P(X ≥ 20)P(Y ≥ 20) = (0.8)(0.6) = 0.48.
(ii) Write F for failing within 60 hours and NF for not failing wihin 60 hours. There is no need to
explicitly record the entries marked *
10000 A F 6000(0.488) = 2928 P(A|F) = 2928/(2928+3136) = 0.4828
6000 NF *
B F 4000(0.784) = 3136
4000 NF *
27. We first obtain E(X) and V(X) by a painless method. Define X1 as 1 if the student chosen from
the 2004 batch has secured a job with a multinational, and 0 otherwise. Similarly, define X2 and X3
with reference to the 2005 and 2006 batches, respectively. Now, if Z equals 1 or 0, with respective
probabilities p and 1 – p, then E(Z) = p, V(Z) = p(1 – p). Taking p = 0.4, 0.2 and 0.3, we get
E(X1) = 0.4, V(X1) = 0.24, E(X2) = 0.2, V(X2) = 0.16, E(X3) = 0.3, V(X3) = 0.21.
Because X = X1 + X2 + X3 and X1, X2, X3 are independent, E(X) = E(X1) + E(X2) + E(X3) = 0.9 and
V(X) = V(X1) + V(X2) + V(X3) = 0.61.
Next, write A for the event that the student chosen from the 2004 batch has secured a job with a
multinational. Similarly, define the events B and C with reference to the 2005 and 2006 batches,
respectively. Then, invoking independence,
P(X=0) = P ( A c B c C c ) = (1 – 0.4)(1 – 0.2)(1 – 0.3) = 0.336,
9
P(X=1) = P ( A B c C c ) + P ( A c B C c ) + P( A c B c C )
= (0.4)(1 – 0.2)(1 – 0.3) +(1 – 0.4)(0.2)(1 – 0.3) +(1 – 0.4)(1 – 0.2)(0.3) = 0.452,
P(X=2) = P( A B C c ) + P ( A B c C ) + P( A c B C )
= (0.4)( 0.2)(1 – 0.3) +(0.4)(1 – 0.2)(0.3) +(1 – 0.4)(0.2)(0.3) = 0.188,
P(X=3) = P( A B C ) = (0.4)(0.2)(0.3) = 0.024.
Check: (a) 0.336+0.452+0.188+0.024 = 1, E(X) = 0(0.336)+1(0.452)+2(0.188)+3(0.024) = 0.9,
2 2 2 2 2 2
E(X ) = 0 (0.336)+1 (0.452)+2 (0.188)+3 (0.024) = 1.42, V(X) = 1.42 – (0.9) = 0.61.
28. (a) Define X1 as 1 if the person chosen from North is tall, and 0 otherwise. Similarly, define X2,
and X3 and X4 with reference to East, South and West, respectively. Now, if Z equals 1 or 0, with
respective probabilities p and 1 – p, then V(Z) = p(1 – p). Hence with p = 0.5, 0.3, 0.3 and 0.4, we get
V(X1) = 0.25, V(X2) = 0.21, V(X3) = 0.21, V(X4) = 0.24.
Because X = X1 + X2 + X3 + X4 and X1, X2, X3, X4 are independent, we get
V(X) = V(X1) + V(X2) + V(X3) = 0.91.
(b) Write N for the event that the person chosen from North is tall. Similarly, define the events E, S
and W with reference to East, South and West, respectively. Then, invoking independence,
P(X=2) = P ( N E S c W c ) + P ( N E c S W c ) + P( N E c S c W )
+ P( N c E S W c ) + P( N c E S c W ) + P( N c E c S W )
= (0.5)(0.3)(1 – 0.3)(1 – 0.4) + (0.5)(1 – 0.3)(0.3)(1 – 0.4) + (0.5)(1 – 0.3)(1 – 0.3)(0.4)
+ (1 – 0.5)(0.3)(0.3)(1 – 0.4) + (1 – 0.5)(0.3)(1 – 0.3)(0.4) + (1 – 0.5)(1 – 0.3)(0.3)(0.4) = 0.335.
(c) Because in each of East and South, the proportion of tall persons is 0.3, the distribution of Y is
binomial with n = 2 and p = 0.3. Hence E(Y) = np = 0.6, V(Y) = np(1 – p) =0.42.
2 2 2 2
As V(Y) = E(Y ) – {E(Y)} , it follows that E(Y ) = V(Y) + {E(Y)} = 0.78.
(d) As in part (b),
P(X=3) = P( N E S W c ) + P( N E S c W ) + P( N E c S W ) + P( N c E S W )
= (0.5)(0.3)(0.3)(1 – 0.4) + (0.5)(0.3)(1 – 0.3)(0.4) + (0.5)(1 – 0.3)(0.3)(0.4) +(1 – 0.5)(0.3)(0.3)(0.4)
= 0.027 + 0.042 + 0.042 + 0.018 = 0.129
Given X=3, the only possible values of Y are 1 and 2. Among the four terms in the above, the second
and third correspond to Y=1, while the first and fourth correspond to Y=2. Hence, given X=3, Y
equals 1 or 2 with respective conditional probabilities (0.042+0.042)/0.129 = 28/43 and
(0.027+0.018)/0.129 = 15/43. Therefore, by the two-point variance formula
2
V(Y| X=3) = (2 – 1) (28/43)(15/43) = 0.2271.
10
30. X=2 Y=1 200 XY=2 (a) P(X=2 | Y=2) = 200/(200+200) = 0.5
400 Y=2 200 XY=4 (b) P(X=3 | Y=3) = 200/(200+100) = 2/3
1000 X=3 Y=2 200 XY=6 Given Y=2, the possible values of X are 2 and 3, each with
400 Y=3 200 XY=9 conditional probability 0.5. Hence
X=4 Y=3 100 XY=12 (c) E(X|Y=2) = 2(0.5) + 3(0.5) = 2.5, and
200 Y=4 100 XY=16 (d) by the two-point variance formula,
2
V(X|Y=2) = (3 – 2) (0.5)(0.5) = 0.25.
(e) The possible values of Y are 1, 2, 3 and 4 with respective probabilities 200/1000 = 0.2,
(200+200)/1000 = 0.4, (200+100)/1000 = 0.3 and 100/1000 = 0.1. Therefore,
E(Y) = 1(0.2) + 2(0.4) +3(0.3) + 4(0.1) = 2.3.
(f) As in (e), E(XY) = 2(200/1000) + 4(200/1000) + 6(200/1000) + 9(200/1000)
+ 12(100/1000) + 16(100/1000) = 7
32. (a) P(2 < X1 < 3) = 23 (1 / 72)(2 3x 2 )dx = (1 / 72)(2 x x 3 ) |32 = (1/72){(6+27) – (4+8)} = 7/24.
(b) E(X1) = 04(1 / 72) x(2 3x 2 )dx = (1 / 72){x 2 (3 / 4) x 4 } |04 = (1/72)(16+192) =26/9.
(c) P(2 < W < 3) = P(W > 2) – P(W > 3). Now,
P(W > 2) = P(X1 > 2 and X2 > 2) = P(X1 > 2)P(X2 > 2) [due to independence], and
P(X1 > 2) = P(X2 > 2)= 24(1 / 72)(2 3x 2 )dx = (1 / 72)(2 x x 3 ) | 42 = (1/72){(8+64) – (4+8)} = 5/6.
Thus, P(W > 2) = (5/6)(5/6) = 25/36.
Similarly, P(X1 > 3) = P(X2 > 3)= (1 / 72)(2 x x 3 ) |34 = (1/72){(8+64) – (6+27)} = 13/24, and
P(W > 3) = P(X1 > 3)P(X2 > 3) = (13/24)(13/24) = 169/576.
Finally, P(2 < W < 3) = P(W > 2) – P(W > 3) = (25/36) – (169/576) = 77/192
(d) P(2 < W < Z < 3) = P(2 < X1 < 3 and 2 < X2 < 3) = P(2 < X1 < 3)P(2 < X2 < 3)
= (7/24)(7/24) [by (a)] = 49/ 576.
(e) Note that X1,…, X36 are independent and that they have the same distribution, each with mean m
= 26/9 [by (b)] and variance, say, 2 . By the central limit theorem with n = 36, then Y = X1+…+ X36
is approximately normal with = nm = 104 and n . Hence
P(Y > 104) = 1 – ((104 ) / ) = 1 – Φ(0) = 0.5. Explicit calculation of δ or was not needed.
33. (a) 0.80 = P(– 6 ≤ X ≤ 6) = 66 ( / 2)e |x| dx = 06e x dx [the integrand is an even function]
11
= e x |60 = 1 – e 6 , i.e., –6θ = loge(0.2) = –loge5, i.e., θ = 0.2682.
–11/ 3
34. (a) P(X > 11) = e = 0.0256. Let Y be the number of hyperactive people among the 100
chosen. Then Y is binomial with n = 100 and p = 0.0256. As n is large and p is small, Y may be
supposed to have the Poisson distribution with λ = np =2.56. Therefore,
–λ –λ –2.56
P(Y ≥ 2) = 1 – P(Y ≤ 1) = 1 – {P(Y = 0) + P(Y = 1)} = 1 – {e + λe } = 1 – 3.56e = 0.725.
2 2
(b) By the variance formula for the uniform distribution, 3 = (1/12){h – (– 4)} , i.e., 36 = (h + 4) , i.e.,
h = 2 (the other root is inadmissible as h > – 4). Hence X is uniform over [–4, 2], i.e., it can never
exceed 2. So, P(X > 0 | – 3 < X < 3) = P(0 < X ≤ 2 | – 3 < X ≤ 2) = 2/5 = 0.4.
(c) First, note that 1 = 44 f ( x)dx = k 44 | x | dx = 2k 04 xdx = kx 2 |04 = 16k, i.e., k = 1/16. Next,
P(– 3 ≤ X ≤ 1) = 13 f ( x)dx = k 13 | x | dx = k{03 ( x)dx 01 xdx} = (k / 2)( x 2 |03 x 2 |10 )
= (k/2)(9 + 1) = 5k = 5/16 = 0.3125.
35. (a) The possible values of Y are 0, 1 and 2, with respective probabilities
–2 –2
P(Y = 0) = P(X = 0) = e , P(Y = 1) = P(X = 1) = 2e ,
–2
P(Y = 2) = 1 – P(Y = 0) – P(Y = 1) = 1 – 3e .
2 2 2 2
Hence E(Y) = 0(e– ) + 1(2e– ) + 2(1 – 3e– ) = 2 – 4e– = 1.459.
(b) Let Y be the number of defectives in the second sample (of two items) when it is drawn. Because
the lot size is large, X is binomial with n = 4, p = 0.2, Y is binomial with n = 2, p = 0.2, and X and Y
are independent. Hence, writing q = 1 – p = 0.8,
P(lot is accepted ) = P(X = 0) + P(X =1 and Y < 2) = P(X = 0) + P(X = 1)P(Y < 2)
4 3
= P(X = 0) + P(X = 1){1 – P(Y = 2)} = q + (4pq )(1 – p2) = 0.4096 + (0.4096)(1 – 0.04) = 0.8028.
(c) Let the life (in hours) distribution of each component be exponential with mean θ. Then any com-
–8/θ
ponent functions for at least 8 hours with probability e . Due to the series arrangement, the system
–8/θ 3
functions for at least 8 hours if and only all three components do so, and this has probability (e ) =
–24/θ –24/θ
e . From the given information, therefore, e = 0.6561. Hence the probability for component A
–6/θ –24/θ 1/4 1/4
to function for 6 hours or more is e = (e ) = (0.6561) = 0.9.
(d) We have Y = g(X), where g(X) = X+1 if X ≤ 10, and g(X) = X2 otherwise. Hence
E(Y) = 812 g ( x) 1218 dx = (1 / 4) 812 g ( x)dx = (1 / 4){810 ( x 1)dx 10
12 2
x dx}
12
= (1 / 4){(1 / 2)( x 1) 2 |10 3 12
8 (1 / 3) x |10 } = (1/4){(1/2)(121 – 81) + (1/3)(1728 – 1000)} = 65.667.
–10/10 –1
36. (a) The probability for any bulb to survive for 10 hours or more is e = e . Hence the num-
ber of bulbs, out of the two selected, that survive for 10 hours or more is binomial with n =2 and p =
–1
e . So, the answer is obtained as 2p(1 – p) = 0.4651.
2
(b) If X is continuous uniform over [a, b] then 10 = E(X) = (a+b)/2 and 12 = V(X) = (b – a) /12, i.e.,
a+b = 20, b – a = 12 (as b > a), i.e., b = 16, a = 4. Thus X is uniform over [4, 16]. Hence P(X ≥ 12) =
(16 – 12)/(16 – 4) = 1/3.
(d) Let Z denote the profit (in Rs). Then Z = 12 – 10 = 2 or Z = –10, according as the bulb is good or
defective, which have probabilities 0.9 and 0.1, respectively. Hence by the two-point variance for-
2
mula, V(Z) = {2 – (–10)} (0.9)(0.1) = 12.96.
(e) Write d for defective and g for good. Because Y is a nonnegative integer, the possible values of Z
are 0, 1, 2, with
P(Z=0) = P(Y=0) = P(d) = 0.1, P(Z=1) = P(Y=1) = P(gd) = (0.9)(0.1) = 0.09,
P(Z=2) = P(Y > 1) = 1 – {P(Y=0) + P(Y=1)} = 1 – (0.1 + 0.09) = 0.81.
Hence E(Z) = 0(0.1) +1(0.09) + 2(0.81) = 1.71.
38. X=2 Y=1 1500 (a) Of the 1500+1000+750 = 3250 cases with Y=2, there are 1500,
3000 Y=2 1500 1000 and 750 cases that correspond to X= 2, 3 and 4, respectively.
9000 X=3 Y=1 1000 Thus, given Y=2, the possible values of X are 2, 3 and 4, with re-
3000 Y=2 1000 spective conditional probabilities 1500/3250 = 6/13, 1000/3250 =
Y=3 1000 4/13 and 750/3250 = 3/13. Hence E(X| Y=2)
X=4 Y=1 750 = 2(6/13) + 3(4/13) + 4(3/13) = 36/13.
3000 Y=2 750 (b) Similarly, given Y=3, the possible values of X are 3 and 4, with
Y=3 750 respective conditional probabilities 1000/(1000+750) = 4/7 and
Y=4 750 750/(1000+750) = 3/7. So, by the two-point variance formula,
2
V(X| Y=3) = (4 – 3) (4/7)(3/7) = 12/49.
(c) Given X=2, the possible values of Y are 1 and 2, each with conditional probability 1/2. Hence
2
by the two-point variance formula, V(Y| X=2) = (2 – 1) (1/2)(1/2) = 1/4.
39. (a) P(Y < 0| Y < 2) = P(Y < 0 and Y < 2)/P(Y < 2) = P(Y < 0)/P(Y < 2).
Now, P(Y < 0) = 1 – P(Y ≥ 0) = 1 – P(all three X-values are nonnegative)
13
3 3
= 1 – {P(any X-value is nonnegative)} =1 – [(3 – 0)/{3 – (–2)}] = 98/125.
3
Similarly, P(Y < 2) =1 – [(3 – 2)/{3 – (–2)}] = 124/125. Hence the answer is obtained as
{(98/125)}/{(124/125)} = 0.7903.
–30/θ
(b) If X is exponential with mean θ, then 0.216 = P(X > 30) = e . Hence
–40/θ –30/θ 4/3 4/3 4
P(X < 40) = 1 – P(X ≥ 40) = 1 – e = 1 – (e ) =1 – (0.216) =1 – (0.6) = 0.8704.
(c) Here k is such that 1 = 14 f ( x)dx = k 14 | 2 x | dx = k{12 (2 x)dx 24( x 2)dx}
= (k / 2){(2 x) 2 |12 ( x 2) 2 | 42 } = (k/2)(1+4), i.e., k = 2/5. Next,
E(X) = 14 xf ( x)dx = k 14 x | 2 x | dx = k{12 x(2 x)dx 24 x( x 2)dx}
= k[{x 2 (1 / 3) x 3 } |12 {(1 / 3) x 3 x 2 } | 42 ] = k[{(4/3) – (2/3)} + {(16/3) – (– 4/3)}]
= (22/3)k = (22/3)(2/5) = 2.9333.
40. (a) Define X1 as 1 if the randomly chosen candidate passes component I, and 0 otherwise.
Similarly, define X2 and X3 with reference to components II and III, respectively. Now, if Z equals 1
or 0, with respective probabilities p and 1 – p, then E(Z) = p, V(Z) = p(1 – p). Taking p = 0.3, 0.2 and
0.1, we get E(X1) = 0.3, V(X1) = 0.21, E(X2) = 0.2, V(X2) = 0.16, E(X3) = 0.1, V(X3) = 0.09.
Because X = X1 + X2 + X3 and X1, X2, X3 are independent, E(X) = E(X1) + E(X2) + E(X3) = 0.6 and
V(X) = V(X1) + V(X2) + V(X3) = 0.46. Hence V(X)/E(X) = 0.7667.
(b) Write A for the event that the randomly chosen candidate passes component I. Similarly, define
the events B and C with reference to components II and III, respectively. Then, invoking
independence,
P(X ≥ 2) = P ( A B C c ) + P ( A B c C ) + P( A c B C ) + P( A B C )
= (0.3)( 0.2)(1 – 0.1) +(0.3)(1 – 0.2)(0.1) +(1 – 0.3)(0.2)(0.1) + (0.3)(0.2)(0.1) = 0.098.
(c) Let T be the number of candidates, among the 100 randomly chosen ones, who pass all the
components. Then T is binomial with n =100 and p = (0.3)(0.2)(0.1) = 0.006. Because n is large and p
is small, T may be supposed to have the Poisson distribution with λ = np = 0.6. Hence
–λ –λ –λ
P(T ≤ 1) = P(T = 0) + P(T = 1) = e + e λ = (1+λ) e = 0.8781.
(d) Write S for passing component I and F for failing in component I. The possible values of Y are
2 2
2, 3 and 4, with P(Y=2) = P(SS) = (0.3) = 0.09, P(Y=3) = P(FSS) + P(SFS) = 2(0.3) (0.7) = 0.126,
P(Y=4) = 1 – {P(Y=2) + P(Y=3)} = 0.784. Hence E(Y) = 2(0.09) + 3(0.126) + 4(0.784) = 3.694.
2
41. (a) If Y is continuous uniform over [a, b] then 10 = E(Y) = (a+b)/2, and 12 = V(Y) = (b – a) /12,
i.e., a+b = 20, b – a = 12 (as b > a), i.e., b = 16, a = 4. Thus Y is uniform over [4, 16], i.e., Y cannot
exceed 16. So, P(12 < Y < 18 | 6 < Y < 18) = P(12 < Y < 16 | 6 < Y < 16) = (16 – 12)/(16 – 6) = 0.4.
(b) If Y is exponential with mean θ, then
0.0784 = P(Y is less than 10 for at least one of the three consumers)
3
= 1 – P(all three consumers have Y-values 10 or more) = 1 – {P(Y ≥ 10)}
–10/θ 3 –30/θ –30/θ
= 1 – (e ) =1– e , i.e., e = 0.9216. Hence
3 –15/ θ 3 –45/ θ –30/θ 3/2
P(all three consumers have Y-values over 15) = {P(Y > 15)} = (e ) =e = (e )
3/2
= (0.9216) = 0.8847.
(c) Note that the integrands considered below are even functions. Here k is such that
1 = 33 f ( y )dy = k 33 (3 | y |)dy = 2k 03(3 y )dy = k{(3 y ) 2 } |30 = 9k, i.e., k=1/9.
It is easy to see that E(Y) = 0, because f ( y ) is symmetric about 0. Hence
2
V(Y) = E(Y ) = 33 y 2 f ( y )dy = k 33 y 2 (3 | y |) dy = 2k 03 y 2 (3 y )dy = 2k{ y 3 (1 / 4) y 4 } |30
14
= (27/2)k = (27/2)(1/9) = 1.5.
42. (a) Write A for the event that the randomly chosen candidate answers question A correctly. Simi-
larly, define the events B and C with reference to questions B and C, respectively. Then
P(at least two questions are answered correctly)
= P ( A B C c ) + P ( A B c C ) + P( A c B C ) + P( A B C )
= (0.1)( 0.3)(1 – 0.4) +(0.1)(1 – 0.3)(0.4) +(1 – 0.1)(0.3)(0.4) + (0.1)(0.3)(0.4)
= 0.018 + 0.028 +0.108 + 0.012 = 0.166
Among the four terms in the above, the first, third and fourth correspond to answering B correctly.
Since these terms sum to 0.018 +0.108 + 0.012 = 0.138, the answer is 0.138/0.166 = 0.8313.
(b) Let T be the number of candidates, among the 100 randomly chosen ones, who answer all three
questions correctly. Then T is binomial with n =100 and p = (0.1)(0.3)(0.4) = 0.012. Because n is large
and p is small, T may be supposed to have the Poisson distribution with λ = np = 1.2. Hence
–λ –λ –λ
P(T ≥ 2) = 1 – P(T ≤ 1) = 1 – {P(T = 0) + P(T = 1)} = 1 – (e + e λ) = 1 – (1+λ) e = 0.3374.
(c) P(At least 6 candidates have to be drawn to get 4 candidates correctly answering question C)
= P(Of the first 5 candidates drawn, at most 3 answer question C correctly)
= P(Z ≤ 3) [here Z is the number of candidates, of the first five drawn, who answer C correctly]
= 1 – {P(Z=4) + P(Z=5)} [note that Z is binomial with n = 5, p = 0.4, q = 1– p = 0.6]
4 5
= 1 – (5p q + p ) = 0.9130.
(d) Note that X is binomial with n =2 and p = P( A c C c ) = (1 – 0.1)(1 – 0.4) = 0.54. Hence
2
P(Y=2) = P(X > 1) = P(X=2) = (0.54) = 0.2916, P(Y=1) = 1 – P(Y=2) = 0.7084.
2
The two-point variance formula now yields V(Y) = (2 – 1) (0.2916)(0.7084) = 0.2066.
2
43. (a) If X is continuous uniform over [a, b] then 13 = E(X) = (a+b)/2 and 12 = V(X) = (b – a) /12,
i.e., a+b = 26, b – a = 12 (as b > a), i.e., b = 19, a = 7. Thus X is uniform over [7, 19], i.e., X cannot
be less than 7. So, P(moderate) = P(6 ≤ X ≤ 12) = P(7 ≤ X ≤ 12) = (12 – 7)/(19 – 7) = 5/12.
–6/θ –6/θ
(b) If X is exponential with mean θ, then 0.16 = P(poor) = P(X < 6) = 1 – e , i.e., e = 0.84.
–12/θ –6/θ 2 2
Hence P(upbeat) = P(X > 12) = e = (e ) = (0.84) = 0.7056.
(c) From the given information,
0.3 = P(poor) = P(X < 6) = ((6 ) / ) = 1 – (( 6) / ) , i.e., 0.7 = (( 6) / ) .
0.2= P(upbeat) = P(X > 12) = 1 – ((12 ) / ) , i.e., ((12 ) / ) = 0.8.
Because Φ(0.5244) = 0.7 and Φ(0.8416) = 0.8, we get ( 6) / = 0.5244 and (12 ) / = 0.8416.
So, 2{( 6) / } + (12 ) / = 2(0.5244) + 0.8416, i.e., / = 1.8904, i.e., / = 0.5290.
44. (a) Let X denote the number of gift coupons that Ayesha gets. We need to find
P(X ≥ 2| X ≥ 1) = P(X ≥ 2 and X ≥ 1)/P(X ≥ 1) = P(X ≥ 2)/P(X ≥ 1).
Because X is binomial with n = 4 and p = 0.2 (i.e., q = 1 – 0.2 = 0.8)
4 3
P(X ≥ 2) = 1 – {P(X=0) + P(X=1)} = 1 – (q + 4pq ) = 0.1808,
4
P(X ≥ 1) = 1 – P(X=0) = 1 – q = 0.5904. Thus the answer is obtained as 0.1808/0.5904 = 0.3062.
(b) Let Z be the number of gift coupons that Hari gets. Then Z is binomial with n =100 and p = 0.01.
As n is large and p is small, Z may be supposed to have the Poisson distribution with λ = np = 1.
λ 1 λ 1
Hence, P(Y=0) = P(Z=0) = e– = e– , P(Y=1) = P(Z=1) = e– λ = e– ,
15
1
P(Y=2) = 1 – {P(Y=0) + P(Y=1)} = 1 – 2e– . As a result,
–1 –1 –1 –1 2 2 1 2 1 2 1 1
E(Y) = 0(e ) + 1(e ) + 2(1 – 2e ) = 2 – 3e , E(Y ) = 0 (e– ) + 1 (e– ) + 2 (1 – 2e– ) = 4 – 7e– ,
2 2 1 1 2 1 2
V(Y) = E(Y ) – {E(Y)} = 4 – 7e– – (2 – 3e– ) = 5e– – 9e– = 0.6214
(c) Write G for a packet with a gift coupon and N for a packet with no gift coupon. Then the answer
is obtained as
P(GG) + {P(NGG) + P(GNG)} + {P(NNGG) + P(NGNG) + P(GNNG)}
2 2 2 2
= (0.2) + 2(0.8)(0.2) + 3(0.8) (0.2) = 0.1808.
Alt: Is the similarity with (a) accidental? No, here is why.
P(at most four packets to get two gift coupons)
= P(at least two gift coupons in the first four packets)
= P(X ≥ 2) [here X is the number of gift coupons in the first four packets]
= 0.1808. [as in (a), because X is binomial with n = 4, p = 0.2]
16