Académique Documents
Professionnel Documents
Culture Documents
(a) The function is non-negative and the area of the rectangle is 0.5 2 = 1 . It is a
probability density function.
(b) Since f ( x ) takes negative values over the interval from 1 to 2, it is not a
probability density function.
(c) The function is non-negative and the area of the triangle is
1
base height =
2
1
2 1 = 1 . It is a probability density function.
2
(d) The function is non-negative, but the area of the rectangle is 1 2 = 2 so it is not
a probability density function.
6.2
(a) P[0 < X < 0.5] = Area under f ( x ) between the points 0 and 0.5
= 0.5 0.5 = 0.25 .
(b) P[0.5 < X < 1] = (1 0.5) 0.5 = 0.25 .
(c) P[1.5 < X < 2] = (2 1.5) 0.5 = 0.25 .
(d) P[ X = 1] = 0 .
6.3
From the picture of f ( x ) we see that the interval 1.5 to 2 has a larger area under
f ( x ) than the interval 0 to 0.5. Consequently, P[1.5 < X < 2] is the larger of the
two probabilities.
181
182
6.4
The points x = 0.5, 1, and 1.5 divide the area under f ( x ) into quarters. Therefore,
the median = 1 , first quartile = 0.5 , and the third quartile = 1.5 .
6.5
6.6
6.7
(a) The median is the time such that there is an equal probability of being earlier or
of being later. Consequently, the median is later than 1:20 p.m.
(b) No. The mean could be larger or smaller than the median depending on the
distribution. The mean is often larger than the median when the distribution has
a long tail to the right.
6.8
(a) The median is larger than the mean so the distribution cannot be symmetric.
The distribution with a long tail to the left has mean (balance point of density)
to the left of the median and is descriptive of this situation.
(b) The median is less than the mean. The density with the long right hand tail is
compatible with the situation.
6.9
. So,
(a) Z = ( X 15) / 4
(b) Z = ( X 61) / 9
(c) Z = ( X 161) / 25 = ( X 161) / 5
6.10 The standardized variable is Z =
(a)
(b)
(c)
(d)
. So,
Z = ( X 70.0) / 3.0
Z = ( X 64.2) / 2.6
Z = (68 70) / 3.0 = 0.667
Z = (68 64.2) / 2.6 = 1.46 The z scores in (c) and (d) are two standard
deviations away from each other.
183
6.11 (a) Z = ( X 7) / 2
(b) Z = ( X 250) / 6
(c) Z = ( X 444) / 81 = ( X 444) / 9
6.12 We use Appendix B Table 3 which gives the area under the standard normal curve
to the left of a z-value.
(a)
(b)
(c)
(d)
P[ Z
P[ Z
P[ Z
P[ Z
6.13 We use Appendix B Table 3 which gives the area under the standard normal curve
to the left of a z-value.
(a)
(b)
(c)
(d)
P[ Z
P[ Z
P[ Z
P[ Z
6.14 The area to the right of a z-value = 1 Area to the left of the z-value.
(a)
(b)
(c)
(d)
6.15 The area to the right of a z-value = 1 Area to the left of the z-value.
(a) P[ Z > 0.83] = 1 P[ Z < 0.83] = 1 0.7967 = 0.2033
(b) P[ Z > 2.83] = 1 P[ Z < 2.83] = 1 0.9977 = 0.0023
(c) P[ Z > 1.23] = 1 P[ Z < 1.23] = 1 0.1093 = 0.8907
184
(d) z = 1.635 is the mid-point between z = 1.63 and z = 1.64 . From the normal
table, we find
P[ Z < 1.64] = 0.9495
P[ Z < 1.63] = 0.9484
difference = 0.0011
The mid-point between 0.9495 and 0.9484 is
1
0.9484 + (0.0011) = 0.9490 (rounded)
2
so P[ Z 1.635] = 0.9490 and P[ Z > 1.635] = 1 0.9490 = 0.0510 .
6.16 (a)
(b)
(c)
(d)
P[0.75 < Z < 0.75] = P[ Z < 0.75] P[Z < 0.75] = 0.7734 0.2266 = 0.5468
P[1.09 < Z < 1.09] = P[ Z < 1.09] P[Z < 1.09] = 0.8621 0.1379 = 0.7242.
P[0.32 < Z < 2.65] = P[Z < 2.65] P[ Z < 0.32] = 0.9960 0.6255 = 0.3705.
We find that
P[ Z < 0.74] = 0.2296
P[ Z < 0.75] = 0.2266
difference
= 0.0030
Therefore,
1
P[ Z < 0.745] = 0.2266 + (0.0030) = 0.2281 (rounded)
2
Also,
P[ Z < 1.25]
= 0.8944
= 0.0019
so
P[ Z < 1.244] = 0.8925 + 0.4(0.0019) = 0.8933 (rounded)
Finally,
P[0.745 < Z < 1.244] = 0.8933 0.2281 = 0.6652
6.17 (a)
(b)
(c)
(d)
P[0.44 < Z < 0.44] = P[ Z < 0.44] P[ Z < 0.44] = 0.6700 0.3300 = 0.3400.
P[1.33 < Z < 1.33] = P[ Z < 1.33] P[ Z < 1.33] = 0.9082 0.0918 = 0.8164.
P[0.40 < Z < 2.03] = P[ Z < 2.03] P[Z < 0.40] = 0.9788 0.6554 = 0.3234.
We find that
P[ Z < 1.41] = 0.9207
P[ Z < 1.40] = 0.9192
difference = 0.0015
Therefore,
1
P[ Z < 1.405] = 0.9192 + (0.0015) = 0.9200 (rounded)
2
Also,
185
= 0.0003
so
P[ Z < 2.306] = 0.9893 + 0.6 (0.0003) = 0.9895 (rounded)
Finally,
P[1.405 < Z < 2.306] = 0.9895 0.9200 = 0.0695
6.18 (a) P[ Z < 0.61] = 0.2709 so the z-value is 0.61 .
(b) We are to find the z-value for which the area to the left is 1 0.35 = 0.65 .
From the normal table, we find that
P[ Z < 0.39] = 0.6517
P[ Z < 0.38] = 0.6480
difference
= 0.0037
We need P[ Z < z ] = 0.65 . Since 0.65 0.6480 = 0.002 , and 0.39 0.38 = 0.01 ,
the required z-value is
0.002
Z = 0.38 + (0.01)
= 0.38 + 0.0054 = 0.3854 (rounded)
0.0037
(c) P[ Z < z ] = 0.5 + 0.38 = 0.88 . From the normal table,
P[ Z < 1.18] = 0.8810
P[ Z < 1.17] = 0.8790
difference
= 0.0020
186
1
P[ Z < z ] = (0.2584) = 0.1292
2
From the normal table, we find P[ Z < 1.13] = 0.1292 . Therefore, the required
z-value is z = 1.13 .
(f) Since P[ Z < 1.82] = 0.9656 , and it is given that P[ z < Z < 1.82] = 0.65 , we must
have
P[ Z < z ] = 0.9656 0.65 = 0.3156
The normal table shows P[ Z < 0.48] = 0.3156 , so that z = 0.48 .
6.19 (a) We are to find the z-value for which the area to the left is 0.20. From the
normal table, we find
P[ Z < 0.84] = 0.2005
P[ Z < 0.85] = 0.1977
= 0.0028
difference
Since 0.2005 0.20 = 0.0005 , the required z-value is
0.0005
Z = 0.84 (0.01)
= 0.842 (rounded)
0.0028
(b) We are to find the z-value for which the area to the left is 1 0.125 = 0.875 so
z = 1.15 .
(c) By the symmetry of the normal curve, we have
2 P[ Z < z ] = 1 P[ z < Z < z ] = 1 0.668 = 0.332
or
P[ Z < z ] = 0.166 , so that z = 0.97 .
(d) We are to find the z-value for which
0.888 = P[ z < Z < 2.0] = P[Z < 2.0] P[Z < z ] = 0.9772 P[ Z < z ]
or
P[ Z < z ] = 0.0892 .
We find that
P[ Z < 1.34] = 0.0901
P[ Z < 1.35] = 0.0885
difference
= 0.0016
0.0009
= 1.34 0.006 = 1.346 (rounded)
0.0016
187
6.20 (a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
0.0003
= 1.996
0.0005
0.0009
= 0.753
0.0030
188
6.22 From the symmetry of the standard normal distribution about 0, the median, or
second quartile, is 0. From the normal table, we find
P[ Z < 0.67] = 0.2514
P[ Z < 0.68] = 0.2483
difference
= 0.0031
Since 0.25 is about halfway between these two probabilities, we could use 0.675 .
More accurately,
0.0014
Z = 0.67 (0.01)
= 0.675 (rounded)
0.0031
so the first quartile = 0.675 and the third quartile = 0.675 .
6.23 (a)
(b)
(c)
(d)
6.24 (a)
(b)
(c)
(d)
0.0015
= 0.524 (rounded)
0.0034
X 50
.
9
189
X 13.0
.
2.4
X 50
.
9
(a) From the normal table we find P[ Z < 1.96] = 0.975 . Therefore,
b 50
= 1.96
9
X 13
.
2.4
(a) From the normal table we find P[ Z < 0.84] = 0.7995 . Therefore,
so b = 13 + 2.4(0.84) = 15.016 .
b 13
= 0.84
2.4
190
(b) Here P[ X < b] = 1 0.002 = 0.998 . From the normal table we find that
b 13
P[ Z < 2.88] = 0.9990 . Therefore,
= 2.88 so b = 13 + 2.4(2.88) = 19.912 .
2.4
(c) Since P[ Z < 2.17] = 0.015 , we must have
b 13
= 2.17 or b = 13 + 2.4(2.17) = 7.792 .
2.4
6.29 Let X denote the score of a randomly selected student. We have = 500 and
X 500
.
= 100 so the standardized variable is Z =
100
650 500
(a) P[ X > 650] = P[Z >
] = P[ Z > 1.5] = 1 0.9332 = 0.0668
100
(b) P[ X < 250] = P[Z < 2.5] = 0.0062
325 500
(c) x = 325 gives z =
= 1.75
100
675 500
= 1.75
x = 675 gives z =
100
Thus, we have
P[325 < X < 675] = P[1.75 < Z < 1.75] = 0.9599 0.0401 = 0.9198.
6.30 (a) The proportion of students who score over 680 is the probability
680 500
P[ X > 680] = P[Z >
] = P[ Z > 1.80] = 1 0.9641 = 0.0359
100
(b) Because the center of the scores distribution is = 500 , 50% of the students
score over 500. The required limit is 500.
(c) We are to find the score b such that
P[ X > b] = 0.15 or P[ X < b] = 0.85
We find P[ Z < 1.04] = 0.8508 0.85 . Therefore,
b 500
= 1.04 or b = 500 + 100(1.04) = 604 .
Z=
100
6.31 The z-value for 32.5 inches is z = (32.5 34.5) / 1.4 = 1.429 and that for 36.5
inches is z = (36.5 34.5) /1.4 = 1.429 so
P[32.5 < X < 36.5] = P[1.429 < Z < 1.429] = 0.9235 0.0765 = 0.847
6.32 The z-value for 62.5 minutes is z = (62.5 64.3) / 1.15 = 1.57 and that for 67.7
minutes is z = (67.7 64.3) / 1.15 = 2.96 so
P[62.5 < X < 67.7] = P[1.57 < Z < 2.96] = 0.9985 0.0582 = 0.9403
191
ounces.
6.37 The arrival time X is distributed as N (17, 3) .
22 17
(a) (i) P[ X > 22] = P[ Z >
] = P[ Z > 1.67] = 1 0.9525 = .0475
3
13 17
(ii) The z-values corresponding to x = 13 and x = 21 are z =
= 1.33
3
21 17
= 1.33 , respectively. Thus, we have
and z =
3
P[13 < X < 21] = P[1.33 < Z < 1.33] = 0.9082 0.0918 = 0.8164
192
15.5 17
18.5 17
= 0.5 , z =
= 0.5 , we have
3
3
P[15.5 < X < 18.5] = P[0.5 < Z < 0.5] = 0.6915 0.3085 = 0.3830
(iii) Since z =
(b) The probability density curve of a normal distribution peaks at the mean.
Therefore, the 1-minute interval that has the highest probability is one that is
centered at = 17 and has length 1, that is, the interval 16.5 to 17.5.
6.38 Let X be the strength of the tube. The standardized variable is Z =
X 32
.
4
25 32
] = P[Z < 1.75] = 0.0401
4
38 32
] = P[ Z < 1.5 = 0.9332
(b) P[ X < 38] = P[ Z <
4
6.39 (a) We use the binomial table (Appendix B Table 2) for n = 25 and p = 0.6 .
(i) P[ X = 17] = P[ X 17] P[ X 16] = 0.846 0.726 = 0.120
(ii) P[11 X 18] = P[ X 18] P[ X 10] = 0.926 0.034 = 0.892
(iii) P[11 < X < 18] = P[ X 17] P[ X 11] = 0.846 0.078 = 0.768
(b) With n = 25 and p = 0.6 , we calculate
np = 15 and
(ii)
193
(iii) Here the end points 11 and 18 are not included. The continuity correction
leads to the interval 11.5 to 17.5.
P[11.5 < X < 17.5] = P[1.429 < Z < 1.020]
= 0.8461 0.0765 = 0.7696.
6.40 (a) We use the binomial table (Appendix B Table 2) for n = 25 and p = 0.4 .
(i) P[ X = 11] = P[ X 11] P[ X 10] = 0.732 0.586 = 0.146
(ii) P[6 X 12] = P[ X 12] P[ X 5] = 0.846 0.029 = 0.817
(iii) P[6 < X < 12] = P[ X 11] P[ X 6] = 0.732 0.074 = 0.658
(b) With n = 25 and p = 0.4 , we calculate
np = 10 and
(ii)
194
so Z =
X 51
is approximately N (0, 1) .
6.51
P[ X 45] P[ Z <
47.5 51
69.5 51
<Z<
]
6.51
6.51
= P[0.5376 < Z < 2.842]
so Z =
X 132
is approximately N (0, 1) .
6.699
195
6.43 (a) Normal approximation is appropriate because n is large and p is not too close to
0 or 1.
(b) Not appropriate because p is too small, np = 3 .
(c) Not appropriate because p is too close to 1, n(1 p ) = 2.4 .
(d) Normal approximation is appropriate because n is large and p is not too close to
0 or 1.
6.44 (a) Normal approximation is appropriate because n is large and p is not too close to
0 or 1.
(b) Not appropriate because n is not large.
(c) Normal approximation is appropriate because np = 6 > 5 and n(1 p) = 294 > 5 .
(d) Not appropriate because p is too close to 1, n(1 p ) = 4.5 .
(e) Appropriate because n is large and np = 71 and n(1 p) = 39 .
6.45 The standardized scale is z =
x np
np (1 p )
n = 5, p = 0.4 :
z = ( x 2 ) 1.095
n = 12,
p = 0.4 :
z = ( x 4.8) 1.697
n = 25,
p = 0.4 :
z = ( x 10 ) 2.449.
The probability histograms of binomial distribution for p = 0.4, n = 5, 12, and 25,
and the corresponding z-scores are given below.
196
197
6.46 Since 35.6 is the median, the probability is p = 0.5 that a randomly selected
resident is under 35.6 years of age.
Let X = the number of residents under 35.6 years old out of a random sample of
200 residents. Then X has the binomial distribution with n = 200 and p = 0.5 . To
use the normal approximation, we calculate
np = 200 0.5 = 100,
npq = 200 0.5 0.5 = 7.071 .
Using continuity correction,
109.5 100
P[ X 110] P[ Z >
] = P[ Z > 1.344] = 1 0.9105 = 0.0895 .
7.071
6.47 Let X = number of unemployed persons in a random sample of 300. Then, the
distribution of X is binomial with n = 300 and p = 0.079 , and a normal
approximation is appropriate. Here
np = 300 0.079 = 23.7
npq = 300 0.079 0.921 = 4.672
Z=
X 23.7
4.672
17.5 23.7
] = P[ Z < 1.327] = 0.0923
4.672
30.5 23.7
(b) P[ X > 30] P[Z >
] = P[ Z > 1.455] = 1 0.9272 = 0.0728 .
4.672
6.48 We do not know the population proportion so we answer using the sample
proportion 48 50 = 0.96 . No, the normal distribution does not give a good
approximation. The sample size n = 50 seems large, but p is too close to 1 so
n (1 p ) = 50 (1 0.96 ) = 2 is small.
6.49 We assume that participants in charity work are independent of one another. The
binomial distribution applies. Since n = 64 and p = 0.25 , the distribution of
X = number who participate in charity work, is approximately normal with mean
= 64 0.25 = 16.0 and sd = 64 0.25 0.75 = 3.464 .
Using continuity correction, we approximate
19.5 16.0
P[ x 20] = P[ Z
] = 1 P[ Z < 1.01] = 1 0.8438 = 0.1562
3.464
198
P[18 X 24] P[
6.52 Let X = the number of students who live in apartments out of a random sample of
200 students. Then, X has the binomial distribution with n = 200 and p = 0.3 . To
use the normal approximation, we calculate
np = 200 0.3 = 60,
199
6.53 We assume the residents make their decision to move independently of one another.
The binomial distribution applies. Since n = 100 and p = 0.33 , the distribution of
X = number who move, is approximately normal with mean = 100 0.33 = 33 and
sd = 100 0.33 0.67 = 4.702
Using continuity correction, we approximate
38.5 33
P[ X 39] P[ Z >
] = 1 P[Z < 1.170] = 1 0.8790 = 0.1210
4.702
6.54 A binomial model is reasonable for X = number of infested trees out of 300.
(a) Since n = 300 and p = 0.2 , the distribution of X is approximately normal with
mean = 300 0.2 = 60, and sd = 300 0.2 0.8 = 6.928
48.5 60
71.5 60
P[49 X 71] P[
<Z<
]
6.928
6.928
= P[1.660 < Z < 1.660] = 0.9515 0.0485 = 0.9030
(b) The observed count x = 72 is at the distance of 72 60 = 12 from the mean.
We reason that the probability of getting this or a more extreme observation is
P X 60 12 P Z > 1.732
200
chapter. If this process is repeated with an even much larger sample, the jumps
between consecutive rectangles will then dampen out, and the top of the histogram
will approximate the shape of a smooth curve. The density curve would be nonnegative and the area under the curve is one.
6.58 (a)
(b)
(c)
(d)
6.59 (a) Scanning the probabilities given in the normal table, we find that
P[ Z < 1.38] = 0.0838. Therefore, z = 1.38.
(b) In the table we look for probabilities close to .047, and find
P[ Z < 1.68] = 0.0465 and P[ Z < 1.67] = 0.0475
Interpolation gives P[ Z < 1.675] = 0.047 , so that z = 1.675.
(c) Area to the left of z is 1 0.2611 = 0.7389. From the table we get
P[ Z < 0.64] = 0.7389 , so that z = 0.64.
(d) Area to the left of z is 1 0.12 = 0.88. The table gives
P[ Z < 1.17] = 0.8790 and P[ Z < 1.18] = 0.8810 .
Since 0.88 is halfway between these results, we interpolate z = 1.175.
6.60 The 20th percentile is the point z such that the area to its left is 0.2 or
P[ Z < z ] = 0.2. Scanning the normal table, we find that
P[ Z < 0.84] = 0.2005 0.2 , so the 20th percentile = 0.84 . (Using interpolation,
we obtain 0.84 (0.01) 285 = 0.842.)
Similarly, we find that
P[ Z < 0.253] = 0.40 , so the 40th percentile = 0.253
By the symmetry of the standard normal distribution around zero, the 60th
percentile is 0.253 and the 80th percentile is 0.84.
6.61 (a) P[ Z > 0.62] = 1 P[ Z < 0.62] = 1 0.7324 = 0.2676
(b) P[1.40 < Z < 1.40] = P[ Z < 1.40] P[ Z < 1.40] = 0.9192 0.0808 = 0.8384
(c) P Z > 3 = P [ Z > 3] + P [ Z < 3]
= 2 P [ Z < 3]
(by symmetry)
= 2 0.0013 = 0.0026
(d) P Z < 2 = P [ 2 < Z < 2] = P [ Z < 2] P [ Z < 2] = 0.9772 0.0228 = 0.9544
Alternatively, we can calculate
201
P Z > 2 = 2 P [ Z < 2]
(as in part (b))
= 2 0.0228 = 0.0456
Hence, P Z < 2 = 1 0.0456 = 0.9544.
6.62 Denoting x = raw score and z = standardized score, we have the relation
x mean x 115
z=
=
, so that x = 115 + 22 z .
sd
22
108 115
(a) For x = 108 we have z =
= 0.318
22
(b) For z = 0.8, we find x = 115 + 22(0.8) = 97.4
(c) The raw scores x = 105 and x = 128 correspond to the standardized scores
105 115
128 115
z=
= 0.455, z =
= 0.5909,
22
22
respectively, so the required interval is -0.455 to 1.045.
(d) Observe that
z = 1.5 gives x = 115 + 22(1.5) = 82
z = 1.5 gives x = 115 + 22(1.5) = 148
Thus, the required interval is 82 to 148.
6.63 The standard normal variable is Z =
X 280
.
58
337 280
] = P[ Z < 0.983] = 0.8372
58
P[ X < 240] = P[Z < 0.690] = 0.2451
P[ X > 230] = P[ Z > 0.862] = 1 0.1944 = 0.8056
P[ X > 90] = P[Z > 3.276] = 1 P[ Z 3.276] = 1 0.005 = 0.995
P[235 < Z < 335] = P[0.776 < Z < 0.948] = 0.8284 0.2188 = 0.6096
P[305 < X < 405] = P[0.431 < Z < 2.155] = 0.9844 0.6669 = 0.3175
X 100
.
5
(a) From the normal table, P[ Z < 0.44] = 0.6700.
b 100
Since P[ X < b] = P[ Z <
], we must have
5
b 100
Z=
= 0.44 or b = 100 + 5(0.44) = 102.2.
5
b 100
(b) Since P[ Z > 2.29] = 0.011, Z =
= 2.29 or b = 100 + 5(2.29) = 111.45.
5
(c) We first find z such that P Z < z = 0.966.
202
X 499
.
120
600 497
(a) P[ X > 600] = P[Z >
] = P[ Z > 0.842] = 1 0.8046 = 0.1954
120
(b) We first find the 90th percentile of the standard normal distribution and then
convert it to the x scale. Indeed, observe that
P[ Z < 1.28] = 0.8997 0.90
The standardized score z = 1.28 corresponds to
x = 497 + 120(1.28) = 651
(c) P[ X < 400] = P[Z < 0.808] = 0.2095 .
6.66 Denote X = lifting capacity. It has N (65,8) distribution. Ability to lift an 80 pound
load requires that X > 80 .
80 65
P[ X > 80] = P[ Z >
] = P[ Z > 1.875] = 1 0.9696 = 0.0304 .
8
So, only about 3% of these workers can lift the load.
6.67 The strength X is distributed as N (100,8) . The bonding will fail if X < 90 . Its
probability is
90 100
P[ X < 90] = P[ Z <
] = P[ Z < 1.25] = 0.1056 .
8
So, about 11% will fail.
X 70
so the distribution scores corresponding to
8
x = 58, 66, 74, and 82 are given by
58 70
66 70
74 70
82 70
= 1.5,
= 0.5,
= 0.5, and
= 1.5 .
8
8
8
8
Thus, we have:
203
204
(b)
6.73 (a)
(b)
npq = 2
8.5 8
] = P[ Z > 0.25] = 1 P[ Z < 0.25]
.
2
= 1 0.5987 = 0.4013
P[ X 9] = P[ Z >
6.74 Let X = number of late payments out of 20,000 tax bills. It has binomial
distribution with n = 20, 000 and p = 0.07. For a normal approximation, we
calculate
np = 20, 000 0.07 = 1400 , npq = 20, 000 0.07 0.93 = 36.08.
The continuity correction is not important because the sd is fairly large.
1350 1400
(a) P[ X < 1350] P[ Z <
] = P[Z < 1.386] = 0.0829
36.08
1480 1400
P[ X 1480] P[ Z >
] = P[ Z > 2.217] = 1 P[ Z < 2.217]
(b)
36.08
= 1 0.9867 = 0.0133
205
102.5 120
] = P[Z < 1.91] = 0.0281
9.165
(b) Under the hypothesis that p = 0.3 , the calculations in part (a) show that the
occurrence of fewer than 103 viewers in a sample of 400 is unlikely (probability
0.0281). Such an observation should strongly support suspicion that p < 0.3.
6.76 (a) n is large and p not too close to 0 or 1 so the normal approximation is
appropriate.
(b) n is not large and p is very small so the normal approximation is not
appropriate.
(c) Although n is large, p being too close to 1, the normal approximation is not
appropriate.
6.77 Out of 400 ticket holders, let X = no. of persons who show up. The distribution of
X is binomial with n = 400 and p = 1 0.10 = 0.90. It is approximately normal with
mean = np = 360 and sd = npq = 6.
(a) One or more reservation holders will not be accommodated if more than 370
passengers show up.
370.5 360
P[ X > 370] P[ Z >
] = P[ Z > 1.75] = 1 0.9599 = 0.0401
6
349.5 360
(b) P[ X < 350] P[ Z <
] = P[ Z < 1.75] = 0.0401.
6
6.78 Let X = number of card users out of 147 customers. It has a binomial distribution
with n = 147 and p = 0.43. For a normal approximation, we calculate
np = 147 0.43 = 63.2 , npq = 147 0.43 0.57 = 6.002
60.5 63.2
(a) P[ X > 60] = P[Z >
] = P[Z > 0.4499] = 0.6736
6.002
59.5 63.2
70.5 63.2
(b) P[60 X 70] = P[
Z
]
6.002
6.002
= P[0.616 Z 1.216] = 0.8882 0.2689 = 0.6191
206
6.79 Let X be the number of new words in the poem. We approximate the probability
that a new word will not be on the list by 14, 376 884, 647 = 0.01625. Then, we
treat X as having a binomial distribution with n = 429 and p = 0.01625.
(a) Expected number of new words = number of words probability
= 429 0.01625 = 6.97
(b) The standard deviation of X is 429 0.01625 0.98375 = 2.619. Using the
continuity correction, we approximate the binomial probability
11.5 6.97
P[ X 12] P[ Z
] = P[ Z 1.73] = 1 0.9582 = 0.0418
2.619
(c) Using continuity correction, we approximate the binomial probability
2.5 6.97
P[ X 2] P[ z
] = P[ Z 1.707] = 0.0439
2.619
2.5 6.97
11.5 6.97
Z
]
2.619
2.619
11.5 6.97
2.5 6.97
= P[Z
] P[Z
]
2.619
2.619
= P[Z 1.73] P[Z 1.707]
= 0.9582 0.0439 = 0.9143
The observed number 9 is close to the expected number 6.97. The difference
9 6.97 = 2.03 is not large according to our probability calculation which
shows that 4 words from 6.97 is not unusual.
6.80 The large values of volume are too large for the normal distribution to hold. They
should be pulled down with a transformation.
207
6.81 The normal scores plot for volume displays a curved pattern. The data are probably
not normal.
6.83 (a) The normal scores plot for the original data set is as follows:
Probability Plot of Lightning Strikes
Normal
200
Mean
StDev
N
AD
P-Value
Lightning Strikes
150
100
50
0
-2
-1
0
Score
83.34
38.37
50
1.158
<0.005
208
(b) The normal scores plot for the fourth-root of the values in the original data set
is as follows:
Probability Plot of fourth roots
Normal
4.0
Mean
StDev
N
AD
P-Value
fourth roots
3.5
2.963
0.3442
50
0.490
0.212
3.0
2.5
2.0
-2
-1
0
Score
(c) The data in both plots are very close to the line, indicating that a normal
approximation in both cases is reasonable.