Professional Documents
Culture Documents
Number of people, x
Number of households,
fx
2
f2
3
20
4
f4
5
18
6
10
7
5
8
6
7
7
8
2
8
5
2
1
6
7
7
8
3
Male
9
3
7
11
9
6
5
6
5
3
4
6
9
6
1
10
1
2
2
3
2
8
3
7
3
7
6
5
2
6
5
4
4
4
5
Female
1
9
1
5
6
5
4
11
9
6
1
4
1
6
2
3
4
8
8
3
2
Male: xM 215 , xM
1, 481 ; Female: xF 168 and xF2 1, 026
The male observations are assumed to be normally distributed with mean 1 and
standard deviation 1 , and independently the female observations are assumed to
be normally distributed with mean 2 and standard deviation 2 .
Suppose that it is known that 1 3.0 days and 2 2.5 days. Construct a 95%
confidence interval for the difference between the mean length of stay for males
and the mean length of stay for females, that is for 1 2 . Comment briefly on
any implications of this confidence interval.
Suppose now that 1 and 2 are unknown. Perform a two-sample t-test to
investigate whether there is a difference between the mean length of stay for
males and the mean length of stay for females, assuming that 1 and 2 are
equal.
Show that the variances in the male and female samples are not significantly
different at the 5% level.
Suppose you are not prepared to assume more than you feel is absolutely
necessary in particular you do not want to assume that 1 and 2 are equal, nor
that the observations necessarily come from normal populations.
6. A random sample of 25 recent claim amounts in a general
insurance context is taken from a population that you may
assume is normally distributed. In units of 1,000, the sample
mean is x = 9.416 and the sample standard deviation is s =
2.105. Calculate a 95% one-sided upper confidence limit (that is,
the upper limit k of a confidence interval of the form (0,k)) for the
standard deviation of the claim amounts in the population.
7. Consider the following questions:a. A random sample of 200 policy surrender values (in units
of 1,000) yields a mean of 43.6 and a standard deviation
of 82.2. Determine a 99% confidence interval for the true
underlying mean surrender value for such policies.
b. Suppose that a random sample of nine observations is taken from a normal
distribution with mean, 0 . Let X and S 2 denote the sample mean and
variance respectively. Determine (to 2 decimal places) the probability that
the value of X exceeds that of S, i.e. determine P( X S).
8. Consider then, the following questions:a. A random sample of size 10 is taken from a normal population
with standard deviation = 15 and the sample standard
versus
H1 : X 105.6
If H 0 is true, then
T
X 105.6
3 X 105.6
~ t35
8
16 / 36
At 5%, t5%,34 1.691 whereas t5%,36 1.688 . Then,
1.691 1.688
t5%,35
1.6895
2
3 100 105.6
Reject null hypothesis if T t5%,35 . T
2.1 . T t5%,35 , reject
8
null hypothesis.
The new education procedure affects the average childrens scores.
#2
2
X
X i ~ N , 2 , then, X ~ N , and thus,
~ tn 1
n
s
/
n
a
Pr tn 1
/ n
/ n / n
1900 1800
Pr X 1900 Pr t89
Pr t89 2.372
400 / 90
Pr X a Pr
#3
7 f 2 20 f 4 18 10 5 100
f 2 f 4 100 60
f 2 f 4 40
(eq 1)
7 1 f 2 2 20 3 f 4 4 18 5 10 6 5 7
4
100
7 2 f 2 60 4 f 4 90 60 35 400
2 f 2 4 f 4 400 252
f 2 2 f 4 74
(eq 2)
(eq 2) (eq 1)
f 4 74 40 34
M
f 2 40 34 6
X 50 X 51 4 4
4
2
2
X i ~ N X 94.2, X2 24.86
24.86
2
Then, X ~ N X 94.2, X
,
15
X ~ N , 2
and thus,
X
15
X
~ N 0,1
24.86
15
At 1% , we may compute Z 0.5% Z 0.5% 2.576 . Then, the 99% confidence
interval,
2.576
X
24.86
15
2.576
24.86
24.86
X 2.576
15
15
24.86
24.86
2.576
X 2.576
15
15
24.86
24.86
X 2.576
X 2.576
15
15
2.576
With, X 94.2 ,
For i 1,...,15 ,
90.8837 97.5163
CI99% 90.8837,97.5163
X i ~ N , 2
2
X
~
N
,
Then,
2
s2
X ~ N , and A n 1
n
2
X
~ N 0,1
Thus, T
s/ n
Then, the confidence interval at 1 100% is formed as follows
t / 2, n 1
t / 2, n 1
t / 2, n 1
X t / 2, n 1
X
s/ n
t / 2, n 1
X t / 2, n 1
n
s
n
s
X t / 2, n 1
X t / 2, n 1
s
n
s
n
s
n
s
s
CI 1 100% X t / 2, n 1
, X t / 2, n 1
n
n
For 99% confidence interval, is set as 1%. Then, for n 15 , t0.5%,14 2.977 . It is set
that, X 94.2 and s 2 24.86
24.86
24.86
CI 1 0.1 100% 94.2 2.977
,94.2 2.977
15
15
CI99% 90.3675,98.0325
#5
Male:-
2 , 22
X i,1 ~ N 1 , 12
i 1,..., 40
Female:X i,2 ~ N
i 1,...,35
X1 X 2 1 2
12
40
2
2
12
then, X1 ~ N 1 ,
40
22
ten, X 2 ~ N 2 ,
35
~ N 0,1
35
40 35
1.96
12 22
2 2
X1 X 2 1 2 1.96 1 2
40 35
40 35
1.96
12 22
2 2
1 2 X1 X 2 1.96 1 2
40 35
40 35
12 22
12 22
X1 X 2 1.96 40 35 1 2 X1 X 2 1.96 40 35
215
168
X1
5.375 and X 2
4.8 ; then;
40
35
32 2.52
32 2.52
40 35
40 35
0.575 1.96 0.225 0.1786 1 2 0.575 1.96 0.225 0.1786
0.6702 1 2 1.8202
CI95% 0.6702,1.8202
Statistically, 1 2 0 .
H 0 : 1 2 0
versus H 0 : 1 2 0
Male:-
X i,1 ~ N 1 , 2
i 1,..., 40
2
X
~
N
,
1
then, 1
40
i 1,...,35
2
X
~
N
,
2
ten, 2
35
Female:-
X i,2 ~ N 2 , 2
X1 X 2 1 2
1
1
40 35
~ N 0,1
2
Male: xM 215 , xM
1, 481 ; Female: xF 168 and xF2 1, 026
s12
s12
1481 215 / 40
39
8.3429
sP2
1026 168 / 35
34
6.4588
7.4654
39 34
73
X1 X 2 1 2
~ t73
1
1
sP
40 35
At 5% ,
t2.5%,70 1.994 and t2.5%,80 1.990 , then,
t2.5%,73 t2.5%,80
73 80
70 80
t2.5%,70 t2.5%,80
7
t2.5%,73 t2.5%,80 t2.5%,70 t2.5%,80
10
7
t2.5%,73 1.990 1.994 1.990
10
t2.5%,73 1.9928
0.575 0
7.4654 7.4654
40
35
0.909
It is observed that, T t2.5%,73 . Do not reject null hypothesis. The mean length of stay
for males and the mean length of stay for females are not different.
s2
2
A 39 1 ~ 39
2
1
s2
2
B 34 2 ~ 34
2
2
and
H 0 : 1 2 versus H1 : 1 2
A / 39 s12 / 2 s12
~ F39,35
If H 0 is true, F
B / 35 s22 / 2 s22
At 5% , F0.975,39,34 0.5204 and F0.025,39,34 1.9508 . The null hypothesis is
true if F0.975,39,34 F F0.025,39,34
8.3429
1.2917 F0.025,39,34 . The null hypothesis is not rejected. There is
6.4588
significantly no difference between variance of male and female waiting length.
F
X1 X 2 1 2
s12 s22
n1 n2
s 2 s 2
1 2
n1 n2
~ tw
8.3429 6.4588
40
35
72.99 73
2
2
2
2
8.3429
6.4588
s 2
s 2
1
2
35
40
n1
n2
39
34
n1 1
n2 1
At 5% ,
t2.5%,73 t2.5%,70
t2.5%,80 t2.5%,70
73 70
80 70
3
10
3
1.9928
10
0.9171
8.3429 6.4588
40
35
X i ~ N , 2
For, i 1,..., 25 ,
s2
2
~ 24
s2
s 2
s 2
1 1
Pr 24
24 Pr A 24 0.05
k
2
k 2
k 2
Pr k Pr
s 2
Pr A 24 0.05
k 2
24
s2
k2
ks
36.42
24
24
2.105
1.709
36.42
36.42
#7
a
2
X ~ N , with X 43.6 and s 82.2 .
200
X
T
~ t199
s / 200
Pr a T a 0.99 . Here, a t0.5%,199 . In statistical table, it is given that
t0.5%,150 2.609 and t0.5%,200 2.601 . Using interpolation,
a t0.5%,150
t0.5%,200 t0.5%,150
199 150
200 150
49
50
49
2.60116
50
a T a
X
a
a
s / 200
as
as
X
200
200
as
as
X
X
200
200
82.2
82.2
43.6 2.60116
43.6 2.60116
200
200
28.481 58.719
b
2
X ~ N 0,
X
~ t8
s/3
X
s
Pr T 3 a
s / 3 s / 3
Given Pr T t1%,8 2.896 and Pr T t0.5%,8 3.355 . By using interpolation,
a 0.5%
3 3.355
#8
a
A9
s2
~ 92
9s 2
9a 2
9a 2
Pr s a Pr
2
15
2
95%,9
3.325
9a 2
152
9a 2
152
3.325
152
a
3.325 9.1173
9
2
X ~ N , with X 130.2 and s 25 .
10
X
T
~ t9
s / 10
Pr a T a 0.9 . Here, a t5%,9 1.833
a T a
X
a
a
s / 200
as
as
X
200
200
as
as
X
X
200
200
25
25
130.2 1.833
130.2 1.833
10
10
115.709 144.691