Sei sulla pagina 1di 34

FIRA FAIZA TERTIA

06111740000098

EXERCISE 8
Lee J. Bain pages 283 – 287
1. Let 𝑿 denote the weight in pounds of a bag of feed, where 𝑿 ~ 𝑵 (𝟏𝟎𝟏, 𝟒). What
is the probability that 20 bags will wight at least a ton?
Answer :
X ~ N (101, 4)
2000
X= = 100 ; 𝑛 = 20 ; 𝜇 = 101 ; 𝜎 2 = 4
2

𝑋− 𝜇
𝑧= 𝜎
⁄ 𝑛

100 − 101
= = −2,23
2⁄
√20
𝑃 [𝑍𝑛 ≥ 𝑧] = 𝑃 [𝑍𝑛 ≥ −2,23]

The Probability from the table is 0, 987

2. S denote the diameter of a shaft and B the diameter of a bearing, where S and B
are independent with 𝑺~𝑵(𝟏, 𝟎. 𝟎𝟎𝟎𝟒) and 𝑩~𝑵(𝟏. 𝟎𝟏, 𝟎. 𝟎𝟎𝟎𝟗).
a. If a shaft and a bearing are selected at random, what is the probability that
the shaft diameter will exceed the bearing diameter?
Answer:
Since both S and B are normal variables we may transform them, via theorem
8.3.1 into a new normal variable. The values we need are 𝜇 and 𝜎 both S and B.
For S we have 𝜇 = 1, 𝜎 2 = 0.0004, for B 𝜇 = 1.01, 𝜎 2 = 0.0009. For the
question, we want the probability that 𝑆 > 𝐵 so in other words Pr (𝑆 − 𝐵 > 0).
This means 𝑆 − 𝐵 = 𝑌 is a new normal variable (by theorem), with values 𝜇 =
−0.01 and 𝜎 2 = 0.0013. We will use the CLT to solve the probability, so we
need 𝜎 = 0.036. Using that we solve:

Pr(𝑆 − 𝐵 > 0) = Pr(𝑌 > 0)


𝑌 − (−0.01) 0 − (−0.01)
= Pr ( > )
0.036 0.036
0.01
= ɸ( )
0.036
≈ 0.39

b. Assume equal variances 𝝈𝟐𝟏 = 𝝈𝟐𝟐 = 𝝈𝟐, and find the value of 𝝈 that will yield a
probability of noninterference of 0.95.
Answer:
FIRA FAIZA TERTIA
06111740000098

We now assume that for S and B that 𝜎 2 are identical for each, but unknown. We
do know our desired probability, 0.95 so we will solve for that instead. Very
similar in approach to part a), we just solve for 𝜎 now. Important fact is that the
𝑁(−0.01, 𝜎 2 + 𝜎 2 ) = 𝑁(−0.01, 2𝜎 2 ) so 𝜎 = 𝜎√2
0.01
ɸ( ) = 0.95
𝜎√2
So we find the value in our table, 1.65 and solve for 𝜎

0.01
= 1.65
𝜎√2
≈ 0.00428

3. Let 𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒏 be a random sample of size n from a normal distribution.


𝑿𝒊 ~𝑵(𝝁, 𝝈𝟐 ) and define 𝑼 = ∑𝒏𝒊=𝟏 𝑿𝒊 and 𝑾 = ∑𝒏𝒊=𝟏 𝑿𝟐𝒊 .
a. Find a statistic that is a function of 𝑼 and 𝑾 and unbiased for the
parameter 𝜽 = 𝟐𝝁 − 𝟓𝝈𝟐 .

Answer:
Find for µ
𝜇 = 𝐸(𝑥̅ )
∑𝑛 𝑋
𝑖
𝜇 = ( 𝑖=1 )
𝑛
1
𝜇 = 𝐸(𝑈)
𝑛

Find for 𝜎 2
𝜎 2 = 𝐸(𝑆 2 )
∑𝑛𝑖=1 𝑋𝑖2 − 𝑛𝑥̅ 2
𝜎2 = ( )
𝑛−1
1 (∑𝑛
𝑖=1 𝑋𝑖 )
2
𝜎 2 = 𝑛−1 𝐸(𝑊 − 𝑛( ))
𝑛2
1 𝑈2
𝜎2 = 𝐸(𝑊 − )
𝑛−1 𝑛
1 𝑈2
𝜎2 = 𝐸 ( (𝑊 − ))
𝑛−1 𝑛

Thus
𝜃 = 2𝜇 − 5𝜎 2
1 1 𝑈2
𝜃 = 2( 𝐸(𝑈)) − 5𝐸 ( (𝑊 − ))
𝑛 𝑛−1 𝑛
2 5 𝑈2
𝜃 = 𝐸( (𝑈) − (𝑊 − ))
𝑛 𝑛−1 𝑛
2
2𝑈 5 𝑈
𝜃= − (𝑊 − )
𝑛 𝑛−1 𝑛
FIRA FAIZA TERTIA
06111740000098

Which is unbiased estimator for 𝜃.

b. Find a statistic that is unbiased for 𝝈𝟐 + 𝝁𝟐 .


Answer:
∑𝑛
𝑖=1 𝑋𝑖
𝜇=( )
𝑛
∑𝑛𝑖=1 𝑋𝑖 2
𝜇2 = ( )
𝑛
∑𝑛𝑖=1 𝑋𝑖 2 ∑𝑛𝑖=1 𝑋𝑖
𝜇 2 = 𝐸( ) − 𝑉𝑎𝑟( )
𝑛 𝑛
𝑛 𝑛
2
1 2
1
𝜇 = 𝐸( 2 (∑ 𝑋𝑖 ) − 2 𝑉𝑎𝑟(∑ 𝑋𝑖 )
𝑛 𝑛
𝑖=1 𝑖=1
2
𝑈 1
𝜇 2 = 𝐸 ( 2 ) − 2 𝑛𝜎 2
𝑛 𝑛
2
2
𝑈 𝜎2
𝜇 = 𝐸 ( 2) −
𝑛 𝑛

From this then,


1 𝑈2
𝜎2 = 𝐸 (𝑊 − ).
𝑛−1 𝑛

Then finally the following :


𝑈2 𝜎2
𝜎 2 + 𝜇2 = 𝜎 2 + 𝐸 ( 2 ) −
𝑛 𝑛
2 2
𝑛−1 2 𝑈2
𝜎 +𝜇 = 𝜎 + 𝐸 ( 2)
𝑛 𝑛
𝑛−1 𝑛 𝑈2 𝑈2
𝜎 2 + 𝜇2 = 𝐸 (𝑊 − ) + 𝐸 ( 2 )
𝑛 𝑛−1 𝑛 𝑛
2 2
𝑊 𝑈 𝑈
𝜎 2 + 𝜇2 = 𝐸 ( − ) + 𝐸 ( 2 )
𝑛 𝑛 𝑛
𝑊
𝜎 2 + 𝜇2 = 𝐸 ( )
𝑛
𝑊
𝜎 2 + 𝜇2 =
𝑛
Which is an unbiased estimator for 𝜎 2 + 𝜇 2.

c. Let 𝒄 be a constant, and define 𝒀𝒊 = 𝟏 if 𝑿𝒊 ≤ 𝒄 and zero otherwise. Find


a statistic that is a function of 𝒀𝟏 , 𝒀𝟐 , … , 𝒀𝒏 and also unbiased for 𝑭𝒙 (𝒄) =
𝝁−𝒄
𝝓( 𝝈 ).

Answer:
𝑃(𝑌𝑖 = 1) = 𝑃(𝑋𝑖 ≤ 𝑐)
FIRA FAIZA TERTIA
06111740000098

𝑋𝑖 − 𝜇 𝑐 − 𝜇
𝑃(𝑌𝑖 = 1) = 𝑃 ( ≤ )
𝜎 𝜎
𝑐−𝜇
𝑃(𝑌𝑖 = 1) = 𝑃 (𝑁(0,1) ≤ )
𝜎
𝑐−𝜇
𝑃(𝑌𝑖 = 1) = 𝜙 ( )
𝜎
𝑃(𝑌𝑖 = 1) = 𝐹𝑥 (𝑐)

𝐸(𝑌𝑖 ) = 1(𝑃(𝑌𝑖 = 1)) + 0(𝑃(𝑌𝑖 = 0))


𝐸(𝑌𝑖 ) = 𝑃(𝑌𝑖 = 1)
𝑐−𝜇
𝐸(𝑌𝑖 ) = 𝜙 ( )
𝜎

𝑛 𝑛
1 1
𝐸 ( ∑ 𝑌𝑖 ) = ∑ 𝐸(𝑌𝑖 )
𝑛 𝑛
𝑖=1 𝑖=1
𝑛
1 1 𝑐−𝜇
𝐸 ( ∑ 𝑌𝑖 ) = 𝑛 𝜙 ( )
𝑛 𝑛 𝜎
𝑖=1
𝑛
1 𝑐−𝜇
𝐸 ( ∑ 𝑌𝑖 ) = 𝜙 ( )
𝑛 𝜎
𝑖=1

1 𝑐−𝜇
Then 𝑛 ∑𝑛𝑖=1 𝑌𝑖 = 𝑌 which is unbiased for 𝐹𝑥 (𝑐) = 𝜙 ( ).
𝜎

4. Assume that 𝑿𝟏 and 𝑿𝟐 are independent normal random variables,


𝑿𝒊~ 𝑵(𝝁, 𝝈𝟐 ), and let 𝒀𝟏 = 𝑿𝟏 + 𝑿𝟐 and 𝒀𝟐 = 𝑿𝟏 − 𝑿𝟐. Show that 𝒀𝟏 and 𝒀𝟐
are independent and normally distributed.

Answer:
𝑋, 𝑋𝑖~ 𝑁(𝜇, 𝜎 2 ) Independent and normally distributed 𝑌 ≔
∑𝑛𝑖=1 𝑎𝑖𝑋𝑖~𝑁( ∑𝑛𝑖=1 𝑎𝑖𝜇𝑖, ∑𝑛𝑖=1 𝑎𝑖 2 𝜎𝑖 2 )

With MGF method :

1
𝑀𝑋(𝑡) = 𝑒𝑥𝑝{𝑡µ + 𝑡 2 𝜎 2 }
2
𝑛1 𝑛1
𝑡 𝑡
𝑌1 = 𝑀𝑋1+𝑋2 (𝑡) = 𝑀𝑥1 (𝑡)𝑀𝑥2 (−𝑡) = ∏ 𝑀𝑋𝑖 ( ) ∏ 𝑀𝑋𝑗 ( )
𝑛1 𝑛1
𝑖=1 𝑗=1
𝑛1 𝑛1
2
𝑡 1 𝑡 2
𝑡 1 𝑡2 2
= ∏ 𝑒𝑥𝑝 { 𝜇1 + 𝜎 } ∏ 𝑒𝑥𝑝 { 𝜇2 + 𝜎 }
𝑛1 2 𝑛1 2 1 𝑛1 2 𝑛2 2 2
𝑖=1 𝑖=1
1 𝜎1 2 𝜎2 2
= 𝑒𝑥𝑝 {𝑡(𝜇1 + 𝜇2 ) + 𝑡 2 ( + )}
2 𝑛1 𝑛2
FIRA FAIZA TERTIA
06111740000098
𝑛
𝜎1 2 𝜎2 2
𝑌1 ≔ ∑ 𝑎𝑖𝑋𝑖~𝑁(𝜇1 + 𝜇2 , + )
𝑛1 𝑛2
𝑖=1

𝑛1 𝑛1
𝑡 −𝑡
𝑌2 = 𝑀𝑋1−𝑋2 (𝑡) = 𝑀𝑥1 (𝑡)𝑀𝑥2 (−𝑡) = ∏ 𝑀𝑋𝑖 ( ) ∏ 𝑀𝑋𝑗 ( )
𝑛1 𝑛1
𝑖=1 𝑗=1
𝑛1 𝑛1
𝑡 1 𝑡2 2 −𝑡 1 𝑡2 2
= ∏ 𝑒𝑥𝑝 { 𝜇1 + 𝜎 } ∏ 𝑒𝑥𝑝 { 𝜇2 + 𝜎 }
𝑛1 2 𝑛1 2 1 𝑛1 2 𝑛2 2 2
𝑖=1 𝑖=1
2
1 𝜎1 𝜎2 2
= 𝑒𝑥𝑝 {𝑡(𝜇1 − 𝜇2 ) + 𝑡 2 ( + )}
2 𝑛1 𝑛2
𝑛
𝜎1 2 𝜎2 2
𝑌2 ≔ ∑ 𝑎𝑖𝑋𝑖~𝑁(𝜇1 − 𝜇2 , + )
𝑛1 𝑛2
𝑖=1

From the answer we know that 𝑌1 and 𝑌2 are normally distributed

5. A new component is placed in service and nine spares are available. The times to
failure in days are independent exponential variables, 𝑻𝒊 ~ 𝑬𝑿𝑷(𝟏𝟎𝟎).

(a) What is the distribution of ∑𝟏𝟎


𝒊=𝟏 𝑻𝒊 ?

(b) What is the probability that successful operation can be maintained for at
least 1.5 years? Hint: Use Theorem 8.3.3 to transform to a chi-square
variable.

(c) How many spares would be needed to be 95% sure of succesful operation
for at leats two years?

Answer:

T = waktu kegagalan (hari), 𝑇𝑖 ~ 𝐸𝑋𝑃(100) , i = 1,2,.......,10 (independent)

(a) 1 1
𝑀𝑇𝑖 = =
1 − 𝜃𝑡 1 − 100𝑡

10
𝑛
𝑀 ∑ 𝑇𝑖 (𝑡) = 𝐸(𝑒 𝑡(∑𝑖=1 𝑇𝑖) )
𝑖=1

= 𝐸(𝑒 𝑡𝑇1+𝑡𝑇2+...+𝑡𝑇𝑛 ) Because 𝑻𝒊 independent

So that
𝑀∑ 𝑇𝑖 (𝑡) = 𝐸(𝑒 𝑡𝑇1 . 𝑒 𝑡𝑇2 … . 𝑒 𝑡𝑇10 )

= 𝐸(𝑒 𝑡𝑇1 ). 𝐸(𝑒 𝑡𝑇2 ) … . 𝐸(𝑒 𝑡𝑇10 )


1 1 1
= (1−100𝑡) . (1−100𝑡) … . (1−100𝑡)
1
= (1−100𝑡 )10
FIRA FAIZA TERTIA
06111740000098

10

∑ 𝑇𝑖 ~ 𝐺𝐴𝑀(100,10) 𝑐𝑎𝑢𝑠𝑒 𝑤𝑒 ℎ𝑎𝑣𝑒 𝜃 = 100 𝑎𝑛𝑑 𝑛 = 10


𝑖=1

10 10

(b) 𝑃 (∑ 𝑇𝑖 ≥ 1.5 𝑡𝑎ℎ𝑢𝑛) = 𝑃 (∑ 𝑇𝑖 ≥ (1.5)(365))


𝑖=1 𝑖=1

From the theorem 8.3.3


2𝑌 2
If 𝑌 ~ 𝐺𝐴𝑀(𝜃, 𝜅) 𝑠𝑜 𝑡ℎ𝑎𝑡 𝑋 = 𝜃
~ 𝑋(2𝜅)

10
2𝑇𝑖 2
∑ 𝑇𝑖 ~ 𝐺𝐴𝑀(100,10) So that 100 ~ 𝑋(20)
𝑖=1

2 ∑10
𝑖=1 𝑇𝑖 2(1.5)(365)
𝑃( > )
100 100

2
𝑃 (𝑋(20) ≥ 10.95)

2
= 1 − 𝑃 (𝑋(20) ≥ 0.95)

= 1 − 0.05 = 0.95

(c) Find the n


𝑛

∑ 𝑇𝑖 ~ 𝐺𝐴𝑀(100, 𝑛)
𝑖=1

𝑃 (∑ 𝑇𝑖 ≥ (2)(365)) = 0.95
𝑖=1

2 ∑10
𝑖=1 𝑇𝑖 (2)(730)
0.95 = 𝑃 ( ≥ )
100 100

2
0.95 = 𝑃 (𝑋(2𝑛) ≥ 14.6)

2
0.95 = 1 − 𝑃 (𝑋(2𝑛) ≤ 14.6)

2
0.05 = 𝑃 (𝑋(2𝑛) ≤ 14.6)

From the table 2n = 25 so that n = 12.5

6. Repeat Exercise 5 assuming 𝑻𝒊 ~𝑮𝑨𝑴(𝟏𝟎𝟎, 𝟏. 𝟐).


Answer:
FIRA FAIZA TERTIA
06111740000098

a) What is the distribution of ∑𝟏𝟎


𝒊=𝟏 𝑻𝒊 ?
𝑘
1
𝑌~𝐺𝐴𝑀(𝜃, 𝑘) 𝑡ℎ𝑒𝑛 𝑀𝑌 (𝑡) = ( )
1 − 𝑡
1.2
1
𝑀𝑇𝑖 (𝑡) = ( )
1 − 100𝑡
10
𝑀∑10 = 𝐸 (𝑒 𝑡(∑𝑖=1 𝑇𝑖) )
𝑖=1 𝑇𝑖 (𝑡)

= 𝐸(𝑒 𝑡𝑇1 + 𝑒 𝑡𝑇2 + 𝑒 𝑡𝑇3 + ⋯ + 𝑒 𝑡𝑇10 ) 𝑀∑10 =


𝑖=1 𝑇𝑖 (𝑡)
𝐸(𝑒 𝑡𝑇1 . 𝑒 𝑡𝑇2 . 𝑒 𝑡𝑇3 . … . 𝑒 𝑡𝑇10 )

= 𝐸(𝑒 𝑡𝑇1 ). 𝐸(𝑒 𝑡𝑇2 ). 𝐸(𝑒 𝑡𝑇3 ). … . 𝐸(𝑒 𝑡𝑇10 )

1.2 1.2 1.2 1.2


1 1 1 1
=( ) .( ) .( ) .… .( )
(1 − 100𝑡) (1 − 100𝑡) (1 − 100𝑡) 1 − 100𝑡
12
1
=( )
1 − 100𝑡

𝑆𝑜 𝑡ℎ𝑎𝑡
10

∑ 𝑇𝑖 ~𝐺𝐴𝑀(100,12)
𝑖=1

b) What is the probability that successful operation can be mainted for at


least 1.5 years? Hint: Use Theorem 8.3.3 to transform to a chi-square
variable.
We search 𝑃(∑10 10
𝑖=1 𝑇𝑖 ≥ 1.5 𝑡𝑎ℎ𝑢𝑛) = 𝑃(∑𝑖=1 𝑇𝑖 ≥ (1.5)(365))

From the theorem 8.3.3


𝑋
𝑖𝑓 𝑋~𝐺𝐴𝑀(, 𝑘), 𝑡ℎ𝑒𝑛 𝑌 = 2 ~2 (2𝑘)

10
∑10
𝑖=1 𝑇𝑖
∑ 𝑇𝑖 ~𝐺𝐴𝑀(100,12) 𝑚𝑎𝑘𝑎 2 ~2 (24)
100
𝑖=1

2 ∑10
𝑖=1 𝑇𝑖 2(1.5)(365)
𝑃( ≥ )
100 100

= 𝑃(2 (24) ≥ 10.95) = 0.99


FIRA FAIZA TERTIA
06111740000098

c) How many spares would be needed to be 95% sure of successful


operation for at least two years?
𝑘

∑ 𝑇𝑖 ~𝐺𝐴𝑀(100, 𝑘)
𝑖=1

𝑃 (∑ 𝑇𝑖 ≥ (2)(365)) = 0.95
𝑖=1

∑𝑘𝑖=1 𝑇𝑖 (2)(730)
𝑃(2 ≥ = 0.95
100 100

𝑃(2 (2𝑘) ≥ 14.6) = 0.95

From the table we get 2𝑘 = 25 𝑡ℎ𝑒𝑛 𝑘 = 12.5

We need 12 spares

7. Five independent tasks are to be performed, where the time in hours to complete
the 𝒊th task is given by 𝑻𝒊 ~𝑮𝑨𝑴(𝟏𝟎𝟎, 𝒌𝒊 ), where 𝒌𝒊 = 𝟑 + 𝒊/𝟑. What is the
probability that it will take less than 2600 hours to complete all five tasks?
Answer:
Assume 𝑦 = ∑5𝑖=1 𝑇𝑖 ; 𝑇𝑖 ~𝐺𝐴𝑀(100, 𝑘𝑖 ), where 𝑘𝑖 = 3 + 𝑖/3

1 10
𝑘1 = 3 + =
3 3
2 11
𝑘2 =3+ =
3 3
3 12
𝑘3 =3+ =
3 3
4 13
𝑘4 =3+ =
3 3
5 14
𝑘5 =3+ =
3 3
5
10 + 11 + 12 + 13 + 14 60
∑ 𝑘𝑖 = = = 20
3 3
𝑖=1

𝑀𝑦 (𝑡) = 𝐸(𝑒 𝑡𝑦 )
5
= 𝐸 (𝑒 𝑡 ∑𝑖=1 𝑇𝑖 )
= 𝐸(𝑒 𝑡𝑇1+𝑡𝑇2+⋯+𝑡𝑇5 )
= 𝐸(𝑒 𝑡𝑇1 ). 𝐸(𝑒 𝑡𝑇2 ). … . 𝐸(𝑒 𝑡𝑇5 )
= 𝑀𝑇1 (𝑡). 𝑀𝑇2 (𝑡). … . 𝑀𝑇5 (𝑡)
FIRA FAIZA TERTIA
06111740000098
10 11 14
= (1 − 100𝑡)− 3 . (1 − 100𝑡)− 3 . … . (1 − 100𝑡)− 3
10 11 12 13 14
= (1 − 100𝑡)−( 3 + 3 + 3 + 3 + 3 )
5
= (1 − 100𝑡)− ∑𝑖=1 𝑘𝑖

5 5

𝑦 = ∑ 𝑇𝑖 ~𝐺𝐴𝑀(100, ∑ 𝑘𝑖 )
𝑖=1 𝑖=1

2 2600.2
𝑃(𝑦 < 2600) = 𝑃 ( 𝑦 < )
𝜃 𝜃
2 2600.2
= 𝑃 (𝜒(∑ 5 < )
𝑖=1 𝑘𝑖 .2) 100
2
= 𝑃(𝜒(40) < 52)
= 0,90

8. Suppose that 𝑿~𝝌𝟐 (𝒎) and 𝒀~𝝌𝟐 (𝒏), and 𝑿 and 𝒀 are independent.
Is 𝑿 − 𝒀~𝝌𝟐 if 𝒏 > 𝒎?
Answer:
No, because the value from 𝑋 − 𝑌 can be negative, whereas the random
variable with the distribution 𝜒 2 cant be negative.

9. Suppose that 𝑿~𝝌𝟐 (𝒎), 𝒀~𝝌𝟐 (𝒏) and X and Y are idenpendent. Use MGF’s to
show that 𝑺 − 𝑿~𝝌𝟐 (𝒏)
Answer:
𝑀𝑠 (𝑡) = 𝐸(𝑒 𝑡𝑠 )

𝑀𝑠 (𝑡) = 𝐸(𝑒 𝑡(𝑠−𝑥+𝑥) )

𝑀𝑠 (𝑡) = 𝐸(𝑒 𝑡(𝑠−𝑥) ∙ 𝑒 𝑡𝑥 )

𝑀𝑠 (𝑡) = 𝐸(𝑒 𝑡(𝑠−𝑥) ) ∙ 𝐸(𝑒 𝑡𝑥 )


(𝑚+𝑛)
We Know 𝑀𝑠 (𝑡) = (1 − 2𝑡)− 2

(𝑚+𝑛) 𝑚
(1 − 2𝑡)− 2 = 𝐸(𝑒 𝑡(𝑠−𝑥) ) ∙ (1 − 2𝑡)− 2
𝑚

(1−2𝑡) 2
𝐸(𝑒 𝑡(𝑠−𝑥) ) = (𝑚+𝑛)
− 2
(1−2𝑡)

𝑛
𝑀(𝑠−𝑥) (𝑡) = (1 − 2𝑡)− 2

Proven that 𝑆 − 𝑋~𝜒 2 (𝑛)


FIRA FAIZA TERTIA
06111740000098

10. A random sample of size n = 15 in drawn from EXP(𝜃). Find c so that P[cX < 𝜃]
= 0,95. Where X in the sample mean.
Answer:

P[c𝑥̂ < 𝜃] = 0,95

P[𝑥̂< 𝜃/c] = 0,95

𝑥̂−𝐸(𝑥̂) 𝜃/𝑐−𝐸(𝑥̂)
𝑃[ < ] = 0,95
√𝑉𝑎𝑟 (𝑥̂) √𝑉𝑎𝑟 (𝑥̂)

𝜃/𝑐−𝐸(𝑥̂)
𝑃 [𝑍 < ] = 0,95
√𝑉𝑎𝑟 (𝑥̂)

∑𝑥
Untuk 𝐸(𝑥̂) = 𝐸 ( 𝑛
)

= 1/n 𝐸 (∑ 𝑥)

= 1/n. n .x = 𝜃

∑𝑥
Untuk 𝑉𝑎𝑟(𝑥̂) = 𝑉𝑎𝑟 ( )
𝑛

= 1/n2 𝑉𝑎𝑟 (∑ 𝑥)

= 1/n2. n. σ2

= σ2/ n

= 𝜃2/n

Sehingga :

𝜃/𝑐−𝜃
𝑃 [𝑍 < 𝜎2
] = 0,95
𝑛

𝜃/𝑐−𝜃
Misal A = 2
√𝜎
𝑛

𝜃/𝑐−𝜃
A= 𝜃
√𝑛

𝜃−𝜃𝑐
𝑐
1,65 = 𝜃
√𝑛

1−𝑐
1,65 = 𝑥 √𝑛
𝑐

1−𝑐
1,65 = 𝑐
𝑥 √15

1−𝑐
0,4260 = 𝑐
FIRA FAIZA TERTIA
06111740000098

0,4260 = 1- c

1 = c + 0,4260
1 = 1,4260 c
1
C = 1,4260

C = 2,347262

11. Let 𝒁 ~ 𝑵 (𝟎, 𝟏)

a. Find 𝑷 (𝒁𝟐 < 𝟑, 𝟖𝟒 ) using tabled values of the standard normal distribution.

b. Find 𝑷 (𝒁𝟐 < 𝟑, 𝟖𝟒 ) using tabled values of the chi-square distribution.

Answer :

a. 𝑃 (𝑍2 < √3,84) = 𝑃 (𝑍2 < 1,959) = 0,9750


𝑋𝑖− µ
b. 𝑍= 𝜎
~ 𝑁 (0,1)
𝑋− 𝜇 2 2
𝑍2 = ( ) ~ 𝑋(1)
𝜎
= 𝑃 (𝑍 2 < 3,84)
𝑋𝑖− µ 2
= P [( 𝜎
) < 3,84]
2
= 0,95 where the distribution of 𝑍 2 ~ 𝑋(1)

12. The distance in feet by which a parachutist misses a target is 𝑫 = √𝑿𝟐𝟏 + 𝑿𝟐𝟐 ,
where 𝑿𝟏 and 𝑿𝟐 are independent with 𝑿𝒊 ~𝑵(𝟎, 𝟐𝟓). Find 𝑷[𝑫 ≤ 𝟏𝟐. 𝟐𝟓 𝒇𝒆𝒆𝒕]!
Answer:

𝑋𝑖 ~𝑁(0,25) 𝑎𝑛𝑑 𝐷 = √𝑋12 + 𝑋22


𝜇 = 0 , 𝜎 2 = 25
𝑋−𝜇
𝑃[𝐷 ≤ 12.25 𝑓𝑒𝑒𝑡] = 𝑃 [𝑍 ≤ ]
𝜎
12.25 − 0
= 𝑃 [𝑍 ≤ ]
√25
12.25
= 𝑃 [𝑍 ≤ ]
5

= 𝑃[𝑍 ≤ 2.45]

= 0.99
FIRA FAIZA TERTIA
06111740000098

̅ be
13. Consider independent random variables 𝒁𝒊 ~ 𝑵(𝟎, 𝟏), 𝒊 = 𝟏, … , 𝟏𝟔, and let 𝒁
the sample mean. Find :
𝟏
a. 𝑷[ ̃𝒁 < 𝟐]

Answer:

𝑍𝑖 ~𝑁(0,1)
𝑖 = 1,2,3, … . . ,16
𝜎𝑥2 1
̃𝑍 ~ 𝑁 (𝜇𝑥, ) = 𝑁 (0, )
𝑛 16
Sehingga,

1 0.5 − 0
𝑃 ( ̃𝑍 < ) = 𝑃 (𝑧 < )
2 1
4
= 𝑃(𝑧 < 2)
= 0.977

b. 𝑷[𝒁𝟏 − 𝒁𝟐 < 𝟐]
Answer:
𝑍1 − 𝑍2 ~ 𝑁(𝜇1 − 𝜇2 , 𝜎12 − 𝜎22 )
2−0
𝑃(𝑍1 − 𝑍2 < 2) = 𝑃 (𝑧 < )
√2
= 𝑃(𝑍 < √2)
= 0.921
c. 𝑷[𝒁𝟏 + 𝒁𝟐 < 𝟐]
Answer:
Berdasarkan definisi 8.3.1 diketahui 𝑎1 = 1, 𝑎2 = 1
2 2

𝑦 = 𝑍1 + 𝑍2 ~ 𝑁(∑ 𝑎𝑖 𝜇𝑖 , ∑ 𝑎𝑖2 𝜎𝑖2 )


𝑖=1 𝑖=1

𝑦 = 𝑍1 + 𝑍2 ~ 𝑁(𝑎1 𝜇1 + 𝑎2 𝜇2 , 𝑎12 𝜎12 + 𝑎22 𝜎22 )

𝑦 = 𝑍1 + 𝑍2 ~ 𝑁(1.0 + 1.0, 1.1 + 1.1)

𝑦 = 𝑍1 + 𝑍2 ~ 𝑁(0, 2)

Sehingga,
FIRA FAIZA TERTIA
06111740000098

𝑃[𝑍1 + 𝑍2 < 2] = 𝑃[𝑥 < 2]~𝑁(0, 2)

𝑋−𝜇 2−𝜇
= 𝑃( < )
𝜎 𝜎
2
= 𝑃(𝑍 < )
√2

= 0,9207

d. 𝑷[∑𝟏𝟔 𝟐
𝒊=𝟏 𝒁𝒊 < 𝟑𝟐]

Answer:
Definisi 𝑌~𝑁(0, 1) ; 𝑍𝑁(0,1)
16
(𝑍𝑖 − 𝜇)2
𝑌=∑ ~ 𝑋 2 (16)
𝜎2
𝑖=1

16
(𝑍𝑖 − 𝜇)2
= 𝑃 (∑ < 32)
𝜎2
𝑖=1

16
(𝑍𝑖 − 0)2
= 𝑃 (∑ < 32)
12
𝑖=1

16

= 𝑃 (∑ 𝑍𝑖 2 < 32)
𝑖=1

16

= 𝑃 (∑ 𝑦 < 32) ~𝑋 2 (16)


𝑖=1

= 0,990

e. 𝑷[∑𝟏𝟔 ̅ 𝟐
𝒊=𝟏(𝒁𝒊 − 𝒁) < 𝟐𝟓]

Answer:
Use theorem 8.3.6 part 3. First divide both by 𝑛 − 1 to get 𝑆 2 . Next,
multiply both sides by 𝑛 − 1 and divide by 𝜎 2 which will give a
distribution of 𝑋 2 (15).
16 ̅ 2
15 ∑ (𝑋− 𝑋 ) 15 ×25
Pr(∑16(𝑋 − 𝑋̅)2 ≤ 25) = Pr ( (15)(1) ≤ ) = Pr( 𝑋 2 (15) ≤
15 ×1
25)

Which we find to be ≈ .95


FIRA FAIZA TERTIA
06111740000098

14. If 𝑻~𝒕(𝒗) give the distribution of 𝑻𝟐


Answer:
𝑍
Given 𝑇~𝑡(𝑣) means 𝑇 = 𝑉 where 𝑍~𝑁(0,1) and 𝑉~𝜒 2 (𝑣)

𝑣

𝑍 𝑍2
From that, 𝑇 2 = ( )= 𝑉 .
𝑉
√ 𝑣
𝑣
From the Theorem, we get 𝑍 2 ~𝜒 2 (1).

𝑊
2 1
Misal 𝑊 = 𝑍 ~𝜒 (1) we get 𝑇 = ( )
2 2
𝑉
𝑣
From the theorem 8.4.4 𝑇 2 ~𝐹(1, 𝑣)
So the distribution is 𝑇 2 ~𝐹(1, 𝑣)

15. Suppose that 𝑿𝒊 ~ 𝑵 (𝝁, 𝝈𝟐 ), 𝒊 = 𝟏, … , 𝒏 and 𝒁𝒊 ~ 𝑵(𝟎, 𝟏), 𝒊 = 𝟏, … , 𝒌 and all


variables are independent. State the distribution of each of the following
variables if it is a “named” distribution or otherwise state “unknown”.

a. 𝑿𝟏 − 𝑿𝟐

Answer:

a. 𝑋𝑖 ~ 𝑁 (𝜇, 𝜎 2 ), 𝑖 = 1, … , 𝑛
𝑋1 ~ 𝑁 (𝜇, 𝜎 2 ) and 𝑋2 ~ 𝑁 (𝜇, 𝜎 2 )
Where : 𝜇1 = 𝜇2 = 𝜇
𝜎1 2 = 𝜎2 2 = 𝜎 2
Then : 𝑌 = 𝑋1 − 𝑋2
Misal : 𝑎1 = 1 (Coef 𝑋1 )
𝑎2 = −1 (Coef 𝑋2)
2 2

𝑌 ~ 𝑁 (∑ 𝑎𝑖 𝜇𝑖 , ∑ 𝑎𝑖 2 𝜎𝑖 2 )
𝑖=1 𝑖=1

𝑌 ~ 𝑁((𝑎1 𝜇1 + 𝑎2 𝜇2 ), ( 𝑎1 2 𝜎1 2 + 𝑎2 2 𝜎2 2 ))

𝑌 ~ 𝑁((1 ∙ 𝜇1 + (−1) ∙ 𝜇2 ) , (12 ∙ 𝜎1 2 + (−1)2 ∙ 𝜎2 2 ))

𝑌 ~ 𝑁( (𝜇1 − 𝜇2 ), ( 𝜎1 2 + 𝜎2 2 ))

𝑌 ~ 𝑁( (𝜇 − 𝜇 ), ( 𝜎 2 + 𝜎 2 ))

𝑌 ~ 𝑁(0 , 2𝜎 2 )

So that, 𝑋1 − 𝑋2 ~ 𝑁(0 , 2𝜎 2 )
FIRA FAIZA TERTIA
06111740000098

b. 𝑿𝟏 + 𝟐𝑿𝟐
Answer:
𝑋𝑖 ~ 𝑁 (𝜇, 𝜎 2 ), 𝑖 = 1, … , 𝑛
𝑋2 ~ 𝑁 (𝜇, 𝜎 2 ) and 𝑋3 ~ 𝑁 (𝜇, 𝜎 2 )
Where 𝜇2 = 𝜇3 = 𝜇
𝜎2 2 = 𝜎3 2 = 𝜎 2
Then : 𝑌 = 𝑋2 + 2𝑋3
Misal : 𝑎2 = 1 (Coef 𝑋2 )
𝑎3 = 2 (Coef 𝑋3)

3 3

𝑌 ~ 𝑁 (∑ 𝑎𝑖 𝜇𝑖 , ∑ 𝑎𝑖 2 𝜎𝑖 2 )
𝑖=2 𝑖=2

𝑌 ~ 𝑁((𝑎2 𝜇2 + 𝑎3 𝜇3 ), ( 𝑎2 2 𝜎2 2 + 𝑎3 2 𝜎3 2 ))

𝑌 ~ 𝑁((1 ∙ 𝜇2 + 2 ∙ 𝜇3 ) , (12 ∙ 𝜎2 2 + 22 ∙ 𝜎3 2 ))

𝑌 ~ 𝑁( (𝜇2 + 2 ∙ 𝜇2 ), ( 𝜎2 2 + 4 ∙ 𝜎3 2 ))

𝑌 ~ 𝑁( (𝜇 + 2 𝜇 ), ( 𝜎 2 + 4𝜎 2 ))

𝑌 ~ 𝑁(3𝜇 , 5𝜎 2 )

So that, 𝑋2 + 2𝑋3 ~ 𝑁(3𝜇 , 5𝜎 2 )

𝐗𝟏− 𝐗𝟐
c.
𝛔 𝐒𝐳 √𝟐

Answer:

𝑋1 ~ 𝑁 (𝜇 , 𝜎 2 )

𝑋2 ~ 𝑁 (𝜇 , 𝜎 2 )

𝑋− 𝜇
𝑍=
𝜎

(𝑋1 − 𝑋2 ) − 𝐸(𝑋1 − 𝑋2 )
𝑍=
𝜎(𝑋1 − 𝑋2 )

For 𝑬(𝑿𝟏 − 𝑿𝟐 )

𝑉𝑎𝑟(𝑋1 − 𝑋2 ) = 𝑉𝑎𝑟 (𝑋1 ) + 𝑉𝑎𝑟 (𝑋2 )

= 𝜎2 + 𝜎2
FIRA FAIZA TERTIA
06111740000098

= 2𝜎 2

𝜎(𝑋1 − 𝑋2 ) = √2𝜎 2

= 𝜎 √2

So that,

(𝑋1 − 𝑋2 ) − 0
𝑍=
𝜎 √2

(𝑋1 − 𝑋2 )
𝑍=
𝜎 √2

Form Theorem 8.3.6 we know that


2
2
∑𝑘𝑖=1(𝑋𝑖 − 𝑋̅)
𝑆 =
𝑘−1
2
∑𝑘𝑖=1(𝑍𝑖 − 𝑍̅)
𝑆𝑧 2 =
𝑘−1

(𝑘 − 1)
𝑆𝑧 2 ~ 𝑋(𝑘−1)
2
𝜎2

2
(𝑘 − 1) ∑𝑘𝑖=1(𝑍𝑖 − 𝑍̅) 2
~ 𝑋(𝑘−1)
𝜎2 𝑘−1

2
∑𝑘𝑖=1(𝑍𝑖 − 𝑍̅ ) 2
~ 𝑋(𝑘−1)
𝜎2
2
∑𝑘𝑖=1(𝑍𝑖 − 𝑍̅ ) 2
~ 𝑋(𝑘−1)
12
𝑘 2

𝑆𝑧2 = ∑(𝑍𝑖 − 𝑍̅) ~ 𝑋(𝑘−1)


2

𝑖=1

2
𝑋(𝑘−1)
𝑆𝑧 = √
𝑘−1

So that

(𝑋1 − 𝑋2 ) 1 𝑍
=
𝜎 √2 𝑆𝑧 2
𝑋(𝑘−1)

𝑘−1

𝑍
~ 𝑡 (𝑘 − 1)
𝑋2
√ (𝑘−1)
𝑘−1
FIRA FAIZA TERTIA
06111740000098

The answer is 𝑡 (𝑘 − 1)

d. 𝒁𝟐𝟏

Answer :
2
From the theorem 8.3.5 we know that if 𝑍 2 ~𝑋(1)
2
So that the 𝑍12 ~𝑋(1)

2
The answer is 𝑋(1)

̅− 𝝁 )
√𝒏 (𝑿
e.
𝑺𝟐
Answer :

Based on theorem 8.4.3,


if 𝑋1 , … , 𝑋𝑛 denotes a random sample from 𝑁(𝜇, 𝜎 2 ), then

𝑋̅ − 𝜇
~ 𝑡 (𝑛 − 1)
𝑆/√𝑛

Since 𝑋𝑖 ~ 𝑁(𝜇, 𝜎 2 ), 𝑖 = 1, … , 𝑛 and 𝑍𝑖 ~ 𝑁(0, 1) , 𝑖 = 1, … , 𝑘

√𝑛 (𝑋̅ − 𝜇 )
So, we get ~ 𝑡 (𝑘 − 1)
𝑆𝑧

f. 𝒁𝟐𝟏 + 𝒁𝟐𝟐

Answer:

Based on theorem 8.3.4,


𝑣 = ∑𝑛𝑖=1 𝑦𝑖 ~ 𝑥 2 (∑𝑛𝑖=1 𝑣𝑖 ) and 𝑍12 ~ 𝑥 2 (1) , 𝑍22 ~ 𝑥 2 (1)
so, 𝑍12 + 𝑍22 ~ 𝑥 2 (1 + 1)
𝑍12 + 𝑍22 ~ 𝑥 2 (2)

g. 𝐙𝟏𝟐 − 𝐙𝟐𝟐
Answer:
Based on theorem 8.3.4 if 𝑍𝑖 ~𝑋 2 (𝑣1 ); 𝑖, … . 𝑛 are independent chi-square variables,
then
𝑛 𝑛

𝑉 = ∑ 𝑍𝑖 ~ 𝑋 2 (∑ 𝑣1 )
𝑖=1 𝑖=1
FIRA FAIZA TERTIA
06111740000098

Where is known that 𝑋𝑖 ~ 𝑁(𝜇, 𝜎 2 ), 𝑖 = 1, … , 𝑛 and 𝑍𝑖 ~ 𝑁(0, 1) , 𝑖 = 1, … , 𝑘. So


𝑍12 ~ 𝑥 2 (1) and 𝑍22 ~ 𝑥 2 (1) and then Z12 − Z22 ~(0)

𝒁𝟏
h.
√𝐙𝟐𝟐

Answer:
Based on theorem 8.3.4 if 𝑍𝑖 ~𝑋 2 (𝑣1 ); 𝑖, … . 𝑛 are independent chi-square variables,
then
𝑛 𝑛
2
𝑉 = ∑ 𝑍𝑖 ~ 𝑋 (∑ 𝑣1 )
𝑖=1 𝑖=1
𝑍1
We get 𝑍1 ~𝑍12 and then 𝑍1 ~ 𝑥 2 (1)
and 𝑍22 ~𝑥 2 (1).
So ~(1)
√Z22

𝒛𝟐𝟏
i.
𝒛𝟐𝟐

Answer:

𝑧2
Y = 𝑧12
2

𝑥(𝑖)2 /1
Y = 𝑥(1)2 /1

Y ~ F(1,1)
𝒛𝟏
j. 𝒛𝟐

Answer:
𝑧1
identik dengan
√𝑧22

𝑧
jadi 𝑧1 identik dengan t(1)
2

̅
𝑿
k. ̅
𝒁
Answer:

𝑋̅ = µ dan 𝑍̅ = 0
Sehingga,
𝑋̅ µ
=
𝑍̅ 0
µ 𝑋̅
Karena 0 , maka 𝑍̅ Tidak diketahui distribusinya
FIRA FAIZA TERTIA
06111740000098

l.

dengan

𝑍̅ µ 𝜎2 1
dan (𝜎2 ) = = 0 , 𝑣𝑎𝑟 (𝑍̅ ) = =
𝜎2 𝑘 𝑘

dari Teorema 8.4.3 kita tahu bahwa :

sehingga,

√𝑛𝑘(𝑋̅ − µ) √𝑛(𝑋̅ − µ)
=
𝑘 2
√∑𝑖=1 𝑍𝑖
𝜎
√∑𝑘𝑖=1 𝑍𝑖2
𝑘
𝜎=1

Menurut Teorema 8.4.1 dengan 𝑇~𝑡(𝑣)

Dan

(𝑍𝑖 −0)2
Serta 𝑉1 = ∑𝑘𝑖=1 1

Maka
√𝑛𝑘(𝑋̅ − µ)
~𝑡(𝑘)
𝜎
√∑𝑘𝑖=1 𝑍𝑖2

∑𝒏
𝒊=𝟏(𝑿𝒊 − 𝝁)
𝟐
m. ̅ )𝟐
+ ∑𝒌𝒊=𝟏(𝒁𝒊 − 𝒁
𝝈𝟐

Answer:
menurut teorema 8.3.4 , maka didapat:
2 2 2
𝜒(𝑛) + 𝜒(𝑘−1) = 𝜒(𝑛+𝑘−1)
FIRA FAIZA TERTIA
06111740000098

̅
𝑿 ∑𝒌𝒊=𝟏 𝒁𝒊
n. +
𝝈𝟐 𝒌
Answer:
𝑍̅ 𝜇
𝐸(𝑥̅ ) = 𝜇 𝐸 (𝜎2 ) = =0
𝜎2

𝑥̅ 𝜇 𝜎 1 2
𝐸 (𝜎2 ) = 𝜎2 𝑣𝑎𝑟(𝑍̅) = 𝑘 = 𝑘

𝜎2
𝑣𝑎𝑟(𝑥̅ ) = 𝑛

𝑋̅ ∑𝑘
𝑖=1 𝑍𝑖 𝑋̅
2
+ = + 𝑍̅ ...............(karena 𝜎 = 1)
𝜎 𝑘 𝜎2

𝑋̅ 𝑍̅
= +
𝜎2 𝜎2

Misal:
𝑋̅
𝑦= + 𝑍̅
𝜎2

𝑥̅ 𝑍̅ 𝜇
𝐸(𝑦) = 𝐸 ( 2 ) + 𝐸 ( 2 ) = 2 + 0
𝜎 𝜎 𝜎

𝑥̅ 𝑍̅ 1 1
𝑣𝑎𝑟(𝑦) = 𝑣𝑎𝑟 ( 2
) + 𝑣𝑎𝑟 ( 2
)= 2
+
𝜎 𝜎 𝑛𝜎 𝑘
𝜇
𝑦 − 𝐸(𝑦) 2+0 𝑑 𝑦−
= 𝜎 → 𝑍 ~𝑁(0,1)
√𝑣𝑎𝑟(𝑦) √ 1 1
+
𝑛𝜎 2 𝑘
𝑋̅ 𝜇 1 1
+ 𝑍̅ ~ 𝑁 (𝜎2 , 𝑛𝜎2 + 𝑘)
𝜎2

̅𝟐
o. 𝑘𝒁

Answer:

𝑋𝑖 ~𝑁(𝜇, 𝜎 2 )

𝑋𝑖 − 𝜇
𝑍𝑖 = ~𝑁(0,1)
𝜎

2
𝑋𝑖 − 𝜇 2
𝑍 =( ) ~ 𝜒 2 (1)
𝜎

Maka 𝑘𝑍̅ 2 ~ 𝜒 2 (1)


FIRA FAIZA TERTIA
06111740000098

(𝒌−𝟏) ∑𝒏 ̅̅̅̅𝟐
𝒊=𝟏(𝑿𝒊 −𝑿)
p. 𝒌 ̅ )𝟐
(𝒏−𝟏)𝝈𝟐 ∑𝒊=𝟏(𝒁𝒊 −𝒁

Answer:

𝑋𝑖 ~𝑁(𝜇, 𝜎 2 )

𝑍𝑖 ~𝑁(0,1)
(𝑘−1) ∑𝑛 ̅̅̅̅2
𝑖=1(𝑋𝑖 −𝑋)
(𝑛−1)𝜎2 ∑𝑘 ̅ 2
𝑖=1(𝑍𝑖 −𝑍 )

∑𝑛 ̅̅̅̅2
𝑖=1(𝑋𝑖 −𝑋)
𝑆2 = (𝑛−1)

̅̅̅2
∑𝑛𝑖=1(𝑋𝑖 − 𝑋)
= (𝑛 − 1) ~𝜒 2 (𝑛 − 1)
(𝑛 − 1)
𝜎2
∑𝑛 ̅̅̅̅
𝑖=1(𝑋𝑖 −𝑋)
= ~𝜒 2 (𝑛 − 1)
𝜎2

So,
(𝑘−1) ∑𝑛 ̅̅̅̅2
𝑖=1(𝑋𝑖 −𝑋) (𝑘−1)𝜒2 (𝑛−1)
𝑘
(𝑛−1)𝜎 ∑𝑖=1(𝑍𝑖 −𝑍̅)2
2
= (𝑛−1) ∑𝑘 ̅ 2
𝑖=1(𝑍𝑖 −𝑍 )

∑𝑘𝑖=1(𝑍𝑖 − 𝑍̅)2

∑𝑘 ̅ 2
𝑖=1(𝑍𝑖 −𝑍 )
𝑆2 = 𝑛−1

∑𝑘 ̅ 2
𝑖=1(𝑍𝑖 −𝑍)
= (𝑛 − 1) 𝑛−1 ~𝜒 2 (𝑘 − 1)
12

∑𝑘 ̅ 2
𝑖=1(𝑍𝑖 −𝑍 )
= = ∑𝑘𝑖=1(𝑍𝑖 − 𝑍̅)2 (𝑃𝑟𝑜𝑣𝑒𝑛)
1

Thus :
(𝑘−1)𝜒2 (𝑛−1)
= (𝑛−1) ∑𝑘 ̅ 2
𝑖=1(𝑍𝑖 −𝑍)

(𝑘 − 1)𝜒 2 (𝑛 − 1)
=
(𝑛 − 1)𝜒 2 (𝑘 − 1)

𝜒 2 (𝑛 − 1)
= 2𝑛 − 1
𝜒 (𝑘 − 1)
𝑘−1
Suppose : 𝑉1 ~𝜒 2 (𝑛 − 1) , 𝑉2 ~𝜒 2 (𝑘 − 1)
FIRA FAIZA TERTIA
06111740000098

with 𝑣1 = 𝑛 − 1 , 𝑣2 = 𝑘 − 1
𝜒2 (𝑛−1)
𝑛−1
let : 𝑋 = 𝜒2 (𝑘−1)
𝑘−1

𝑉1
𝑣1
𝑋= 𝑉2 ~𝐹(𝑣1 , 𝑣2 )
𝑣2

~𝐹((𝑛 − 1), (𝑘 − 1))

16. Let X1, X2, … , X9 be a random sample from a normal distribution Xi ~ N(6, 25),
and denote by 𝑿̅ dan S2 the sample mean and sample variance. Use table from
Appendix C to find each of the following:
̅<7]
a. P [ 3 < 𝑿
Answer:
𝑋̅ − 𝜇
P [ 3 < 𝑋̅ < 7 ], ingat dimana Z = 𝜎 . Sehingga
⁄ 𝑛

P [ 3 < 𝑋̅ < 7 ]
3−𝜇 𝑋̅ − 𝜇
7−𝜇
=P[𝜎 <𝜎 <𝜎 ]
⁄ 𝑛 ⁄ 𝑛 ⁄ 𝑛
√ √ √
3−6 7−6
=P[ 5⁄ <𝑍< 5 ]

√9 √9
= P [-1,8 < Z < 0,6 ]
= P [ Z < 0,6 ] – P [ Z < -1,8 ]
= (0,5 + 0,2257) – (0,5 – 0,4641)
= 0,6898

̅ – 6)/S ]
b. P [ 1.860 < 3(𝑿
Answer:
P [ 1.860 < 3(𝑋̅ – 6)/S ]
𝑋̅ −𝜇 𝑋̅ −6
Bawa 3(𝑋̅ – 6)/S ke bentuk 𝑆 ~ t(n – 1), sehingga menjadi 𝑆⁄ ~ t(8),
⁄ 3
√𝑛
Sehingga
P [ 1.860 < 3(𝑋̅ – 6)/S ]
𝑋̅ −6
= P [ 1.860 < 𝑆⁄ ]
3
= P [ 1.860 < T ], dengan v = 8
= 0,9

c. P [𝑺𝟐 ≤ 31, 9375]


Answer:
(𝑛−1)𝑠2 (𝑥𝑖 − 𝑥̅ )2
We know that V = 𝜎2
~ 𝜆(𝑛−1) , from which S2 = ∑𝑛𝑖=1 𝑛−1
.

Therefore
FIRA FAIZA TERTIA
06111740000098

(𝑛−1)𝑠2 (𝑛−1)(31.9365) (𝑛−1)(31.9365)


P[ ≤ ] = P[𝑉 ≤ ] ~ 𝜆(8) 2
𝜎2 252 252

= P[𝑉 ≤ 10.22] ~ 𝜆(8) 2

P = 0,75

17. Use table values from Appendix C to find the following:


a. 𝑷[𝟕. 𝟐𝟔 < 𝒀 < 𝟐𝟐. 𝟑𝟏]𝒊𝒇 𝒀 ~ 𝝌𝟐 (𝟏𝟓)
Answer:
𝑃[7.26 < 𝑌 < 22.31] = 𝑃[𝑌 < 22.31] − 𝑃[𝑌 < 7.26]
= 𝑃[𝜒 2 (15) < 22.31] − 𝑃[𝜒 2 (15) < 7.26]
From appendix C, we can conclude that:
= 0.90 – 0.05
= 0.85
The answer is 0.85
b. The value 𝒃 such that 𝑷[𝒀 ≤ 𝒃] = 𝟎. 𝟕𝟓 𝒊𝒇 𝒀 ~ 𝝌𝟐 (𝟏𝟓)
Answer:

𝑃[𝑌 ≤ 𝑏] = 0.75

𝑃[𝜒 2 (15) < 𝑏] = 0.75

From appendix c, we can conclude that the value of 𝑏 is 27.14

The answer is 27.14

𝒀 𝟏𝟏
c. P (𝟏+𝒀 > 𝟏𝟔) if 𝒀 ~ 𝑿𝟐 (𝟔)
Answer:
𝑌 11 𝑌+1 16
Pr (1+𝑌 > 16) = Pr ( < 11)
𝑌

1 11
= Pr (𝑌 < 16 − 1)

11
= Pr (𝑌 ≥ )
5

= 1 – 0,1 = 0,90

d. P [ 𝟎, 𝟖𝟕 < 𝑻 < 𝟐, 𝟔𝟓] if 𝑻~ 𝒕(𝟏𝟑)


Answer:

T(2,65) – T(0,87) = 0,99 – 0,8 = 0,19


FIRA FAIZA TERTIA
06111740000098

e. The value b such that P[T < b] = 0.60 if T ~ t(26)


Answer:
P[T < b] = 0.60 if T ~ t(26)
From that we know that v = 26 and 𝛾 = 0.60
Based on the table of distribution T with v = 26 and 𝛾 = 0.60,
we get b = 0.256

f. The value c such that P[|T| ≥ c] = 0.02 if T ~ t(23)


Answer:
P[|T| ≥ c] = 0.02 if T ~ t(23)
From that we know that v=23

𝑇≥𝑐, 𝑇≥0
|T| ≥ c {
𝑇 ≤ −𝑐 , 𝑇 < 0 (𝑛𝑜𝑡 𝑒𝑥𝑖𝑠𝑡)

P[T ≥ c] = 0.02

1- P[T ≥ c] = 1-0.02

P[T < c] = 0.98

So 𝛾=0.98 ≈0.975

Based on the table of distribution T with v = 23 and 𝛾 = 0.975,


we get c = 2.069
g. 𝑷[𝟐, 𝟗𝟏 < 𝑿 < 𝟓, 𝟓𝟐] if 𝑿~𝑭(𝟕, 𝟏𝟐)
Answer:
𝑃[2,91 < 𝑋 < 5,52] if 𝑋~𝐹(7,12)
𝑃[2,91 < 𝑋 < 5,52] = 𝑃[𝑋 < 5,52] − 𝑃[𝑋 < 2,91]
= 0,995 − 0,95 = 0,045

𝟏
h. 𝑷 [𝑿 > 𝟎, 𝟐𝟓] if 𝑿~𝑭(𝟐𝟎, 𝟖)
Answer:
1
𝑃 [ > 0,25] if 𝑋~𝐹(20,8)
𝑋
1 1
𝑃 [ > 0,25] = 𝑃 [ > 𝑋]
𝑋 0,25
= 𝑃[𝑋 < 4]
= 0,975

18. Assume that Z, V1, and V2 are independent random variabel with Z ~ N(0,1), V1 ~
χ2(5), and V2 ~ χ2(9). Find the following :

a. P [ V1 + V2 < 8,6 ]
Answer:
P [ V1 + V2 < 8,6 ]

Menurut sifat distribusi chi-square χ2(m + n) = χ2(m) + χ2(n) sehingga


FIRA FAIZA TERTIA
06111740000098

P [ V1 + V2 < 8,6 ] = P [χ2(5) + χ2(9) < 8,6 ]


= P [χ2(14) < 8,6 ]
P [ V1 + V2 < 8,6 ] = 0,9

b. P [ Z/√𝐕𝟏/𝟓 ]
Answer:
P [ Z/√V1/5 ] = P [ T < 2,015 ] , T ~ t(5) (diasumsikan distribusi t)
= P [t(5) < 2,015 ] (dicari melalui tabel distribusi t)
= 0,95


c. P Z  0.611 V2 
Answer:


P Z  0.611 V2 
 Z 
= P  0.611
 V 
 2 
 Z 
= P 3   3  0.611
 V2 
 
 Z 
= P  1.833
 V 9 
 2 
= P T  1.833
 (9) 

= 1 − P T  1.833
 (9) 
= 1 − 0.95
= 0.05

d. PV1 V2  1.450
Answer:
PV1 V2  1.450
V1 ~  2 (v1 ) V1 v1
→ X = ~ F (v1 , v2 )
V2 ~  (v 2 )
2
V2 v2

Karena V1 ~  2 (5) , dan V2 ~  2 (9)

V1 5 9V1
Y= = ~ F (5,9)
V2 9 5V2

Sehingga
FIRA FAIZA TERTIA
06111740000098


P V1 V2  1.450 
 9V 9 
= P  1   1.450
 5V2 5 
= PY  2.61
= 0.90

𝑉1 𝑉1 +𝑉2 1
e. Pr (𝑉 +𝑉 ≤ 𝑏) = Pr ( 𝑉1
≥ 𝑏)
1 2

𝑉2 1
= Pr (1 + ≥ )
𝑉1 𝑏

𝑉2 1
= Pr ( ≥ − 1)
𝑉1 𝑏
5 𝑉2 5 1
= Pr ( ≥ ( − 1))
9 𝑉1 9 𝑏
5 1
= Pr (𝐹(9, 5) ≥ ( − 1))
9 𝑏
We now find our value for an 𝐹(9, 5) = 0.90 which turns out to be 2.61. We must
take the inverse of this since we are ≥ and our table is in format for ≤. We now set
that equal to our percentile and solve for 𝑏.
1 5 1
= ( ( − 1))
2.61 9 𝑏
9 1 1
+1=
5 2.61 𝑏
𝑏 ≈ 5.92

19. if T~t(1), then show the following:


𝟏 𝟏
a. The CDF of T is 𝑭(𝒕) = 𝟐 + 𝝅 𝒂𝒓𝒄𝒕𝒂𝒏𝒕(𝒕)
Answer:
The provided detail is:

𝑇~𝑡(1)

Therefore,

Γ(1)
𝑓𝑟 (𝑡) = (1 + 𝑡 2 )−1
1
Γ (2) √𝜋

1
= (1 + 𝑡 2 )−1
√𝜋 × √𝜋
1
=
𝜋(1 + 𝑡 2 )
FIRA FAIZA TERTIA
06111740000098
1 1
The CDF of T is 𝐹(𝑡) = 2 + 𝜋 𝑎𝑟𝑐𝑡𝑎𝑛𝑡(𝑡)

𝐹(𝑡) = ∫ 𝑓𝑟 (𝑡)𝑑𝑡
1
𝐹(𝑡) = ∫ 𝑑𝑡
𝜋(1 + 𝑡 2 )
1 1
= + 𝑎𝑟𝑐𝑡𝑎𝑛𝑡(𝑡)
2 𝜋

𝟏
b. The 100 × 𝜸𝒕𝒉 percentile is 𝒕𝜸(𝟏) = 𝐭𝐚𝐧[𝝅 (𝜸 − 𝟐)]
Answer:
1
The 100 × 𝛾𝑡ℎ percentile is 𝑡𝛾(1) = tan[𝜋 (𝛾 − 2)]
𝑡𝛾 (𝑣)

𝛾 = ∫ 𝑓(𝑡; 𝑣)𝑑𝑡
−∞
𝑡𝛾 (1)
1
𝛾= ∫ 𝑑𝑡
𝜋(1 + 𝑡 2 )
−∞
1 1 (1)
𝛾= + 𝑎𝑟𝑐𝑡𝑎𝑛𝑡(𝑡) |𝑡𝛾
2 𝜋 −∞
1
𝑡𝛾(1) = tan[𝜋 (𝛾 − )]
2

20. Show that if 𝑿~𝑭(𝟐, 𝟐𝒃), then


𝒙 −𝒃
a. 𝑷[𝑿 > 𝒙] = (𝟏 + 𝒃) ,for all x > 0

Answer:
2 + 2𝑏 2+2𝑏
Γ( ) 2𝑏 2⁄2 2 −1 2 −(
𝑏
)
𝑓(𝑥, 𝑣1 , 𝑣2 ) = 𝑏 ( ) 𝑥 ( ⁄2) (1 + 𝑥)
2 2𝑏
Γ (2) Γ ( 2 ) 2 2𝑏

Γ(1+𝑏) 1 1 −(1+𝑏)
= Γ(1)Γ(𝑏) 𝑏 (1 + 𝑏 𝑥)

Γ(1+𝑏) 1 𝑥 1 −(1+𝑏)
𝑃[𝑋 > 𝑥] = Γ(1)Γ(𝑏) 𝑏 ∫0 (1 + 𝑏 𝑥) 𝑑𝑢

𝑥
b Γ(𝑏) 1 1 𝑢 −𝑏
= 1 Γ(𝑏) 𝑏 [𝑏 (1 + 𝑏 ) 𝑏]
0

𝑥 −𝑏
= (1 + 𝑏)

𝟏
b. The 100 x 𝜸th percentile is 𝒇𝜸 (𝟐, 𝟐𝒃) = 𝒃[(𝟏 − 𝜸)−𝒃 − 𝟏]
Answer:
FIRA FAIZA TERTIA
06111740000098

𝑥 −𝑏
(1 + ) = (1 − 𝛾)
𝑏
𝑥 1⁄
1+𝑏 = (1 − 𝛾)− 𝑏

1
𝑥 = 𝑏[(1 − 𝛾)−𝑏 − 1]

21. Show that if 𝑭(𝒙; 𝝁) is the CDF of 𝑿 ~𝑷𝑶𝑰(𝝁) and if H(𝒚; 𝒗) is the CDF of a
chi-square distribution with v degrees of freedom, then 𝑭(𝒙; 𝝁) = 𝟏 −
𝑯[𝟐𝝁; 𝟐(𝒙 + 𝟏)]. Hint: Use Theorem 3.3.2 and the fact that 𝒀 ~𝑿𝟐 (𝒗)
𝒗
corresponds to 𝒀 ~𝑮𝑨𝑴 (𝟐, 𝟐)
Answer:

Teorema 3.3.2

If 𝑋 ~𝐺𝐴𝑀(𝜃, 𝑛), where n is a positive integer, then the CDF can be written
𝑛−1
(x/θ)i −x/θ
𝐹(𝑥; 𝜃, 𝑛) = 1 − ∑ e
i!
𝑖=0

Akan dibuktikan 𝐹(𝑥; 𝜇) = 1 − 𝐻[2𝜇; 2(𝑥 + 1)]

misal μ= x/θ, k= x+1


𝑛−1
(μ)i −μ
𝐹(𝑥; 𝜇) = 1 − ∑ e
i!
𝑖=0

CDF of 𝑋 ~𝑃𝑂𝐼(𝜇)
𝑘
μi e−μ
𝐹(𝑥; 𝜇) = ∑
i!
𝑖=0

And fact that 𝑌 ~𝑋 2 (𝑣) corresponds to 𝑌 ~𝐺𝐴𝑀(2, 𝑣/2)

2𝑥
𝐹𝑥 (𝑥) = 𝐻( ; 2𝑘)
𝜃
Maka

𝐹(𝑥; 𝜇) = 1 − 𝐹𝑥 (𝑥)

2𝑥
𝐹(𝑥; 𝜇) = 1 − 𝐻( ; 2𝑘)
𝜃
FIRA FAIZA TERTIA
06111740000098

𝐹(𝑥; 𝜇) = 1 − 𝐻[(2𝜇; 2(𝑘 + 1)]

22. If X BETA(p, q), derive E(Xn).

Answer:
1
Pdf : 𝑓(𝑥) = 𝐵(𝑝 , 𝑥 𝑝−1 (1 − 𝑥)𝑞−1
𝑞)

Where p > 0 and q > 0 and the function of B (p , q) is


1
B(𝑝 , 𝑞) = ∫ 𝑥 𝑝−1 (1 − 𝑥)𝑞−1 𝑑𝑥
0


𝑛)
E(𝑋 = ∫ 𝑥 𝑛 𝑓(𝑥) 𝑑𝑥
−∞

∞ 1
= ∫−∞ 𝑥 𝑛 . 𝑥 𝑝−1 (1 − 𝑥)𝑞−1 𝑑𝑥
𝐵(𝑝 , 𝑞)

1 𝑞
=
𝐵(𝑝 ,
∫ 𝑥 𝑛+𝑝−1 (1 − 𝑥)𝑏−1 𝑑𝑥
𝑞) 𝑝

1
= 𝐵((𝑛 + 𝑝), 𝑞)
𝐵(𝑝 , 𝑞)
Γ(𝑝+𝑞) Γ(𝑛+𝑝)Γ(𝑞)
=
Γ(𝑝)Γ(𝑞) Γ((𝑛+𝑝)+𝑞)

Γ(𝑝+𝑞)Γ(𝑛+𝑝)
=
Γ(𝑝)Γ(𝑛+𝑝+𝑞)

23. Consider a random sample from a beta distribution, Xi BETA(1,2). Use the CLT
(Theorem 7.3.2) to approximate P[ X ≤ 0.5 ] for n = 12
Answer:

Xi BETA (1,2)

a = 1 dan b = 2
𝑎 1 1
E(Xi) = 𝑎+𝑏 = 1+2
= 3

𝑎𝑏 2𝑋1 1
Var (Xi) = (𝑎+𝑏+1)(𝑎+𝑏)2 = (2+1+1)(2+1)2 = 18

𝑋𝑖− 𝜇
P[ Xi ≤ 0.5 ] = [𝑍𝑛 ≤ 𝜎/√𝑛
]
FIRA FAIZA TERTIA
06111740000098

1
0.5 −
3
= [𝑍𝑛 ≤ 1
]
√ /√12
18

0.5 − 3
= [𝑍𝑛 ≤ 1 1
]
√ /2√3
3 2

1
6
= [𝑍𝑛 ≤ 1 1
]
√ /2√3
3 2

= [𝑍𝑛 ≤ 2.45]

= 0.9929

𝒀𝒏 − 𝒏
24. Let 𝒀𝒏 ~𝑿𝟐 (𝒏). Find the limiting distribution of ⁄ as 𝒏 → ∞, using
√𝟐𝒏
moment generating function
Answer:

𝑌𝑛 ~𝑋 2 (𝑛)

𝑀𝑌𝑛 = (1 − 2𝑡)−𝑛⁄2
𝑡
𝑀(𝑌𝑛 −𝑛) (𝑡) = 𝑀(𝑌𝑛 − 𝑛)
√2𝑛
𝑡
(𝑌𝑛 −𝑛)
= 𝐸 (𝑒 √2𝑛 )

𝑡 𝑛𝑡
𝑌𝑛 −
= 𝐸 (𝑒 √2𝑛 𝑒 √2𝑛 )

𝑡 𝑛𝑡
𝑌𝑛 −
= 𝐸 (𝑒 √2𝑛 ) 𝐸 (𝑒 √2𝑛 )

𝑛𝑡 𝑡
− 𝑌𝑛
=𝑒 √2𝑛 𝐸 (𝑒 √2𝑛 )

𝑡√𝑛
− 𝑡
=𝑒 √2 𝑀𝑌𝑛 ( )
√2𝑛


𝑛𝑡
𝑡 −𝑛⁄2
=𝑒 √2𝑛 (1 − 2 )
√2𝑛
FIRA FAIZA TERTIA
06111740000098


𝑛𝑡
2𝑡 −𝑛⁄2
=𝑒 √2𝑛 (1 − )
√2𝑛

−𝑛⁄2

𝑡 √2
2𝑡 −𝑛⁄2
= (𝑒 √𝑛 ) (1 − )
√2𝑛

𝑡√2 𝑡√2
−𝑛⁄2
2𝑡
= (𝑒 √𝑛 − 𝑒 √𝑛 )
√2𝑛

𝑡√2 𝑡√2
−𝑛⁄2
𝑡 √2
= (𝑒 √𝑛 − 𝑒 √𝑛 )
√𝑛

We know that

𝑡√2 2
𝑡 √2 1 𝑡 √2
𝑒 √𝑛 = 1 + + ( ) +⋯
√𝑛 2 √ 𝑛

So the equation become :

−𝑛⁄2
𝑡2 𝑑
𝑀(𝑌𝑛 −𝑛) (𝑡) = (1 − + )
𝑛 𝑛

Where

√2𝑡 3 𝑒 𝜆(𝑛) √2𝑡 3 2𝑡 4 𝑒 𝜆(𝑛)


𝑑= − −
3√ 𝑛 √𝑛 3𝑛

So

−𝑛⁄2 𝑡 2⁄
𝑡2 𝑑
lim (1 − + ) =𝑒 2
𝑛 𝑛

25. Rework Exercise 5(b) and (c) using normal approximation, and compare to the
exact results.
Answer:
𝑇𝑖 ~𝐸𝑋𝑃(100)
1
MGF from the exponential distribution : 1−𝜃𝑡
10

𝑌 = ∑ 𝑇𝑖
𝑖=1

𝑀𝑌 (𝑡) = 𝐸(𝑒 𝑡𝑌 )
FIRA FAIZA TERTIA
06111740000098
10
= 𝐸 (𝑒 𝑡(∑𝑖=1 𝑇𝑖 ) )

= 𝐸(𝑒 𝑡(𝑇1+𝑇2+⋯+𝑇10 ) )
= 𝐸(𝑒 𝑡𝑇1+𝑡𝑇2+⋯+𝑡𝑇10 )
= 𝐸(𝑒 𝑡𝑇1 . 𝑒 𝑡𝑇2 … 𝑒 𝑡𝑇10 ) cause independent
= 𝐸(𝑒 𝑡𝑇1 ). 𝐸(𝑒 𝑡𝑇2 ) … 𝐸(𝑒 𝑡𝑇10 )
= 𝑀𝑇1 (𝑡) … 𝑀𝑇10 (𝑡)
1 1 1
= . …
1 − 𝜃𝑡 1 − 𝜃𝑡 1 − 𝜃𝑡
1
= ~𝐺𝐴𝑀(100,10)
(1 − 𝜃𝑡)10
from 𝑌~𝐺𝐴𝑀(100,10) we get
𝜇 = 𝐸(𝑌) = 𝑘𝜃 = 10.100
𝑉𝑎𝑟(𝑌) = 𝑘𝜃 2 = 10. 1002

With normal distribution


𝑃[𝑌 ≥ 1,5 𝑦𝑒𝑎𝑟𝑠] = 1 − 𝑃[𝑌 < 1,5 𝑦𝑒𝑎𝑟𝑠]
= 1 − 𝑃[𝑌 < 547,5]
𝑌 − 𝜇 547,5 − 𝜇
= 1−𝑃[ < ]
𝜎 𝜎
547,5 − 10.100
= 1 − 𝑃 [𝑍 < ]
√10. 1002
= 1 − 𝑃[𝑍 < −1,4309]
= 1 − 0,0764
𝑃[𝑌 ≥ 1,5 𝑦𝑒𝑎𝑟𝑠] = 0,9236
We know that P = 0,9236.

𝑌−𝜇 770−100𝑛
c. 𝑃[𝑌 > 770] = 𝑃 [ > ]
𝜎 100√𝑛

7,7 − 𝑛
= 𝑃 [𝑍 > ]
√𝑛
Dengan menggunakan tabel distribusi normal, Z yang mempunyai peluang 95%,
maka didapat:
7,7 − 𝑛
= 1,645
√𝑛
7,7 − 𝑛 = 1,645√𝑛
FIRA FAIZA TERTIA
06111740000098

(7,7 − 𝑛)2 = 2,71𝑛


𝑛 ≈ 13
We get 𝑛 = 13

26. Let 𝑿𝟏 , 𝑿𝟐 , … , 𝑿𝒏 be a random sample from a distribution whose first four


𝑷
̅ )𝟐 /(𝒏 − 𝟏) Show that 𝑺𝟐𝒏 → 𝝈𝟐 as n → ∞.
moments exist, and let 𝑺𝟐𝒏 = ∑𝒏𝒊=𝟏(𝑿𝒊 − 𝑿
Hint: Use Theorem 8.2.2 and the Chebychev inequality.
Answer:
∑𝑛 ̅ 2
𝑖=1(𝑋𝑖 −𝑋 )
𝑆𝑛2 = (𝑛−1)

𝐸(𝑆 2 ) = 𝜎 2
𝑛−3 4
(𝜇 4 − 𝜎 )
𝑉𝑎𝑟 (𝑆 2 ) = 𝑛−1
;n>1
𝑛
1 𝑛−3
= 𝑛 (𝜇 4 − 𝑛−1 𝜎 4 )

𝜇 4 = 𝐸(𝑆 2 − 𝜎 2 )4
1
𝑃[|𝑆 2 − 𝜎 2 | < 𝑘𝜎] ≥ 1 − 𝑘 2
𝑛−3 4
√𝜇 4 − 𝜎
𝑛−1
𝑘𝜎 = 𝑘 =𝜀
𝑛
𝜀 √𝑛
𝑘= 𝑛−3 4
√𝜇 4 − 𝜎
𝑛−1

𝜀2 𝑛
𝑘2 = 𝑛−3 4
𝜇4 − 𝜎
𝑛−1
𝑛−3 4
(𝜇 4 − 𝜎 )
𝑃[|𝑆 2 − 𝜎 2 | < 𝜀] ≥ 1 − 𝑛−1
2
𝜀 𝑛
𝑛−3 4
(𝜇 4 − 𝜎 )
2 2| 𝑛−1
lim 𝑃[|𝑆 − 𝜎 < 𝜀] = lim 1 −
n→∞ n→∞ 𝜀2 𝑛
2 2|
lim 𝑃[|𝑆 − 𝜎 < 𝜀] = 1
n→∞

27. Compare the Wilson-Hilferty approximation (Equation 8.5.2) to exact tabled values
of 𝝌𝟐𝟎.𝟗𝟓 (𝟏𝟎) and 𝝌𝟐𝟎.𝟎𝟓 (𝟏𝟎).

Answer:
Given Wilson-Hilferty approximation
FIRA FAIZA TERTIA
06111740000098

2 2
𝜒𝛾2 (𝑣) = 𝑣[1 − + 𝑧𝛾 √ ]3
9𝑣 9𝑣

Based of the formula, so

2 (10)
2 2
𝜒0.95 = 10[1 − + 1.645√ ]3
90 90

2 (10)
𝜒0.95 = 18.307

And

2 (10)
2 2
𝜒0.05 = 10[1 − − 1.645√ ]3
90 90

2 (10)
𝜒0.05 = 3.940

Conclution :
2 (10)
The Wilson-Hilferty Approximation 𝜒0.95 is equal to the Table Chi-Square
if we use degrees of value is 10 and area to the right of critical value is 0,05. The
2 (10)
value is 18.307. And The Wilson-Hilferty Approximation 𝜒0.05 is equal to
the Table Chi-Square if we use degrees of value is 10 and area to the right of
critical value is 0.95. the value is 3.940.

Potrebbero piacerti anche