SOME TEهنا المعادلة اكتب.
CHNIQUES FOR FINDINGINTERVAL ESTIMATORS FOR PARAMETERS
[Link] interval for mean (µ known)
Let x 1, x 2,… ., xnbe a random sample from N (µ,σ 2). Construct a (1−α )%
confidence interval for μ.
α α
Ζ=¿(x̄− μ)/(σ −√ n), p (−Ζ ≤ Ζ ≤ Ζ )¿ 1−α
2 2
Using the definition of Ζ and solving for μ, we obtain
α σ α σ
Ρ (x̄− Ζ ≤ μ ≤x̄+ Ζ
2 √n )
¿ 1−α
2 √n
[Link] interval for mean ( σ unknown)
Ζ=¿(x̄− μ)/ (s/√ n) , Ρ −Ζ ( α
2
α
≤ Ζ ≤ Ζ =1−α
2 )
Using the definition of Ζ and solving for μ, we obtain
α s α s
Ρ (x̄− Ζ ≤ μ ≤x̄+ Ζ
2 √n )
¿ 1−α
2 √n
Example: Two statistics professors want to estimate average scores for an
elementary statistics course that has two sections. Each professor teaches one
section and each section has a large number of students. A random sample of
50 scores from each section produced the following results:
a) Section1: x̄1 = 77.01, S1=10.32 b) Section 2: x̄1 =72.22, S2=11.02
Calculate 95% confidence intervals for each of these three samples.
Solution: Because n=50 is large, we could use normal approximation. For α =¿
0.05, from the normal table:
α
Ζ =Ζ 0.025 ¿ 1.96 , The confidence intervals are:
2
α S1 α S1
We have Ρ(x̄1 −Ζ 2 ≤ μ ≤ x̄1+ Ζ
2 √n)
¿ 1−α
√n
(
a) Ρ 77.01−1.96
10.32
√ 50
≤ μ≤ 77.01+1.96
10.32
√ 50
=0.9
) Which gives a 95 %
confidence interval Ρ ( 74.149≤ μ ≤79.871 )=0.95
α S2 α S2
Ρ (X̅ 1−Ζ ≤ μ ≤X̅ 2 + Ζ
2 √n)
b) We can compute ¿ 1−α
2 √n
(
Ρ 72.22−1.96
11.02
√ 50
≤ μ≤ 72.22+1.96
√50)
11.02
=0.95
Which gives the interval Ρ ( 69.165≤ μ ≤75.275 )=0.95
Example: Let population is given to be normal with standard deviation σ 3.2 =
the sample size is not large , X =73.3 and σ = 3.2. Here, n = 15, and α = 0.010.
α
thus , ᴢ 2 = z 0.05 =1.645. Hence, a 90% confidence interval for μ is given by
(
p ¿ 73.3−1.645
3.2
≤ μ ≤ 73.3+1.645
√ 15
3.2
=0.90
√15 )
P ( 71.681≤ μ ≤74.919 ) =0.90
Interpretation: We are 90% confident that the true mean μ is between
71.681 and 74.919.
[Link] interval for proportion, p
Consider a binomial distribution with parameter p. Let X be the number of
X
successes in n trials. The maximum likelihood estimator ^P of p is ^p = n . It can
be shown, using the procedure outlined at the beginning of this section, that
an approximate large sample ( 1 - α ) 100% confidence interval for p is
( √
P ^p −z α
2
^p (1−^p )
n
≤ p ≤ ^p + z α
2√^p (1− ^p )
n )=1-α
Example : let X = 20 and n = 60 , find a 95% confidence interval for the true
proportion, p.
Solution Here we need to find a 95% confidence interval for the true
20 1
proportion, p. Here ^p = 60 = 3 .
z α =z = 1.96. The confidence interval is :
0.025
2
( √
P ^p −z α
2
^p (1−^p )
n
≤ p ≤ ^p + z α
2
^p (1− ^p )
n √
=1-α )
(
p 1
−1.96
3 √ 60 3 √
( 13 )( 23 ) ≤ p ≤ 1 + 1.96 ( 13 )( 23 )
60
) = 0.95
which gives the confidence interval as p (0.21405 ≤ p ≤ 0.45262) = 0.95
Example : The scores of a random sample of 16 people who took the TOEFL
(Test of English as a Foreign Language) had a mean of 540 and a standard
deviation of 50. Construct a 95% confidence interval for the population mean μ
of the TOEFL score, assuming that the scores are normally distributed.
Solution : Because n = 16 is small, for 15 degrees of freedom, and for α = 0.05,
from the t-table ,t 0.025,15=2.131. Hence, a 95% confidence interval for μ is
p( x−t α
2
,n−1√ n
s ≤ μ≤ x +t α
2
, n−1√ n
s
)=1- α
(
p 540−2.131
50
√ 16
≤ μ ≤540+2.131
50
√ 16
= 0.95 )
p(513.36 ≤μ ≤ 566.64 ¿=0.95
5. LARGE SAM
5.1 σ1 and σ2 are known . The (1 – α) 100% large sample confidence interval
for µ1 -µ2 is given by
√ √
2 2 2 2
σ1 σ2 Z σ1 σ1 z α
P]¿=( 1−α ) ( + ¿)¿ α2 + ) X 1 −X 2( ≤ μ 1−μ2 ≤ ( − ) – )¿
n1 n2 n1 n 2 2
5.2 If σ1 and σ2 are unknown , σ1 and σ2 can be replaced by the respective
sample standard deviations S1 and S2 when ni ≥ 30 , i =1,2. Thus , we can
write
√( ) √( )
2 2 2 2
S1 S2 S1 S2
P ]( X 1−X 2 ) −z α + ≤ μ1 −μ 2 ≤ ( X 1 −X 2 ) + z σ + [ = (1 – α)
2
n1 n2 2
n1 n2
Assumptions :
EXAMPLE :