You are on page 1of 7

Chapter 9

Cases:
Case # Known
Case I normal population
σ known
Case II Large sample
σ unknown
Case III small sample
σ unknown

General Procedure:

Step 1: H 0 :θ=θ 0

H a :θ>θ0 or H a :θ<θ0 or H a :θ ≠θ 0

Step 2: identify sample statistic, θ^

Sampling distribution of θ^ : eg:

Calculation of error possiblities:

Risk1: P ( type I error )

¿ P ( H 0 reject| H 0 true ¿=α

Risk2: P ( type II error )

¿ P ( H 0 not rejected|

Rejection Region:
Step 3: Calculate test Statistic
Step 4: Decision
Compare test statistic to critical value

reject H 0 or do not reject H 0

Step 5: Conclusion in terms of the problem stated


Case I:

Step 1: H 0 :μ=μ0

H a : μ> μ0 or H a : μ< μ0 or H a : μ≠ μ0

Step 2: identify sample statistic, X́

X́−μ0
Z= N ( 0,1 )
Sampling distribution of X́ : σ
√n
Decision Rule:

α given
rejection region(RR):
z ≥ zα or z ≤ zα or ¿ or z ≥ z α / 2 ¿
Step 3: Calculate test Statistic
x́−μ0
z=
σ
√n
Step 4: Decision
Compare test statistic to critical value

reject H 0 or do not reject H 0

Step 5: Conclusion in terms of the problem stated


Case II: (large random sample, any distribution, σ unknown)

Step 1: H 0 :μ=μ0

H a : μ> μ0 or H a : μ< μ0 or H a : μ≠ μ0

Step 2: identify sample statistic, X́

Sampling distribution of X́ :

X́−μ0
Z= N ( 0,1 )
S (By CLT)
√n
Decision Rule:

α given
rejection region(RR):
z ≥ zα or z ≤ zα or ¿ or z ≥ z α / 2 ¿
Step 3: Calculate test Statistic
x́−μ0
z=
s
√n
Step 4: Decision
Compare test statistic to critical value

reject H 0 or do not reject H 0

Step 5: Conclusion in terms of the problem stated


Case III: (small random sample, normal population distribution, σ unknown)

Step 1: H 0 :μ=μ0

H a : μ> μ0 or H a : μ< μ0 or H a : μ≠ μ0

Step 2: identify sample statistic, X́

Sampling distribution of X́ :

X́−μ 0
T= t ( n−1 )
S
√n
Decision Rule:

α given
rejection region(RR):

t ≥ t α ,n−1 or t ≤−t α , n−1 or ¿or t ≥ t α ,n−1 ¿


Step 3: Calculate test Statistic
x́−μ0
t=
s
√n
Step 4: Decision
Compare test statistic to critical value

reject H 0 or do not reject H 0

Step 5: Conclusion in terms of the problem stated


Population proportion (Assumption: Large sample)

Step 1: H 0 : p= p0

H a : p > p0 or H a : p < p0 or H a : p ≠ p0

Step 2: identify sample statistic, X

Sampling distribution of ^p= X ∕ n:

^p − p0
Z= N ( 0,1 )
√ p ( 1− p ) ∕ n
0 0

Decision Rule:

α given
rejection region(RR):

z ≥ zα or z ≤ zα or ¿ or z ≥ z α / 2 ¿
Step 3: Calculate test Statistic
^p− p0
z=
√ p ( 1−p ) ∕ n
0 0

Step 4: Decision
Compare test statistic to critical value

reject H 0 or do not reject H 0

Step 5: Conclusion in terms of the problem stated

Type II error Calculations: β ( p')

p0− p' + z α √ p0 q0 ∕ n
Φ
[ √ p' q' ∕ n ]
p0 −p ' −z α √ p0 q0 ∕ n
1−Φ
[ √ p' q' ∕ n ]
p0− p' + z α √ p0 q0 ∕ n p0− p' −zα √ p 0 q 0 ∕ n
Φ
[ √ p' q' ∕ n ] [ −Φ
√ p' q' ∕ n ]
Population proportion (Assumption: n ≤ 25)

Step 1: H 0 : p= p0

H a : p > p0 or H a : p < p0 or H a : p ≠ p0

Step 2: identify sample statistic, X

Sampling distribution of X :

X BIN ( n , p 0 )

Decision Rule:

α given
rejection region(RR):

x ≥ c ∋ P ( X ≥ c ) =1−B ( c−1 ; n , p ) ≤ α
x ≤ c ∋ P ( X ≤ c ) =B ( c ; n , p ) ≤ α
( x ≤ c∨x ≥ c ) ∋[P ( X ≤ c ) + P ( X ≥ c ) ]≤ α
Step 3: Calculate test Statistic
x = number of successes
Step 4: Decision
Compare test statistic to critical value

reject H 0 or do not reject H 0

Step 5: Conclusion in terms of the problem stated

Type II error Calculations: β ( p')

P ( X < c when X BIN ( n , p' ) ) =B ( c−1 ; n , p' )

P ( X > c when X BIN ( n , p' ) ) = 1−B ( c ; n , p ' )

P ( X > c when X BIN ( n , p' ) ) =¿ !!!!!!!!!!!!!!!!!!HELP!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!


P-values

 probability
 calculated assuming H 0 is true
 decide which values of the test statistic are at least as contradictory to H 0 as value obtained
from our sample
1−ϕ ( z )


for z tests: P=

for t tests:
{ ϕ( z)
2 [ 1−ϕ (| z|) ]

You might also like