Professional Documents
Culture Documents
Lecture 50 Hypothesis Testing
Lecture 50 Hypothesis Testing
General Procedure:
Step 1: Formulation of null and alternative
hypotheses
Possible null and alternative hypotheses:
𝐻0 : 𝑝 = 𝑝
0 𝐻1 : 𝑝 ≠ 𝑝0
𝐻 0 : 𝑝 ≥ 𝑝0 𝐻1 : 𝑝 < 𝑝0
𝐻 0 : 𝑝 ≤ 𝑝0 𝐻1 : 𝑝 > 𝑝0
Step 2: Specification of level of significance
denoted by 𝛂
Step 3: Test to be used to test the null hypothesis
𝑝−𝑝0
𝑍= 𝑝 0 𝑞0
𝑛
H1 p ≠ 20
General Procedure:
Step 1: Formulation of null and alternative
hypotheses
Possible null and alternative hypotheses:
𝐻 0 : 𝑝 = 𝑝 𝐻1 : 𝑝 ≠ 𝑝
1 2 1 2
𝐻 0 : 𝑝1 ≥ 𝑝2 𝐻1 : 𝑝 1 < 𝑝2
𝐻 0 : 𝑝1 ≤ 𝑝2 𝐻1 : 𝑝 1 > 𝑝2
Step 2: Specification of level of significance
denoted by 𝛂
Step 3: Test to be used to test the null
hypothesis
𝑝1 −𝑝2
𝑍= 1 1
𝑝𝑐 𝑞𝑐 𝑛 +𝑛
1 2
𝑛1 𝑝1 +𝑛2 𝑝2
Where 𝑝 =
𝑐 𝑛1 +𝑛2
Step 4: Defining critical region / rejection region
Critical region is always based on alternative hypothesis.
If alternative hypothesis is 𝐻1 : 𝑝1 ≠ 𝑝2 the critical region
will be 𝑍 ≥ 𝑍𝛼/2
If alternative hypothesis is 𝐻1 : 𝑝1 < 𝑝2 the critical region
will be 𝑍 ≥ 𝑍𝛼
If alternative hypothesis is 𝐻1 : 𝑝1 > 𝑝2 the critical region
will be 𝑍 ≤ −𝑍𝛼
Where, 𝑍𝛼/2 𝑎𝑛𝑑 𝑍𝛼 are critical values that are calculated
using t-distribution table of critical values.
Note: critical region tells us how to reject a null hypothesis.
Step 5: Draw a random sample and using sample
information calculate the value of Z test.
Step 6: We draw our conclusion whether we accepted or
rejected our null hypothesis.
EXAMPLE 4
Consider the hypothesis test
H0: p1 – p2 = 0
H1: p1 – p2 ≠ 0