Scribd
Upload
143 views

Lesson 4 - Testing A Population Proportion

This document discusses hypothesis testing for population proportions. It provides a formula for testing a population proportion which involves calculating a test statistic (Zc) and comparing it to critical values based on the hypothesized proportion (p0) and alternative hypothesis. Two examples are given to demonstrate applying the steps of hypothesis testing using this formula to determine if there is sufficient evidence to reject the null hypothesis.

Uploaded by

Yolanda Descallar
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
143 views

Lesson 4 - Testing A Population Proportion

This document discusses hypothesis testing for population proportions. It provides a formula for testing a population proportion which involves calculating a test statistic (Zc) and comparing it to critical values based on the hypothesized proportion (p0) and alternative hypothesis. Two examples are given to demonstrate applying the steps of hypothesis testing using this formula to determine if there is sufficient evidence to reject the null hypothesis.

Uploaded by

Yolanda Descallar
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Hypothesis Testing

Lesson 4: Testing a Population Proportion


Tests of hypotheses concerning proportions are required in many areas. Like for instance, the politician is
certainly interested in knowing what fraction of the voters will favor him in the next election; or the gambler
depends upon knowledge of the proportion of outcomes that he considers favorable.
Hypothesis testing that involves proportion can be considered as a binomial experiment when there are
only two outcomes and the probability of success does not change from trial to trial. When np ≥ 5 and nq ≥ 5, the
normal distribution is used to approximate the binomial distribution and so the standard normal distribution can
be used to test hypotheses for proportions.
Table below provides a formula in testing a population proportion.
FORMULA FOR TESTING A POPULATION PROPORTION
Test Statistic 𝑯𝟎 𝑯𝒂 Rejection Region
𝑝̂ − 𝑝0 𝑃 ≤ 𝑝0 𝑃 > 𝑝0 𝑍𝑐 > 𝑍𝛼
𝑍𝑐 =
𝑝 𝑞 𝑃 ≥ 𝑝0 𝑃 < 𝑝0 𝑍𝑐 < −𝑍𝛼
√ 0 0
𝑛
where
𝑥
𝑝̂ = 𝑃 = 𝑝0 𝑃 ≠ 𝑝0 𝑍𝑐 < −𝑍𝛼 or 𝑍𝑐 > 𝑍𝛼
𝑛 2 2
𝑞𝑜 = 1 − 𝑝𝑜

Example 1. The president of a certain firm, concerned about the safety record of the firm’s employees, sets aside
PhP 500,000 a year for safety education. The firm’s accountant believes that more than 75% of similar firms
spend more than PhP 500,000 a year on safety education. When the president asks the accountant for evidence to
support this belief, the accountant responds with the following hypothesis test. The accountant decides to obtain
information from random sample of 60 firms, 50 of which state that they spend more than PhP 500,000.00 on
safety education. Is there reason to believe the accountant’s claim? Use 𝛼 = 0.05.
Given:
𝑝𝑜 = 75% 𝑜𝑟 0.75,
𝑥 = 50,
𝑛 = 60,
𝑥 50
𝑝̂ = = = 0.83,
𝑛 60
𝑞𝑜 = 1 − 𝑝𝑜 = 1 − 0.75 = 0.25.

Solution: Following the steps in hypothesis testing, we have:


Steps:
1. Hypotheses:
𝐻0 : 𝒑 ≤ 𝟎. 𝟕𝟓
𝐻𝑎 : The firm’s accountant believes that more than 75% of similar firms spend more than PhP 500,000
a year on safety education, that is, 𝒑 > 𝟎. 𝟕𝟓.
2. Significance Level: 𝜶 = 𝟎. 𝟎𝟓
3. Test Statistic: The appropriate test statistic is
̂ − 𝒑𝟎
𝒑
𝒛𝒄 =
𝒑 𝒒
√ 𝟎 𝟎
𝒏
4. Critical Regions: Since the alternative hypothesis is 𝑝 > 0.75, we will used 𝑧𝑐 > 𝑧𝛼 , a right-tailed
test, where 𝑧𝛼 = 𝑧0.05 = 1.645.
Thus, we reject 𝑯𝟎 if 𝒛𝒄 > 𝟏. 𝟔𝟒𝟓.
5. Computation: Using the formula in step 3, the actual value of the test statistic is:
̂ − 𝒑𝟎
𝒑 𝟎. 𝟖𝟑 − 𝟎. 𝟕𝟓
𝒛𝒄 = = = 𝟏. 𝟒𝟑𝟏
𝒑 𝒒
√ 𝟎 𝟎 √(𝟎. 𝟕𝟓)(𝟎. 𝟐𝟓)
𝒏 𝟔𝟎

6. Statistical Decision: Since 𝒛𝒄 = 𝟏. 𝟒𝟑𝟏 is NOT greater than 𝟏. 𝟔𝟒𝟓 (meaning, it is NOT in the critical
region), the null hypothesis 𝑯𝟎 is NOT rejected.
7. Conclusion: The test result does not provide sufficient evidence to support the accountant’s claim
that more than 75% of similar firms spend more than PhP 500,000.00 on safety education at
𝜶 = 𝟎. 𝟎𝟓.

Example 2. A survey of 1000 household shows that 54% own personal computers at home. Is there reason to
believe that the proportion of households who own a computer is not 50%? Use 𝛼 = 0.05.

Given:

𝑝̂ = 54% 𝑜𝑟 0.54,
𝑝𝑜 = 50% 𝑜𝑟 0.50,
𝑞𝑜 = 1 − 𝑝𝑜 = 1 − 0.50 = 0.50.
𝑛 = 1000

Solution: Following the steps in hypothesis testing, we have:


Steps:
1. Hypotheses:
𝐻0 : The proportion of households who own a computer is 50%, that is, 𝒑 = 𝟎. 𝟓𝟎.
𝐻𝑎 : The proportion of households who own a computer is not 50%, that is, 𝒑 ≠ 𝟎. 𝟓𝟎.
2. Significance Level: 𝜶 = 𝟎. 𝟎𝟓
3. Test Statistic: The appropriate test statistic is
̂ − 𝒑𝟎
𝒑
𝒛𝒄 =
𝒑 𝒒
√ 𝟎 𝟎
𝒏
4. Critical Regions: Since the alternative hypothesis is 𝒑 ≠ 𝟎. 𝟓𝟎, we apply two-tailed test, that is,
𝑍𝑐 < −𝑍𝛼 or 𝑍𝑐 > 𝑍𝛼 , where 𝑍𝛼 = 𝑍0.05 = 𝑍0.025 = 1.96.
2 2 2 2
Thus, we reject 𝑯𝟎 if 𝒁𝒄 < −𝟏. 𝟗𝟔 or 𝒁𝒄 > 𝟏. 𝟗𝟔.
5. Computation: Using the formula in step 3, the actual value of the test statistic is:
̂ − 𝒑𝟎
𝒑 𝟎. 𝟓𝟒 − 𝟎. 𝟓𝟎
𝒛𝒄 = = = 𝟐. 𝟓𝟑
𝒑 𝒒
√ 𝟎 𝟎 √(𝟎. 𝟓𝟎)(𝟎. 𝟓𝟎)
𝒏 𝟏𝟎𝟎𝟎

6. Statistical Decision: Since 𝒛𝒄 = 𝟐. 𝟓𝟑 is greater than 𝟏. 𝟔𝟒𝟓 (meaning, it is in the critical region), the
null hypothesis 𝑯𝟎 is rejected.
7. Conclusion: There is sufficient evidence to believe that the proportion of households who own a
computer is not 50% at 𝜶 = 𝟎. 𝟎𝟓.

Reference: Supe, A., et. al., (2013). Elementary Statistics. Central Book Supply Inc.

Prepared by:

JOBELLE S. SIMBLANTE
Stat 26 Instructor

You might also like