Professional Documents
Culture Documents
Hypothesis Testing – a statistical procedure using sample data to assess or determine which hypothesis
is more acceptable as true which hypothesis is more likely to be true.
Significance Test – performed to determine the validity of a claim regarding the population parameter.
Ex. A manufacturer claims that a new type of metal has a breaking strength of 448kg per square
cm. A sample of 50 such metals yields a mean breaking strength of 446.5 kg/cm2
Errors in Hypothesis
Type I - rejection of the NULL hypothesis which is actually true
Type II –decision to retain a NULL Hypothesis which is actually false
NULL Hypothesis (Ho) – it is assumed as the true value of the parameter being testes and is usually
“hoped” to be rejected
Ho: μ = μ o; σ2 = σo2; μ≤ 𝜇o
ALTERNATIVE Hypothesis (Ha) – a claim that disagrees with the NULL Hypothesis
Ha: μ ≠μ o; σ2≠ σo2; μ> 𝜇o
MATH 8A
Engr. Mildred M. Martinez Page 1
Hypothesis Testing
2. One-Tailed Test – the critical region is in one of the tail of the distribution
- directional test
example: The mean monthly household income in rural areas is less than ₱9,800.00.
Ho: μ = ₱9,800.00
Ha: μ ≠ ₱9,800.00
α = 0.05
left-tailed right-tailed
Exercises:
1. A researcher wants to determine if the mean monthly household expenditure is really
₱10,760.00 as published.
Ho: μ = ₱10,760.00
Ha: μ ≠ ₱10,760.00
2. The mean age of professors is more than 30 years.
Ho: μ ≤ ₱10,760.00
Ha: μ > ₱10,760.00
3. The government anti-pollution spokesperson asserts that more than 78% of the
industries in the region meet the anti-pollution standards.
Ho: Π ≤ 0.78
Ha: Π > 0.78
4. The mean height of females is 155 cm.
Ho: μ = 155cm
Ha: μ ≠155cm
5. DTI receives many complaints from consumers that the bottles of cooking oil sold by a
company contains less than 1 liter of oil advertised.
6. A health practitioner wants to know if the mean height of girls at birth is at least 3.2kgs
as per record of a particular hospital.
CRITICAL REGION
- also known as the REJECTION Region
- area under the curve that contains all values of the statistics that allow rejection of the NULL
Hypothesis
- region on the extreme tails of the sampling distribution
- it has equivalent to α, the level of significance
MATH 8A
Engr. Mildred M. Martinez Page 2
Hypothesis Testing
TEST STATISTIC – a value derived or computed from a sample data set that is used to determine the
relative position of the mean in the hypothesized probability distribution of the sample mean.
Example. If α =0.01 (right-tailed z test) then the critical region for rejecting Ho is the area under the tail
of the curve that is equal to 0.01
0.5-0.01 = 0.49 ; 0.4901 from the z-table is 2.33, thus Zcritical = 2.33
0.5-0.01 = 0.49 ; 0.4901 from the z-table is 2.33, thus Zcritical = -2.33on the left side
c. Two-tailed test
α = 0.01
0.5-0.005 = 0.495 ; 0.4901 from the z-table is 2.575, thus Zcritical = ±2.575 on the both sides
MATH 8A
Engr. Mildred M. Martinez Page 3
Hypothesis Testing
_ _
x–μ x–μ _
z = -------- and for a t-test, t = --------- ; with: x as the mean
σ/√𝑛 s/√𝑛 μ as the average
σ as the standard deviation
s as the estimate sample
deviation
5. Evaluate the test statistic. Make the decision and draw conclusion.
If the computed value (the test statistic) lies within the rejection region, then the NULL
Hypothesis is rejected, otherwise, fail to reject.
The decision rule is
Reject the NULL Hypothesis if I zcomp I ≥ I zcrit I; I tcomp I ≥ I tcrit I; I χ2comp I ≥ I χ2crit I;
I Fcomp I ≥ I Fcrit I. Otherwise, retain the NULL Hypothesis.
Example.
Example.
In 2006, the average annual family income in NCR was ₱310,860.00 with a standard deviation of
₱176,870.00. With the increased price of commodities, it is hoped that the mean annual family income
has also increased. A survey is conducted to verify this contention. A random sample of 320 families has
MATH 8A
Engr. Mildred M. Martinez Page 4
Hypothesis Testing
a mean annual income of ₱325,255.00 Is the sample result sufficient to demonstrate that the mean
annual family income in NCR is significantly higher than ₱310,860.00 at the 5% level of significance.
Solution:
1. State the NULL and ALTERNATIVE Hypothesis
Since the term “higher” is used, one-sided ALTERNATIVE Hypothesis is chosen .
Ho: μ ≤ ₱310,860.00
Ha: μ > ₱310,860.00
4. Compute the test statistic. We are using the z-test since deviation is given.
_
x – μ 325,255 – 310,860
zc = ------- = --------------------------- = 1.456
σ/√𝑛 176,870/√320
Exercises:
1. A random sample of 64 apartments in the city shows that the average monthly water consumption
is 55 cubic meters with a standard deviation of 5 cubic meters. Test the hypothesis that the
average monthly water consumption is 53 cu.m. Use 1% level of significance.
2. A softdrink dispensing machine was designed to discharge, on the average, 7 ounces of beverage
per cup. An experiment was conducted to test if it is really working effectively as designed.
Sixteen cupfuls of beverage were drawn from the machine and measured. The mean and
standard deviation of the 16 measurements were 7.15 oz and 0.12 oz respectively. Is there
sufficient evidence to indicate that the mean discharge differs from 7 oz. Test at the 5% level of
significance.
_
3. A sample of n=100 has a mean x = 15. Assume that the population is approximately normally
distributed a population standard deviation of 4. The hypothesis desired are:
Ho: μ=12.5 Ha: μ≠ 12.5
Test the hypothesis at α =0.02
MATH 8A
Engr. Mildred M. Martinez Page 5
Hypothesis Testing
Ho is true Ho is false
Reject Ho Type I error (α) Correct Decision (1-β)
Fail to reject Ho Correct Decision (1-α) Type II error (β)
MATH 8A
Engr. Mildred M. Martinez Page 6