HYPOTHESIS

❏ a tentative statement about relationships that exist between two or more variables

❏ an assumption or conjecture about a population parameter which may or may not be true

HYPOTHESIS TESTING
❏ widely used method of statistical inference to make a decision about a parameter; that is, given two competing
claims about an unknown parameter, to decide which of them is more plausible

❏ always consist of two contradictory hypotheses or statements, a decision based on sample data, and a
conclusion.

❏ The purpose of testing a hypothesis is to make inferences about a population parameter by analyzing
differences between observed sample statistics and the results expected to be obtained if some underlying
assumptions are true.

2
 TYPES OF HYPOTHESES
Null Hypothesis, 𝐻𝑜 Alternative Hypothesis, 𝐻𝑎 or 𝐻1

It shows that there is significant difference, an effect,

It shows no significant difference, no changes, nothing
change, relationship between a parameter and specific
happened, no relationship between two parameters.
value.

The independent variable has no effect on the The independent variable has an effect on the
dependent variable. dependent variable.

𝐻𝑜: 𝜇1 = 𝜇2, 𝐻𝑜: 𝜇1 ≤ 𝜇2, 𝐻𝑜: 𝜇1 ≥ 𝜇2 𝐻𝑎: 𝜇1 ≠ 𝜇2, 𝐻𝑎: 𝜇1 > 𝜇2, 𝐻𝑎: 𝜇1 < 𝜇2

not equal, differ from, changed from, not the same

= equal to, the same as, not changed from, is ≠ as

greater than, above, higher than, longer than,

≤ less than or equal to > bigger than, increased, at least

less than, below, lower than, smaller than,

≥ greater than or equal to < shorter than, decreased or reduced from, at most

3
 NULL HYPOTHESIS AND ALTERNATIVE HYPOTHESIS

EXAMPLE 1: EXAMPLE 2:
The average age of bus drivers in Metro Manila The average number of calories of a
is 37 years. low-calorie meal is at most 250.

(Null) 𝐻𝑜: 𝜇 = 37 years (Null) 𝐻𝑜: 𝜇 = 250 calories

(Alternative) 𝐻𝑎: 𝜇 ≠ 37 years (Alternative) 𝐻𝑎: 𝜇 < 250 calories

EXAMPLE 3:
A DTI representative wants to test at 99% confidence level whether the average content of Soda Z
is less than 400ml. as indicated in the label.

Null and Alternative Hypothesis in words Null and Alternative Hypothesis in symbol

𝐻𝑜: The average content of Soda Z is greater than or equal to 400ml. 𝐻𝑜: 𝜇 ≥ 400

𝐻𝑎: The average content of Soda Z is less than 400ml. 𝐻𝑎: 𝜇 < 400

4
 TYPES OF ALTERNATIVE HYPOTHESIS

Two-Tailed Hypothesis One-Tailed Hypothesis

(non-directional hypothesis) (directional hypothesis)

An alternative hypothesis in which the An alternative hypothesis in which the researcher predicts the
researcher predicts that the groups being direction of the expected difference between the groups.
compared differ but does not predict the
direction of the difference.

Left-tailed:
𝐻𝑎: 𝜇1 < 𝜇2

Right-tailed:
𝐻𝑎: 𝜇1 > 𝜇2
𝐻𝑎: 𝜇1 ≠ 𝜇2

5
EXAMPLE:
The school registrar believes that the average number of enrollees this school year is not the same as the previous
school year.
Let 𝜇𝑜 be the average number of enrollees last year.

𝐻𝑜: 𝜇 = 𝜇𝑜
Use the two-tailed test
𝐻𝑎: 𝜇 ≠ 𝜇𝑜
if 𝐻𝑎 uses the symbol ≠.

However, if the school registrar believes that the average number of enrollees this school year is less than the
previous year, then you will have:

𝐻𝑜: 𝜇 = 𝜇𝑜
Use the left-tailed test
𝐻𝑎: 𝜇 < 𝜇𝑜
if 𝐻𝑎 uses the symbol <.

On the other hand, if the school registrar believes that the average number of enrollees this school year is greater
than the previous school year, then you will have:

𝐻𝑜: 𝜇 = 𝜇𝑜 Use the right-tailed test

𝐻𝑎: 𝜇 > 𝜇𝑜 if 𝐻𝑎 uses the symbol >.

6
 TYPE I AND TYPE II ERROR IN HYPOTHESIS TESTING

● Type I Error
○ An error in which the null hypothesis is rejected when it is true
○ Also known as “false positive error”

● Type II Error
○ An error in which there is a failure to reject the null hypothesis when it is false
○ Also known as “false negative error”

NULL HYPOTHESIS (Ho) REJECT Ho FAIL TO REJECT Ho

Ho is TRUE Type I Error Correct Decision

Ho is FALSE Correct Decision Type II Error

8
 LEVEL OF SIGNIFICANCE
- It is the predetermined threshold value to decide whether to reject the null hypothesis or not
- It is the probability of rejecting the null hypothesis when it is true(Type I error).
- Denoted by Greek symbol alpha(a)
- 0.05 or 5% is the most commonly used significance level, meaning that there is a 5% chance of rejecting the
null hypothesis when it is actually true.
- Significance level is set or chosen by the researcher before conducting the hypothesis test
- A level of significance of 0.01(1%) or 0.10(10%) can also be used depending on the confidence level being
required.

Confidence Level -
Alpha Level Error Confidence Level measures how
confident we are that
our conclusions are
correct
0.01 1% 99% CL= 1-a

0.05 5% 95%

9
Note:
In symbol, if null hypothesis is used, it is written as:
a=0.01, a=0.05, a=0.10
If alternative hypothesis used is ≠, then alpha will be divided by 2
a/2=0.005, a/2=0.025, a/2=0.05

Example:
1. A researcher uses 5% level of significance in proving that there is no significant change in the average
number of enrollees in the 10 sections for the last two years.
- It means that the chance of null hypothesis would be rejected when it is true is 5%
a = 0.05

1. A sample of 100 private school students were surveyed whether they planned to transfer to public school
in the incoming school year. Results showed that 40% of these students will transfer to public school.
Test the average number of private school students that will transfer to public school is not 40% at 10%
level of significance.
a/2= 0.05

https://www.youtube.com/redirect?event=video_description&redir_token=QUFFLUhqbmFnQVd4dXJGZUljd3FJUlVxWjJXalNhTDRSUXxBQ3Jtc0tsazhXbEhCejNHMkNtRzB1bkpXaHlVVW5MbUlGV21JWFA2ajhBeTBZWUR6YnVGTUtxYnFqZkQ5Sld2UUpiUmF3ZHk2MUpGMTVMUm5BbGJYUTlsVjNGTGZrSDZOVW5qUE5IeWhsNUQ5YUNOWDdENW9sQQ&q=
https%3A%2F%2Ftinyurl.com%2Fna6wmffu&v=D3qFmPn1pgM
(Example Source:https://youtu.be/D3qFmPn1pgM?si=Z9oMgJ9pG4_lkDFM=≠≠)
10
 REJECTION OR ACCEPTANCE OF NULL HYPOTHESIS
1. CRITICAL VALUE APPROACH
Test Statistic ● is a numerical value that is calculated from a statistical test of a hypothesis
● is a number that describes how much your test result differ from the null hypothesis
● there are different types of test statistics, such as z-test, t-test, F-test, chi-square test, etc.
vs

Critical Value ● is the value from the distribution that separates the rejection region from the non-rejection region

General Rule:
If the absolute value of:

● Test Statistic (Computed Value) > Critical Value = Reject the H0

● Test Statistic (Computed Value) < Critical Value = Accept/Fail to Reject
the H0

11
What are critical values?

❏ are specific values used in hypothesis testing and confidence intervals to determine the boundary
between accepting or rejecting a null hypothesis

❏ are closely tied to the significance level, often denoted by α (alpha).

❏ determined based on the chosen significance level and the distribution of the data.
❏ The most well-known critical values are those associated with the normal distribution, often referred to
as z-scores
❏ Critical values for other distributions, like the t-distribution or chi-squared distribution, can also be found
in statistical tables or using software.

13
Critical values are used in various statistical tests, including:

Confidence ANOVA (Analysis of

Hypothesis testing: Chi-squared tests:
intervals: Variance):

Comparing a Critical values are Critical values are Critical values help
sample statistic to a used to establish used to determine assess whether
critical value helps the margin of error the significance of there are
determine whether around a sample relationships in statistically
to accept or reject a statistic. contingency tables. significant
null hypothesis. differences between
groups.

14
Examples:

1. According to a study conducted by the Grade 12 students, ₱155 is the average monthly expense for cell
phone loads of high school students in their province. A Statistics student claims that this amount has increased
since January of this year. Do you think his claim is acceptable if a random sample of 50 students has an
average monthly expense of ₱165 for cell phone loads? Using 5% level of significance, assume that a
population standard deviation is ₱52

Given: 𝑥̅ = 165 𝜇 = 155 𝜎 = 52 ; 𝑛 = 50 𝛼 = 0.05

Step 1: State the null and alternative hypotheses.

Ho : 𝜇 = 155 Ha : 𝜇 > 155

Step 2: Determine the test statistic, then compute its value. Since the population mean is being tested,
the population standard deviation 𝜎 is known, and 𝑛 > 30, the appropriate test statistic is the z-test.

Assuming that the z-computed value = 1.361

15
Step 3:

Find the critical value and draw the critical

region. Use the z-critical value table. The
alternative hypothesis is directional. Hence,
the one-tailed test (right-tailed test) shall be
used.

From the z-value table at 0.05 level of

significance:

Critical Value is 1.645

16
 1.645
1.361

Step 4:
Draw a conclusion. Critical Value is 1.645
The z-computed value is 1.361 and it lies within the non-rejection region, so we fail to reject the null
hypothesis.
Therefore, there is no enough evidence to support the claim that the average monthly expense for cell
phone loads is more than ₱155. This result is significant at 𝛼 = 0.05 level. 17
Examples:

2. The average IQ of Senior High School students is 99 with a standard deviation of 15. A researcher believes
that the average IQ of Senior High School students is lower. A random sample of 40 students was tested and
got an average of 95. Is there enough evidence to suggest that the average IQ is lower? Test the hypothesis at
0.05 level of significance.

Given: 𝑥̅ = 95 𝜇 = 99 𝜎 = 15 𝑛 = 40 𝛼 = 0.05

Step 1: State the null and alternative hypotheses.

Ho : 𝜇 = 99 Ha : 𝜇 < 99

Step 2: Determine the test statistic, then compute its value. Since the population mean is being tested,
the population standard deviation 𝜎 is known, and 𝑛 > 30, the appropriate test statistic is the z-test.

Assuming that the z-computed value = -1.68

18
Step 3:

Find the critical value and draw the critical

region. Use the z-critical value table. The
alternative hypothesis is directional. Hence,
the one-tailed test (left-tailed test) shall be
used.

From the z-value table at 0.05 level of

significance:

Critical Value is -1.645

19
 -1.688

Step 4:
Draw a conclusion. Critical Value is -1.645
The z-computed value is -1.688 and it lies within the rejection region, so we reject the null hypothesis.
Therefore, there is enough evidence to support the claim that the IQ level of Senior High School students
is lower than 99. This result is significant at 𝛼 = 0.05 level.
20
Examples:

3. Blood glucose levels for obese teenagers have a mean of 120. A researcher thinks that a diet high in
raw cornstarch will have a positive or negative effect on blood glucose levels. A sample of 25 patients
who have tried the raw cornstarch diet has a mean glucose level of 135 with a standard deviation of 38.
Test the hypothesis at 𝛼 = 0.10 that the raw cornstarch had an effect.

Given: 𝑥̅ = 135 𝜇 = 120 𝑠 = 38 𝑛 = 25 𝛼 = 0.10 𝑑𝑓 = 24

Step 1: State the null and alternative hypotheses.

Ho : 𝜇 = 120 Ha : 𝜇 ≠ 120

Step 2: Determine the test statistic, then compute its value. Since it is the population mean being
tested, the population standard deviation is unknown, and 𝑛 < 30, the appropriate test statistic is the t-
test.

Assuming that the t-computed value = 1.974

21
Step 3:
Find the critical value and draw the critical region.

The alternative hypothesis is non-directional. Hence, the two-tailed

test shall be used.

From the t-value table at 0.10 level of significance, 𝑑𝑓 = 24 :

Critical Value is ±1.711.

22
 -1.711 1.711

1.974
Step 4:
Draw a conclusion. Critical Value is ±1.711.
Since the t-computed value is 1.974 which is greater than the critical value of 1.711, we reject the null
hypothesis and support the alternative hypothesis.

We can conclude that there is enough evidence to support the claim that the raw cornstarch had an
effect on blood glucose levels.
23
2. P-VALUE APPROACH
● the probability that an event will happen purely by chance assuming the null hypothesis is
P - Value
true
● the probability that, if the null hypothesis is true, the results from another randomly selected
vs sample will be as extreme or more extreme as the results obtained from the given sample

Significance level ● is a threshold value – it is the maximum value allowed for a Type I error

● If p-value < statistical significance (0.01/0.05/0.10)

Reject null hypothesis (H0)
Example: .03 < .05

● If p-value > statistical significance (0.01/0.05/0.10)

Accept null hypothesis (H0)
Example: .07 > .05

24
What are p-value?

❏ The p-value represents the probability of observing a test staFsFc as extreme as, or more extreme than,
the one calculated from your data, assuming the null hypothesis is true.
❏ The exact method for finding the p-value depends on the chosen staFsFcal test and distribuFon. In some
cases, you may directly find it from staFsFcal tables or use soJware for complex calculaFons.
❏ is a numeric representaFon of probability, specifically tailored to gauge the plausibility of observed results
under the assumpFon of a null hypothesis.

26
Calculating P-Values
The process involves several key elements:

Determine the
Formulate Choose a Test
Collect Data Probability
Hypotheses Statistic
Distribution

Compare the P-
Calculate the Test
Draw Conclusions Value to the Find the P-Value
Statistic
Significance Level
Example:

Given:

Zc = ±2.31 0.025 0.95 0.025

α = 1-C = 0.05
2.31 Z score corresponds to Area Value of 0.98956

0.01044 0.01044
1 minus 0.98956 = 0.01044
- 2.31 + 2.31

P-value is the sum of the shaded area:

Ho : μ = 168 Ha : μ ≠ 168
P-VALUE = 0.02088
We conclude that with a 95% level of confidence we
cannot accept the Ho that the average weight of all
0.02088 < 0.05 residents in town BNM is 168 lbs.
P-VALUE is < α or
In 95% level of confidence we believe that there is
enough evidence to discard the null hypothesis. 28
Z-Test and T-Tests

● Parametric Tests, where Null Hypothesis is less than, greater than or

equal to some value.
● compares the means of two data group to determine if there is a
significant difference between them.

30
Z-TEST

❏ Z test is a statistical way of testing a Null Hypothesis when either:

❖ We know the population variance, or
❖ We do not know the population variance, but our sample size is large n≥30
❏ z-statistic follows a standard normal distribution with a degree of freedom equal to
n-1, where n is the sample size.
❏ samples used for z-test of t-test must be independent sample, and also must have
a distribution identical to the population distribution.

31
Example:

(Two-tailed Z-test)

The average weight of all residents in town BNM is 168 lbs. A nutritionist believes the true mean to be different.
She measured the weight of 36 individuals and found the mean to be 169.5 lbs with a standard deviation of 3.9.
(a) State the null and alternative hypotheses (b) At a 95 % confidence level, is there enough evidence to discard
the null hypothesis?

Ho : μ = 168 Ha : μ ≠ 168

Solution:

Given: z=
n = 36

X̄ = 169.5 =

σ = 3.9
= 2.31
μ = 168
2.31 > 1.96, Reject the Ho 32
One-tailed Z-test

A factory manufactures cars with a warranty of 5 years on the engine and transmission. An engineer believes
that the engine or transmission will malfunction in less than 5 years. He tests a sample of 40 cars and find the
average time to be 4.8 years with a standard deviation of 0.50. (a) State the null and alternative hypotheses.
(b) At a 2% significance level, is there enough evidence to support the idea that the warranty should be
revised?

Solution:
Ho : μ ≥ 5 Ha : μ < 5

Given: z=

n = 40
=
X̄ = 4,8

σ = 0.50
= -2.53
μ=5
P-value = 0.0057 < 0.02, reject the Ho
33
T-TEST

❏ T-Test is a statistical way of testing hypothesis when:

❖ We do not know the population variance
❖ Our sample size is small, n<30

34
Example:

Blood glucose levels for obese teenagers have a mean of 120. A researcher thinks that a diet high in raw
cornstarch will have a positive or negative effect on blood glucose levels. A sample of 25 patients who have
tried the raw cornstarch diet has a mean glucose level of 135 with a standard deviation of 38. Test the
hypothesis at 𝛼 = 0.10 that the raw cornstarch had an effect. (a) State the null and alternative hypotheses. (b)
Determine the test statistic, then compute its value.

Since it is the population mean being tested, the population standard deviation is unknown, and .. < 30, the
appropriate test statistic is the t-test.

Since the t-computed value is 1.974

which is greater than the critical value of
1.711, we reject the null hypothesis and
support the alternative hypothesis. We
can conclude that there is enough
evidence to support the claim that the raw
cornstarch had an effect on blood
glucose levels.

35
So when should we perform the z-test, and when should we perform t-Test?

If the sample size is large enough, then the z-Test and t-test will conclude with the same results. For a
large sample size, Sample Variance will be a better estimate of Population variance, so even if
population variance is unknown we can use the z-test using sample variance.

Similarly, for a Large Sample, we have a high degree of freedom. And since t-distribution approaches
the normal distribution, the difference between the z score and t score is negligible.

36