STATS 4TH QUARTER

LESSON 1.1: HYPOTHESIS TESTING > Null and Alternative Hypotheses

> Hypothesis  The H0 asserts the value to which the population
 Simply a statement that something is true parameter is equal and is presumed to be true while
 A tentative explanation, claim, or assertion Ha negates H0
about people, objects, or events  E.g.
H0: The mean general weighted average of the
> Examples of Hypothesis college students in a particular university is 85.28.
 Is there a significant relationship between the Ha: The mean GWA of the college students in a
mathematics attitude and competency levels of particular university is not equal to 85.28.
grade 11 SHS students of SSAP? Ha: The mean GWA of the college students in a
 Is there a significant difference in the mean life particular university is greater than 85.28.
span between the Eveready and Energizers Ha: The mean GWA of the college students in a
batteries? particular university is less than 85.28.
 Is the mean grade of grade 11 SHS of SSA-SF in
General Mathematics for the First Semester of the > Directional vs. Non-Directional Test
School Year 2016-2017 equal to 88.75?  Directional Test (One-Tailed)
– when the alternative hypothesis utilizes the > or
> Hypothesis Testing < symbol
 Decision-making process for evaluating claims – a test that determines the relationship between
about a population based on the characteristics of a the variable in only one direction (left or right tail)
sample purportedly coming from that population  Non-Directional Test (Two-Tailed)
 Decision: whether acceptable or not – when the alternative hypothesis utilizes the ≠
 Process that involves making a decision between symbol
opposing hypotheses – a standard test used in many researches and it
 Procedure in making decisions based on a sample compares population parameter in both direction
evidence (left and right)

> Two Types of Hypothesis ONE-TAILED TWO-TAILED

 Null Hypothesis (H0) H 0 : μ0=μ H 0 : μ0=μ
– H0 stands for hypothesis and the subscript 0 H a : μ0 < μ (left-tailed) H a : μ0 ≠ μ
implies “no difference” H a : μ0 > μ (right-tailed)
– usually designated by a “not” or “no” term, H 0 : μ0 ≥ μ
which means there is no change
H a : μ0 < μ (left-tailed)
– states that two groups come from the same
sample H 0 : μ0 ≤ μ
– e.g. H a : μ0 > μ (right-tailed)
o There is no statistically significant difference
between two groups on the dependent > Examples:
variable. * State the alternative hypothesis of the following if the
o There is no difference between a parameter test is:
and a specific value. a. Right-tailed
o There is no difference between two b. Left-tailed
parameters. c. Two-tailed
 Alternative Hypothesis (Ha) 1. The mean age of the patients in hospital is equal to
– “H sub a” 26.
– states that two groups come from different a. Ha: The mean age of the patients in hospital is
samples greater than 26. (μ>26) – right-tailed
– e.g. b. Ha: The mean age of the patients in hospital is
o There is a statistically significant difference less than 26. (μ<26) – left-tailed
between two groups on the dependent c. Ha: The mean age of the patients in hospital is
variable. not equal to 26. (μ ≠ 26) – two-tailed
o There is a difference between a parameter and 2. The average height of 20 grade 11 students in cm
a specific value. is 160.
o There is a difference between two parameters.
 a. Ha: The average height of 20 grade 11 students Decision
in cm is greater than 60. ( μ>60 ) – right-tailed Do not Reject Reject
b. Ha: The average height of 20 grade 11 students Null hypothesis Correct Type I Error
in cm is less than 60. (μ<60) – left-tailed is true Decision
c. Ha: The average height of 20 grade 11 students Null hypothesis Type II Error Correct
in cm is not equal to 60. (μ ≠ 60) – two-tailed is false Decision

> Formulating Null and Alternative Hypothesis If the H0 is false and you reject Correct Decision
*Write the null and alternative hypothesis in symbols for it
each of the following: If the H0 is true and you did not Correct Decision
1. The net weight of a packet of snack brand is 130g. reject it
A sample of 80 packets yielded a sample mean
weight of 122g with a standard deviation of 15g. > Examples:
H 0 : μ=130 g 1. Maria insists that she is 30 years old when, in fact,
H a : μ<130 g she is 32 years old. What error is Mary
committing?
2. A teacher wants to know if listening to popular
Answer: Mary is rejecting the truth. She is
music affects the performance of the students. A
committing Type I Error.
class of 50 grade 7 students was used in the
2. A man plans to go hunting the Philippine monkey-
experiment. The mean score was 83 and the
eating eagle believing that it is proof of his mettle.
standard deviation is 5. A previous study revealed
What type of error is this?
that μ=82and the σ =10.
Answer: Hunting the Philippine eagle is prohibited
H 0 : μ=82 by the law. It is a Type II Error.
H a : μ ≠82
3. A farmer believes that using organic fertilizers on LESSON 1.2: HYPOTHESIS TESTING
his crop will yield greater income. His average > Significance Level (α )
income from the past was P200,000 per year.  The decision to reject or not to reject the Ho is
H 0 : μ=P 200,000 called significance
H a : μ> P 200,000  The most frequently used levels of significance are
4. A treasurer of a certain institution claims that the 0.05 (5%), 0.10 (10%), and 0.01 (1%)
mean monthly salary of their teacher is P21,750. A
researcher takes a random sample of 75 teachers > Rejection Region
and found out that they have a mean monthly
salary of P19,375. Do the 75 teachers have lower
salaries than the rest?
H 0 : μ=P 21,750
H a : μ< P 21,750
5. The mean weight of the baggage carried into an
airplane by individual passengers at Clark
International Airport is 15.8kg. A statistician takes
a random sample of 110 passenger and obtains a > Hypothesis Testing Using Z-Test
sample mean weight of 13.8kg. 1. Define the H0 and Ha
H 0 : μ=15.8 kg 2. Write the significance level, type of test, and
critical value
H a : μ<15.8 kg
Type of Test Level of Significance
> Types of Errors a=1 % a=5 % a=10 %
* Errors in hypothesis testing is referred to as the One-Tailed Test c=± 2.33 c=± 1.645 c=± 1.28
possibility of committing an error in deciding whether to Two-Tailed Test c=± 2.575 c=± 1.96 c=± 1.645
accept or reject hypothesis. 3. Compute the test statistic
 Type I Error X−μ
– rejecting the null hypothesis when in fact the null z=
σ
hypothesis is true
 Type II Error
√n
4. Decide whether to reject or to accept H0
– not rejecting the null hypothesis when in fact the
5. Give interpretation
null hypothesis is false
Decision Rule: The sample does not belong to the population
 Reject H0 if the computed test statistic ≤ negative whose mean μ is 26. So, there is a significant
critical value or if the computed test statistic ≥ difference between the sample mean and the
positive critical value. Otherwise, do not reject (or population mean.
accept) H0  The production manager of a large manufacturing
company estimates that the mean age of his
> Examples employees is 22.8 years. The treasurer of the firm
 Sol administered a mathematics achievement test needs more accurate employee mean age figure in
to a random of 50 graduating pupils. In this order to estimate the cost of an annuity benefit
sample, X =90. The population parameters are program being considered for employees. The
μ=83 and σ =15. Is the performance of the treasurer takes a random sample of 70 workers and
sample above average? Use α =0.05 . finds out that the mean age of the employees
Step 1: sampled is 24 years with a standard deviation of
H 0 : μ=83 4.6 years. At the 0.01 level of significance, what
can the treasurer conclude about the accuracy of
H a : μ>83 the production manager’s estimate?
Step 2: Step 1:
α =0.05 , one−tailed , ¿−tailed , H 0 : μ=22.8
c=1.645 H a : μ>22.8
Step 3: Step 2:
X−μ 90−83 α =0.01 , one−tailed , ¿−tailed ,
z= = =3.30
σ 15 c=2.33
√n √50 Step 3:
Step 4: X−μ 24−22.8
3.30 ≥ 1.645 z= = =2.18
σ 4.6
The null hypothesis is rejected (Reject H0)
Step 5: √n √70
The sample does not belong to the population Step 4:
whose mean μ is 83. So, there is a significant 2.18 ≤ 2.33
The null hypothesis is not rejected (Do not reject
difference between the sample mean and the
H0)
population mean.
Step 5:
 A manufacturer of cellphone batteries claim that
The sample belong to the population whose mean
when fully charged, the mean life of his products
lasts for 26 hours with a standard deviation of 5 μ is 22.8. So, there is no significant difference
hours. Mr. DG, randomly picked 35 of the between the sample mean and the population
batteries. His test showed that the average life of mean.
his sample is 24.3 hours. Is there a significant
between the average life of all the manufacturer’s > Hypothesis Testing Using T-Test
batteries and the average battery of his sample? X−μ X−μ
t= t=
Use two-tailed test at α =5 % . σ s
Step 1: √n √n
H 0 : μ=26 * Use t-test if n ≤ 30. Refer to t-table for critical
H a : μ ≠26 values
Step 2:
α =5 % ,two−tailed , c=± 1.96 > Examples
Step 3:  The school directress claims that the grade school
X−μ 24.3−26 teachers work an average of 40 hours per week
z= = =−2.01 during the school year. A principal of a certain
σ 5 province surveyed 20 randomly selected teachers
√n √ 35 and found that they work for an average of 41.5
Step 4: hours a week with a standard deviation of 2.5
–2.01 ≤ –1.96 hours. Test if the mean number of the teachers’
The null hypothesis is rejected (Reject H0) work hours recorded by the principal differs from
Step 5: the record of the directress. Use level of
significance of 0.05.
 Step 1: H a : μ> 4
H 0 : μ=40 Step 2:
H a : μ ≠ 40 α =10 % , one−tailed ,¿−tailed ,
Step 2: c=1.895
α =0.05 , two−tailed , c=± 2.093 (df =8−1=7)
(df =20−1=19) Step 3:
Step 3: Solve for mean and standard deviation.
X−μ 41.5−40 ΣX
z= = =2.68 μ= =4.375
s 2.5 N
√n √20 X X −μ (X −μ)
2

Step 4: 2 –2.375 5.641

2.68 ≥ 2.093 3 –1.375 1.891
The null hypothesis is rejected (Reject H0) 4 –0.375 0.141
Step 5: 5 0.625 0.391
The sample does not belong to the population 6 1.625 2.641
whose mean μ is 40. So, there is a significant 7 2.625 6.891
difference between the sample mean and the 8 3.625 13.141
population mean. 9 4.625 21.391
 The accounting department claims that the mean
monthly salary of their teacher is P21,750 with a
∑ (X −μ)2 =52.128
2
standard deviation of P6,000. A researcher takes a 2 Σ ( X −μ ) 52.128
σ = =
random sample of 25 teachers and found out that N 6.516
they have a mean monthly salary of P19,375. Do σ =2.55
the 25 teachers have a significant difference 4.375−4
between their salary and the accounting claims? t= =0.42
2.55
Use of 0.01 level of significance.
Step 1: √8
H 0 : μ=21,750 Step 4:
0.42 ≤ 1.895
H a : μ ≠21,750
The null hypothesis is not rejected (Do not reject
Step 2: H0)
α =0.01 , two−tailed , c=± 2.797 Step 5:
(df =25−1=24) The sample belongs to the population whose mean
Step 3: μ is 4. So, there is no significant difference
X−μ 19,375−21,750 between the sample mean and the population
z= = =−1.98
s 6,000 mean.
√n √ 25
Step 4: LESSON 1.3: HYPOTHESIS TESTING
–1.98 ≥ –2.797 > Apply the p-Value Method
The null hypothesis is not rejected (Do not reject  Another way of conducting hypothesis
H0)  p-Value (p): probability of committing a Type I
Step 5: error
The sample does belong to the population whose  Corresponding decision rule
mean μ is 21,750. So, there is no significant o If the obtained probability is less than or equal to
difference between the sample mean and the a critical probability level, called alpha (α ) level,
population mean. then we reject the null hypothesis
 The average family size in the Philippines was Reject H0 if p ≤ α
reported as 4. A random sample of families in a o If the obtained probability level is greater than
particular barangay resulted in the following the alpha level, we do not reject H0. Accept H0
family sizes: 5, 6, 7, 4, 3, 8, 2, and 9. At 10% level Do not reject H0 if p>α
of significance, is the average family size more
than the national average? > Solving for p-Value
Step 1:  In a test of H0: μ=100 against Ha: μ>100 , the
H 0 : μ=4 sample data yielded the sample statistic z = 2.15.
Find the p-value for the test
 Ha: μ>83
– The p-value is the area to the right of the z. From Step 2:
the z-table, the probability between z = 0 and z = α =0.05 , right-tailed
2.15 is 0.4842. Step 3:
p-value = 0.5000 – 0.4842 = 0.0158 90−83
z= =3.30
15
 In a test of H0: μ=100 against Ha: μ ≠100 , the
sample data yielded the sample statistic z = 2.15. √ 50
Find the p-value for the test. Since 3.30 is above the maximum z-value found in
the table, we get the maximum value of 0.4999
– Since the test is two-tailed, the p-value is p-value = 0.5000 – 0.4999 = 0.0001
multiplied by 2. Step 4:
p-value = 0.5000 – 0.4842 = 0.0158 x 2 = 0.0316 0.0001 < 0.05
The null hypothesis is rejected (Reject H0)
> Test for Significant Difference Step 5:
 A random sample of 200 school managers was The sample belongs to another population with a
administered a developed Leadership Skills Test. mean that is above average. So, there is a
The sample mean and the standard deviation were significant difference between the sample mean
78 and 4.2, respectively. In the standardization of and the population mean.
the test, the mean was 73 and the standard
deviation was 8. Test from the significant > Hypothesis Testing Population Proportion
difference using 0.05 level of confidence utilizing  Inferences involving proportions are made in the
the p-value. context of probability, p, of ‘success’ for a
Step 1: binomial distribution
H0: μ=73  When testing about a proportion, a percentage, or a
Ha: μ ≠73 probability, the following assumptions must be
Step 2: made:
1. The conditions for binomial experiment are
α =0.05 , two-tailed
met. That is, there is a fixed number of
Step 3:
independent trials with constant probabilities,
78−73
z= =8.77 and each trial has two outcomes that we usually
8 classify as “success” and “failure.”
√200 2. The conditions np ≥ 5 and nq ≥ 5 are both
Since 8.77 is way above the maximum z-value satisfied so that the binomial distribution of
found in the table, we get the maximum value of sample proportions can be approximated by a
0.4999. normal distribution with μ=npand σ =√ npq.
p-value = 0.5000 – 0.4999 = 0.0001 x 2 = 0.0002
Step 4: > Test Statistic z for Proportions
0.0002 < 0.05 ^p − p
z=

√
The null hypothesis is rejected (Reject H0)
pq
Step 5:
The sample belongs to another population with a n
mean that is above average. So, there is a ^p=sample proportion
significant difference between the sample mean p= population proportion ( % )
and the population mean. n=sample ¿ ¿
 For a one-tailed test:
 Teacher Lyn administered a mathematics H 0 : p= p0
achievement test to a random sample of 50 H a : p> p0 and the rejection region is z >+ z α
graduating students. In this sample, the sample
mean is 90 and the sample standard deviation is 10. Or H a : p< p0 and the rejection region is z ← z α
The population parameter is 83 for the population  For a two-tailed test:
mean and 15 for population standard deviation. Is H 0 : p= p0
the performance of the sample above average? Use H a : p ≠ p0 and the rejection region is z ← z α or
0.05 and p-value method. 2
Step 1: z>zα
H0: μ=83 2
> Example Step 3:
 Out of 200 students, 120 said YES to campaign for ^p − p
z=

√
the CLEAN GREEN project. pq
X
^p= ∧ q^ =1−^p n
n 0.225−0.17
120 z= =2.07

√
^p= =0.6 (0.17)(0.83)
200
q^ =1−0.6=0.4 200
 A recent survey done in the Philippine Housing Step 4:
Authority found that 35% of the population owns Using p-value method:
their house. In a random sample of 240 heads of The area between z = 0 and z = 2.07 is 0.4808
households, 78 responded that they owned their p-value = 0.5 – 0.4808 = 0.0192
house. At the 0.01 LS, does that indicate a 0.0192 < 0.05 The null hypothesis is rejected
difference from the national proportion? (Reject H0)
Step 1: Step 5:
There is a significant difference between the
H 0 : p=0.35
sample mean and the population mean.
H a : p ≠0.35
Step 2:
a=0.01, two-tailed
Step 3: LESSON 2: CORRELATION AND REGRESSION
^p − p > Correlation Analysis
z=

√ pq  Statistical method used to determine whether a

relationship between two variables exists
n
 E.g.
74
^p= =0.325 o IQ scores and scores in Math Examinations
240 o Teachers Salary and Performance of the Teachers
0.325−0.35 o Daily Allowance and Daily Expenses
z= =−0.81

Step 4:
√
(0.35)(0.65)
240
> Regression Analysis
 Statistical method used to describe the nature of
the relationship between variable
Using p-value method: > Nature of Bivariate Data
The area between z = 0 and z = –0.81 is 0.2910  Univariate Data
p-value = 0.5 – 0.2910 = 0.2090 x 2 = 0.4180 – involves only one variable
0.4180 > 0.01 The null hypothesis is not rejected  Bivariate Data
(Do not reject H0) – involves relationship between two variables
Using the traditional method: > Relationship Using Scatter Plots
–0.81 > –2.575 The null hypothesis is not rejected * Scatterplots – depicts the relationship between two
(Do not reject H0) variables
Step 5:  Perfect Positive Correlation
The is no significant difference between the sample
mean and the population mean.
 The professional organization for private colleges
and universities professors reported that more than
17% of professors attended a national convention
in the past years. To test the claim, a researcher
surveyed 200 prof. and found out that 45 had
attended a national convention for the past years.
At 0.05 LS, test the claim that this figure is correct
 Positive Correlation
using p-value.
Step 1:
H 0 : p=0.17
H a : p>0.17
Step 2:
a=0.05, one-tailed, right-tailed
 > Correlation
 Statistical method to determine existence, strength,
and direction of relationship between two variables
> Pearson Product – Moment Correlation Coefficient
(Pearson r)
 Formula for Pearson r
n Σ xy−Σ x ∙ Σ y
r=
 Negative Correlation √ [ n Σ x − ( Σ x ) ] [n Σ y −( Σ y ) ]
2 2 2 2

 Interpretation
Pearson r Qualitative Description
±1 Perfect Correlation
± 0.75 ¿<± 1 Very High Correlation
± 0.50 ¿<± 0.75 Moderately High Correlation
± 0.25 ¿<± 0.50 Moderately Low Correlation
¿ 0 ± ¿<± 0.25 Very Low Correlation
 No Correlation 0 No Correlation

> Example
 The following data shows the term test scores of 10
students in Pagbasa at Pagsulat and General
Chemistry. Determine if there is a relationship
between the scores in Pagbasa at Pagsulat and
General Chemistry. Interpret the results.
Students Term Test Scores Term Test Scores
in PagPag x in GenChem y
> Relationship Using Graph
1 81 83
 Positive Association Correlation
2 84 88
3 87 82
4 88 85
5 90 84
6 78 75
7 93 90
8 86 89
9 80 79
10 91 96
 Negative Association/Correlation
Solution:

 No Association/Zero Correlation

n Σ xy−Σ x ∙ Σ y
r=
√ [ n Σ x − ( Σ x ) ] [n Σ y −( Σ y ) ]
2 2 2 2

(10)(73 , 21 9)−(858)(851)
r=
√ [ (10)( 73,840)−( 858 ) ] [ ( 10 ) (72,741)−( 851 ) ]
2 2
 2,032
r=
2 , 67 5 .3176
r =0.7 6
This indicates positive very high correlation.