DESCRIPTIVE

HYPOTHESIS TEST
By: Group 2 BESP 2019

Eunike M (4193141007)
Fadhillah R (4193342004)
Gresia P (4193342003)
Maria N (4193141022)
 CONTENT
01 Definition 02 Requirements

03 Parametric test 04 Non-

Parametric test
 01

Definition
• Hypothesis (hypothesis) comes from Greek. Hypothesis comes from
2 word fragments, "hypo" means temporary and "thesis" means
conclusion.

• A descriptive hypothesis is a temporary answer to a descriptive

problem, which is a test based on only one sample or an
independent variable and does not aim to make comparisons or look
for a relationship.
• The descriptive hypothesis to be
tested with parametric statistics is
the estimated value in a sample
(sample unit), compared to the
standard, while the descriptive
hypothesis to be tested with
nonparametric statistics is an
assumption that there is a
significant difference in the value
between groups in one sample.
For example, if the formulation of the research problem is as follows, then the
hypothesis (temporary answer) that is formulated is a descriptive hypothesis.
a. How durable is the X brand lamp?
b. How high is the productivity of rice in Klaten district?
c. How long will the lamp life of brands A and B be?
From the three statements, among others, a hypothesis can be formulated as
follows:
a. Durability of X brand lamp = 800 hours
b. Rice productivity in Klaten Regency is 8 tons / ha.
c. The durability of the lamp brand A = 450 hours and brand B = 600 hours.
 The steps in testing the descriptive hypothesis:
1. Calculate average data
2. Calculate standard deviation
3. Calculate the t / z price
4. Look at the T / Z price table
5. Draw curves
6. Putting the position in the t / z count and t / z table in the curve
drawing
7. Make a hypothetical decision
8. Analysis.
 02

Requirements
• two-party hypoteric test:
• H0 is accepted if -Z table <Z count <Z table.
• When using the formula t:
• H0 is accepted if -table <t count <t table

• For the one-sided (right-hand) hypothesis:

• H0 is accepted if zcount ≤ z0,5-α and H0 is rejected if zcount> z0,5-α.
• When using the formula t:
• H0 is accepted if tcount ≤ t table and H0 is rejected if tcount> ttable
• one-party hypothesis test (left party)
• H0 is accepted if z ≥ -z0,5-α H0 is rejected if z <-z0,5-α.
• When using the formula t:
• H0 is accepted if tcount ≥ttable and H0 is rejected if tcount <ttable
 03

Statistic
Parametric
Parametric statistics is a statistical technique that can be used
to test hypotheses involving population parameters.
Parametric statistics have limitations in the use of data types,
namely at least using interval and ratio data.
 Test T 1 Sample
T Test 1 Sample is the t test procedure for a single sample
when the average of a single variable is compared to a given
value. T test is used if the sample data count ≤ 30 (small
sample)
 t = calculated t value = t count

𝑥 = average of x
𝑥−μo
𝑡= μo = hypothesized value
s/√n
s = Standard Deviation

n = number of samples
 Stage 1 Determining hypotheses

Stage 2 Determining significance

Testing Procedure
Stage 3 Determining Test Statistics

Stage 4 Determining Testing Criteria and

Making Conclusions
E x a m p l e : No
Study Time ( Hours
/Week)
N
o
Study Time (Hours
/Week)
1 30 16 17
2 18 17 17
3 18 18 25
There is a hypothesis that the average study
4 26 19 15
time of PGMI Students of Teladan University is 5 25 20 25
6 23 21 15
more than 20 Hours / Week.
7 16 22 28
When randomly taken 30 students were 8 16 23 25
9 27 24 20
obtained data that the average student study
10 17 25 14
time is 22 hours / week with a standard 11 21 26 15
12 20 27 24
deviation of 5.83 Hours. Is the hypothesis
13 15 28 31
correct? 14 26 29 25
15 35 30 31
Stage 1 : Determining hypotheses

Ho : Average study time of PGMI Students of Teladan

University Less than equal to 20 hours / week Statistical hypotheses are:
Ho : μ ≤ 20
H1 : Average study time of PGMI Students of Teladan H1 : μ > 20
University is more than 20 hours/week

Stage 2 : Determining significance

The significance level used is 5 % or 0.05

In one-way testing, Table t Value obtained, t : 1,699

Stage 3 : Determining Test Statistics

𝑥−μo
Formula : 𝑡 = s/√n

22−20
𝑡 = 5,83/√30 = 1,879 So obtained Statistical value test t is 1,879

Stage 4 : Determining The Criteria for Exposure and making

conclusions
Testing criteria : using one-way test

Right direction : Test Statistical Values > Table Values

Test Statistical Value is 1,879 more than Table Value is 1,699 .

Stage 4 : Determining The Criteria for Exposure and making
conclusions
Statistical hypotheses are:
Ho : μ ≤ 20
H1 : μ > 20

Thus, H0 Rejected and H1 Accepted

Conclusion: With a signification rate of 5% or a confidence level of 95% that the average
study time of PGMI University students exemplary more than 20 Hours / week.
 T-Test dengan SPSS

One-Sample Statistics

N Mean Std. Deviation Std. Error Mean

Rata-rata 30 22.00 5.831 1.065

One-Sample Test
Test Value = 20

95% Confidence Interval

Sig. (2- Mean of the Difference
t df
tailed) Difference
Lower Upper
Rata-rata 1.879 29 .070 2.000 -18 4.18
Z-test
 Z Test statistics is a statistical procedure used to test an alternative
hypothesis against the null hypothesis. It is any statistical
hypothesis used to determine whether two samples' means are
different when variances are known and the sample is large.

the data must be normally distributed, then the population variance is known (σ2), then the sample
• the data must be normally distributed,
size (n) is large, ≥ 30, and is usually used only to compare 2 observations.
CHARACTERISTIC
• the population variance is known (σ2),
• the sample size (n) is large, ≥ 30,
• is usually used only to compare 2
observations.
 Basis for comparison T-test Z-test

z-test is a statistical tool used for the comparison or

The t-test is a test in statistics that is used for testing
determination of the significance of several statistical
hypotheses regarding the mean of a small sample taken
Definition measures, particularly the mean in a sample from a
population when the standard deviation of the
normally distributed population or between two
population is not known.
independent samples.

The t-test is usually performed in samples of a smaller z-test is generally performed in samples of a larger size
Sample size
size (n≤30). (n>30).

t-test is performed on samples distributed on the basis z-test is performed on samples that are normally
Type of distribution of population
of t-distribution. distributed.

A t-test is not based on the assumption that all key z-test is based on the assumption that all key points on
Assumptions
points on the sample are independent. the sample are independent.

Variance or standard deviation is not known in the t-

Variance or standard deviation Variance or standard deviation is known in z-test.
test.

The sample values are to be recorded or calculated by In a normal distribution, the average is considered 0
Distribution
the researcher. and the variance as 1.

In addition, to the mean, the t-test can also be used to

In addition, to mean, z-test can also be used to
Population parameters compare partial or simple correlations among two
compare the population proportion.
samples.

t-tests are less convenient as they have separate critical z-test is more convenient as it has the same critical
Convenience
values for different sample sizes. value for different sample sizes.
 Z Test Formula
Z Test = (x̄ – μ) / (σ / √n)
Z Test = (x̄ – μ) / (σ / √n)
With Here,
With Here,
x̄ = Mean of Sample
x̄ =ofMean
μ = Mean Population
of Sample
σ = Standard Deviation of Population
n = Number
μ = Meanof Observation
of Population

σ = Standard Deviation of Population

n = Number of Observation
Example with the case using Z test as solver

Two-tailed Z-Test
01 (Two Tail Test)

One-party Z test 02
(One Tail Test)

• The left side test

• The right side test
 Two-tailed Z-Test (Two Tail Test)

The two-part test is used when the null

hypothesis (Ho) reads "is the same as" and
the alternative hypothesis (Ha) says "is not
the same as". H0 = ; Ha ≠
 Example of hypothesis formulation:

Zero hypothesis: student learning endurance

in class per day = 5 hours
Alternative hypothesis: student learning
endurance in class per day ≠ 5 hours
If written more concisely:
H0: μ = 5 hours
Ha: μ ≠ 5 hours
 Example:

 If a sample of 400 male workers has a mean height of 67.47 inches, is it reasonable to regard the
sample as a sample from a large population with a mean height of 67.39 inches and a standard
deviation of 1.30 inches at a 5% level of significance?

Taking the null hypothesis that the mean height of the population is equal to 67.39 inches, we can
write:

Found : H0: µ = 67.39“

Ha: µ ≠ 67.39“

x̄ = 67.47“, σ = 1.30“, n = 400

Ask : Ho accepted or not..?

Answer: Assuming the population to be normal, we can work out
the test statistic z as under:

As Ha is two-sided in the given question, we shall be applying a

two-tailed test for determining the rejection regions at a 5% level
of significance which comes to as under, using normal curve area
table: Rα: | z | > 1.96 .The observed value of t is 1.231 which is in
the acceptance region since R: | z | > 1.96, and thus, H0 is
accepted.
 One-party Z test (One Tail Test)

• Left side test • Right side test

The left side test is used if: the null The right side test is used when
• Test one party left
hypothesis (Ho) reads "greater or the"greater
null hypothesis (H0)
The left side test is used if: the null hypothesis (Ho) reads or equal to (≥)" reads
and the
equalhypothesis
alternative to (≥)" and thesays
(Ha) alternative
"smaller (<)", the word"smaller
is greateror
or equal
equal to
to the
(≤)"synonym
and the"the
word hypothesis
is the least or the smallest".
(Ha) says "smaller (<)", alternative hypothesis (ha) reads
the word is greater or equal to the "greater (>)". Sentences smaller
synonym "the word is the least or than or equal to synonyms with
the smallest". the word "largest".
Ho is accepted if Z ≥ -Z0,5-α, Ho is rejected if Z < -Z0,5-α Ho is accepted if Z ≤ -Z0,5-α, Ho is rejected if Z> -Z0,5-α
 Example of the formulation of the left- Example of the formulation of the
party Test hypothesis: right-party Test hypothesis:

Null hypothesis: student learning Null hypothesis: the largest fruit

endurance at home (outside the seller can sell 100kg of oranges
classroom) is at least 60 minutes (greater per day
than or equal to (≥ 60 minutes)
Alternative hypothesis: fruit
Alternative hypothesis: student learning traders can sell more than 10 kg
endurance at home is less than (<) 60 of citrus fruit per day.
minutes.
or it can be written briefly:
Or it can be written briefly: H0: μ ≤ 100 kg / day
H0: μ ≥ 60 minutes Ha: μ> 100 kg / day
Ha: μ <60 minutes
Example:

Urea has 2 forms, namely granules and tablets. The granular form came first while the
tablet form was a new form. It is known that the yield of paddy grain fertilized with grain
urea has an average of 40 resistance. One researcher believes that tablet urea is better than
granular urea. Then he conducted research with n = 30 replications and the results were as
follows:

Paddy grain yield in t / ha

4,0 5,0 6,0 4,2 3,8 6,5 4,3 4,8 4,6 4,1

4,9 5,2 5,7 3,9 4,0 5,8 6,2 6,4 5,4 4,6

5,1 4,8 4,6 4,2 4,7 5,4 5,2 5,8 3,9 4,7
Hypothesis:

Ho: = (the average yield of rice grain fertilized with urea

tablet is the same as rice fertilized with granulated urea)

Ha: > (the average yield of paddy grain fertilized with urea
tablet is higher than rice fertilized with granulated urea)
Analysis = 4.0 t / h = 4.9 t / h S = 0.78 is used as an estimate

σcount = (yt - yb) / (σ / √n) = (4.0 - 4.9) / (0, 78 / √30) = - 6.4286

Ztabel = Zα = Z0,05 = 1,645 S.

Conclusion Criteria

If | Zcount | <| Ztabel |, then Ho is accepted

If | Zcount | ≥ | Ztabel |, then Ho is reject and Ha is accepted

Because of the price | Zcount | = 6,4286 > price | Ztabel | = 1.645, then reject Ho,
aka accept Ha. So, the average yield of paddy grain fertilized with urea tablet was
significantly higher than rice fertilized with granulated urea.
 04

Statistic Non
Parametric
Estimation of parameters or population values is called
parametric statistics while non-parametric statistics do
not involve estimating population values. Non-
parametric statistics is a statistical method that can be
used as an alternative if parametric statistical methods
cannot be used.
BINOMIAL
TEST
 1. The binomial test is
used to test a hypothesis
if the population consists If the price P <α → Ha
of two class / category
is accepted
groups and the number
of samples is small (less
than 25).
The
characteristic
2. In practice, to prove
Ho, the binomial test If the price P> α → Ho
can be done by is accepted
comparing the P value Ho is a hypothesis
in the table based on that shows no
N from the smallest X difference in
value with an error frequency
rate of 1%.
 Example:
An automotive company produces two types of
cars, namely cars that run on gasoline and diesel
fuel. The company wants to see if the public prefers
cars running on gasoline or diesel. Based on 24
sample members who were randomly selected, it
turned out that 14 people chose gasoline-powered
cars and 10 people chose diesel-fueled cars
 The hypothesis proposed is
Ho: The number of people who choose
gasoline and diesel cars is no different
Ha: The number of people who choose cars
with gasoline and diesel fuel is different

Ho: p1 = p2 = 0
Ha: p1 ≠ p2 = 0.5
RESEARCH RESULT

Alternative Options Which frequency to choose

Gasoline type car 14

Diesel type car 10
amount 24

Statistical techniques for testing hypotheses

Based on the formulated question
hypothesis, is a descriptive hypothesis (one
variable / one sample) the data is nominal in
form with a sample size <25, the statistical
technique used is the binomial test.
 From the case it is known that N = 24, the smallest
frequency (x) = 10
By looking at the table, the binomial coefficient value
= 0.271
If the error level (α) = 0.01
Then P = 0.271> 0.01 → Ho is accepted
The conclusion: the frequency of people choosing two
types of cars is the same
 Chi Kuadrat
 The statistical technique used to test the descriptive
hypothesis when the population consists of two or more
classes, the data is nominal in form with a large sample.

 The descriptive hypothesis is an assumption of whether or

not there is a difference in frequency between one
category and another in a sample
Rumus dasar Chi Kuadrat adalah sbb:
k
( fo  fh) 2
Х2 = 
i 1 fn

X2 = Chi Kuadrat
fo = Frekuensi yang diobservasi
fh = Frekuensi yang diharapkan
 Example of two categories
A women's organization wants to find
out whether women have the same
opportunities as men to become
members of the legislature. This
requires research. The study population
Example was the people of the city of Medan.
Candidate I is a woman and candidate
II is a man. The sample as a data source
was taken randomly as many as 300
people. From this sample, it turns out
that 200 people chose men and 100
people chose women.
 Based on this, then:
• The title of the research:
Opportunities for men and women to
become members of the legislature

• Research variant: Member of the

Legislature
Example
• Sample
The study used one community
sample group with 300 members
consisting of two categories, namely
200 male voters and 100 female
legislative voters.

• Place of research: Medan City

 The results of data collection are
shown in the table:
Legislative Candidate Voters Data

Alternative Obtained frequency The expected

legislative frequency
candidates
Male Candidates 200 150
Female Candidates
100 150
Amount 300 300

The proposed hypothesis:

Ho: The frequency of people who elect male and female
candidates is the same
Ha: The frequency with which people elect male and
female candidates is not the same
Ho = p1 = p2 = 0.5
Ha = p1 ≠ p2 ≠ 0.5
 Hypothesis testing conditions:
If Chi Square count ≥ Chi Square Table → Ha is accepted
If Chi Square count <Chi Square table → Ho is accepted

Hypothesis test
Alternative fo fh ( fo –fh) (𝒇𝒐 − 𝒇𝒉)𝟐 (𝒇𝒐 − 𝒇𝒉)𝟐
choice
fh
Male 200 150 50 2500 16,67
Female 100 150 -50 2500 16,67
Amount 300 300 0 5000 33,33
 • To be able to make a decision, the Chi square price obtained must be
compared with the Chi Square table price with dk and a certain level of error
dk = n - 1 = 2 - 1 = 1
Error rate = 5%
then the price of the Chi Square table = 3.481
It turns out that the calculated chi square value = 33.33> chi squared table =
3.481 then Ha is accepted

• Conclusion:
The null hypothesis that is proposed that men and women have equal chances to
be elected as legislative members in the city of Medan is rejected. The results
showed that people in the city of Medan tend to elect men to become members
of the legislature

• Suggestion :
Women's groups do not need to run as members of the legislature
 RUN TEST

1. Run test is used 2. Tests are carried out by 3. Ho test is done by 4. If the
measuring population comparing the number observation run is
to test the
concordance based on of runs in the between the small
descriptive
observational data observation with the table and the large
hypothesis (one
through sample data by values in the table for the run then Ho is
sample) if the data
measuring the number of run test with a certain accepted
is in ordinal form.
"runs" in an event. level of significance
Example:
In one canteen in a company there is a group of female
employees who are having lunch. There were 24 people
taken randomly and then interviewed when to take
maternity leave. In the question, 2 alternative answers are
provided, namely going to take great leave before giving
birth or after giving birth.

• Based on the foregoing, then:

1. Research title:
2. Time to take the employee's major leave
3. The variables studied:
4. Time to take employee leave
5. The study sample was a group of 24 women, (N = 24)
6. Place of research: In a company
7. Research hypothesis:
8. Ho: The chances of taking leave before and after giving
birth are the same
9. Ha: Opportunities to take leave before and after giving
birth are not the same
 • Hypothesis testing criteria
If the observation run is between the
small run table value and the large run,
then Ho is accepted

• Presentation of data
No Answer no Answer
1 R 1 C
2 R 2 R
3 C 3 R
4 R 4 C
5 C 5 R
6 R 6 C
7 C 7 C
8 C 8 R
9 C 9 C
10 R 10 C
11 R 11 R
12 C 12 R
Jlh
run
• Hypothesis test
From the table it can be seen that the number of run = 15
With N = 24 d
then based on the run test table, the price of small r = 7 and large r
= 19
The number of runs = 15 turns out to be in numbers 7-19, namely in
the receiving area of Ho. Thus Ho was accepted and Ha was rejected

• Conclusion
In taking maternity leave, the female employees of the company vary,
some are before giving birth, some are after giving birth. The
opportunity to take leave before and after giving birth is the same,
namely 50%

• Suggestion
Companies can make a policy that major leave can be taken before
and after delivery
THANKS
From Group 2