BUSINESS STATISTICS
PRACTICUM WEEK 2 MB-48-INT-1
NEXT SLIDE
MB-48-INT-01
Topics Inferential
Statistics
F-Test Independent Paired Samples
NEXT SLIDE
Samples T-Test T-Test
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
INFERENTIAL
STATISTICS
STATISTIC BUSINESS
Inferential Statistics
Purpose of Inferential Statistics
This technique is used to compare the results obtained from a sample with those
of the entire population to determine the level of similarity.
Hypothesis in Statistical Inference
Null Hypothesis (H0): This hypothesis is tested using a statistical procedure and
is usually a statement that there is no difference or no relationship in the
population.
Alternative Hypothesis (H1): This is the opposite of the null hypothesis. It
asserts that there is a difference or a relationship, and is used to show that the
data strongly supports this conclusion.
F Test
Purpose of F-Test
The F test is a test that is used to determine how the effect of the variance
between the independent variable samples and the dependent variable in the two
samples.
Formula
Where:
2
S : Sample Variance
Xi: Each Individual Data
- : Mean (Average) of all data points
X
n: Number of data points in the sample
Sample Case
Mc. Donald and KFC are two restaurants that are constantly competing to
improve their service. Mc. Donald advertises that they serve faster compared to
KFC. To determine if this ad is valid, you and your friends have decided to order 10
food packages from Mc. Donald and 10 food packages from KFC. The service time
in minutes is as follows:
If the significance level is
0.05, is there a difference
in variance between the
service time Mc. Donald
and KFC?
MB-48-INT-01 HOME ABOUT SERVICE
DATA
NORMALITY
TEST
STATISTIC BUSINESS
Remember
Decision Making
If sig < 0.05 then the data is not normally distributed.
If sig > 0.05 then the data is normally distributed.
MB-48-INT-01 HOME ABOUT SERVICE
Open SPSS
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Edit the Data in Variable View
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Click Values Column
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Fill Values Column
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Fill Values Column
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Fill Values Column
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Fill the Data View
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
CLick Descriptive Statistics - Explore
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Make sure the Dependent
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Click the Requirements
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Result
NEXT SLIDE
STATISTIC BUSINESS
Result
Based on these data that the service time Mc. Donald with KFC
is included in the Shapiro-Wilk test, because the data is less
than 50.
From Shapiro-wilk normality results can be seen sig. as big
as 0.121
0.121 > 0.05 (so it can be concluded that the data Mc Donald
and KFC service times are distributed normal).
MB-48-INT-01 HOME ABOUT SERVICE
F TEST
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Back to Data View
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Click One Way Anova
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Seperate the Variable
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Pointing the Criteria
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Result
NEXT SLIDE
STATISTIC BUSINESS
MB-48-INT-01
F Test Step Creating Hypothesis
Analysis
Determine F table Decision Rule Decision-
and F count making NEXT SLIDE
STATISTIC BUSINESS
Make a Hyphotesis
Determine F table and F count
F-Table
Level significanceis 5%, (️=0.05).
Degree of Freedom:
df numerator (df1): n (total data in one company) – 1 = 9
df denominator (df2) : n total (total data in two companies) - 2 = 18
FU (F upper) = F 0.025 ; 9;18 = 2.93
FL (F Lower) = 1/F 0.025 ; 9;18= 1/2.93= 0.341
F count: from the output result SPSS = 4.592
Rule Decision
2 Ways
Withcompare f tableand f count
By comparing the value of sig
If F count > F table uppers, so H0 rejected
If score significance < 0.05, then H0 is rejected (there is a difference)
If the significance value is > 0.05 then variable H0 is accepted (no difference)
Taking Decision
From rule decision in on so:
Because the value of sig (0.046) < 0.05 and
F count (4.592) > F table upper (2.93), so H0 rejected, it means that
there is difference variance Among results model I with model II or
normal said variance assumed different.
MB-48-INT-01 HOME ABOUT SERVICE
T TEST
STATISTIC BUSINESS
T test
Definition Test are divided into 2 types:
T test is a comparative Independent Samples t-test
test to assess the Paired Samples t-test (Dependent-
difference between a Samples t-test)
certain value and the
average population
group.
Sample Case (Independent Samples T test)
Mc. Donald and KFC are two restaurants that
are constantly competing to improve their
service. Mc. Donald advertises that they
serve faster compared to KFC. To determine
if this ad is valid, you and your friends have
decided to order 10 food packages from Mc.
Donald and 10 food packages from KFC. The
service time in minutes is as follows:
If the significance level is 0.05,
is there sufficient evidence to conclude
that the average service time MC.D faster
than KFC? what the conclusion?
Here, the sample data is unrelated or
independent.
Analyze => Compare Means =>
Independent-Samples T Test
Click Define Groups => Type “1”
for “Group 1”, and type “2” for
“Group 2” => Continue
Click Options => Type “95” for
Confidence Interval Percentage
(or it is the same with a = 0.05)
means that the level of confidence
in making decisions is 95%.
Result
Making Analysis
Step Analysis
1. Creating a hypothesis
The hypotheses for the t-test case are:
H0: ️m1 ️≤ m2 or ️m1 - m2 ≤ 0
This means that the average service time [Link] NOT faster than KFC.
H1: ️m1 > ️m2 or ️m1 - ️m2 > 0
This means that the average service time MC Donald faster than KFC.
2. Determine t table and t count.
t table:
The significance level is 5%, (️= 0.05), while degrees of freedom (df) of 20 – 2= 18 (n = 10).
So for ️=0.05, t(0.05;18) on a one-sided t table (one tail) obtained -1,734.
t count:
Determination of t count, can be seen in column 't' in the result. Because in the F-test it is known
that the variances are assumed to be equal (equal variances assumed), then the t value is taken
as -2.143.
Statistics Table
Degree of Freedom = 18
The significance level is 5%, (️= 0.05)
So, the t table is 1.734
Making Analysis
Step Analysis
3. Decision rule:
By comparing t table and t count:
if t count ≥ t table, then H0 accepted. – [Link] no more fast
if t count < t table, then H0 rejected – Mc. Donald more fast
4. Decision-making:
From the decision rules above, then:
Because t count (-2,143) < t table (-1.734), then H0 meaning that the average service time
[Link] faster than KFC.
Sample Case (Paired Samples T test)
Company Indonesia Jaya has just
implemented a new marketing strategy for
one of their products which was started at
the beginning of the fiscal year. After closing
the book, the sales data is known as follows:
Are sales after the new marketing strategy
increasing?
(️= 0.05)
Here we see paired or related samples,
between the data before and after
implementing the new marketing strategy.
Entering the data =>
Analyze => Compare
Means => Paired
Samples T Test
Move “sebelum” to “variable 1”
and “sesudah” to “variable 2”
Click Options => Type “95” for
Confidence Interval Percentage
(or it is the same with a = 0.05)
means that the level of
confidence in making decisions
is 95%.
Click Options => Type “95” for
Confidence Interval Percentage
(or it is the same with a = 0.05)
means that the level of
confidence in making decisions
is 95%.
Result
Making Analysis
Step Analysis
1. Creating a hypothesis
The hypotheses for the t-test case are:
H0: ️md ≤0
That is, that the average volume sale after update strategy marketing not increase compared to
beforeupdate strategy marketing.
H1: ️md > 0
It means, that the average volume sale after update strategy marketing increased compared to
before update strategy marketing
2. Determine t table and t count.
t table:
The significance level is 5%, (️= 0.05), while degrees of freedom (df) of n – 1 = 11 (n = 12).
So for ️=0.05, t(0.05;11) on a one-sided t table (one tail) is obtained -1,7959.
t count:
From the results of the SPSS output above, in the column 't' the t value is calculated as -4.292.
Statistics Table
Degree of Freedom = 11
The significance level is 5%, (️= 0.05)
So, the t table is 1.7959
Making Analysis
Step Analysis
3. Decision rule:
By comparing t table and t count:
if t count ≥ t table, then H0 accepted.
if t count < t table, then H0 rejected
4. Decision-making:
From the decision rules above, then:
Because t count (-4.292) < t table (-1.7959), then H0 rejected, meaning that the average volume
sale after update strategy marketing increased compared to before pupdate strategy marketing.
MB-48-INT-01 HOME ABOUT SERVICE
ASSIGNMENT
STATISTIC BUSINESS
MB-48-INT-01 HOME ABOUT SERVICE
Task & Rules
The output of spss is pasted into word
Interpret the output
Sent with the format nim_name_class_week 2
The 5 fastest people get a bonus score
NEXT SLIDE
STATISTIC BUSINESS
Task 1 The following is sales data per week which is calculated in boxes:
Cora Corp. wants to analyze the
sales performance of the two
products it produces, namely
Cora Smartphone and Cora
Tablet.
Is there a difference variance
sales of the Cora smartphone
and the Cora tablet? Use the 5%
significance level!
To find out the data variance of
the two products is the same or
not, then you need to do the F
test first.
Task 2
A telecommunications company sent seven
employees to participate in a marketing
training program. The marketing abilities of the
seven employees were evaluated before and
after participating in the training program.
The following table shows the results of
the evaluation before and after the
training program of each employee in the
form of a score (consisting of a scale of 1
to 15, where 1 indicates the lowest score
and 15 the highest score).
Using a significance level of 0.01, is there
sufficient evidence that the training
program for the seven employees can
improve their marketing skills?
MB-48-INT-01 HOME ABOUT SERVICE
POINT
TIME !!!
STATISTIC BUSINESS
