Source: Primary Data, 2013 Hypothesis Ho: Data variable Facebook

CHAPTER 4
RESULTS
4.1 History of Full Color
Full Color event organizer is engaged in the party organizer service business.
Full Color was first established in 2000. The company's history is obtained by direct
interviews with David Ananda as the owner of Full Color on December 20, 2013 at
the office of Full Color.
At first, Full Color management structure is very simple. Along with the
development of the company, then David Ananda, as the owner of the company,
decided to recruit new personnel to join in Full Color. Starting from a small team,
Full Color has now developed into a party organizer to have about 100 employees
and is well known by the public as a credible party organizer. Full Color name
chosen by David Ananda based on a premise that every person has a different color
on each character and personality of each.
These colors represent the personality of each person different then the team
Full Color Party Organizer strives to achieve every event and party in accordance
with the wishes and dreams of each client who uses the services of Full Color. David
Ananda, as the owner and manager of Full Color event organizers to build this with
the principle of kindship and togetherness. He considers his team is a family,
therefore a Full Color can survive and grow rapidly enough until today. Full Color is
an event organizer that takes into account the satisfaction of its clients; therefore all
efforts are made to ensure that the client desired event could be done well and a
maximum in accordance with the character and desires of the client.
Based on the author's observation, 90% of the party or event held by Full
Color is a birthday party for the teen, who turned 17 years old or so-called sweet
seventeen party. At the beginning, if a client wants to hold a party activity, then a full
color team will hold a meeting with visiting directly the client's home to discuss the
overall concept and all of the activities they want.
In mid-2007, along with the growing number of clients, David Ananda
decided to open an office located at Mal Taman Anggrek, West Jakarta since before
all the activities carried out in the home of David Ananda.
52
53
4.2.
Company Profile
In this chapter the author will discuss the profile of Full Color. Full Color has
a Full Color Party Organizer name located in Mall Taman Anggrek lt. P2 no. 19,
West Jakarta. The company was established in 2000 and employs today
approximately 100 people consisting of 30 permanent employees and 70 temporary
employees. Communication with the Full Color often done by customers by visiting
the website www.fullcolorparty.com, via twitter or facebook of Full Color on @
fullcolorparty and can also be contacted via telephone at 021-56390543 (source:
Interviews with David Ananda as the owner of the day December 20, 2013 in the
Office Full Color)
4.2.1 Logo, Vision, Mission
Have a festive concept, cheerful and vibrant, full color logo has a wide range
of bright colors as the basis of the logo
Figure 4.1 Full Color Logo
Source: www.fullcolorparty.com
Vision: The total party and entertainment for you
Mission: 1 ) To be the best party organizer especially in Teenagers Party ( sweet
seventeen event ) in town
54
2) Give the service excellently
3) Creative and Uniqueness in every evenT
4) Brings the color that be the identity of the clients
4.2.2
Organizational Structure
Full Color organizational structure, is structurally composed of several
divisions that are divided in accordance with the interests of the company that helps
the operational of the company, with the following composition:
Managing Director
Senior Advisor, Public
Relations, Event
Captain
Creative Director
Captain
Marketing Director
Captain
Creative Design
Finance
Captain
Creative Invitation
Figure 4.2 Full Color Party Organizer Organization Structure
Source: Interview From Full Color Management
The job description of each division based company profile is as follows:
Senior Advisor, Public Relations, Event
Senior Advisor, Public Relations, Event headed by a Public Relations manager, who
supervises several captains. Captain is a person who plays the lead events
55
implemented by Full Color. The captain also serves as the Public Relations of Full
Color in charge of providing services to customers regarding the event will be held.
In addition to the job description of Senior Advisor also ensure that the event runs
smoothly. As a senior advisor also play a role in providing advice and making
decisions related to the company. And the job description as an event in Full Color,
also have a duty to prepare for the event either internal employees ( source:
Interviews with Full Color Management on December 10, 2013 at the office of Full
Color ).
Finance
Finance headed by the finance manager was in charge of all matters relating to the
cost of a full color ( cost ). Take into account, records and also controls all
expenditures and revenues of the Full Color, collect repayment of the event, and also
set the expenditure and income of Full Color and reported to the owner ( source:
Interviews with Full Color Management on December 10, 2013 at the office of Full
Color ).
Marketing Director
Duty to offer implementation services and also the birthday of school farewell party
or prom night. Marketing director in charge to ensure that the customer is getting
what is promised by Full Color -payment at the time had a down payment, make an
appointment with the captain continued to discuss the event, and also take care of the
rest of the payment and coordinated with the finance ( source: Interviews with Full
Color Management on December 10, 2013 at the office of Full Color ).
Creative Director
Creative director in charge of creating new concepts to satisfy customers who want a
party that is different from the other parties and the concept is reported to be a
captain for the concept of the show. In addition it is also the creative director in
charge of making videos that will be shown when the event took place ( source:
Interviews with Full Color Management on December 10, 2013 at the office of Full
Color ).
Creative Design
56
Tasked to design the location of the event, not only the layout of the stage and the
audience. Creative design has a role to decorate or decorate the event location in
accordance with the concept that the captain agreed to help build a different party
atmosphere ( source: interview with Full Color Management on December 10, 2013
at the office of Full Color ).
Creative Invitation
A creative invitation has a duty to assist customers in preparing the invitations to be
given to the guests. Creative invitation not only helps to think about the design you
want made and in accordance with the theme, a creative invitation also assist
customers in selecting materials to realize the invitation invitations so ready for
distribution ( source: interview with Full Color Management on December 10, 2013
at the office of Full Color ).
4.3
Respondent Profile
Table 4.1 Social Media Users
No.
Social
of
Media
Percentage Respondents
Facebook
31%
31
Twitter
42%
42
Instagram 27%
27
Total
100
Source: Primary Data, 2013
57
Social Media
42%
50%
31%
40%
27%
30%
20%
10%
0%
Facebook
Twitter
Figure 4.3 Social Media Users Bar Chart
Source: Social Media Questionnaire
Table 4.2 Usage Frequency
No.
Usage
Frequency
Percentage
Respondents
36%
36
52%
52
12%
12
Total
100
1-3
times/day
4-5
times/day
>5
times/day
of
Source: Primary Data, 2013
Instagram
58
Usage Frequency
52%
60%
50%
36%
40%
30%
12%
20%
10%
0%
1-3 kali sehari
4-5 kali sehari
Figure 4.4 Usage Frequency Bar Chart
Source: Frequency Usage Questionnaire
Table 4.3 Usage Frequency (in hours)
No.
of
No.
of
Hours
Percentage Respondents
< 1 hour
28%
28
37%
37
23%
23
12%
12
Total
100
1-2
hours
3-4
hours
>
hours
5
Source: Primary Data, 2013
> 5 kali sehari
59
No. of Hours
37%
40%
35%
30%
25%
20%
15%
10%
5%
0%
28%
23%
12%
< 1 jam
1-2 jam
3-4 jam
> 5 jam
Figure 4.5 Usage Frequency Bar Chart ( in Hours )
Source: Social Media Usage Frequency Questionnaire (in hours)
Table 4.4 Ads in Social Media
Whether
respondent
has
seen
ads
in
social
No.
of
media
Percentage Respondents
Yes
82%
82
No
18%
18
Total
100
60
Ads in Social Media
Tidak
18%
Ya
82%
Figure 4.6 Ads in Social Media Bar Chart
Source: Ads in Social Media Questionnaire
Table 4.5 Full Color Ads
Whether respondent has seen
Full Color’s Ads in Social
No.of
Media
Percentage
Respondents
Facebook
32.93%
27
Twitter
43.90%
36
Instagram
23.17%
19
Total
82
Source: Primary Data, 2013
61
Ads Full Color
43.90%
50.00%
40.00%
32.93%
23.17%
30.00%
20.00%
10.00%
0.00%
Facebook
Twitter
Figure 4.7 Full Color Ads Bar Chart
Source: Full Color Ads Questionnaire
Instagram
62
4.4
Validity Test
Test Validity shows the extent to which a hypothesis is measured accurately what it
intends to measure (Sekaran 2010:157). Regression results from Pre – Test ( 30
respondents ) for Facebook variable can be seen below:
Table 4.6 Validity Results of Pre – Test for Facebook Variable
Indicator
Corrected Item-Total
Correlation
Description
F1
.739
Valid
F2
.786
Valid
F3
.692
Valid
F4
.858
Valid
F5
.571
Valid
F6
.698
Valid
F7
.648
Valid
F8
.846
Valid
F9
.824
Valid
F10
.674
Valid
Source: ( Primary Data, 2013 )
If r - count which is the value of Corrected Item-Total Correlation > r - table, then it
is valid. r table for a sample of 30 respondents is 0.361 with a 95% confidence level.
In the table above it can be seen that the entire value of Corrected Item-Total
Correlation greater than 0.361, so that all indicators are declared invalid. (Supranto,
2012:93).
Validity test results of Pre-Test ( 30 respondents ) to a variable Twitter can be seen
in the table below:
63
Table 4.7 Validity Results of Pre – Test for Twitter Variable
Indicator
Corrected Item-Total
Correlation
Description
T1
.773
Valid
T2
.852
Valid
T3
.407
Valid
T4
.565
Valid
T5
.716
Valid
T6
.556
Valid
T7
.744
Valid
T8
.554
Valid
T9
.614
Valid
T10
.773
Valid
Source: ( Primary Data, 2013 )
If r - count which is the value of Corrected Item-Total Correlation > r - table, then it
is valid. r table for a sample of 30 respondents is 0.361 with a 95% confidence level.
In the table above it can be seen that the entire value of Corrected Item-Total
Correlation greater than 0.361, so that all indicators are declared invalid. (Supranto,
2012:93).
Validity test results from Pre-Test ( 30 respondents ) for Instagram variables can be
seen in the table below:
Table 4.8 Validity Results of Pre – Test for Instagram Variable
Indicator
Corrected Item-Total
Correlation
Description
I1
.719
Valid
I2
.762
Valid
I3
.486
Valid
64
I4
.403
Valid
I5
.471
Valid
I6
.573
Valid
I7
.626
Valid
I8
.747
Valid
I9
.662
Valid
I10
.723
Valid
Source: ( Primary Data, 2013 )
If r - count which is the value of Corrected Item-Total Correlation > r - table, then it
is valid. r table for a sample of 30 respondents is 0.361 with a 95% confidence level.
In the table above it can be seen that the entire value of Corrected Item-Total
Correlation greater than 0.361, so that all indicators are declared invalid. (Supranto,
2012:93).
Validity test results from Pre-Test ( 30 respondents ) for Purchase Decision variables
can be seen in the table below:
Table 4.9 Validity Results of Pre – Test for Purchase Decision Variable
Indicator
Corrected Item-Total
Correlation
Description
PD1
.769
Valid
PD2
.882
Valid
PD3
.437
Valid
PD4
.648
Valid
PD5
.704
Valid
PD6
.536
Valid
PD7
.749
Valid
PD8
.555
Valid
PD9
.614
Valid
65
PD10
.769
Valid
PD11
.855
Valid
PD12
.648
Valid
Source: ( Primary Data, 2013 )
If r - count which is the value of Corrected Item-Total Correlation > r - table, then it
is valid. r table for a sample of 30 respondents is 0.361 with a 95% confidence level.
In the table above it can be seen that the entire value of Corrected Item-Total
Correlation greater than 0.361, so that all indicators are declared invalid. (Supranto,
2012:93).
4.5
Reliability Test
Once all the indicator is valid, then it will do the reliability test. Reliability is a term
used to indicate the extent to which a measurement result are relatively consistent if
the measurement is repeated twice or more (Sekaran, 2010:157). Reliability test
results for 30 of the respondents can be seen in the table below.
Table 4.10 Reliability Results of Pre – Test of Facebook, Twitter,
Instagram, Purchase Decision Variable
Variable
Cronbach’s Alpha
Description
Facebook
.929
Reliable
Twitter
.894
Reliable
Instragram
.884
Reliable
Purchase
Decision
.917
Reliable
Source: ( Primary Data, 2013 )
From the table above, it can be seen that all the Cronbach's Alpha value of all the
above variables above 0.60, then it can be declared reliable. It can be said to be in
66
accordance with the requirements of reliable measuring scale with a Cronbach's
Alpha value of at least 0.60 shall be declared reliable. (Sekaran, 2010:157)
4.6
Normality Test
How to test for normality with the graph approach is to use the Normal Probability
Plot. The normal distribution is described by a straight diagonal line from the lower
left to the upper right. Normality test occurs if the test results indicate that the data is
spread around the diagonal line and follow the direction of the diagonal line. To test
the normality of population data, it can be seen from the picture the normal p-plot,
when the data spread around the diagonal line and follow the diagonal direction, the
data can be called normal (Suliyanto, 2011:69).
Normality test is done by distributing questionnaires to 100 respondents. This
test is conducted to determine the data that gathered is distributed normally. The
distribution can be normal or abnormal, normal distribution if the sig. kolmogorovsmirnov test is greater than 0.05, Data is not distributed normally if the value of sig.
kolmogorov-smirnov smaller than 0.05. The following are results of normality test
Facebook (X1), Twitter (X2), Instagram (X3), and purchase Decision (Y)
4.6.1
Normality Test Facebook Variable (X1)
To determine the data distribution of variable Facebook normal or not, the
data is analyzed through software SPSS for windows. This is the following result:
67
Table 4. 11 Normality Test Facebook Variable (X1)
Source: Primary Data, 2013
Hypothesis
Ho: Data variable Facebook is distributed normally
Ha: Data variable Facebook isn’t distributed normally
Basis for Decision Making
If sig Kolmogorov-Smirnov > 0.05 then the Ho is accepted
If sig Kolmogorov-Smirnov < 0.05 then the Ho rejected
Decision
Sig of variable
Facebook is 0.052, and then it’s concluded the data is
distributed normally.
4.6.2 Normality Test Twitter Variable (X2)
To determine the data distribution of variable Twitter normal or not, the data
is analyzed through software SPSS for windows. This is the following result:
68
Table 4. 12 Normality Test Twitter Variable (X2)
-
Source: Primary Data, 2013
Hypothesis
Ho: Data variable Twitter is distributed normally
Ha: Data variable Twitter isn’t distributed normally
Basis for Decision Making
If sig Kolmogorov-Smirnov > 0.05 then the Ho is accepted
If sig Kolmogorov-Smirnov < 0.05 then the Ho rejected
Decision
Sig of variable frontline displayed work is 0.090, and then it’s concluded the
data is distributed normally
4.6.3
Normality Test Instagram Variable (X3)
To determine the data distribution of variable Instagram normal or not, the
data is analyzed through software SPSS for windows. This is the following result:
69
Table 4. 13 Normality Test Instagram Variable (X3)
Source: Primary Data, 2013
Hypothesis
Ho: Data variable Instagram is distributed normally
Ha: Data variable Instagram isn’t distributed normally
Basis for Decision Making
If sig Kolmogorov-Smirnov > 0.05 then the Ho is accepted
If sig Kolmogorov-Smirnov < 0.05 then the Ho rejected
Decision
Sig of variable brand image is 0.085, and then it’s concluded the data is
distributed normally.
4.6.4 Normality Purchase Decision Variable
To determine the data distribution of Purchase Decision variable normal or
not, the data is analyzed through software SPSS for windows. This is the following
result:
70
Table 4. 14 Normality Test Purchase Decision ( Y )
Source: Primary Data, 2013
Hypothesis
Ho: Data variable Purchase Decision is distributed normally
Ha: Data variable Purchase Decision isn’t distributed normally
Basis for Decision Making
If sig Kolmogorov-Smirnov > 0.05 then the Ho is accepted
If sig Kolmogorov-Smirnov < 0.05 then the Ho rejected
Decision
Sig of variable Purchase Decision is 0.121, and then it’s concluded the data is
distributed normally.
71
Figure 4.8 Normal P-P Plot of Regression Standardized Residual
Source: Primary Data, 2013
Normality seen from the graph above (Normal P-Plot of Regression Standardized
Residual) seen some dots spread around the diagonal line and follow the direction of
the diagonal then the data can be considered normal. (Suliyanto, 2011:69).
72
4.7
Transformation Test
Facebook
Table 4.15 Transformation Test Facebook
Ordinal Variabel X1
1
2
3
4
5
Interval Variabel X1
1
1.51
2.27
3.44
4.78
Source: Primary Data, 2013
Twitter
Table 4.16 Transformation Test Twitter
Ordinal Variabel X2
1
2
3
4
5
Interval Variabel X2
1
1.59
2.65
3.83
4.95
Source: Primary Data, 2013
Instagram
Table 4.17 Transformation Test Instagram
Ordinal Variabel X3
1
2
3
4
5
Interval Variabel X3
1
1.56
2.73
3.92
5.02
Source: Primary Data, 2013
73
Purchase Decision
Table 4.18 Transformation Test Purchase Decision
Ordinal Variabel Y
1
2
3
4
5
Interval Variabel Y
1
1.71
2.79
3.89
5.01
Source: Primary Data, 2013
4.8
Simple Regression Analysis
4.8.1. Facebook Influence on Purchase Decision
Table 4.19 Model Summary
Model Summaryb
Model
1
R
.948a
Adjusted R
Std. Error of the
Square
Estimate
R Square
.899
.898
.20858
a. Predictors: (Constant), FACEBOOK
b. Dependent Variable: PURCHASE_DECISION
Source: ( Primary Data, 2013 )
In the R column in the above table for 0.948 shows that the relationship between
variables facebook and very strong purchasing decisions. R Square value of 0.899 for
89.9 % means that purchasing decisions are influenced by the facebook and the
remaining 10.1 % is influenced by other variables.
74
Table 4.20 Anova b
ANOVAb
Model
1
Sum of Squares
Regression
Residual
Total
df
Mean Square
F
37.808
1
37.808
4.264
98
.044
42.071
99
Sig.
868.993
.000a
a. Predictors: (Constant), FACEBOOK
b. Dependent Variable: PURCHASE_DECISION
Source: ( Primary Data, 2013 )
Sig value on F test of 0.000 of < 0.05, it can be said facebook has a significant
influence on purchasing decisions.
Table 4.21 Coefficients a
Coefficientsa
Standardized
Unstandardized Coefficients
Model
1
B
Std. Error
(Constant)
.241
.119
FACEBOOK
.935
.032
Coefficients
Beta
t
.948
Sig.
2.018
.046
29.479
.000
a. Dependent Variable: PURCHASE_DECISION
Source: ( Primary Data, 2013 )
Sig on t test of 0.000 < 0.05, we can say facebook has a significant influence on
purchasing decisions.
Hypothesis:
Ho = There is no significant effect of Facebook on purchasing decisions
H1 = There is a significant effect of Facebook on purchasing decisions
Basic Decision Making:
Ho is accepted if the Sig > 0.05
Ho is rejected if the Sig < 0.05
75
Conclusion:
Sig value on F test and t test for 0.000 of < 0.05, then Ho is rejected, which means
there is a significant effect of Facebook on purchasing decisions
4.8.2. Twitter Influence on Purchase Decision
Table 4.22 Model Summary b
Model Summaryb
Model
R
Std. Error of the
Square
Estimate
R Square
.750a
1
Adjusted R
.562
.558
.43356
a. Predictors: (Constant), TWITTER
b. Dependent Variable: PURCHASE_DECISION
Source: ( Primary Data, 2013 )
In the R column in the above table for 0.750 shows that the relationship between
variables twitter and purchase decisions are strong. R Square value of 0.562 means
that 56.2% of purchase decisions are influenced by the twitter and the remaining 43.8%
is influenced by other variables.
Table 4.23 Anova b
ANOVAb
Model
1
Sum of Squares
df
Mean Square
Regression
23.650
1
23.650
Residual
18.421
98
.188
Total
42.071
99
F
125.816
Sig.
.000a
a. Predictors: (Constant), TWITTER
b. Dependent Variable: PURCHASE_DECISION
Source: ( Primary Data, 2013 )
Sig value on F test of 0.000 of < 0.05, we can say twitter has a significant influence
on purchasing decisions.
76
Table 4.24 Coefficients a
Coefficientsa
Standardized
Unstandardized Coefficients
Model
1
B
Std. Error
(Constant)
.606
.279
TWITTER
.832
.074
Coefficients
Beta
t
.750
Sig.
2.169
.032
11.217
.000
a. Dependent Variable: PURCHASE_DECISION
Source: ( Primary Data, 2013 )
Sig on t test of 0.000 < 0.05, we can say twitter has a significant influence on
purchasing decisions.
Hypothesis:
Ho = There is no significant effect of Twitter on purchasing decisions
H1 = There is a significant effect of Twitter on purchasing decisions
Basic Decision Making:
Ho is accepted if the Sig > 0.05
Ho is rejected if the Sig < 0.05
Conclusion:
Sig value on F test and t test for 0.000 of < 0.05, then Ho is rejected, which means
there is a significant effect of Twitter on purchasing decisions
77
4.8.3 Instagram Influence on Purchase Decision
Table 4.25 Model Summary b
Model Summaryb
Model
R
.739a
1
Adjusted R
Std. Error of the
Square
Estimate
R Square
.547
.542
.44121
a. Predictors: (Constant), INSTAGRAM
b. Dependent Variable: PURCHASE_DECISION
Source: ( Primary Data, 2013 )
In the R column in the above table for 0.739 shows that the relationship between
variables twitter and purchase decisions are strong. R Square value of 0.547 means
that 54.7% of purchase decisions are influenced by the twitter and the remaining 45.3%
is influenced by other variables.
Table 4.26 Anova b
ANOVAb
Model
1
Sum of Squares
df
Mean Square
Regression
22.994
1
22.994
Residual
19.077
98
.195
Total
42.071
99
F
Sig.
118.119
.000a
a. Predictors: (Constant), INSTAGRAM
b. Dependent Variable: PURCHASE_DECISION
Source: ( Primary Data, 2013 )
Sig value on F test of 0.000 of < 0.05, we can say Instagram has a significant
influence on purchasing decisions.
Table 4.27 Coefficients a
Coefficientsa
Standardized
Unstandardized Coefficients
Model
1
B
(Constant)
Std. Error
.559
.292
Coefficients
Beta
t
1.913
Sig.
.059
78
INSTAGRAM
.841
.077
.739
10.868
.000
a. Dependent Variable: PURCHASE_DECISION
Source: ( Primary Data, 2013 )
Sig on t test of 0.000 < 0.05, we can say Instagram has a significant influence on
purchasing decisions.
Hypothesis:
Ho = There is no significant effect of Instagram on purchasing decisions
H1 = There is a significant effect of Instagram on purchasing decisions
Basic Decision Making:
Ho is accepted if the Sig > 0.05
Ho is rejected if the Sig < 0.05
Conclusion:
Sig value on F test and t test for 0.000 of < 0.05, then Ho is rejected, which means
there is a significant effect of Instagram on purchasing decisions
4.9 Multiple Regression Analysis
Table 4.28 Model Summary b
Model Summaryb
Model
1
R
R Square
.957a
Adjusted R
Std. Error of the
Square
Estimate
.916
.913
a. Predictors: (Constant), INSTAGRAM, FACEBOOK, TWITTER
b. Dependent Variable: PURCHASE_DECISION
Source: ( Primary Data, 2013 )
.19181
Durbin-Watson
2.633
79
In the R column in the above table for 0.957 shows that the relationship between the
independent variables (facebook, twitter and instagram) with a very strong buying
decision. R Square value of 0.916 means that 91.6% of purchase decisions are
influenced by facebook, twitter and instagram and the remaining 8.4% is influenced
by other variables.
Table 4.29 Anova b
ANOVAb
Model
1
Sum of Squares
Regression
Mean Square
F
38.539
3
12.846
3.532
96
.037
42.071
99
Residual
Total
df
Sig.
.000a
349.175
a. Predictors: (Constant), INSTAGRAM, FACEBOOK, TWITTER
b. Dependent Variable: PURCHASE_DECISION
Source: Primary Data, 2013
Sig value on F test of 0.000 of < 0.05, it can be said facebook, twitter and instagram
jointly have a significant influence on purchasing decisions.
Table 4.30 Coefficients a
Coefficientsa
Standardized
Unstandardized Coefficients
Model
1
B
Std. Error
(Constant)
.263
.129
FACEBOOK
.924
.046
TWITTER
.875
-.866
INSTAGRAM
Coefficients
Beta
Collinearity Statistics
t
Sig.
Tolerance
VIF
2.041
.044
.936
20.117
.000
.404
2.477
.197
.789
4.443
.000
.028
36.056
.207
-.761
-4.194
.000
.027
37.690
a. Dependent Variable: PURCHASE_DECISION
Source: Primary Data, 2013
Sig on facebook variable value of 0.000 < 0.05, we could say that facebook has a
significant influence on purchasing decisions.
80
Sig on twitter variable value of 0.000 < 0.05, we could say that twitter has a
significant influence on purchasing decisions.
Sig on instagram variable value of 0.000 < 0.05, we could say that instagram has a
significant influence on purchasing decisions.
Regression equation
Y = a + b1X1 + b2X2 + b3X3
Y = 0.263 + 0.924X1 + 0.875X2 - 0.866X3
Where:
Y = Purchase decision
X1 = Facebook
X2 = Twitter
X3 = Instagram
Regression coefficient = 0.263 means that if there is no facebook, twitter and
instagram, then the constants to the purchasing decision is 0.263
Regression coefficient of X1 = 0.924 means the addition of each variable Facebook,
then the constants to the purchasing decision of 0.924
Regression coefficient of X2 = 0.875 means the addition of each variable Twitter,
then the constants to the purchasing decision of 0.875
Regression coefficient of X3 = - 0.866 means the addition of each variable Instagram,
then the constants to the purchasing decision of - 0.866
Hypothesis:
Ho = There is no significant effect of facebook, twitter and instagram on purchasing
decisions
H1 = There is a significant effect of facebook, twitter and instagram on purchasing
decisions
Basic Decision Making:
81
Ho is accepted if the Sig > 0.05
Ho is rejected if the Sig < 0.05
Conclusion:
The results of the Sig on the F test and t test of < 0.05, then Ho is rejected, which
means there is a significant effect of facebook, twitter and instagram on purchasing
decision
4.10 Summary of Data Processed
Table 4.31 Summary of Data Processed
Influence
The amount
Influence
of Regression equation
Result
X1 to Y
89.9 %
Y = 0.241+0.935X
There
is
a
significant effect
X2 to Y
56.2 %
Y = 0.606 +0.832X
There
is
a
significant effect
X3 to Y
54.7%
Y = 0.559+ 0.841X
There
is
a
significant effect
X1,X2&X3 to Y
91.6%
Y= 0.263+0.924X1+ There
is
a
0.875X2-0.866X3
significant effect
Source: ( Primary Data, 2013 )
The influence of Facebook (X1) on the purchase decisions is 89.9%
The influence of Twitter (X2) on the purchase decisions is 56.2.%
The influence of Instagram (X3) on the purchase decisions is 54.7%
The influence of Facebook, Twitter and Instagram jointly to the purchasing decision
is 91.6%.