Вы находитесь на странице: 1из 29

|

2
Research Analysis – What attracts you to facebook?
Client – Mrs Lopamudra Ghosh

Researchers –

Neeraj Singh – 09BS0001377


Neha Gupta – 09BS0001399
Nilankan Dey – 09BS0001439
Nishant Vora – 09BS0001462
Prakash Krishnamoorthy – 09BS0001615
Preha Sharma – 09BS0001667
Rachit Sharma – 09BS0001742
Ramkumar Venkiteswaran – 09BS0001822

3
Acknowledgement

I like to take this opportunity to thank all those who have


contributed to make this finding successful.

Special thanks to my peers and colleagues who have given


valuable input.

I am also indebted to Mrs Lopamudra Ghosh to have


given this project of great magnitude to this respectable group.

4
Letter of Authorization:

IBS Mumbai
Date: 5th January, 2010

TO WHOM SO EVER CONCERNED

This is to authorize Section ACE group to conduct a research on


FACEBOOK and interact with the faculty and the students of
the college.

Please assist in their endeavor.

IBS MUMBAI

5
TABLE OF CONTENTS

Sr No TOPIC Page No

1 List of Tables 6

2 List of Graphs 10

3 Executive Summary 11

4 Problem Definition 12

5 Approach to the Problem 13

6 Research Design 14

7 Data Analysis 16

8 Results 21

9 Limitations 23

10 Conclusions and Recommendations 24

11 Exhibits 26

6
LIST OF TABLES

• Correlation Matrix

• KMO and Bartlett’s Test

• Communalities

• Total Variance Explained

• Scree Plot

• Component Matrix

• Rotated Component Matrix

• Component Transformation Matrix

7
Correlation Matrix(a)
Chec
k out
how
Play your Upda
non- frien te
interacti ds your To Community
ve Cha are profil pass Simplici Horosco Personalizati Flirtin Flexibili
games t doing e time ty pe on g ty
Correlati Play non-
on interactive 1.000 .465 .375 .443 .365 .321 .324 .318 .264 .222
games
Chat .465 1.000 .791 .524 .540 .540 .321 .313 .392
Check
out how
your
.375 .791 1.000 .527 .599 .413 .126 .244 .229
friends
are
doing
Update
1.00
your .443 .524 .527 .558 .459 .391 .447 .406
0
profile

To pass .365 .540 .599 .558 1.000 .446 .174 .314 .090
time

Simplici .321 .540 .413 .459 .446 1.000 .436 .449 .253
ty

Horosco .324 .321 .126 .391 .174 .436 1.000 .647 .399
pe

Commu
nity .318 .313 .244 .447 .314 .449 .647 1.000 .252
Personal
ization
.264 .392 .229 .406 .090 .253 .399 .252 1.000
Flirting

Flexibili .222 .487 .440 .329 .306 .502 .315 .443 .232
ty

a Determinant = .010

KMO and Bartlett's Test


Kaiser-Meyer-Olkin Measure of Sampling
Adequacy. .820

Bartlett's Test of Approx. Chi-Square 366.124


Sphericity Df 45
Sig. .000

8
Communalities
Initial Extraction
Play non-interactive
games 1.000 .364
Chat 1.000 .764
Check out how your
friends are doing 1.000 .811
Update your profile 1.000 .588
To pass time 1.000 .638
Simplicity 1.000 .533
Horoscope 1.000 .791
Community
Personalization 1.000 .673
Flirting 1.000 .352
Flexibility 1.000 .408
Extraction Method: Principal Component Analysis.

Total Variance Explained


Extraction Sums of Squared
Component Initial Eigenvalues Loadings Rotation Sums of Squared Loadings
% of Cumulative % of Cumulative % of Cumulative
Total Variance % Total Variance % Total Variance %
1 4.55 4.55 3.32
45.543 45.543 45.543 45.543 33.292 33.292
4 4 9
2 1.36 1.36 2.59
13.677 59.221 13.677 59.221 25.929 59.221
8 8 3
3 .932 9.322 68.543
4 .813 8.131 76.673
5 .616 6.156 82.829
6 .502 5.023 87.853
7 .438 4.381 92.233
8 .331 3.313 95.546
9 .288 2.875 98.421
10 .158 1.579 100.000
Extraction Method: Principal Component Analysis.

Component Matrix(a)
Component
1 2
Play non-interactive
games .603 -.032
Chat .819 -.304

9
Check out how your
friends are doing .733 -.523
Update your profile .766 -.040
To pass time .674 -.428
Simplicity .726 .076
Horoscope .591 .665
Community
Personalization .646 .506
Flirting .492 .331
Flexibility .637 .052
Extraction Method: Principal Component Analysis.
a 2 components extracted.

Rotated Component Matrix(a)


Component
1 2
Play non-interactive
games .493 .349
Chat .831 .269
Check out how your
friends are doing .899 .044
Update your profile .625 .444
To pass time .794 .083
Simplicity .523 .510
Horoscope .051 .888
Community
Personalization .193 .797
Flirting .181 .565
Flexibility .468 .436
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.
a Rotation converged in 3 iterations.

Component Transformation Matrix


Component 1 2
1 .785 .620
2 -.620 .785
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.

10
LIST OF GRAPHS

• Scree Plot
1
3
4
5
6
7
8
9
0
2
C
E
Scree
0
Plot
i
o
m
g
p
e
o
n
n
v
e
a
ln
t
u
e
N
u
m
b
e
r

11
EXECUTIVE SUMMARY

Facebook is one of the leading networking sites in the


world. We plan to analyze why it is so!!

Various features and application attracts a user. We have


undertook a research and found some key factors that act as
influencers for users to stay loyal to the site and on the contrary
rope in more users. In our research we included 12 factors that
are there to attract to people to use face book. We used a
questionnaire where in we asked the respondents to rate those
factors which they thought were important factors to use the site.

After getting all the respondents answers we then did the


data entry using SPSS. Here we used factor analysis to interpret
the data by using KMO, Bartler Test, Communalities and Scree
Plot

Our analysis showed that there were basically 2 main


factors in which we could analyze all the other 12 factors with
the help of data reduction.

12
PROBLEM DEFINITION

a)Problem Statement: Here the basic problem that the research


will probe is whether the factors which are attractive to use face
book are correlated to each other or are independent. By
understanding these factors we will be able to see whether we
can combine 2 or more factors into one which would be easy for
us to interpret. So the null hypothesis in this problem would be
that the factors are not correlated to each other.

b)Background of the Problem: Facebook as we all know is one


of the world’s largest social networking site. Social networking
in the earlier years only was about interaction between people.
But due to tough competitions from many networking sites it is
necessary that there need to be some extra facilities on the basis
of technology as well as on the basis of needs of the people.
Facebook has been successful in doing that. Facebook has so
many new and different features that it is very difficult to
understand what kind of things does people like in Facebook. To
understand this problem we are doing a research where in we
analyze whether some factors really are correlated to
attractiveness and whether those factors are related to each other
or not.

13
Approach to Problem

Once we understood what our problem was, we started to plan


on how should we proceed with the research. We decided to
start our survey from college itself. Here we provided our
questionnaire to people at specific areas where we may get a
large amount of respondents for e.g: Library, Reading Hall,
Canteen etc. We also tried to do this research with different set
of age group people also that may help in analyzing the problem
better. While creating questionnaire our first criteria was to find
the factors that are relevant to the problem. After a lot of
research we decided to analyze on the basis of following 12
factors
1)New Friends
2)Interactive Games
3)Non Interactive Games
4)Chat
5)Wall
6)Profile Update
7)Passing Time
8)Simplicity
9)Horoscope
10)Community
11)Flirting
12)Flexibility

We prepared a questionnaire relating to the rating of these


factors.

14
We decided to use Factor Analysis along with Semantic Scale to
rate the factors and then analyze
Research Design

a)Type of Research Design: Factor Analysis


Here we used factor analysis because our main problem talks
about correlation among various factors. To understand
attractiveness towards facebook, we need to understand the
interdependence among variables. All the factors could be
grouped together with respect to specific characteristics and then
analyze it well.

b)Information needs:There were various types of information


that was needed while taking the survey. First information was
whether they use facebook or not. If the answer was yes then
only we used to proceed with the other informations. Among
them was there name, age, contact number and how often do
you use facebook. This was some of the basic information along
with the rating of factors that was needed.

c)Data Collection from Secondary Sources:


Here secondary sources include data entry of primary sources
into SPSS and then using factor analysis interpreting the
relationship between the primary data.

d)Data Collection from primary sources: Primary sources


included most of the students of IBS who were the respondents
of this survey along with others who generally use face book.

15
e)Scaling Techniques: Semantic Scale
Here we needed ratings. Hence we used semantic scale where
we asked the respondents to rate the factors from 1-7. Here 1
being least important factor and 7 being most important factor.
Rating will help us to understand the importance of these
factors.

f)Questionnaire development and pretesting: The most


important factor that is necessary for a research is to prepare a
good questionnaire. Before starting our questionnaire, we made
the necessary research to find the factors involved .Here the
most important which the respondents may relate to has to be
used. The most important question being asked in the
questionnaire was whether they use facebook. If no, then there is
no point in proceeding with further questions. There were some
other relevant question as to how often do you use facebook.
Also some basic questions like respondents age, contact number
etc.

g)Sampling Techniques : Simple Random sampling.


All items in the population had an equal chance of being chosen
in the sample. This was done to avoid bias.

16
h)Field Work:To conduct a research we need to do some field
work. Hence even this research involved certain field work. Our
initial respondents were IBS students. Hence we went to main
areas of IBS campuses like library, canteen, reading hall,
classroom etc to take the survey. Others were been also given
personally.

17
Data Analysis

a)Methodology:

i)Correlation Matrix: First set of analysis that can be done is


through correlation matrix. From our table, we can see that there
is correlation between individual variables (e.g. chat has 1
correlation with chat itself but it has .487 with flexibility). From
this method we are able to reject the hypothesis that correlation
matrix is an identity matrix since the correlation with other
variables is not 0. Hence accept the alternative hypothesis that
there is a correlation between variables.

ii)KMO and Bartlett’s Test: According to this case we test the


significance and adequacy level of samples.
KMO checks the adequacy level of samples. The condition to
check the adequacy level is that if the level is greater than 0.5 it
is fit. More it is closer to1 better would be the research. In our
case the level is .820. Hence we can say the sampling is good
enough since it is more than .5. It is conducive enough to
conduct factor analysis. Hence we can proceed further.

The next set of test is Bartlett’s test of significance: Here the


condition is if significance level is less than 0.05, we can say
that there exits significant correlation between variables. In our
case it was 0.00.Hence the case is significant enough to find
correlation.

18
iii)Communalities :Communalities help us to explain how
much the research has explained the factors. Here we can see
that some have more than 50% explanations.
E.g. New friends has the maximum 82% explanation in this
research while flirting has just 35.2% explained in the case.
Hence we can say that more different data need to be taken in
the case of flirting because it is even less than 50%.

iv)Total Variances explained: This is the case where we


analyze the variances of various factors and determine eigen
values of that factors. Eigen value explains us how much all the
attitudes are been explained by individual factors. Here the
condition is that eigen value should be at least one. Other wise it
is irrelevant if it is goes to less than one. Here there is a method
to calculate the number of factors to be obtained finally. When
eigen value becomes less than 1,we need to stop and number of
factors would be equal to previous number when it was at least
one. In our case till the second factor the eigen value is more
than 1.After that it reduces to less than 1.Hence we can
determine the number of factors as 2 since after that the eigen
values reduces to less than 1. Cumulative here is 62% in case of
2 factors.

19
v)Scree Plot: This is another way to find the number of factors
used. Scree plot is a graph where you are able to find the number
of factors with the help of flatness of the graph. In our case the
graph becomes flat after 2 factors. Hence according to the
condition in our case due to its flatness we can say that the
number of factors here is 2.

vi)Component Matrix: Component Matrix is used for


reloading. This explains how much one attitude has explained
each of the 2 factors.
Here some of the attitudes are best explained by factor 1 while
some factors are best explained by factor 2.Highest loading for
most of the attitudes is attitude 1 while for horoscope highest
loading is factor 2.

Factor 1 Factor 2
Chat- 81% Horoscope- 66%
New Friends- 73%
Update Profiling -76%
Pass Time -65%
Simplicity -74%
Community
Personalization- 65%
Flirting -49%
Flexibility -65%

20
vii)Rotation Component Matrix: If almost all the factors are
same then we tend to rotate to get an accurate picture of the
sample. Maximum rotation can be done 25 times .Here we have
used Varimax method for rotation. Here we have some changes
in the loading factor. Even the percentage of loading has also
increased after rotation. Eg: Communality Personalization and
Flirting in the previous condition was best explained by factor1
but then after rotation now it is best explained by factor 2.

viii)Component Transformation Matrix: Component


transformation matrix tells us the way the overall attitudes have
been explained by 2 components. Here it 77% and 63%

21
Plan of Data Analysis:

1. After completing our questionnaire, we did the statistical


analysis of our research. We used SPSS, factor analysis to carry
forward our research.
2. First process was data entry. While adding our data, we first
described our variables in Variable View, which included Name,
Type, Width, Decimal etc.
3. Then data was entered into the Data View.
4. To obtain the output, we selected Data Reduction, where we
selected factor.
5. We selected all the variable and choose Initial and KMO from
descriptives.
6. Then Principal component method was selected from the
extraction option. Correlation matrix was used to analyze and
unrotated factor solution and scree plot was used to display. The
value for the Eigen value was given over 1.
7. Varimax rotation was used.
8. Listwise cases were excluded.

22
Results:

We rejected the hypothesis that correlation matrix is an identity


matrix since the correlation with other variables is not 0. Hence
accept the alternative hypothesis that there is a correlation
between variables.
KMO checks the adequacy level of samples. In our case the
level is .820. Hence the sample was good enough, since it is
more than .5. It is conducive enough to conduct factor analysis.
Hence we proceeded further.
The next set of test was Bartlett’s test of significance. In our
case it was 0.00. Hence the case was significant enough to find
correlation.
Communalities help us to explain how much the research has
explained the factors. New friends has the maximum 82%
explanation in this research while flirting has just 35.2%
explained in the case. Non interactive games has 36.4%. Chat
was 76.4%. Checking out how friends are doing was 81%.
Profile update was 58.8%. Passing time constituted 63.8%.
Simplicity accounted to 53.3%. Horoscope was about 79%.
Community personalization and Flexibility accounted to 67.3%
and 40.8% respectively.
Then we analyzed the variances of various factors and determine
eigen values of that factors. In our case till the second factor the
eigen value was more than 1. After that it reduced to less than 1.
Hence we determined the number of factors as 2 since after that
the eigen values was reduced to less than 1. Cumulative here
was 62% in case of 2 factors.
Some of the attitudes were best explained by factor 1 while
other factors by factor 2. Highest loading for most of the
23
attitudes is attitude 1 while for horoscope highest loading is
factor 2.

Factor 1 Factor 2
Chat- 81% Horoscope- 66%
New Friends- 73%
Update Profiling -76%
Pass Time -65%
Simplicity -74%
Community
Personalization- 65%
Flirting -49%
Flexibility -65%
Component transformation matrix tells us the way the overall
attitudes have been explained by 2 components. Here it was
77% and 63%.

24
Limitations

1. We have taken 12 question in our survey to analyze


facebook. And because of using 12 questions only we have
limited our research. We have considered various factors
like chat, friends, games etc through which we have come
to a conclusion.

2. Our survey was limited to Mumbai only. So the results


would be biased to a metro city where such networking
sites are very common so this might not give us the right
survey results as small cities are totally ignored.

3. As our survey was mostly among the students only so this


survey lags the interest of other age group people like
teenagers and people above the age of 30.

25
Conclusions

Here our main aim was to find the significance of each and
every factor that was related to attractiveness to facebook. With
factor analysis using SPSS we were able to find the following
conclusions:

1)Using KMO we measured the sample adequacy and we


proved that the samples were fit to do the research.
2) Using Bartlett we concluded that there is a significant
correlation between the variables.
3)Scree plot and Variances helped us to reduce the factors to 2
and then help us to analyze the data.
4)Using Component matrix we saw that factors like non
interactive games, chat, new friends, pastime, simplicity etc
were all explained most by factor1 due to high loading while
Horoscope was best explained by factor 2.

26
Recommendations:

• The privacy security should be increased:


The current features can be upgraded with certain latest
technologies.For eg Currently the photo which are
uploaded can be easily edited by any person and use it in
any manner.This may lead to some undesirable
consequences.

• Facebook can also introduce calling facility:


Upgarding the facebook with calling facility would
increase the usage of it to a larger extent.The calling
facility which is availbe in Skype pr GTalk can be
incorporated which would lead to overall improvement in
the facebook features.

• Videos and songs download can be added:


Songs are backbone of each and very individual.When
ever people are online they do always work while listening
to music.So inclusion of songs and videos would entyertain
the people using facebook.

• Repair and maintenance regularity

Exhibits:

QUESTIONNAIRE ON “ATTRACTION TO FACEBOOK”

Do you use Facebook?

27
Yes No

How frequently do you log into Facebook?


Several times a day Daily
Weekly Monthly
Less than monthly

What do you use Facebook for? Please rate the following 7


activities according to the frequency of usage with 1 being the
least frequent and 7 being the most.
1 2 3 4 5 6 7
1 New friends
Play
2 interactive
games
Play non-
3 interactive
games
Chat
(including
4
comments and
wall)
Check out how
your friends
5 are
doing(photos,
walls etc)
28
Update your
6
profile
7 To pass time
8 Simplicity
9 Horoscope
Community
10 personalizatio
n
11 Flirting
12 Flexibility

Name:

Mobile:
Gender:
Age:

29