A Research proposal on Satisfaction Of Internet Service In Bule Hora University

Research proposal
Submitted to
Department of Statistics for the partial fulfillment Bsc in degree of statistics

Prepared by:
-Negashu Birmaji
-Dina sheraf
-Dawit Biranu
-Abbas Siraj
-Fikadu Kinde
-Shallema Dime

Advisor. Tayu Nugusie (Bsc)

BULE HORA, ETHIOPIA

May 2017

1
Table of contents
1 INTRODUCTION
1.1Background of the study
1.2 Statement of the problem
1.3 Objective of the study
1.3.1 General objective of the study
1.3.2 Specific objective of the study
1.4 Significancy of the study
1.5 Limitations of the study
2. LITERATURE REVIEWS

2
 CHAPTER ONE
1. INTRODUCTION

1.1 Back Ground of Study

The history of communication dates back to the earliest signs of life. Communication can range from every subtle
process of exchange to till conversation and mass communication.
The Internet is a "network of networks" that consists of millions of smaller domestic, academic, business, and
government networks. Internet is also described as the worldwide publicly accessible network of interconnected
computer networks that transmit data by packet switching using the standard (IP). Internet is the transport vehicle
for the information stored in files or documents on another computer. It carries together various information and
services, such as electronic mail, online chat, file transfer, and the interlinked Web pages and other documents of the
(WWW). The Internet itself does not contain information, it is a slight misstatement to say a "document was found
on the Internet." It would be more correct to say it was found through or using the Internet. What it was found in (or
on) is one of the computers linked to the Internet every aspect of our day to day life is affected by the Internet.
Whether it is shopping, business, banking, communication, paying bills, social gathering, party, learning, education
etc. Internet is everywhere, knocking at our door, making our life easier and smooth. Moreover, when it comes to
education and research Internet is paving way for a great leap and sure library and information centers has no
exception The Internet made the information on our finger tips. The libraries of the developed world has adopted
the Internet facilities to provide the fast and better library services to its patron but this is not the case with many
developing nations and third world countries. The libraries of the third world countries still do not have the basic
Internet access facilities in many cases because of the poor funding and budget crisis, while we talking Ethiopia has
the second lowest Internet penetration rate in sub-Saharan Africa (only Sierra Leon is lower) and is currently
attempting a broad expansion of access throughout the country. These efforts have been hampered by the largely
rural makeup of the Ethiopian population and the government’s refusal to permit any privatization of the
telecommunications market. Only 360,000 people had Internet access in 2008, a penetration rate of 0.4%. The state-
owned (ETC) is the sole (ISP) in the country (Mark L.S, 200). Internet cafes are the main source of access in urban
areas, and an active community of bloggers and online journalists now plays an important role in offering alternative
news sources and venues for political dialogue. However, three-quarters of the country’s Internet cafés are in the
capital city, Addis Ababa, and even there access is often slow and unreliable. A test conducted by a Media Ethiopia
researcher in July 2007 determined that the average connectivity speed was 5 KBPS and that Internet service in most
cafés was unavailable between 10 and 20 percent of the time.( Sanford, W. (2005). In Ethiopia announced plans to
spend hundreds of millions of dollars over the next three years to connect all of the country’s schools, hospitals, and

3
government offices, and most of its rural population, to broadband Internet via satellite or fiber-optic cable. Between
2005 and 2007, the government spent US$40 million to install Woreda Net and School NET, two nationwide
networks meant to increase connectivity. Woreda Net provides e-mail, videoconferencing and (VoIP) services to
local governments, and School Net provides streaming audio and video through a downlink-only (VSAT) satellite.
The government has pledged to dedicate 10% of its annual budget to the development and maintenance of these
networks, which are managed by the government-run (EICTDA) (Sanford, W. (2005).
Today, the Internet plays a vital role in the teaching, research and learning process. This study has also tried to
explore broadly the importance of Internet with regard to access of information sources and its utilities for library
patrons in academic organizations and institutions. It is assumed that the University students feel more dependent on
the Internet for their class assignments and for the latest information of their subject areas than conventional
resources of information.
Definition of crucial terms
Student satisfaction: is the customers’ fulfillment response. It is a judgment that a product or service feature or the
products of service itself, Direction Bhu recognized
A standard measure for the satisfaction of internet user in2007E.C academic year it is assumed that if a student
uses the internet two or more than two hours per a day he or she is to be satisfied otherwise it is assumed to be
unsatisfied
1.2 Statement of the Problem
Internet services have become key social networking tools because of the high level of penetrations. It is found that
the Internet has become a vital instrument for education, research and learning process of those users in higher
education particularly in universities. Students use the Internet to keep in touch with friends, instructors, groups,
organizations and associates. However, customers are not satisfied that much expected within the Internet service
provider due to a lot of barriers. The problems are no network coverage or absence of internet connection, line busy
or trafficking repeatedly displayed on internet, slow speed and improper instillation on network and lack of
computer access and limitation of time to use internet service are the major ones.

Research questions
 How much proportion of customers is satisfied in Internet service provider?
 What are the major problems of customers faced while using the Internet?
 What measures should be made to minimize the problems that users are disposed on the Internet service?

1.3 Objective of the Study

1.3.1 General Objective

The1 main objective of this study is to analyze the student satisfaction on Internet service at Bule Hora University.

1.3.2 Specific Objectives

The study will be aims to carry out the following specific objectives:
 To know the association between independent variables and the dependent variable.
 To identify major problems of customers on Internet service.

4
 To forward suggestions to improve problems of customers on Internet service

1.4 Significance of the Study

It is found that the Internet has become a vital instrument for teaching, research and learning process of those
university students in the modern world. The study was conducted particularly to find an answer to the question of
students’ some problems associated in Internet service provider. So far the survey gives information about the core
problems and measures the satisfaction level of customers and dig out suggested possible solutions. This survey
enables that to point out major problems faced on Internet for Internet server agencies and user, to set suggestions to
reduce the problems that users are disposed on the Internet, to know for which purpose users use internet service in
their professional life. Internet is essential for many things such as for education of students, research, for register
result, for recreation and entertainment so by this case internet is back bone of campus education.

1.5 Limitation of Study

During problem identification and proposal preparation, some problems faced were;
 lack of time,
 Addiction
 Negative effect on family communication
 Loneliness
 Lack of conflict resolution
 Shortage of computers

CHAPTER TWO

2. LITERATURE RIVIEW

According to Nwagwu et al (2009) the Internet serves as a source of information for literature review, authors’
search, subject search, and research. In another instance, Adeogun (2003) reported that the convergence of
computers and telecommunications technologies has made possible the activities which were considered impossible
in the past. Those activities include information retrieval and transfer which were hampered by time and distance.
The Internet has emerged as an important component in academic institutions as it plays a pivotal role in meeting
information and communication needs of institutions. According to them, the Internet makes it possible to access a
wide range of information that is up-to-date. The Net enables scholars and academic institutions to disseminate
information to a wider audience through hosting websites and search facilities. Furthermore, students and lecturers
can communicate via the Internet irrespective of geographical boundaries. Distant learning has also been facilitated
by the Internet (Luambano&Nawe, 2004).

5
Similarly, Remand and Ramzy (2004) opined that the Internet has established a place in the personal and
professional lives of researchers and scholars through their daily use of the Net for serious work and personal
communication.
Luambano and Nawe (2004) investigated the Internet use by students of the University of Dares Salaam. Their
findings revealed that the majority of the students were not using the Internet due to the inadequacy of computers
with Internet access, lack of skills in Internet use and slow speed of computers. It was also revealed that most
students who used the Internet did not use it for academic purposes. It was suggested that more computers connected
to the Internet should be provided and that training should be given to the students on the use of Internet.
The Internet enhances skills and capabilities of students, which assist them in studies and in professional life.
Students with high CGPA use the Internet more for their studies and gain more knowledge and information across
the world (Awais, Bilal, Usman, Waqas, &Sehrish. n.d, 2010).
Ethiopia is the second most populous country in Africa, but poor infrastructure and a government monopoly over the
telecommunications sector have notably hindered the growth of information and communication technologies
(ICTs). Consequently, Ethiopia has one of the lowest rates of Internet and mobile telephone penetration on the
continent. Despite low access, the government maintains a strict system of controls and is the only country in Sub-
Saharan Africa to implement nationwide Internet filtering.

Ethiopia is connected to the international Internet via satellite, a fiber-optic cable that passes through Sudan and
connects to its international gateway, and another cable that connects through Djibouti to an international undersea
cable. In an effort to expand connectivity, the government has reportedly installed several thousand kilometers of
fiber-optic cable throughout the country in recent years. There are also plans in place to connect Ethiopia to a global
undersea cable network through the (Essay) project. The Essay project itself was completed and launched in
Thomson, S.K July 2010,

CHAPTER THREE

3. METHODOLOGY

3.1Study Area
The study will conduct in Bule Hora University (Bhu). Bhu is located in Buld Hora town which is the capital of
West Guji Zone located in Oromia region. It is located around 467 kilometers far from Addis Ababa which is the
capital city of Ethiopia.

3.2 Study Population

The study populations in this project are all students in the Faculty of natural science in bhu.

6
3.3 Methods of Data Collection and Sampling Design
This proposal will be done using primary data, which is collected by self-administered questionnaire. The
questionnaire has been distributed to sample of applied natural science students in ASTU. There were a total of 777
students in our sampling frame. The sampling design used was stratified random sampling using the target
population (frame). To use this method of sampling design first we stratified the population in to 1 st year, 2nd year, 3rd
year and 4th year. Under this we have selected the samples from each strata by using simple random sampling and
the sample was allocated by proportional allocation.

Sample size can be estimated by using the following formula.

n
n 0
n
1 0
N
2
 z 
 
n0  2  * pq
 d 
 
Notation: α = 0.05 level of significance = 5%
no = initial sample size
N = population size
n = sample size
P = probability of success = 0.5
q = probability of failure
d = Margin of error =0.1
Z
no ( 2
) 2 * pq Z  1.96
n=no/ (1+ (no)/N) where: d , 2 then

1.96 2
n0 ( ) * (0.5)( 0.5)
0 .1
=96.04
no 96
Finite population correction: no/N= 96/777=12%, which is greater than 5% implies that it is not satisfactory
sample size approximation to the required. So that we use the correction formula:
n = no/ (1+ (no)/N)
Then n= 96/ (1+96/777) = 85
The sample size of each stratum would be calculated by using proportional allocation as follows with respect to the
number of students in each stratum.

7
 Nh * n
nh 
N

Where:

n1 57, n2 160, n3 505, n4 55

57
n firstyear ( ) * 85 6
777
160
nsec ondyear ( ) * 85 18
777
505
nthirdyear ( ) * 85 55
777
55
n fourthyear ( ) * 85 6
777
Where nh = sample size to be selected from hth strata

Nh = Over all population size under hth strata

N = Over all ample size

h = 1,2,3,4

Table 3.1 the table below shows summary of population size and sample size in each stratum.

Batch Total number of units in batch (Nh) Total number of sample unit (nh)
1 57 6
2 160 18
3 505 55
4 55 6
Total N=777 n=85

3.4 Variables considered in the study

 Dependent Variable
Student satisfaction on Internet service
 Independent Variables
 Sex
 Age
 Religion
 Family residence
 Batch

8
  CGPA
3.5 Method of Data Analysis

The collected data has been analyzed by using statistical software (SPSS) and the out put has been presented in the
form of tables (univariate, multivariate), figures like bar chart, pea chart and histogram with interpretation and also
the analyzed data has been fitted to an appropriate model

3.6 Methods of statistical data analysis

In this study both descriptive and inferential statistical method were used to analyze the data.

3.6.1 Descriptive Statistics

Deals with any method or procedures used to present, organize, summarize and interpret the masses of the numerical
data in to meaning full form, but do not infer properties of the Population from which the sample is drawn. In this
part we can use various statistical techniques like frequency distribution table, graphs and diagrams to express the
descriptive part.

3.6.2 Inferential statistics

Inferential statistics is used to give inference to population that we study by using sample data. Inferential statistics
is consists of generalizing from samples to populations,performing hypothesis testing, determining relationships
among variables and making predictions. There are different kinds of statistical models, among them chi-square and
logistic regression are used for this study.

3.6.2.1 Chi-square Test

The chi square test is used to find out whether there is an association between row and column variables and to test
the homogeneity that exists between the factors and dependent variable.
The chi-square test statistic is given by

(Oij  E ij ) 2
i 1  j 1
R C
 2
cal
E ij
Where: Oij: observed frequency in cell (i.j),
Eij: expected frequency in cell (i.j)
Ri = number of rows
Ci = Number of columns
N = Grand total.
Hypothesis test
Ho: There is no association between the dependent variable and independent variables.
H1: There is association between the dependent variable and independent variables.
The statistic has (R-1) (c-1) degree freedom.
Decision rule:
If χ2cal > χ2(r-1)(c-1),α , then reject H0; other wise accept H0 and Conclude based on the decision.

Assumptions of chi square test of independence

 The sample must be randomly selected from population.

9
 The observation must be independent of each other.
 Each number and each individual (object) is independent of each other.
 Each member qualities for one and only one cell in the table.
 The expected frequency for each category must be 5 or greater than 5.
 The sample size is large enough.

3.6.2.2 Logistic Regression

Logistic regression is used when the dependent variable is qualitative. Qualitatve respnse variable are either binary
(dichotomous variable) or multiple category. Binomial or binary logistic regression is the form of regression which
is used when the dependent variable is dichotomous and the independent variables are of any type. Multinomial
regression can handle the case of dependent variables with more than two classes. The regression model when the
dependent variable is qualitative our aim is to estimate the expected or (mean), value agiven the values of the
regresses E(yi/x1i,…,xki). But in logistic regression our objective is to find the probability of thing happening
(probability of success).

The most appropriate analytical methodology for this study is binary logistic regression. Consider collection of K
independent variables will be denoted by the vector X=(x 1,x2,…,xk). Let the conditional probability that the
outcomes of interest in a study “is present” denoted by P(Y=1/X=x) = P(X).

Then the logit of binary logistic regressions is given by the equation

g(x) = β0+β1X1+β2X2+…..+βkXk

and odds in favor of success for the binary logistic regression will be

ln[P/1-P]= ln(e(g(x)))= β0+β1X1+β2X2+…..+βkXk

exp( g ( x))
p (Y 1 / X  x)  p ( x) 
In which case 1  exp( g ( x))

p
Model: 1  p = exp(β0+β1X1+β2X2+…..+βkXk)

Where: P- the probability of success (probability of something happened)

1-P- the probability of failure

β0- is constant term

βj -coefficients of independent variables

Xi- independent variables

10
  The ratio of probability success to probability of failure was (P/1-P) is odd ratio of success.
 Exp (βj), where j=1,2,….k is a factor by which the odds of occurrence of success change by a unit increase
in the jth independent variables.
 If L= ln(P/1-P), is positive, it means that the value of the regressor(s) increases, the odds that the regressand
equals 1 (meaning some event of interest happens) increases. If ln is negative, the odds that the regressand
equals 1 decrease as the value of x increases.

Method of parameter estimation

The most commonly used method of estimating the unknown parameters of a logistic regression model is the
method of Maximum Likelihood (ML). ML methods seek to maximize the log likelihood (LL) which reflects how
likely it is (the odds) that the observed values of the dependent variable may be predicted from the observed values
of the independents variables.

In logistic regression, the likelihood equations are non-linear explicit function of unknown parameters. Therefore,
we use a very effective and well known Newton-Raphson iterative method to solve the equations which is known as
iteratively reweighted least squares algorithm. Since the observations are assumed to be independent, the likelihood
function is given by:

The estimation of,require the maximization of the likelihood function or equivalently the maximization of the
natural logarithm of the likelihood function denoted by:

One approach to the maximization of the expression in equation (**) involves the differentiation of with respect to,
and setting the k+1resulting equations to zero. The most effective and well known Newton-Raphson iterative
method can solve the equations.

The coefficients can be interpreted as the change in the log-odds associated with a one unit change in the
corresponding independent variable or the odd increases multiplicatively by for every one unit change increase in x.
As a result, the model involving categorical variables essentially involves a change in the reference category as we
change from one category to another.
Assumption of logistic regression
 Logistic regression does not assume a linear relationship between the dependent variable and independent
variables.
 The dependent variable need not to be normally distributed but does assume its distribution is within the
range of the exponential family such as normal, poison, binomial and gamma.
 The distribution of the parameter is asymptotically normally distributed.
 The dependent variable need not to be homoscedastic for each level of independent variables that is there
is no homogeneity.
 Normality distributed errors are not assumed.
 Logistic regression does not require that the dependent variables be continuous.

11
  Meaningful coding. Logistic coefficients will be difficult to interpret if not coded meaningfully. The
convention for binomial logistic regression is to code the dependent class of greatest interest as 1 and the
other class as 0.
 Exclusion of all irrelevant variables and inclusion of all relevant variables in the regression model

3.7 Assessing the fit of the model

A) The Wald Test Statistic
The Wald test is a way of testing the significance of particular explanatory variables in a statistical model. The Wald
statistic is an alternative test which is commonly used to test the significance of individual logistic regression
coefficients for each predictor variable.

The Wald test statistic is:

W   j / s.e j  = 
2 2

Each Wald statistic is compare with a distribution with one degree of freedom. For each explanatory variable in the
model there were associated parameters. If for a particular explanatory variable, or group of explanatory variables,
the Wald test is significant, then we would conclude that the parameters associated with these variables are not zero,
so that
variables should be included in the model. If the Wald test is not significant then these explanatory variables can be
omitted from the model.
j j
Ho: =0 VS HA: ≠ 0 at α level of significance.

Decision rule: We reject the null hypothesis if the p value <α value=0.05.

B) Hosmer - Lemeshow goodness of fit test

The goodness of fit or calibration of a model measures how well the model describes the response variable.
Assessing goodness of fit involves investigating how close values predicted by the model with that of observed
values. The Hosmer and Lemeshow test is commonly used to test for assessing the goodness of fit of the model and
allows for any number of explanatory variables. The Hosmer–Lemeshow test is a commonly used to test for
assessing the goodness of fit of a model and allows for any number of explanatory variables, which may be
continuous or categorical. The test is similar to a χ 2 goodness of fit test and has the advantage of partitioning the
observations into groups of approximately equal size, and therefore, there are less likely to be groups with very low
observed and expected frequencies. In this case, better model fit is indicated by a smaller difference in the observed
and predicted classification.

Hosmer- Lemeshow goodness of fit is:

Where, g denotes the number of groups, is the number of observations in the k th group, is the sum of the Y values
for the kth group and is the average of the ordered for the kth group.

12
Ho: The model is good to fit the data.
H1: The model is not good to fit the data.
Decision rule: We have no evidence to reject the null hypothesis if the p value >α value=0.05.

CHAPTER FOUR

4. RESULTS AND DISCUSSION

4.1 Descriptive statistics

Table 4.1 Percentage distributions of level of satisfaction of customers by predictor variables

Variable Category Satisfaction of customers on internet service

Yes No Total
Count % Count % Count %
Sex Male 16 19.7 29 33.5 45 52.5
Female 15 17.8 25 28.9 40 47.5
17-19 9 10.6 16 15.3 25 25.9

20-22 10 11.76 12 27.04 28 38.8

Age
23-25 8 9.4 12 11.8 20 21.2

>25 5 5.9 5 8.2 12 14.1

Orthodox 13 15.3 20 23.5 33 38.8

Religion Muslim 8 9.44 10 11.76 18 21.2

Protestant 10 11.76 12 16.44 24 28.2

Catholic 0 0 5 5.9 5 5.9

Other 2 2.4 3 3.5 5 5.9

Residence Rural 19 22.4 26 30.5 45 52.9

Urban 18 21.2 22 25.9 40 47.1
1st 0 0 6 7.1 6 7.1
Batch 2nd 5 5.9 12 15.3 18 21.2
3rd 20 23.5 35 41.2 55 64.7
4th 2 2.4 4 4.7 6 7.1
<2.75 10 11.76 15 17.64 25 29.4
CGPA 2.75_3.25 13 15.3 15 17.6 28 32.9
3.26_3.75 7 8.2 13 15.3 20 23.5

13
 >3.75 4 4.7 8 9.4 12 14.1
Gaming 0 0 0 0 0 0
Purpose of File sharing 0 0 0 0 0 0
using Face book 17 20 23 27.1 40 47.1
internet Music 6 7.05 6 7.05 12 14.1
News 11 13 12 14.1 23 27.1
Other 3 3.5 7 8.3 10 11.8
Searching Google 24 28.2 32 37.7 56 65.9
method Yahoo 6 7.06 8 9.44 14 16.5
Bing 5 5.9 10 11.6 15 17.6
Amazon.com 0 0 0 0 0 0
msn searching 0 0 0 0 0 0
Twitter 0 0 0 0 0 0
Others 0 0 0 0 0 0
Rate of skill Very advanced 6 7.05 10 11.75 16 18.8
Advanced 13 15.3 21 24.7 34 40
Average 9 10.6 13 15.3 22 25.9
Basic 4 4.7 6 7.1 10 11.8
Very basic 0 0 3 3.5 3 3.5
How often Daily 20 23.5 19 22.4 39 45.9
do you use Weekly 10 11.76 13 15.34 23 27.1
internet? Twice week 7 8.2 10 11.8 17 20
Monthly 0 6 7.1 6 7.1
How do From teacher 4 4.7 6 7.1 10 11.8
you get Self learning 23 27.05 27 31.75 50 58.8
training? From friend 5 5.9 10 11.7 15 17.6
From external 4 4.7 6 7.1 10 11.8
training

From the above Table we see that the lowest percentage (17.7%) of female students who are satisfied on internet
service, whereas the highest percentage (33.5%) of male students who are not satisfied on internet service and as the
whole 36.3% of the students are satisfied to internet service whereas the highest percentage(63.7) of the students are
not satisfied.

In case of religion we see that the highest percentages (23.5) of orthodox religion followers students who are not
satisfied on internet service, whereas the lowest percentage (2.4%) of other religion followers students who are
satisfied on internet service. In other way when we see about residence 58.8% of students are not satisfied to
internet. Out of this 30.6% of students are comes from rural areas whereas 28.2% of students comes from urban
areas.
When we see according to CGPA only 39.96% of students are satisfied to internet use and out of this the lowest is
the 4.7% students whose CGPA is >3.75 and 60.04% of students are not satisfied to internet use. Out of this the
highest percentage (17.64%) of students are whose CGPA is <2.75.

14
Generally, as we can see from the above table the students who are not satisfied to internet use are higher than the
students who are satisfied to internet use.

k
Figure 4.1 histogram of age of students

Interpretation

From the above histogram we see that there are different ages distributions in ASTU we had collected sample data
from 85 students. i.e. 25.9% age interval 17-19, 38.8% from age interval 20-22, 21.2%, from age interval 23-25 and
14.1 from above 25 age year.

.
Figure 4.2 bar chart religions of
students
Interpretation

From the above bar graph we see that

there are different distributions religions
in ASTU we had collected sample data

15
 from 85 students. i.e 38.8% from orthodox ,21.8% from muslim,28.2% from protestant and 5.9% from catholic
religion followers students.

.
Figure 4.3 bar chart of family residence of students
Interpretation

From the above bar graph we see that there are different distributions religions in ASTU we had collected sample
data from 85 students. i.e 52.9% from students came rural and 47.1% from students came from urban.

Pie charts

Figure 4.4 pie chart of income of students

Interpretation:

From the above pie chart we see that there are different ages distributions in ASTU we had collected sample data
from 85 students. i.e 25.9% age interval 17-19, 38.8% from age interval 20-22, 21.2%, from age interval 23-25 and
14.1 from above 25 age year.

16
Figure 4.5 pie chart of religion of students

Interpretation:

From the above pie chart we see that there are different distributions religions in ASTU we had collected sample
data from 85 students. i.e 38.8% from orthodox ,21.8% from muslim,28.2% from protestant and 5.9% from catholic
religion followers students.

Figure 4.6 pie chart of income of students

Interpretation:

17
From the above pie chart we see that there are different distributions religions in ASTU we had collected sample
data from 85 students. i.e 52.9% from students came rural and 47.1% from students came from urban.

4.2 Inferential statistics

4.2.1 Chi-square test of independence

Ho: There is no association between explanatory variable and satisfaction of customer on Internet service.
H1: There is association between explanatory variable and satisfaction of customer.
Table 4.2 The cross tabulation for chi square test of association between explanatory variable and students
satisfaction on internet service

Variables Pearson Chi- df Asymp.sig(2sided)

square

Students satisfaction vs sex 0.226 1 0.635

Students satisfaction vs age 7.879 3 0.04

Students satisfaction vs religion 12.5858 4 0.014

Students satisfaction vs family residence 4.912 1 0.027

Students satisfaction vs batch 1.017 3 0.797

Students satisfaction vs CGPA 0.934 3 0.817

Decision rule: Reject the null hypothesis Ho if the p value < 0.05
 Since p-value = 0.635 > 0.05 we accept the null hypothesis. i.e there is no association between dependent
and independent variable or there is no association between sex of respondents and the satisfaction of
students on internet service.
 Since p-value = 0.04 < 0.05 we reject the null hypothesis. i.e there is an association between dependent and
independent variable or there is an association between age of respondents and the satisfaction of students
on internet service.
 Since p-value = 0.014 < 0.05 we reject the null hypothesis. i.e there is an association between dependent
and independent variable or there is an association between religion of respondents and the satisfaction of
students on internet service.
 Since p-value = 0.027 < 0.05 we reject the null hypothesis. i.e there is an association between dependent
and independent variable or there is an association between family residence of respondents and the
satisfaction of students on internet service.
 Since p-value = 0.797 > 0.05 we accept the null hypothesis. i.e there is no association between dependent
and independent variable or there is no association between batch of respondents and the satisfaction of
students on internet service.

18
  Since p-value = 0.817 > 0.05 we accept the null hypothesis. i.e there is no association between dependent
and independent variable or there is no association between CGPA of respondents and the satisfaction of
students on internet service.

4.2.2Logistic regression

Table 4.3 Variable in equation

B S.E. Wald df Sig. Exp(B)

Step 0 Constant .505 .224 5.080 1 .024 1.656

Under Variables in the Equation, we see that the intercept-only model. If we exponentiation both sides of this
expression we find that our predicted odds [Exp (B)] = 1.656That is, the predicted 0.505.
Table 4.4 Hosmer and Lemeshew test

Step Chi-square Df Sig.

1 3.494 7 .836

The Hosmer and Lemeshow Test measures the goodness of model fit for the given data.
The test hypothesis can be formulated as:
Ho: The model is good to fit the data.
H1: The model is not good to fit the data.

Table 4.5 Omnibus of model coefficients

Chi-square Df Sig.
Step 1 Step 19.954 6 .003
Block 19.954 6 .003
Model 19.954 6 .003

Decision rule: We have not enough evidence to reject the null hypothesis if the p value >α value=0.05. The result
shows, we do not reject the null hypothesis for the model, since p-Value =0.836 > 0.05 indicates that the model is
good to fit the data.
When we look the Omnibus Tests of Model Coefficients gives us a Chi Square value of 19.954 with 6 df, significant
at 0.05 level. This is a test of the null hypothesis that adding the predictors to the model has not significantly

19
increased our ability to predict the satisfaction made on our Subjects. From the above result of omnibus tests, it is
significant at 0.05 levels implying that adding the predictors has significantly increased our ability to predict internet
satisfaction.

Table 4.6 Variables in equations

95% C.I.for
EXP(B)
B S.E. Wald df Sig. Exp(B) Lower Upper
Step sex .317 .535 .349 1 .554 1.372 .481 3.919
a
1 age -.689 .307 5.037 1 .025 .502 .275 .916
religion -.535 .225 5.661 1 .017 .586 .377 .910
residence -1.463 .569 6.616 1 .010 .232 .076 .706
batch -.024 .377 .004 1 .950 .977 .467 2.044
CGPA .171 .269 .405 1 .524 1.187 .700 2.012
Constant 2.501 .969 6.670 1 .010 12.198
a. Variable(s) entered on step 1: sex, age, religion, residence, batch, CGPA.

a. Variable(s) entered on step 1: sex, age, religion, residence, batch, CGPA.

Log (p(y=1)) =βo+βx1 (sex) +βx2(age) +βx3 (religion) +βx4 (residence) +βx5 (batch) +βx6(CGPA)
Logit (p(y=1) = 2.501+0.317x1-0.689x2-0.535x3-1.463x4-0.024x5+0.171x6

Interpretation by variables in equation

 β1= 0.317 indicates that the sex has positive relation ship with response variable. The log odds that y=1,
the p-value 0.554 indicates that there is no sufficient evidence to conclude that sex does affect the students
satisfaction on internet service.
 β 2 = -0.689 refers the age of respondents negative relation ship with response variable. The p-value 0.025
indicates that there is sufficient evidence to conclude that the age category significantly affects the
student’s satisfaction on internet service.
 β 3 =-0.535 refer the religion group has negative relation ship with the response variable students
satisfaction on internet service. The students satisfaction on the log odds that y=1, the p-value 0.017
indicates that there is sufficient evidence to conclude the religion group significantly affect the students
satisfaction on internet service.
 β 4 = -1.463 refers the family residence has negative relation ship with response variable. The p-value
0.010 indicates that there is sufficient evidence to conclude the family residence significantly affects the
students’ satisfaction on internet service.

20
  β 5 = -0.024 refers the batch has negative relation ship with response variable. The p-value 0.950 indicates
that there is no sufficient evidence to conclude the batch category significantly affects the students’
satisfaction on internet service.
 β 6 = 0.171 indicates that the CGPA of students has positive relation ship with response variable. The log
odds that y=1, the p-value 0.524 indicates that there is no sufficient evidence to conclude that the CGPA
significantly affect the students’ satisfaction on internet service.

Interpretation by Wald test

 Since Wald of sex =0.349 is less than


2
0.05,1=3.84, we accept null hypothesis and we conclude that the
predictor sex is insignificant or it has no effect on response variable students’ satisfaction on internet
service.

 Since Wald of age =5.037 is greater than

 2
0.05, 1=3.84, we reject null hypothesis and we conclude that the
predictor age is significant or it has effect on response variable students’ satisfaction on internet service.

 Since Wald of religion = 5.661 is greater than

2
0.05,1=3.84, we reject null hypothesis and we conclude that
the predictor religion is significant or it has effect on response variable students’ satisfaction on internet
service.

 Since Wald of family = 6.616 is greater than 20.05, 1=3.84, we reject null hypothesis and we conclude that
the predictor family residence is significant or it has effect on response variable students’ satisfaction on
internet service.

 Since Wald of batch = 0.004 is less than


2
0.05,1=3.84, we accept null hypothesis and we conclude that the
predictor batch is insignificant or it has no effect on response variable students’ satisfaction on internet
service.

 Since Wald of CGPA = 0.405is less than 2 

0.05,1=3.84, we accept null hypothesis and we conclude that the
predictor CGPA is insignificant or it has no effect on response variable students’ satisfaction on internet
service.

21
 CHAPTER FIVE

5. CONCLUSION AND RECOMMENDATION

5.1 Conclusion
After observing all the results of the study, we can reach the following conclusion:
 The result of this study shows that among 85 respondents, only 37.6% of the students are satisfied on the
Internet service in the school of natural and computational science in BHU. That means 62.4% of the students
are not satisfied on the Internet service in the BHU.
 The variables such as, religion and place of residence of the students have significant effect on satisfaction of
the students on internet service.
 The Variables such as sex, batch and CGPA of respondents have no significant effect on the satisfaction of
students on internet service.

5.2 Recommendation
 Depending on the outcome of the study the we recommends to the Internet server agencies (Universities, users
and ETC) to adjust for challenge conditions (lack of enough computer access and Internet reliability, absence of
Internet connection, slow speed on searching files and trafficking or bad signals) make attraction for online
activity for the customers.

22
 The users should adjust their devoting time on using the social media, face book properly.
 From the narrative view of customers’ suggestions the researcher recommends that the University or other
concerned body should take the following actions in order to improve Internet service for customers in BHU.
 The satisfaction of the internal customer is not as expected as to be. Therefore, the university management and
the government should work more cooperatively in satisfying their internet service of students.
 The government and university management should pay attention in providing computer laboratory in each
department of BHU and related equipments in the right place.
 Increase the period of time and availability of computer in the University library.
 The members of the Internet technical staff should be trained to improve bad challenges on using Internet
in the area.
 High speed Internet and installations of new systems should be practiced.

REFERENCES:
 Agresti, A. (2002). Categorical analysis wilerinterscice. New York, ISBN 0-471-36093-7.
 Alan, A. (2007). 2nd ed. Categorical Data Analysis. London, UK..
 Blueman, Allag, Elementary statistics MC Gaw Hill(2004). 1st ed. a step by step approach.
 Casella, G. and Berger, R. (1990).Statistical Inference. Pacific Grove: Wadsworth & Brooks-Cole
 Chochran, W.G. (1997). 3rd ed. Sampling Techniques,. No.89, New York.
 Hanauer, David, (2004). Internet Use among Community College Students: Implications in
Designing Health care Interventions. Journal of American College Health, v52, i5, pp197-202.
 Hauck, W.W, and A.D. (1977). Neald’s test as applied to hypothesis in 169 if analysis 3 .AM
statistical Assac, 71.851-853.
 Hosmer, D.and Lemeshow, S. (2000). 2nd ed. Applied Logistic Regression. Wiley, New York.
 Mark, L.S. (2001). Managing The Internet Controversy. Neal-Schuman publishers, Inc. New
York.
 Sanford, W. (2005). 3rd ed. Applied Linear Regression. ISBN 0-471-66379-4.New York.
 Thomson, S.K. (2002).2nd ed. Sampling. Wiley, New York (Rahman, M.S, 2012.History of
communication Wikipedia, the free encyclopedia.Htm).

23
 QUESTIONNIARE
Bule

Bule hora university and

DEPERTMENT OF STATISTICS

First of all, I would like to say thank you for your collaboration to fill this questionnaire voluntarily. This
questionnaire is prepared to collect data about student’s satisfaction on Internet service in BHU. Please, read every
question and put your response in the provided space by check mark [√] for each question.
1. Sex: Female Male
2. Age: 17-19 20-22 23-25 >25
3. Religion: Orthodox Muslim Protestant
Catholic Other
4. Family residence Rural Urban

5. Batch: 1st year 2nd year

3rd year 4th year

6. Your CGPA: < 2.75 2.75 -3.25 3.26-3.75 > >3.75

7. How often do you use the Internet service?
Daily Weekly
Twice a Week Monthly
8. on average, how many hours per day do you spend on-line? __________
9. From where do you get your regular service of the Internet?
University Library Internet Cafe from Mobile
Direction: BHU has recognized a standard measure for the satisfaction of Internet users for education purpose. It is
assumed that if a student uses the Internet two or more than two hours per a day, then he/she is said to be satisfied.
Otherwise, it is assumed to be unsatisfied. Based on this information, answer question number 10 below.
10. Are you satisfied with your current Internet service provider according to the university?
Yes No
11. What is your primary purpose of using the Internet?
Education Communication
Entertainment Research Others

24
12. What do you like doing the most on-line? Gaming File sharing
Face book Music News others
13. Which method of searching do you use most in daily use of Internet service?
Google Yahoo Bing
Amazon. Com MSN Search Twitter
Others
14. How do you get the training for Internet use?
From teacher Self-learning
From friend from external training institutions
15. How do you rate your skill as an Internet user?
Very Advanced Advanced Average
Basic Very Basic
16. Is there enough computers for Internet use in ASTU?
Yes No
17. Have the Internet use served you to score good CGPA?
Yes No

25