DEPARETMENT OF STATISTICS

FACTORS AFFECTING JOB SATIFACTION AMONG

HEALTH WORKERS IN JIMMA SPECILIZED
HOSPITAL, JIMMA ZONE SOUTH WEST OF ETHIOPIA.

PREPARED BY:

AYENEW MOLLA
FIKIRIE MESAFINT

ADIVSOR:
SOLOMON BUKE (MSC.)

ARESEARCH PAPER SUBMITTED TO COLLEGE OF NATURAL

SCIENCE IN PARTIAL FULFILMANT FOR THE REQUIRMENT OF
THE Bsc. IN APPLIED STATISTICS IN JIMMA UNIVERSITY

MAY 2011
JIMMA, ETHIOPIA
TABLE OF CONTENTS
 Page
CONTENTS
ACKNOWLEDGEMENTS ………………………………………….#
ABSTRACT………………………………………………………….#
CHAPTERONE …..………………………………………………………...#
1.1 Introduction…………………………………………………….....#
1.2 Background……………………………………………………….#
1.3 Statement of the problem……………………………....................#
1.4 Objectives of the study……………………………………………#
1.4.1 Generalobjectives………………………………………...#
1.4.2 Specifiobjectives…………………………………………#
1.5 Significance of the study…………………………………………#
1.6 limitation of the study…………………………………………….#
CHAPTER TWO
Literature review …………………………………………………..#
CHAPTER THREE
Methodology………………………………………………………#
3.1. Study area……………………………………..............................
3.2 Target population…………………………………………………#
3.3 Sample size and sampling technique……………………………#
3.4 Method of data collection……………………………………….#
3.5 study variables………………………………………….....#
3.5.1 Dependant variable…………………………………….#
3.5.2 Independent variable…………………………………..#
3.6 Method of statistical data analysis……………………………..#
3.6.1 Descriptive statistics……………………………………#
3.6.2 Chi-square test…………………………………………#
3.6.3 Logistic regression…………………………………….#
CHAPTER FOUR
Result and discussion……………………………………….#
4.1 Analysis of descriptive statistics………………………………#
4.2 Analysis of chi-square………………………………………….#
4.3 Analysis of logistic regression ……………………………..#
CHAPTER FIVE
Conclusion and recommendation
5.1Conclusion
5.2 recommendation………………………………#
Appendix……………………………………………………………….#
Questionnaires………………………………………………………….#
Bar graph……………………………………………………………..#
Reference……………………………………………………………..#

List of tables

List page

Table 3.1: The categories and their corresponding numbers……………………………….

Table 3.2: The number of each category population and strata sample…………………....
Table 4.1(2): Frequency table for characteristics of the respondent ………………....
Table 4.3: Chi-square test for sex Vs marital status of the respondent…………………….
Table 4.4: Chi-square test for sex Vs salary of the respondent…………………………….
Table 4.5: Chi-square test for age Vs job satisfaction of the respondent…………………
Table 4.6: Chi-square test for sex Vs job satisfaction of the respondent…………………
Table 4.7: Omnibus tests of model coefficient…………………………………………
Table 4.8: Variables in the Equation……………………………………………………………

Acronym
SPSS-Statistical Package for Social Science
Snnp-South Nation Nationality and People
 ACKNOWLEDGEMENTS

First of all we would like to thank my family next to God for their unlimited assistance
throughout my education still now.
Also our deepest gratitude and sincere appreciation goes to our advisor Ato Solomon
Buke for this invaluable and unceasable help, advice follow up encouragement from the
date of topic up to the end of the research.
Secondly our thanks go to health professionals for their willingness to fill the
questionnaires.

At last but not least our gratitude goes to the department of statistics who give us the
access of computers and some budgets to accomplish this research and for all who help
us during collection of data.
 Abstract

The purpose of this study is to determine health professionals’ job satisfaction and
identify associated factors that affect level of satisfaction among health workers in Jimma
specialized hospital. Out of 318 health workers working in this place 174 were consented
to participate in the study. The data has been collected using self administered
questionnaire prepared in English that included socio-demographic characters, service
year and factors for job satisfaction. Data has been also cleared and fed into computer
analyzing using SPSS 16.0 window. The result show that 47.7% of the respondents have
not satisfied in a high level of dissatisfaction with their job. The most significant factors
are variables age, educational status, monthly salary and government policy.
It is also believed that this study can be used as a base line data for further study of the
problems on large scale to regional or national level.

Key word: SPSS (statistical Package for Social Science

 CHAPTER ONE

INTRODUCTION

1.1. Background of the study

Jimma zone is located in south western part of Ethiopia .It has352km distance from
Addis Ababa and has 13woredas and 855 kebeles.

According to the 1994 population and housing census, its population is estimated as
1,961,262 with annual-growth rate of 2.9%, crude birth rate of 32.2/1000, crude death
rateof18.1/1000 and infant mortality rate of 142/1000 and live births with life expectancy
of 45.5 years. It had also 4.7/1000 total fertility and 18.4% literacy status.

The extent to which health workers can be attracted to and retained in remote areas
depends on two interacted aspects.
 The factors which contribute to health workers decisions to accept and they stay
in remote post; and
 The strategies employed by government to respond to such factors.
In the recent literature on health work force mobility relating both to international
and internal migration, factors have commonly been categorized in to “pull” and
“push” factors (9).

Pull factors are those which attract an individual to anew destination. These might
include improved employment opportunities and/or career prospects, higher income,
better living conditions or amore stimulating environment.

Push factors are those factors which act to repel the individual from allocation. They
often mirror pull factors and might include loss of employment opportunity, low wages,
poor living conditions, etc.

Both push and pull factors impact an individual who makes a decision about moving to,
leaving or staying in a job in many different ways as described above.

Generally, job satisfaction is a state of being satisfied by individual for opportunities that
he/she achieves success in his/her life.

1.2. Statement of the problem

In Ethiopia, health services are affected by so many factors like human resources,
delivery system, health infrastructure, etc. Among these human resource is vital
component that delivering health services, this is usually due to miss distribution health
personals in favor of urban centers, nobody wants to go to the rural areas (1).

The Ethiopian government made an effort to allocate the health professional to rural and
remote areas irrespective of his status despite this effort it seems that health service
programmers are still being faced by several problems which threats the quality of life in
several rural and remote areas .Many of Ethiopian health indicators currently are bellow
the average for eastern African countries. As same basic indicators show Eastern African
countries are the serious ness of the situation. The high level of morbidity and mortality
rate indicates the unacceptable low of health and nutritional situation that prevails in the
country. For example basic health indicators of Ethiopian like,

 Life expectancy at birth

 Child mortality rate
 Infant mortality rate
 Total fertility rate etc are very large compared to European countries.

In addition to general economic distress, the indictors are also affected by the status of
the health service in the country. Almost 60% of the population did not have local access
to health facilities and less than 1/3 of the Ethiopian children are fully vaccinated (1).

Thus besides other recourses constraint, health professional job satisfaction plays the
central role in the improvement of health care services to the community.

1.3. OBJICTIVE OF THE STUDY

1.3.1. General objective

To determine the factors which affect job satisfaction and degrees/ levels of job
satisfaction among health workers in the selected health service.

1.3.2. Specific objectives

 To determine health professional’s job satisfaction and level of job

satisfaction is relation to monthly income, location and profession.
 To identify the factors that influences the job satisfaction of workers.
 To collect valuable suggestions that has been given by the respondents to
alleviate the existing problems of health workers in the current time.
 To provide information for the concerned bodies.

1.4. Significance of the study

Satisfaction of health personnel is basic for good health care and to meet the health care
need of the society. Health professionals are vital to health care system. There for, it was
important to study staffs moral motivation and job satisfaction in order to provide
information that can be used to minimize the existing problems.

1.5. Scope of the study

This study has specifically focused on analyzing or investigating factors affecting job
satisfaction among health care workers. The study area has been limited to Jimma
specialized Hospital.

Limitations of the study

 There was great problem in the data collection process because some respondents
were not willingness to fill the questionnaire as it requires,
 Shortage of time to collect the data and to analysis it,
 Shortage of computer access,
 Financial problem,
 Shortage of internet access,

CHAPTER TWO
 Literature review

In Africa, public sector health service, salaries have typically been less than a living wage
and staff has received only a small proportion of their total earning from their service
work; in this situation, the power of managers over work motivation is much lower, de-
motivators are crucial and the significance of health service job in relation to other work
is important items of social status or time.

Looking at the number of staffs lost from health care institution

during last five years, rural hospitals were affected (33.3%) followed by regional
hospitals, health centers and central hospital each of which suffered 20% loss. The main
causes for attrition were low salary followed by lack of educational opportunities and
poor career structure (1).

The work conducted by the Joint Learning Initiate in recent years (6) has confirmed that
global inequities in the distribution pf health personnel hit those countries hardest which
can least afford it Asia, which has about half the worlds population, has access to only
about 30% of the world’s health professionals. Africa, with a predominance of countries
with sever shortage of health professionals, has the highest disease burden of any
continent, thus under scoring the perverse relationship between need for health care and
distribution of health personnel. The misdistribution of health personnel has its roots in
long standing global inequalities. It has been aggravated more recently by the
disintegration of health systems in low-income countries and by global policy
environment (6, 7, and 8).

According to survey done in our country by the Policy and Human Resource Project
Office a sample of 330 health workers were asked to provide information whether they
were satisfied with their job. Accordingly 74.6% of medical doctors, 62.s%of
pharmacists, 50.6%of nurses, 50.0%of sanitarian, 36.4%of pharmacy technicians,
45.0%of laboratory technicians and 34.2%of health assistants responded that they were
not satisfied with their job (1).

The main reason for these respondents were dissatisfied

 Low salary (60.3%)
 Narrow opportunity for further education (24.8%)
 Inadequate facilities and supplies (20.1%)
Among those report satisfaction with their job, the main reasons were;
 Satisfaction in helping others (43.0%)
 Professional gratification (32.0%)
 Salary (18.1%)

According to research done on job satisfaction of female Saudi nurses, more than 87.0%
and 92.0% of nurses were satisfied with their work place and the role assigned
respectively, to increase their satisfaction, there is a need to improve the social attitude
towards the nursing professional and to provide more comfortable working condition
(2).

According to research done in Nigeria about job satisfaction of doctors in 1997, 3.0%
had global job satisfaction, with the majority dissatisfied. Intention to resign was high
among 32.7%of the respondents. Factors found to be significantly were age, job
designation, income, duration of service and intention to resign. Intention to resign was
found to be significant associated with ethnicity and income (3).

According to research done on factors affecting job satisfaction of health workers in

Jimma zone in 1999 by health officer graduated students, it was revealed that 97.0%of
249 health professionals working in five health facilities were dissatisfied with their
job(4).

From the research done in West Vergina, USA in 2003, the major predictors for in
tenting health workers to leave their job was job dissatisfaction, and the major predictors
of these; psychological empowerment were hardiness, transformation leadership style,
nurse/physician collaboration and group cohesion (5).

Logistic regression is a variation of ordinary regression which is used when the

dependent (response) variable is dichotomous variable (i.e. it takes only two values,
which usually represent the occurrence and non- occurrence, or success or failure, or
satisfy or dissatisfy of some outcome events usually coded as 0 or 1) and the independent
(input) variables are continuous, categorical or both.(10)

For instance in a medical study, the health workers satisfy or dissatisfy, the patient
survives or dies etc.(10)

Unlike ordinal linear regression, logistic regression does not assume that the relationship
between the independent variables and the dependent variable is a linear one. Nor it
assumes that the dependent variable or the error terms are distributed normally.(10)
 CHAPTER THREE

Methods and materials

3.1. Study area

This study has been conducted in Jimma specialized Hospital which is located in North
East of Jimma university main campus.

3.2. Method of the study

A study has been conducted to asses the factors that affect job satisfaction and level of
job satisfaction among health workers.

3.2.1. Sampling technique and sample size determination

Since a cross sectional of the population is heterogeneous, stratified random sampling has
been conducted to do the research.

The sample size “n” is determined by proportional sampling by using the formula,

Let , then

Where: α be the level of significance (assume α =0.05)

d be the margin of error (for getting more accurate information
assume d= 5%)
P be the estimated proportion of the health workers in the hospital and (assume
p=0.5 and thus, q=1-p=1-0.5=0.5, this is because the population proportion is unknown)

Zα/2 is the normal distribution at 5% level of significance. Thus, from standard table,
Z0.05/2=Z0.025=1.96

N be the total population size whose value has got from the hospital human wealth office
is 318.
Hence, =

= =1.21 which is greater than 0.05. Thus, we must adjust the sample size
as follow.

So, =

Thus, the sample size for this study is 174 health care workers and using this sample size
we have allocated it into nine (9) strata and from each stratum we have selected a sample
by a random case. These strata are called categories.

The categories and their corresponding numbers are listed as follow.

S.NO category Total numbers(Nh(i))

1 Doctor 64
2 Pharmacy 15
3 Laboratory 20
4 Radiographer 9
5 Nursing 200
6 Physiotherapy 2
7 Health assistant 5
8 Optometrist 1
9 sanitarian 2
total 318

Table 3.1: The categories and their corresponding numbers

In order to select a sample of male and female from each category we follow the
following procedure.
General formula for male, nhm (i) =

and for female, nhf(i) =

The number of each category population and strata sample are summarized as follow.

S.No Category Nh(i) Nhm(i) Nhf(i) nhm(i0 nhf(i)

1 Doctor 64 57 7 31 4
2 Pharmacy 15 9 6 5 3
3 Laboratory 20 15 5 8 3
4 Radiographer 9 9 0 5 0
5 Nursing 200 90 110 49 60
6 Physiotherapy 2 2 0 1 0
7 Health assistant 5 5 0 3 0
8 Optometrist 1 1 0 1 0
9 sanitarian 2 1 1 0 1
Total 318 189 129 103 71
Table 3.2: Stratification table

nh(i) =

Key: Ntotal = total number of population (health care workers) in Jimma

specialized hospital.

Nh(i) =total number of population (health care workers) in each category called strata
population.

ntotal =total number of samples selected from the total population.

nh(i) =number of samples selected from each category or strata population.

nhm(i) =number of male selected from each category

nhf(i) =number of female selected from each category

Nhm(i) =total number of male population in each category

Nhf(i) = total number of female population in each category

3.3. Source of population and study population

All health workers of the population who work in Jimma Hospital is source of
population. Workers that has been selected from the source of population in the study
area are the study /or target population.
 3.4. Study variables

3.4.1. Independent variables

The most important variables that are used in analyzing the research are
 Age
 Sex
 Ethnicity
 Religion
 Martial status
 Monthly salary
 Position
 Level of education
 Training in service
 Future plan in the profession
 Service in year
 Access of drugs and materials
 Future plan
 Relation to other team members
.
3.4.2. Dependent variable

 Job satisfaction of health workers

3.5. Method of data collection and measurement

The data has been collected by using structured self administered questionnaire and direct
personal interview. The questionnaire has included socio-demographic characteristics,
level of satisfaction and dissatisfaction variables, level of education, future plan of the
workers, training service in year and things which has been taken to alleviate the
problems.

3.6. Data quality control

 The questionnaire is prepared in simple English words make as much as clear and
easy to understand by any medical personnel.
 Clear and adequate instruction is put on the first page
 Data has been cleaned manually and analyzed with association tables and charts.

3.7. Method of data processing, analyzing and presenting

After the data being collected only a daily basis, it has been checked for completeness
and its internal consistency and analyzed by using SPSS, soft ware, descriptive statistics,
binary logistic regression and logistic regressions. The analyzed data is described and
presented by using different charts, graphs and tables after appropriate frequency
distribution is made for the variables.

3.8. Method of statistical analysis

There are different statistical techniques that have been used in analysis of the collected
data. For example, descriptive statistics, chi-square test statistic, binary logistic regression
and multiple logistic regression models are the most appropriate techniques.

3.8.1. Chi-square test

We have used a chi-square test of independency whether there is an association between

variables or not. The test statistics is given by
χ²=

Where: χ²=chi-square
Oij =observed frequency
Eij=the corresponding expected frequency.

Hypothesis testing
ΗΟ: the given attributes are not significantly associated
Η1: the attributes are significantly associated.

Decision
1. If χ²cal>χ²α(r-1) (c-1), where α is a given level of significance, and r and c are number of
rows and columns respectively,
2. If Р value<α,

We conclude that Ho will reject and conclude that the given attributes are significantly
associated and thus, their association can be analyzed by correlation which indicates
whether these variables have positively or negatively associated.

3.8.2. Logistic regression model

One of the statistical techniques for this study is the logistic regression model. The model
for p-explanatory variables (i.e. x1, x2,…, xp) is given by

Logit (Π(x)) =loge =loge ( )

Where π is the probability of success and 1-π is the probability of failure;
legit (Π(x)) =α+β1x1+β2x2+…+βpxp=α+ βiXi where i=1,2,…,p

Where π(x) is the conditional probability of success and is given by

Π(x) = = exp (α+β1x1+β2x2+…+βpxp)

1+ exp (α+β1x1+β2x2+…+βpxp

That is it is assumed that probability of success is more or less likely depending on

combinations of values of predictor variables.

The parameter βi refer to the effects of Xi on the log odds that y=1, controlling the other
Xs.

Logistic regression is a variation of ordinary regression which is used when the

dependent (response) variable is dichotomous variable (i.e. it takes only two values,
which usually represent the occurrence and non- occurrence, or success or failure, or
satisfy or dissatisfy of some outcome events usually coded as 0 or 1) and the independent
(input) variables are continuous, categorical or both.

For instance in a medical study, the health workers satisfy or dissatisfy, the patient
survives or dies etc.

Unlike ordinal linear regression, logistic regression does not assume that the relationship
between the independent variables and the dependent variable is a linear one. Nor it
assumes that the dependent variable or the error terms are distributed normally.

Logistic regression thus forms a predictor variable (loge ( )) which is a linear

combination of the explanatory variables. The value of these predictor variables are then
transformed into probabilities by a logistic function.

Logistic regression also produces odds ratio (O.R) associated with each predictor values
of a variable. The odds of an event are defined as the probability of an outcome event
occurring divide by the probability of the event not occurring. In general the odds ratio is
one set of odds divide by another.

The odds ratio for a predictor is defined as the relative amount by which the odds of the
outcome increase (O.R greater than 1) or decrease (O.R less than 1) when the value of the
predictor variable is increased by 1 unit.

The odds of response 1 (i.e. the odds of “success”) are

Odds= = exp (α+β1x1+β2x2+…+βpxp) =eβo *(eβ1) x1 *…* (eβp) xp

This implies the odds increase multiplicatively by eβi for every increase in xi. i.e. the odds
at level xi +1 equal the odds at xi multiplied by eβi. When all βi=0, i=1, 2… p the odds
don’t change as the explanatory variable xi changes.

Assumption of logistic regression

1. Dependent variable should be categorical (binary) variable

2. Absence of multi-co linearity
 Chapter four

Data analysis

4.1 Descriptive statistics

Frequency table for characteristics of the respondent

variables Frequenc Valid Cumulative

y Percent Percent Percent

Age age>30"adult" 49 28.2 28.2 28.2

group
18 -30 125 71.8 71.8 100.0

Sex female 40.8

40.8 40.8
71

male 100.0
59.2 59.2
103

Religio Orthodox 43.1

43.1 43.1
n 75

Muslim 64.4
21.3 21.3
37

Protestant 96.6
32.2 32.2
56

Other 100.0
3.4 3.4
6

Ethnicit Oromia 87 50.0 56.1 56.1

y
Amhara 25 14.4 16.1 72.3

Tigray 16 .6 .6 72.9

Snnp 26 14.9 16.8 89.7

Other 35 20.5 10.3 100.0

Table 4.1: Frequency table of the respondent
variables Frequ Percent Valid Percent Cumulative Percent
ency
Educational Certificate 12 6.9 6.9 6.9
status Diploma 75 43.1 43.1 50.0
BSC 55 31.6 31.6 81.6
MSC and above 32 18.4 18.4 100.0
Marital status Single 91 52.3 52.3 52.3
Married 83 47.7 47.7 100.0
Monthly Salary >2200 76 43.7 43.7 43.7
salary Salary <= 2200 98 56.3 56.3 100.0
Training in Long term(year >11 14 8.0 8.0 8.0
service Short term(year<=11) 160 92.0 92.0 100.0
position Clinical 131 75.3 75.3 75.3
Preventive 21 12.1 12.1 87.4
Administrative 5 2.9 2.9 90.2
Neither 17 9.8 9.8 100.0
Part-time No 130 74.7 74.7 74.7
work yes 44 25.3 25.3 100.0
Job No 83 47.7 47.7 47.7
satisfactio yes 91 52.3 52.3 100
n
Human power level of motivation 34 19.5 19.5 19.5

further education 85 48.9 48.9 68.4

others 55 31.6 31.6 100.0

Government access of promotion

policy health and different
123 70.7 70.7
service resources 70.7
founder
long term and short 100.0
29.3 29.3
term training 51
Relation to
Participation of the 100.0
other team
member individual& 70.7 70.7
123
professionals

Interaction with other 100.0

29.3 29.3
team members 51
Table 4.2: Frequency table of the respondent

The above two table show (table 4.1 and 4.2), the frequency and percent of respondents’
characteristics.

Thus, on table 4.1, we can see that 125(71.8%) of the respondents have age between 18
and 30 are youth and 49(28.2%) have age between 31 and above are adult, 103(59.2%)
male, 71(40.8%) are females; 75(43.1%) orthodox, 37(21.3%) are Muslim, 56(32.2%) are
protestant and the rest 6(3.4%) are others religion follower. 87(50.0%) of the respondents
ethnicity are Oromia, 25(14.4%) are Amhara, 1(0.6%) are Tigray, 26(14.9%) are Snnp
and 35(20.1%) are others.
From table 4.2, we can see that 12(6.9%) of the respondent are certificate, 75(43.1%) are
diploma, 55(31.6%) are BSC and 32(18.4%) are MSC and above; 91(52.3%) of them are
single while 83(47.7%) are married; 76(43.7%) of the them have gotten a salary greater
than 2200 and 98(56.3%) have gotten a salary less than or equal to 2200.

Most of the respondents have worked eleven and less than eleven years and a few of them
are worked more than eleven; 131(75.3%) of respondents are mostly involved in clinical,
21(12.1%) are preventive, 5(2.9%) are administrative and 17(9.8%) are in others, most of
them have not part time work, 83(47.7%) of the respondents are not satisfied with their
job while 91(52.3%) are satisfied (see table 4.2).

Table 4.2 also shows the respondents attitude in, which 34(19.5%) have need level
motivation, 85(48.9%) need further education, 55(31.6%) have other plan, 123(70.7%)
have not gotten access of promotion and other resource, 51(29.3%) have not taken long
term and short term training; 123(70.7%) need participation of the individual and
professionals and 51(29.3%) need interaction with other team members.

Their corresponding bar graph also shown on the appendix that shows the same result
that of the frequency table (see appendix B).
 4.2 Analysis of chi-square

Chi-square analysis is used to test whether the attributes (covariates) are significantly
associated or not. The hypothesis is given by;
ΗΟ: the given attributes are not significantly associated
Η1: the attributes are significantly association.

Chi-square analysis for the selected variable

Asymp. Sig. (2- Exact Sig. (2- Exact Sig. (1-

Value df sided) sided) sided)

Pearson Chi-Square 14.038 1 .000

Continuity Correction 12.905 1 .000

Likelihood Ratio 14.214 1 .000

Fisher's Exact Test .000 .000

Linear-by-Linear Association 13.957 1 .000

N of Valid Cases 174

Table 4.3: Chi-square test for sex Vs marital status of the respondent

HO: sex and marital status are not significantly associated;

H1: not Ho.
From the above table 4.3, where (chi-sq= 14.038, df=1, p-value <0.001), which shows
that there is a significance association between sex and marital status.

Asymp. Sig. (2- Exact Sig. (2- Exact Sig. (1-

Value df sided) sided) sided)

Pearson Chi-Square 60.505a 1 .000

Continuity Correctionb 58.110 1 .000

Likelihood Ratio 68.102 1 .000

Fisher's Exact Test .000 .000

Linear-by-Linear Association 60.157 1 .000

N of Valid Casesb 174

Table 4.4: Chi-square test for sex Vs salary of the respondent

HO: sex and salary are not significantly associated;
H1: not Ho.
From the above table 4.4, where (chi-sq= 60.505, df=1, p-value <0.001), which shows
that there is a significance association between sex and monthly salary.

Asymp. Sig. (2- Exact Sig. (2- Exact Sig. (1-

Value df sided) sided) sided)

Pearson Chi-Square 13.605a 1 .000

Continuity Correctionb 12.993 1 .000

Likelihood Ratio 13.614 1 .000

Fisher's Exact Test .000 .000

Linear-by-Linear Association 13.584 1 .000

N of Valid Casesb 174

Table 4.5: Chi-square test for age Vs job satisfaction of the respondent

HO: age and job satisfaction are not significantly associated;

H1: not Ho.
From the above table 4.5, where (chi-sq= 13.605, df=1, p-value <0.001), which shows
that there is a significance association between age and job satisfaction.

Asymp. Sig. (2- Exact Sig. (2- Exact Sig. (1-

Value df sided) sided) sided)

Pearson Chi-Square 1.629a 1 .202

Continuity Correction 1.258 1 .262

Likelihood Ratio 1.630 1 .202

Fisher's Exact Test .220 .131

Linear-by-Linear Association 1.619 1 .203

N of Valid Casesb 174

Table 4.6: Chi-square test for sex Vs job satisfaction of the respondent

Ho : sex and job satisfaction are not significantly associated;

H1: not Ho.
Thus, from the above table 4.6, where (chi-sq= 1.629, df=1, p-value >0.001), which
shows that there is no significance association between sex and job satisfaction.
 4.3 Analysis of Logistic Regression

Case processing summary

The total number of selected cases are174 and the number of rejected to be zero and
number of cases included in the analysis is 174(see appendix D, table D.1).
Job satisfaction is the dependent variable and is coded 0 for no and 1 for yes.

The Omnibus Tests of Model Coefficients table indicates that, we consider all six
predicators, the model or equation is significant (chi-sq= 13.334, df= 6, p-value
<0.001). for the model is less than the commonly used level of significance value
0.05.reject the null hypothesis which states that the entire model coefficients are
zero( Ho: βi=0, where i=1,2,3,4,5,6).

Table for the significant covariates

covariat 95.0% C.I.for EXP(B)

e
variable
s B S.E. Wald df Sig. Exp(B) Lower Upper

Step Age .689 .372 3.424 1 .034 1.991 1.960 4.129

a
1
Educ.stat
.729 .292 6.218 1 .013 2.073 1.169 3.677
us

Mont.
.414 .491 .712 1 .025 1.513 1.578 3.960
salary

G.policy .262 .352 .556 1 .016 1.300 1.653 2.589

Constant -2.637 .998 6.982 1 .008 .072

Table 4.7: Variables in the Equation

The model of logistic regression

logit (Π(x)) =α+β1x1+β2x2+β3x3 + β4x4

y = -2.637+0.689X1+0.729X2+0.414X3+0.262X4

Where y= logit (Π(x))

X1= age of the respondent
X2= educational status of the respondent
X3 = monthly salary of the respondent
X4=government policy.
 Interpretation of logs odds

The coefficient B1= 0.689 represents the change in the logit for a change of one unit in
the professional age. This indicates that the probability of satisfy with job is changed
(increased by 99.1% for a unit change in profession age).

The coefficient B2= 0.729 represents the change in the logit for a change of one unit in
the professional educational status. This indicates that the probability of satisfy with job
is changed (increased by 107.3% for a unit change in educational status of the
professional).

The coefficient B3= 0.414 represents the change in the logit for a change of one unit in
the monthly salary. This indicates that the probability of satisfy with job is changed
(increase by 51.3% for a unit change in professionals’ monthly salary).

The coefficient B4= 0.262 represents the change in the logit for a change of one unit in
the government policy. This indicates that the probability of satisfy with job is changed
(increase by 30% for a unit change in government policy).

As indicated in the Table 4.8 above, among the explanatory variables age, educational
status, monthly salary and government policy are significant predictors when all six
variables are considered together.

As well the p-value for the parameter B of the coefficients of the variables show the
variable is significant if it is less than 0.05 and not significant if it is greater than 0.05.
Thus age, educational status, monthly salary and government policy have significant
effect on the job satisfaction, while sex and position have not.

The expression Exp (B) is called odds ratio for the response variable job satisfaction,
shows the effect of each variable on job satisfaction. This implies the odds increase
multiplicatively by eβi for every increase in xi. i.e. the odds at level xi +1 equal the odds at
xi multiplied by eβi.

The formula for the odds ratio for job satisfaction is

Odds= = exp (α+β1x1+β2x2+…+βpxp) =eβo *(eβ1) x1 *…* (eβp) x

=exp (-2.637+0.689X1 +0.729X2+0.414X3+0.262X

Interpretation of odds

The odds ratio for professionals age is 1.991, for educational status is 2.073, for monthly
salary is1.513, and for government policy is 1.300. This indicates that the odds of
estimating correctly who satisfy with their job by 99.1% if one knows the age group,
107.3% if one knows the educational status, 51.3% if one knows monthly salary and 30%
if one knows the government policy.

The variables that have not used for the analysis of logistic regression are multi-collinear
with those variables which are used for this analysis and have no important significant on
professionals job satisfaction. Thus, the analysis for logistic regression are used the above
four variables and that have used infer their effects on the job satisfaction.
When all βi=0, i=1, 2, 3, 4, the odds don’t change as the explanatory variable xi changes.

Chapter five

Conclusion and recommendation

5.1 Conclusion

From this study we can conclude that 47.7% of health professionals who work in Jimma
specialized hospitals are not satisfied (most of them are highly) with their job.
The analysis of logistic shows age, sex, educational status, and government policy are
most prominent factors for job satisfaction. Not only these factors but annual leave and
quality of education are also some of the most determinant factor affecting job
satisfaction.

Although there are many factors that make the professional dissatisfy, there are few
factors, like professional gratification and helping in the needy, has make the professional
satisfy with his/her job low level of satisfaction.

As the professional have indicated, their main future plan apply for further education in
the existing profession, work in private sector and continue to work in the health sector in
these place. A few professionals want to change the profession and ask for transfer and
work in somewhere else.

Although there are new plans to improve the career structure and improve salary scale of
workers, much more need to be done to improve the working environment and the
existing health care delivery system.
The result of the collected data indicates some doctors have part time work and have got
payment for it and are most of them are satisfied with their job.

Generally, the Ethiopian quality health coverage and satisfaction of health professionals
is medium. The main reasons that make the country’s health quality below the expected
standard are poor transparent relationship between the staff members, committed
responsibility of the professionals and management structure, misdistribution of well
organized and facility based on quality, lack of god working habit and lack of well
educated professionals.

5.2 Recommendation

 Policy makers must acknowledge existing problems and seek solutions by

forming partnership with practicing health professionals.
 Health service managers should encourage and re-ward health professionals
for better performance.
 Government sectors must give due attention to improve infrastructure to better
retain health professionals and benefit every one.
 All levels of health personnel must be provide with incentives by considering
the relative isolation and difficulty of the condition they live and work than
other professionals, these incentives should be adopted to local situation and
may take such forms as better working conditions, opportunity for further
training and continuing education.
 Wider the chance of further education ,
 Increase supply of medical instruments and essential drugs,
 Moderate transparent relationship between staff members and management
structure,
 The government must avoid biasness between professions,
 Quality education in recent situation in health care system is deteriorating for
unspecified reasons.
Finally, depending on the collected information, government policy makers and health
service managers should work in collaboration with health professionals in order to
improve the quantity and quality of health care service. It is also believed that this study
can be used as a base line data for further study of the problems on large scale to regional
or national level.
 Appendixes

Appendix A: Questionnaire

Jimma University
College of natural science
Department of statistics

The questionnaire is prepared for collecting information on job satisfaction of health

workers in Jimma Hospital.
Dear participants:

The general purpose of this study is to determine health workers job satisfaction and to
describe factors which affect their level of satisfaction and to provide information that
can be use for action.

There for,
1. You are kindly requested to answer each question honestly.
2. Mark “Ö” in front of your choice number on the space provided.
3. For certain questions when additional information is required, thus, please write
your opinion on the space provided freely and honestly.
Thank you!
Part I; General information on socio-demographic characteristics

1. Age: --------------
2. Sex: Male „ Female „
3. Religion: Orthodox „ Muslim „ Protestant „ Other------------
4. Place of birth: --------------
5. Educational status: Certificate „ Diploma „ BSC „ MSC and above „
6. Martial status: Single „ Married „ Divorce „ Widow „
7. Monthly salary (in Ethiopian birr):------------
8. Your family size: -------------

Part II; Information on education and work experience

9. What is your profession or qualification? ----------------------

10. For how long have you served in health professional? -----------------
11. Position you have mostly involved
1. Clinical „ 2. Preventive „
3. Administrative „ 4. Other, specify-----------------

12. Do you have part-time work? Yes „ NO „

13. If your reply “Yes” for question 12, have got a payment from it?
Yes No
14. What do you think about the level of quality Ethiopian health coverage?
 Very high „ High „ Medium „ Low „ Very law „
15. If your answer is “Neither very high nor high” for question 14, what are the main
causes and measurements do you think?
Cause ---------------------------------------------------------------------------
---------------------------------------------------------------------------
---------------------------------------------------------------------------
Measurement -----------------------------------------------------------------
-----------------------------------------------------------------
-----------------------------------------------------------------
16. Are you satisfied with the job you are providing? Yes  No 
17. What is your level of satisfaction if your answer is “Yes” for question 16?
Very high „ High „ Medium „ Low „ Very low „
18. If your answer is “No” to question 16, what make you dissatisfied?
i) Problems with related to health human power
1, level of motivation „ 2, annual leave „
3, farther educations „ 4, transfer „
5, inappropriate maintenance of stock „
6, other, specify -------------------------------------
--------------------------------------
ii) Problems related to policy makers and health service founders:
1, lack of promotion „
2, inefficient resources of
 Medical instruments „
 Essential drugs „
 Chemical regents „
 Poor infrastructure „
3, inadequate salary „
4, insufficient long term training „
5, inefficient short term training „
6, inefficient in service training „

iii) Problems related to other team members:

1, poor participation: œ
2, shortage of man power (professionals) : œ
3, poor interaction with other team members: œ
iv) Other factors;
1) Fear of possible infections including HIV/AIDS: œ
2) Difficult in getting recent information: œ
3) Or specify others: -------------
19. What is the degree of your dissatisfaction if your answer in “No” for question 16?
Very high œ High œ Medium œ Low œ Very low œ
20. What is your future plan? You can select more than one
1) Continuing to work in the health sector in this place œ
2) Ask for transfer and work in some where else
3) Change the profession œ
 4) Work in private sector œ

5) Apply for further education in the existing profession œ

6) Other, specify -------------------------------------------

21. What do you think should be done to minimize work dissatisfaction among health
workers?

1) Increase salary œ

2) Increase supply of medical instruments and essential drugs œ

3) Re-evaluate the health policy œ

4) Wider the chance of future education œ

5) Other, specify --------------------------------------------------------------
---------------------------------------------------------------
---------------------------------------------------------------
22. Comment ---------------------------------------------------------------------
----------------------------------------------------------------------
----------------------------------------------------------------------

THANK YOU!
Appendix B: Bar graph for the respondents’ characteristics

Figure B.1: Bar graph for the respondent’s age

Figure B.2: Bar graph for the respondents’sex.
Figure B.3: Bar graph for the respondents’ religion
Figure B.4: Bar graph for the respondents’ehtnicity.
Figure B.5: Bar graph for the respondents’ educational status
Figure B.6: Bar graph for the respondents’ marital status.
Figure B.7: Bar graph for the respondents’ monthly salary (in Ethiopia birr).
Figure B.8: Bar graph for the respondents’ training in service (in year)
Figure B.9: Bar graph for the position of respondents’ who mostly involved.
Figure B.10: Bar graph for part-time respondents
Figure B.11 Bar graph for the job satisfaction of respondents.
Figure B.12 Bar graph for the attitude of the respondent.
Figure B.13 Bar graph for the needs of the respondent
Figure B.14 Bar graph for the relative relation of the respondent
 Appendix C: Cross tabulation

Cross tabulation for some selected covariate variables

Sex of the respondent * Marital status of the respondent

Marital status of the respondent

single married Total

Sex of the respondent female Count 25 46 71

Expected Count 37.1 33.9 71.0

male Count 66 37 103

Expected Count 53.9 49.1 103.0

Total Count 91 83 174

Expected Count 91.0 83.0 174.0

Table C.1: Cross tabulation for sex versus marital status

Sex of the respondent * Monthly salary of the respondent

Monthly salary(in Ethiopian birr) of

the respondent

salary > 2200 salary <=2200 Total

Sex of the respondent female Count 6 65 71

Expected Count 31.0 40.0 71.0

male Count 70 33 103

Expected Count 45.0 58.0 103.0

Total Count 76 98 174

Expected Count 76.0 98.0 174.0

Table C.2: Cross tabulation for respondents’sex versus monthly salary.

 Cross tabulation for some selected covariate variables and dependent
variable, job satisfaction
Age of the respondent * Job satisfaction of the respondent

Job satisfaction of the respondent

no yes Total

Age of the age>30"adult" Count 29 20 49

respondent
Expected Count 23.4 25.6 49.0

18 <=age<= 30 ="youth" Count 54 71 125

Expected Count 59.6 65.4 125.0

Total Count 83 91 174

Expected Count 83.0 91.0 174.0

Table C.3: Cross tabulation for age Vs job satisfaction of the respondent

Sex of the respondent * Job satisfaction of the respondent

Job satisfaction of the respondent

no yes Total

Sex of the respondent female Count 38 33 71

Expected Count 33.9 37.1 71.0

male Count 45 58 103

Expected Count 49.1 53.9 103.0

Total Count 83 91 174

Expected Count 83.0 91.0 174.0

Table C.4: Cross tabulation Sex of the respondent * Job satisfaction of the respondent Cross tabulation
Appendix D: Logistic analysis table

Unweighted Cases N Percent

Selected Cases Included in Analysis 174 100.0

Missing Cases 0 .0

Total 174 100.0

Unselected Cases 0 .0

Total 174 100.0

Table D.1 Case processing summary

Predicted

Job satisfaction of the respondent

Percentage
Observed no yes Correct

Step 0 Job satisfaction of the no 0 83 .0

respondent
yes 0 91 100.0

Overall Percentage 52.3

Table D.2 Classification table

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

Step 0 Constant .092 .152 .368 1 .544 1.096

Table D.3 Variable in the equation when only constants are used

Score df Sig.

Step 0 Variables Age 3.605 1 .018

sex 1.629 1 .002

Educ.status 7.018 1 .008

Mont. salary .991 1 .020

G.policy .602 1 .038

Position .437 1 .009

Overall Statistics 12.829 6 .006

 Table D.4: Variables not in the equation

Cox & Snell R Nagelkerke R

Step -2 Log likelihood Square Square

1 227.513a .274 .098

Table D.5: Model Summary

Table for variables encoding

Variables Value
0 1 2 3 4 5
Age >30 18-30
Sex female Male
Religion orthodox Muslim protestant Other
Ethnicity Oromia Amhara Tigray SNNP other
Educational status certificate Diploma BSC MSC
and
above
Marital status single married
Monthly salary >2200 birr <=2200
Service >11 years <=11years
Position clinical preventive administrative other
Part time no yes
Human power Level of Further others
motivation education
Government policy Access of promotion and Long term
different resources and short
term
training
Relation Participation of the Interaction
individual&professionals with other
team
members
Job satisfaction No Yes
Table D.6: Variable encoding system

Reference

1. Government of Ethiopia and the World Bank (1996) Health sector review:
Ethiopia social sector studies, Addis Ababa.
2. Medline: EI-gilary; Awehaty (2001) job satisfaction of female Saudi nurse:
Eastern Mediterranean health Journal, vol-7, no ½, 2001.\
3. Medline: Saranakas; Janawddis: job satisfaction of doctors in Nigeria: Medical
Journal of Malaysia, 52/3 September 1997.
 4. Unpublished research: Lema Tefera, H.O factors affecting job satisfaction of
health workers who working in Jimma zone, South West of Ethiopia January
1998.
5. Medline Larnasee JH, Janny MA, Ostraw cl, with raw ML, Hobbs, GRJV, Burant
C: major predictor of intent to level and psychological empowerment of nurse in
West Vergina; Medical journal may 2003.
6. JLI In: Human resource for health over coming the crises joint learning-
Inventive, 2004.
7. Sanders D, Lehman U, Ferrinho P. health sector reform: Some implications for
human resources: December 2004.
8. Labonte R, Schrecker T, Sanders D, Meeus W. Fatal Indifference: the G8, Africa
and Global Health Cape Town, UCT and IDRC, 2004.
9. Zurn P, Dal PazMR, still Well B, Adams O. Imbalance in the health work force.
Human resource health 2004.
10. Internet explorer “en.Wikipeda.org/wikilogstic-regress”

BIBLIOGRAPHICAL SKETCH

The authors were born in Mekane-Eyesus and Estie special woredas on July 7, 1988 and

1887 respectively. They attended elementary and junior secondary schools from 1996 to

2008 in Mekane-Eyesus primary School Ginna Memcha and completed secondary school
at Mekane-Eyesus Senior Secondary School in 2008. They joined in Jimma University

and graduated with BSc. degree in statistics in 2011.

DECLARATION

We declare that this research paper is the result of my work and that all

sources of materials used are duly acknowledged. This work has not been

submitted to any other university for achieving any academic degree awards.

Name: Ayenew Molla

Signature ______________

Name: Fikire Mesafnint

Signature ______________

Place: Jimma University

Date of submission: May 31,2011