BINDURA UNIVERSITY OF SCIENCE EDUCATION

THE RELATIONSHIP OF MOTHER’S LEVEL OF EDUCATION TO CHILD

HEALTHCARE SERVICES UTILIZATION THE CASE FOR ZIMBABWE FROM 2000 TO

2014.

Submitted by

WERENGANI SIMON

B0823136

RESEARCH SUPERVISOR : MR DHAMIYANO

A DISSERTATION / THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE

REQUIREMENTS FOR THE HONOURS DEGREE IN ECONOMICS OF BINDUIRA

UNIVERSITY OF SCIENCE EDUCATION. FACULTY OF COMMERCE

FEBRUARY 2016

i
 RELEASE FORM

NAME OF AUTHOR : WERENGANI SIMON

STUDENT REG : BO823136

TITLE PROJECT: The relationship of mother’s level of education to child healthcare services
utilization the case for Zimbabwe from 2000 to 2014.

DEGREE FOR WHICH : BACHELOR OF SCIENCE (HONOURS) DEGREE IN

ECONOMICS PROJECT WAS PRESENTED

YEAR GRANTED : 2016

Permission is hereby granted to the Bindura University Library to produce single copies for
private, scholarly or scientific research purpose only. The author does not reserve other
publication rights and the project nor may extensive extracts from it be printed or otherwise
reproduced without the author’s written permission.

SIGNED ……………………………………………

PERMANENT ADDRESS : Mutumba Primary Sch, P Bag 1026, Madziwa

DATE : 29 February 20160

ii
 APPROVAL FORM

The undersigned certify that they have supervised, read and recommend to the Bindura
University of Science Education for acceptance a research project entitled: The relationship of
mother’s level of education to child healthcare services utilization the case for Zimbabwe
from 2000 to 2014, submitted by Werengani Simon in partial fulfilment of the requirements for
the Bachelor of Science (Honour’s) Degree in Economics

……………………………………………… …/…………/…………/
(Signature of Student) Date
........................................................................ …/…………/…………/
(Signature of Supervisor) Date
……………………………………………… …/…………/…………/
(Signature of the Chairperson) Date
……………………………………………… ……/…………/…………/
(Signature of the Examiner (s)) Date

DECLARATION

I, Werengani Simon, declare this research project herein is my own work and has not been
copied or lifted from any source without the acknowledgement of the source.

…………………… ……/…………/…………/
Signed Date

iii
 DEDICATION

I dedicate this dissertation to God Almighty without whom anything is possible, to my dearest
mother Denga, my friend Charity and Vimbai for their support and to my parents who brought
me up and taught me the value of hard work and love.

iv
 ABSTRACT

The study investigated the determinants of demand for health care services in rural Zimbabwe
using household data from Shamva District in Mashonaland Province. Specifically the paper
studied the influence of socio- economic and institutional features such as household income,
distance to the health centre, education of the mother, religion of the mother and age of the
mother on the probability of seeking of child health care services utilisation from health
facilities. From 29 wards which were selected for the survey, 290 households were randomly
selected in cluster and interviewed ten per every ward in 29 wards. A self-administered
questionnaire was used to collect data between January and February 2014.

The study used a logit model to find the determinants of demand for health care services based
on 290 households that had reported sickness of a member within the last three months before
the survey. The study revealed the statistical significance of education of the mother, household
income and religion as determinants of demand for child health care services utilisation.
Religions of the mother, age of the mother and Distance to the nearest health facility were found
to negatively affect the demand for child health care services utilisation while household income
and education of the mother were found to positively influence demand for health care services.
To increase the demand for health care services, the study acclaims policies that target to educate
the Apostolic sector on the importance of Child healthcare services and shorten the distance
mothers’ travel to health facilities such as acquaint with community based travelling clinics.
Other major approvals of the study include increasing government funding to rural health
facilities through employing income creating projects to improve rural household incomes.

v
 ACKNOWLEDGEMENTS

I would like to pass my heart-felt gratitude to my academic supervisor, Mr Dhamiyano for his
patience in assisting and guiding me through this dissertation .Special thanks also to Dr Kairiza
for his effort in making this project data to be completely analysed and to Bindura University of
Science Education for permitting me an opportunity to study with the institution. I am
industriously thankful to all the lecturers, who imparted their knowledge in me and encouraged
me to soldier on through the degree programme. I also acknowledge the pioneers of this work
whose work have guided me in writing of this research paper they did a great job without them
this paper would not have been possible. I would like to extent grateful acknowledgements to my
family for support, concern, love, tolerance and the financial hardships they underwent in my
education up to this level. To my fellow students your help is greatly acknowledged and I am
very grateful and will remain indebted to you all for the assistance you gave me.

vi
 TABLE OF CONTENTS.

RELEASE FORM ........................................................................................................................... ii

DECLARATION ........................................................................................................................... iii

DEDICATION ............................................................................................................................... iv

ABSTRACT .................................................................................................................................... v

ACKNOWLEDGEMENTS ........................................................................................................... vi

TABLE OF CONTENTS. ............................................................................................................. vii

LIST OF TABLES ......................................................................................................................... ix

LIST OF ACRONYMS ................................................................................................................. ix

INTRODUCTION .......................................................................................................................... 1

1.0 Introduction ........................................................................................................................... 1

1.1 Background to the study ....................................................................................................... 1

1.2 Problem statement ................................................................................................................. 3

1.3 objectives of study ................................................................................................................ 4

1.4 Research questions ................................................................................................................ 4

1.5 Hypothesis............................................................................................................................. 4

1.6 Significance of the study....................................................................................................... 5

1.6.1To the economy................................................................................................................... 5

1.6.2 Public health sector ............................................................................................................ 5

I.7 Assumptions .......................................................................................................................... 6

1.9 Delimitation of the study ...................................................................................................... 6

1.11 Definition of terms .............................................................................................................. 7

1.12 Summary ............................................................................................................................. 8

CHAPTER 11 ............................................................................................................................. 9

LITERATURE REVIEW ........................................................................................................... 9

vii
 2 .0 Introduction .......................................................................................................................... 9

2.7.5 Empirical literature review .............................................................................................. 19

3.1 THEORETICAL MODEL .................................................................................................. 25

3.3.0 RESEARCH DESIGN ..................................................................................................... 26

3.3.1 Model specification .......................................................................................................... 26

3.4 DESCRIPTION AND JUSTIFICATION OF VARIBLES ................................................ 27

3.4.1 Child healthcare utilization (ch)....................................................................................... 28

3.4.3Mother’s or household Income (Y)................................................................................... 28

3.4.4 Mother’s Age ................................................................................................................... 29

3.4.5 Distance to the health services\0 ...................................................................................... 29

3.5 RESEARCH INSTRUMENTS ........................................................................................... 30

3.5.1 Mathematical tools ........................................................................................................... 31

Econometric package ............................................................................................................ 31

3.5.2 Summary statistics ........................................................................................................... 32

3.5.3 Diagnostic checking ......................................................................................................... 32

3.5.4Multicollinearity ............................................................................................................... 33

3.5.5 F Value and Prob (F)........................................................................................................ 33

CHAPTER 1V .......................................................................................................................... 36

DATA ANALYSIS AND RESULTS PRESENTATION ........................................................ 36

4.0 INTRODUCTION .............................................................................................................. 36

4.1 SUMMARY STATISTICS ................................................................................................. 36

4.2 Diagnostic test ........................................................................................................... 37

Summary ............................................................................................................................... 43

viii
 LIST OF TABLES

Table 4.1 summary statistics ……………………………….34

Table 4.2 unit root test……………………………………….35
Table 4.3 regression analysis…………………………………36

LIST OF ACRONYMS

ADF Augmented Dickey-Fuller

BLUE Best, Linear, Unbiased, Estimator
CSO Central Statistical Office
DW Durbin Watson
MDG Millennium Development Goal
WHO World Health Organization
UNESCO United Nations
UNFPA United Nations
NHS National Health Survey
DHS Demographic and Health Survey
ix
ZIMSTAT National Statistical Agency of Zimbabwe

x
 CHAPTER 1

INTRODUCTION

1.0 Introduction

It is important to have an understanding of the general profile of Zimbabwe in relation to

the health sector. It helps to bring out issues of importance in the analysis, so the
following section gives an overview of the country’s general profile

1.1 Background to the study

The Millennium Development Goal (MDG) four targets are to reduce child mortality by
two thirds in the year 2015. This is relevant as the progress and future of any country
depends on how healthy the children are. WHO, (2008) revealed that this is reflected in
their access to basic health care, nutritious food and a protective environment, and
if these are not available, the country’s mortality rates would increase and economic
potentials diminish. WHO , (2005) portrayed that the highest rates of child mortality
are still in Sub-Saharan Africa-where 1 in 8 children dies before the age of 5 years, more
than 20 times the average for industrialized countries despite action plans, interventions
and broad approaches toward improving child’s health in the region . United Nations
Zimbabwe progress report, ( 2012) said that under-five mortality rate decline is 102
deaths per 1,000 live births in 2000 to 89 per 1,000 in 2014, and infant mortality rate
declined is 65 deaths per 1,000 live births in 2000 to 57 deaths per 1,000 live births in
2010/2011. MDG Progress Report Zimbabwe, (2012) said this decline is not fast enough
to achieve the MDG 4 target of 34 per 1, 000. World Bank, (1993) revealed that
investments in women’s education are important for lowering infant and child mortality
and improving child health.

1
Zimbabwe Health Europe Projects, (2012) revealed that chronic malnutrition limits the
life prospects of more than one third of the country’s children. Zimbabweans continue to
experience a heavy burden of disease dominated by preventable diseases such as Human
Immune Virus (HIV) infection and Acquired Immune Deficiency Syndrome (AIDS),
malaria, tuberculosis and other vaccine-preventable diseases, diarrheal diseases and
health issues affecting pregnant women and neonates. ZMDGs Report, (2012) stressed
that 20% of Zimbabwean households lacking access to safe drinking water and 35%
having no access to improved sanitation; children are regularly exposed to water-borne
diseases. The situation is particularly bad for rural households (30% and 50% without
access to safe drinking water and improved sanitation respectively) and poor households
(45% and 89% without access to safe drinking water and improved sanitation
respectively) compared to their urban and wealthier counterparts.

The number of infants in the 12-23-month age group who were vaccinated against
measles by 12 months of age (69.3%) is relatively lower than that for other major
vaccines (BCG 87%, DPT185%, DPT2 80%, Polio1 87%, Polio2 80%).Hong, (2006)
showed that levels of infant and child mortality in many developing countries remain
unacceptably high. Kevany et al., (2011) revealed that Saharan Africa in recent years,
the utilization of health care services in the region has remained low . NHS, (2009-
2013) said Zimbabwe is not an exception to this as the demand for of health care services
has been reported to be very low. Caldwell, (1979) suggested that female education
influence children’s health and survival with regard to pregnancy, childbirth,
immunization, and management of childhood diseases.

Singh et al., (2003) revealed that prenatal care and the ability to avoid high-risk births
help prevent infant and Child deaths. UNFPA, (2004) revealed that reductions in child
mortality require, inter alia, attention to neonatal health including nutrition and
immunization as well as avoidance of high-risk pregnancy and attention to the care and
the wellbeing of women during pregnancy, delivery, and the post-partum period. UN

2
Millennium Project,( 2006) exposed that breastfeeding protects babies and infants
from infectious and chronic disease – diarrhoea and acute respiratory diseases helping
to recover from illness.. An HIV-positive mother may reduce the risk of postnatal HIV-
transmission when she breastfeeds her child exclusively. Mosley and Chen, (1984)
suggested that the mother’s educational level increases her skills in health care
practices related to disease treatment, preventive care and hygiene.

Michael Grossman, ( 1972) depicted that the utilization of health services is influenced
by several factors being social class, work environment, employment status, income,
housing conditions, heating; education, diet and lifestyle. One of the factors that are
supposed to stimulate child health services is the education level of the mother. Several
studies have found that improvement in mother level of education improves the utilization
of child health services hence reduce child mortality. Thomas, Strauss and Henriquez,
(1991) ,Kassouf and Senauer ,(1996) agreed that parental education had both a direct
impact and indirect effects, via wages and full income, on child health, particularly, the
mother’s education.

UNESCO, (2012) revealed that Zimbabwe Literacy rate has increased from 91% in 2009
to slightly above 99% in 2011. There is an improvement in mother education but not
accompanied by child health, therefore the research would like to investigate the veracity
of the findings on the relationship of mother education and child health.

1.2 Problem statement

Despite the fact that UNESCO, (2012) revealed that Zimbabwe Literacy rate has
increased from 91% in 2009 to slightly above 99% in 2011 .The CIA World Fact Book
,(2014) and UN data Statistics division World Bank, (2014) revealed that Zimbabwe
under-five mortality rates have seen slight improvements from 94 per 1000 in 2011 to 89
per 1000 in 2014. The current under-five mortality rate of 89 per 1000 is at a distant from
the 2015 target of 34 per 1,000. Health staff, clinics and hospitals have been made to cater

3
for mother – child care .There is still a rising concern on child mortality. Children are
professionals of tomorrow, high child death rate leads to unfavourable economic growth.

The demand for healthcare is crucial for the design of effective health policy in
developing countries. It is against this background that the researcher would like to find
out the significance of mother level of education on child healthcare demand as a way of
improving child healthcare consumption and demand.

1.3 objectives of study

1. To deduce if mother’s education stimulate child healthcare demand and consumption.

2. To determine whether the age of the mother influences the relationship between mother
education and demand for healthcare
3. To determine whether income of the mother influences the relationship between
mother education and demand for healthcare.

1.4 Research questions

1. Does the mother level of education affect the child healthcare services?
2. Does the age of the mother age affect child health services?
3. Does the income of the mother age affect child health services?

1.5 Hypothesis

Ho: The mother education level determines the demand for child health services.
Hi: The mother education level does not determine the demand for child health services.

4
1.6 Significance of the study

To date very few studies have given a clear magnitude of education level of the mother as
an explanatory in the Grossman model (Grossman, 1972). Several studies among them
being Van Door etal, (2013), Case and Deaton, (2005) and Lieras- Muney, (2005) gave a
les controversy on education being a powerful determinant of health status, however, little
work, has been done in the case of developing countries concerning the impact of the
mother level of education to child healthcare services utilisation. It is the aim of this study
therefore to provide further empirical evidence on the impact of mother level education to
child health services utilization in Zimbabwe from 2000 to 2014 and other determinants
of child healthcare demand will be considered although much being on mother level of
education.

1.6.1To the economy

After realizing many children losing their lives daily it is important in understanding what
is happening on the health sector and how resources can be allocated. It helps in the
designing of the policies that influence effective and efficient utilization of healthcare
services. This will help to create future talents in the economy.

1.6.2 Public health sector

Knowing the effects of mother level of education on child healthcare services utilization
the public health sector will be in a better position to decide policies that minimizes the
cost of child life as well as maintaining good health care services.

1.6.3 Researcher

The researcher will gain a valuable experience in the practical knowledge in the field of
research and also being able to do informed decisions in the healthcare services

5
1.6.4 Scholars

Future researchers will access a platform to further studies improving the knowledge
gained in this research

1.6.5 Mothers

The study will exposes weaknesses and areas which need improvement in individual
healthcare delivery, mothers to be equipped with adequate information so as reduce infant
mortality rate, thus continue to make every pregnancy wanted and every child birth safe

I.7 Assumptions

 The study is based on the assumption that:

 Participants would supply the researcher with the required information.
 Data and information gathered from the sample was reliable and representing the
true population.
 Mother affects child health services utilisation.

1.9 Delimitation of the study

 Secondary data was used in the study which reduces the costs of carry out field
surveys.
 The study has been carried out on the case of Zimbabwe (2000-2014) only so it
was time consuming.
 All the data required was obtained from National Statistical Agency of Zimbabwe
(ZIMSTAT)

6
  The study was specific since it looks on effect of the mother level of education to
child healthcare services utilisation

1.10 Limitations

 It should be said that there are other areas of concern that the current study did not
cover.
 The study did not give much interest to the effects of public policies on
restructuring of child healthcare services utilisation.
 Another shortfall of concern is that the study used secondary data that has its own
limitations. It is not a disputable phenomenon that, data problems in terms of
quality, consistency, accuracy and reliability are very acute in less developed
countries, Zimbabwe included.
 Due to confidentiality of information that concerns the nation, it was difficult for
the researcher to find the data since a lot of procedures has to be followed .

 These tinted problems in this study are likely to decrease the precision of the
parameter

1.11 Definition of terms

Child HealthCare: consist of immunisation growth in terms of weight and age thus care
for the sick children against diarrhoea among others. Demographic and Health Survey
(DHS), 2010-11
Neonatal mortality: the probability of dying within the first month of life. (DHS), 2010-
11
Post-neonatal mortality: the difference between infant and neonatal mortality. (DHS),
2010-11

7
Infant mortality: the probability of dying before the first birthday. (DHS), 2010-11
Child mortality: the probability of dying between the first and fifth birthday. (DHS),
2010-11
Under-5 mortality: the probability of dying between birth and the fifth birthday. (DHS),
2010-11
Literacy rate is the number of persons, who are aged 15 and over and able to read and
write English. (DHS), 2010-11

1.12 Summary

This chapter involved the background of the study which provides the context of study,
the statement of problem which shows the problem to be solved by the study. It also
includes the purpose of the study, research questions and the statement of hypothesis
which is the probable answer of the study. Chapter one involved the significance of the
study that is the importance of the study, assumptions, limitations and delimitations of the
study. Chapter leads the second chapter of the study which presents the literature review.

8
 CHAPTER 11

LITERATURE REVIEW

2 .0 Introduction

This chapter explores and various literature in order to ascertain what other authorities
have contributed to the child healthcare demand .Reviewing literature enable to identify
any research gaps to be filled by the present study as well as compare findings with those
of other researchers to bring the research problem into focus, the study include theoretical
fold dealing with conceptual elements and empirical folds dealing with assessing previous
studies. Hart, (1998) said that literature review is the selection of available documents on
the topic that contains data, ideas, evidence, and information written from a particular
standpoint to fulfil certain aims on the nature of the topic.

2.1 Theoretical perspective

The study will first discuss the Grossman’s model which provides a theoretical
basis upon which many health care demand studies have been done since Grossman’s
seminal paper in 1972. The extension of the Grossman’s model will be discussed under
the household model and Andersen’s model of health care demand. The predictions of
these models will be relevant in this study.

2.2. Demand for Health services

Grossman’s demand for health model (1972)

Education Effect

9
Grossman, (1972a, 1972b) asserts that those who have obtained higher levels of education
(schooling) are more efficient producers of health. Given that, health is a function of
medical care, environment, and lifestyle, the health production function shifts up and
becomes steeper with more education because people can produce it more efficiently.

The Grossman’s model of demand for health is based on certain underlying assumptions
the central assumptions that individuals are producers of health, stock of health that
depreciates over time at an increasing rate, at least after some stage in the life cycle.
Stock of health can also be increased through acts of investment (health is
endogenous).

Education in the model is positively related to demand for health stock in both variants.
Increasing in education captures increasing knowledge in the production of health and it
enables individuals to choose more health consumption decisions which lower the rate of
health capital depreciation. The implication of education in the model is that those
individuals with more number of school years are likely to effectively utilize
resources to produce health and demand less health care services. Education increases
both the demand for health and the productivity of health inputs; the effect of rising
education on the demand for health care is directly proportional. Those who choose to
obtain higher levels of education (schooling) possess other characteristics that increase
their likelihood of being healthier. Thus, this outside factor confounds the attainment of
the direct causation of education (schooling) on health.

Health economists have done extensive analysis regarding the correlation between
education and stock of health. Researchers have found that education itself is very
difficult to measure and quantify. This is due to the fact that education can be obtained in
many ways, both conventionally and informally, therefore as in this study, health
economists use formal schooling as a measure of educational attainment or level. The
effect of years of schooling in terms of the production of health model can be easily
studied through statistical analysis (Grossman, 1972a, 1972b).

10
2.3 Income Effect

Grossman (1972) constructed a model where individuals use medical care and their own
time to produce health. Individual is assumed to invest in health production until the
marginal cost of health production is equal to the marginal benefits of improved health
status. Health status was assumed to affect utility directly by the value that individual
place on good health per se and indirectly through increasing healthy time and, hence,
labour income. The demand for health increases with the income, because as income
increases the value of the labour time increases. Cropper (1977) assumed that savings are
possible if all income takes the form of earnings and that consumers are risk-neutral in the
sense that their objective is to maximize the expected discounted value of lifetime wealth.
This paper used wages as measured through monthly produce from informal and formal
sector employment.

2.4 Wage Effect

Increasing in education increases the chances of getting a better job hence better wages.
The effect of wages on the demand for health is different in the two variants of the model.
In the pure consumption case, wages reduce the demand for health because the higher the
future wage, the higher the marginal cost of holding health stock as consumption good.
Wages are positively related to demand for health as increases in wages increase
the incentive for individuals to work and incentive to be healthy by increasing the
returns to health capital. Thus, higher wage workers will tend to increase their
optimal health stock.

2.5 Age Effect

The demand for health decreases when people grow old because age reduces payoffs
from investment in health. Older people are less efficient at turning health investment in
health stock. If age increases, the marginal cost of holding an additional unit of health

11
stock also increase. Thus, age reduces demand for health in both the consumption and
investment variants. However, the model predicts that health stock decreases with age and
people are expected to demand more health care as they grow up.

Grossman defines health broadly to include longevity and illness free days in a given year
This is demanded and produced by consumers (Grossman, 1972). Health is a source of
Utility and it enters directly into the utility function. The consumer preferences are
represented by the following utility function:

U = U (Ht , Zt ) , t = 0 ,1,2,…, n , ........................................................ (1)

Where Ht is the stock of health at age t. is the service flow per unit stock, h =Ht is
The total consumption of health services and Zt is the consumption of other commodities.
The stock of health in the initial period (H0) is given, but the stock of health at any other
age is endogenous. The length of life as a planning date (n) is also endogenous. Death
occurs in period t if the level of health in that period equals or is below the lowest level of
health stock that can sustain life.

In the model, an individual inherits an initial health stock which depreciates with age and
increases with investment in health. Net investment into health equals gross investment ( )
t I minus depreciation as given by;

H t+1- H= I H t................................................................................... (2)

Where t is the rate of depreciation during period t

Consumers produce gross investment in health and other commodities in the utility
according to a set of household production functions

It= It(Mt,THt;Et)
Z = Zt(Xt,Tt;E) ………………………………………….. (3)

12
Where It is gross investment which is a function of medical care Mt, the time input in the
investment function THt and the stock of human capital given by E. Zt is the consumption
of other goods and is a function of t X which is a vector of goods input that contribute into
the production of commodity Zt, Tt which is time inputs for Zt and E which is the stock of
human capital (education level).

An individual may defines his optimal stock of health capital by the choice he or
she makes. Grossman’s model of health stock is very important because it gives important
factors which influence demand for health and health care. As demand for health
increases, the demand for health care services should also increase hence demand for
health care is a derived demand. An individual may determines his optimal stock of
health capital by the choice he or she makes. From this model, health care utilization is
seen as choice decision by individuals who demand and produce health. In summary,
demand for health care is hypothesized in the theory to depend on age, sex, education,
time variables, in addition to price and income. In my model I am going to ignore
variables such as sex child healthcare is assumed to be cantered to mothers only in this
paper.

However, the model concentrated on individual decision which is not true for all
individuals some individuals are influenced by Household characteristics or head .In
addition Hjortsberg (2002) discussed that Grossman’s model lacked the fact that
individuals are household members and they take much influence from other
household members. In several studies, household characteristics have been found
to influence health seeking decisions.

2.6. Household’s model of health care demand

The household theories of health care demand appraised household characteristics such as
household size, sex of household head, education of household head and household

13
income affect health care utilization. In the context of rural areas, the mother of the
household makes most of the health utilization decisions. The model included sex of
household head which this paper ignored because mother represent the only head in child
healthcare utilization and are the only child healthcare decision makers. Jacobson (1998)
as cited in Hjortsberg, (2002), revealed that the family is viewed as the producer of
health. Jacobson’s model viewed a household with same preferences and the
household’s combined resources are used in the production of health.

Jacobson B, Flowers J, Johnstone F, etal, (2003) revealed that the higher the income the
increases the opportunity for better health through availability of variety food choices and
health facilities. The call for medicine is based on need; therefore utilization of healthcare
facility is high.

Health care demand studies have since taken a point of departure from Grossman’s model
to analyse health care utilisation. Many studies have used household as a unit of analysis
rather than using an individual. Hjortsberg (2002) argued that Grossman’s model lacked
the fact that individuals are household members and they take much influence from other
household members. In several studies, household characteristics have been found to
influence health seeking decisions. Jacobson (1998) (as cited in Hjortsberg, 2002),
extended the Grossman’s model into a model in which the family is viewed as the
producer of health. Jacobson’s model viewed a household with same preferences and the
main conclusion was that not only the individual’s own income, but the household’s
combined resources are used in the production of health.

The model of demand for health developed by Grossman put the individual as a sole
decision maker. In most rural areas of Zimbabwe, as in most African countries, the
decision of an individual to visit a health care facility is not made by that individual
alone. The head of the household plays an important role in influencing seeking of health
care services in the household. From discussions on household theories of health care
demand, it is likely that household characteristics such as household size, sex of

14
household head, education of household head and household income affect health care
utilisation.

A household production function for health inputs and outcomes

It assumed that the household maximizes a utility function, the arguments of which
consist of health of the infant (H) and consumption of a composite good (G), conditional
on (Z), a set of taste and preference shifters

U= U (H, G; Z). (1)

Health of the index child is produced by combining inputs in the manner implied by the
health production function. This function is modelled as a relation between the health
outcome and a set of health input choices; its shape will depend on the underlying health
technology. The production function is written

H= H(C, F; S, M, E, õ), (2)

Kelly Hallman, (1999) revealed that such a unitary model of decision making, in which
households are assumed to make decisions that maximize household utility, does not
allow one to explore the processes of intra - household decision making. The unitary
approach is used because information is not available in these data on individual incomes
or ownership of assets within the household. Where the first two arguments are
endogenous inputs into health: C is the quantity and quality of health care chosen and F
consists of other health inputs, such as food. The model managed to validate other
determinants like education which this study used. Some of the variables like quality of
the health are being left since there are not being determined inside the model. The model
also included the essence of the infant which is of great use whilst analysing child health
demand and mother level of education.

15
2.7 Andersen’s behavioural model of health care demand (1968)

Andersen’s behavioural model was first developed in 1968 to help in explaining the
differences in access to health services in the United States of America (Satayavongthip,
2001). The model is the most widely used analytical model to explain health care
utilisation behaviour. Andersen and Newman, (1973) and Becker et al., (1993) (as cited
in Sunil et al.2000), argued that the use of health care services in any given society is a
complex behavioural phenomenon. The model gives an overview of relevant social
determinants for seeking health care services. The theoretical framework describes the
process of health care utilisation as a causal interaction of three different levels which are
societal, health care system (programme factors) and individual determinants.

The societal and system determinants are postulated to influence individual determinants
which in turn directly influence the use of health care services. The societal determinants
include the current state of knowledge as well as people’s attitude and beliefs about health
and illness. The health care system in turn allocates available resources to health care
institutions and forms the organisational framework to provide health care services
(Stefan and Markus, (2010). The system factors include structures and activities through
which health care and health education are provided. For example, system factors would
consider the availability of information, education and communication (IEC) activities in
a village to educate people on health care services (Sunil et al., 2000). The organisation
component of the system factors addresses how services are delivered to people who are
in need. These factors include distance to the nearest health facility, access to village
health workers and user fees.

The theoretical framework hypothesizes that the individual’s decision to seek health care
services is a function of three sets of variables namely the predisposing, enabling and
need factors. These factors are explained below:

Predisposing factors

16
The model postulates that there are certain factors that predispose people towards health
care service utilisation. These factors influence an individual to seek health care services.
For example, the basic demographic characteristics such as age, sex and past illness may
have an influence on the demand for health care services. The social structure factors
such as education, household size, occupation and race are also important predisposing
factors. More so, beliefs, values and knowledge about health and medical care services
can affect a decision to seek health care services.

Enabling factors

Enabling conditions make health service resources available to an individual. Even if an

individual may be predisposed to the use of health care services, some means must be
available for him or her to do so (Andersen and Newman., 1974). These factors include
both individual and household resources (income and health insurance). The availability
of the health care services is also an enabling factor. Attributes of the community or
region in which people live such as place of residence are important since they indicate
geographic proximity to the source of care as well as local attitudes about health care
services.

Need factors

The need for a service (illness) is perhaps the most important factor which influences
healthcare service utilisation. Even with the existence of predisposing and enabling
factor, the individual seeking health care services must still perceive the need for health
care before seeking it. A perception of illness is necessary for the use of health care
services. The need for care may be perceived by the individual and reflected in reported
symptoms or disability days.

17
2.7.1. Education Effect

Andersen model of healthcare demand insisted that Education in the model is positively
related to demand for health stock in both variants. Increasing in education captures
increasing knowledge in the production of health and it enables individuals to choose
more health consumption decisions. The implication of education in the model is that
those individuals with higher degrees of schooling are likely to effectively utilize
resources to produce health and demand less health care services.

Education increases both the demand for health and the productivity of health inputs; the
effect of rising education on the demand for health care is directly proportional. Those
who choose to obtain higher levels of education possess other characteristics that increase
their likelihood of being healthier. Thus, this outside factor confounds the attainment of
the direct causation of education on health. Education is believed to increase knowledge
and attitude to use health services. Education also increase the opportunity of getting
better jobs hence better income to buy healthcare service.

2.7.2Income

According to Andersen and Newman 1973 revealed that the model insisted that there are
enabling factors which are financing and organizational factors. Individuals obtain
income or wealth in order to pay for healthcare services. It looks at the resources
available within the community as per capita community income, insurance coverage, and
relative price of goods. One of the weaknesses of Andersen’s model is that it does not
directly consider distance as a factor which affects health care service use (Rajaramet al.,
1999). The model nonetheless provides a good theoretical framework in analysing health
care seeking behaviour especially in rural areas. The model will be used to establish
explanatory variables to be included in the study

18
2.7.3. Child healthcare

Child health and survival, including characteristics of the neonate (birth weight and size),
the vaccination status of young children, and treatment practices—particularly contact
with health services—among children suffering from three childhood illnesses: acute
respiratory infection (ARI), fever, and diarrhoea. MIMS, (2009) revealed that in
Zimbabwe 49% of children up to 23 months old were entirely immunized - BCG,
Pentavalent and Measles. There has been a decline in EPI coverage since 2000.

2.7.4 Infant Mortality Rate

DeVanzo (1998) in ―Infant mortality and economic development the case of Malaysia‖
postulates that there is a trade-off between infant mortality and economic development. In
other words, this implies that higher levels of infant mortality are associated with lower
levels of development of a nation as a whole. An increase in mothers’ education (low
literacy rate), improved maternal services and good sanitation are the most key
fundamentals that accompany improvements in economic development

2.7.5 Empirical literature review

Grossman (1991) have come out with the empirical evidence that mother level of
education increases child healthcare services utilisation.

2.7.6 Health outcomes associated with education

A strong positive relationship exists between education and health outcomes whether
measured by death rates (mortality), illness (morbidity), and health behaviours or health
knowledge. Kraal, (2004) investigated the effects of the educational attainment of
mothers and other women in the community on child mortality in India using the National
Family Health Survey of 1998-1999. Child mortality was specified as a discrete–time

19
hazard model and some of the explanatory variables considered were education of the
mother, average education of women) and women’s autonomy variables (economic,
physical, decision-making and emotional) which were incorporated as potentially
mediating or confounding factors.

Berger and Leigh, (1989),suggested that controlling observable factors as race, ethnic
background, family characteristics, and finding means of measuring some unobservable
factors such as personal time preference, both sample results showed that there was a
direct positive correlation between schooling and health

The signalling approach by Roland Akbazee, (2009) revealed that the level of education
attained serves as a signal of individual productivity. Only those with high productivity
potential can achieve higher levels of education.

Lieras-Muney, (2002) conducted a study in which she used compulsory education laws as
an instrument for education. The study analysed the health and survival patterns of people
who were both affected and unaffected by the laws. The study determined that the
correlation between education and health was pure, causal and positive. Bicego and
Boerma (1993), hypothesizes that particular education-induced behaviours are the
underlying factors of the maternal education-child survival link. Bicego and Boerma’s
(1993) hypothesis that mothers with an education seek out health services more than
uneducated mothers. An ILO study carried out in Bolivia, Egypt and Kenya showed that
access of women to maternal and child health services were strongly influenced by their
level of education and rural-urban status (ILO, 2000).

In a study by Rajaramet al., (1999) education was found to be significant in explaining

maternal care utilisation in rural India. Their objective was to examine the factors
influencing utilisation of maternal care services among women in rural India using
Andersen’s behavioural model. Data collected through the National Family Health

20
Survey was used in the study, education of woman, household standard of living index,
distance to the nearest government health facility and women’s exposure to education,
information and communication (IEC) from village health workers during pregnancy had
strong and statistically significant influence on the utilisation of maternal care services in
rural India.

2.7.8Income effect

Sarma,(2000) investigated the determinants of demand for outpatient health care and the
choice of health care provider. The objective of the study was to examine the impact
income and a variety of individual and household specific characteristics on the demand
for health care in rural India. Sarma, (2000) used variable choice set based on
geographical location, price, income and severity of the illness to reflect true health care
behaviour in rural India. The empirical evidence is vital to the study since it verifies
income as an independent variable to healthcare services.

Lawson, (2004) used a discrete choice model to examine the determinants of health care
seeking behaviour in rural Uganda. The study was motivated by the introduction of user
fees for most health care services in Uganda. Hence, the study sought to justify
econometrically the impact of user fees and income on health seeking behaviour in
Uganda. The study used secondary data from two Demographic Health Surveys and a
1999-2000 Ugandan National Household survey (UNHS). Demand for health care was
dichotomous, measured by the probability of a person seeking health care for any illness
over the past thirty days 7 and the choice of health care provider. Variables such as sex,
age, household size, personal education, religion and income were used in the study as
explanatory variables. These factors were analysed across gender and for all ranges of age
including adults, school aged children and preschool children.

21
2.7.9Religion

Those social determinants are usually conceptualized as socioeconomic, political, or

racial/ethnic inequalities. Despite a substantial research literature showing religious
involvement to be a significant predictor of health outcomes, especially as a protective
factor against all-cause mortality, religion has rarely been considered among the set of
social determinants. An interdisciplinary group of scholars at Emory University has
conducted a three-year seminar on religion as a social determinant of health. In their
edited volume (Oxford, 2014) scholars from diverse religious traditions describe daily,
weekly, and annual religious practices that involve the body and may have implications
for health. Historians detail the origins of public health institutions in religious concern
for the community. Public health researchers review evidence for the lifelong impact of
religion on physical and mental health. Global health experts identify the role of religious
institutions in providing health care and influencing health policy. Clinical researchers
describe the role of religion in providing social support, social control, and social capital
in the context of three epidemics of our time: HIV/AIDS, pandemic influenza, and
Alzheimer's disease.

Clinical trial of intercessory prayer by Randolph Byrd (1988) is often cited as proof of
―the power of prayer.‖ In this study, patients in a coronary intensive care unit were
randomly assigned to be prayed for by teams of ―participating Christians‖ who did not
know the patients. The researcher analysed data with 63 outcomes; there were statistically
significant effects in six cases, with the prayed-for group having the better outcomes, and
for 57 outcomes there were no differences

In another study by Rajaramet al, (1999) individual and programme factors were found to
be significant in explaining maternal care utilisation in rural India. Their objective was to
examine the factors influencing utilisation of maternal care services among women in
rural India using Andersen’s behavioural model. Data collected through the National
Family Health Survey was used in the study. A set of variables such as religion,
education, household income and other individual variables were used as explanatory

22
variables. Using an ordered logit model to estimate maternal care demand, the study
found out that there was a substantial variation in the likelihood of utilisation of maternal
care services by religion. Only 18% of the birth to Muslim women received excellent
maternal care services compared to 38% of the births to women belonging to religions
other than Hindu and Islam.

2.8 Distance to the healthcare facility

Using a logistic regression, Hutchinson (1999) in Uganda found out distance is the major
factor that contributes to low utilisation of health care services in rural areas. The study
used data from Ugandan Household Survey to produce micro econometric work on
determinants of health care demand in the context of rural settings. In his study, he found
out that the majority of the population in rural areas had to walk to the health facility
since transport was not readily available. As a result poor families relied more on self-
treatment and use of traditional healers. More specifically, the study found out that for
each extra one kilometre travelled to the health unit, the use of health care fall
approximately by 1%. In the same study, other factors which were found to affect health
care utilisation were household size, age, income and an individual’s level of education.

2.8.0Conclusion on empirical literature review

From empirical literature discussed, there is evidence to suggest that demand for health
care services especially in rural contexts is influenced by socio demographic, economic
and institutional factors. These factors include age, education, household size, household
income, religion, distance among others. The findings from Sub Saharan Africa reveal a
positive relationship between mother’s level of education and demand for child health
services. Despite evidence from other countries on the factors that influence health care
demand in, there is little knowledge in the context of Zimbabwe.

23
Studies on health care utilization in Zimbabwe concentrated in other factors not much on
mother education. A healthcare system that meets the needs of a developing country such
as Zimbabwe requires a thorough understanding of why the present healthcare system
fails to deliver healthcare to the child and also the dynamics of healthcare decision
processes of parents which is supposed to be determined by the level of parent’s
education. Such a good understanding will help in redefining the objectives of national
healthcare policy and those of the active players in the healthcare delivery system;
therefore, guided by empirical literature, this study seeks to investigate the significance of
mother’s level of education on influencing the demand of child health care services using
data from Shamva District Mashonaland Central in Zimbabwe.

24
 CHAPTER 111

METHODOLOGY

3.0INTRODUCTION

The main motive of the chapter is to highlight the methodology used by the researcher in
gathering the relevant data for the study .It emphasizes on the research site and design,
sampling procedure, Research instrument data collection and data presentation
procedures. It seeks to justify methodology used in the study. This chapter revealed the
discussion on the methodology and model specification. The researcher will bring to light
the research design. The study makes use of primary data collected from Shamva District.
Data for child healthcare was collected from ZIMSTAT health statistics section. Stata is
going be used in estimate the model specified

3.1 THEORETICAL MODEL

Several macro models have normally concentrated on estimating the impact of income on
child healthcare utilization services in developing countries (Sarma, 2000). In general, the
studies have shown that macro variables such as mother’s level of income, age and
education have a significant impact on child healthcare services utilization. It is through
this transmission mechanism; according to Grossman, (1972) that investment in
Education has direct effect on child health through increased chances of getting a better
paying job.

25
3.3.0 RESEARCH DESIGN

3.3.1 Model specification

The primary focus of this paper is to bring to light the impact of mother level of education
as the determinant of the demand for child health care service utilisation ,the econometric
model is specified to facilitate the test of the hypothesis that whether explanatory
variables manipulate the demand for healthcare .The demand for healthcare is the derived
demand from the demand for health .Healthcare is demanded as a means to achieve a
large stock of health capita .Grossman, (1972) revealed that healthcare is a consumer
good as well as an investment good .The paper borrowed the econometric model from
Grossman , (1972) and Jacobson B , (2003) A household production model making an
extension of distance and religion as other variables that affect child healthcare demand
in Zimbabwe. The basic methodological approach and model follows the shoes of
Grossman (1972) and Jacobson B, (2003) a household production model. According to
Fredrickx (1998), the demand for health care can be measured by the probability of
visiting a health a centre in case of illness or by the type of the treatment (choice of health
care provider). In this study, the demand for health care services is defined as the
probability of seeking health care services from a clinic or hospital conditional on illness
in the household.

The study has the choice of using the linear Probability Model (LPM), the Logit or Probit
models because the dependent variable is binary. However, the study avoids the use of
the LPM which is estimated using Ordinary Least Square (OLS) since it suffers from a
number of econometrics problems. The problems associated with using LPM are as
below:

A Logit model has a relatively simple form for the first order conditions and asymptotic
distribution (Cameron and Trevedi, 2005). Hence, the study will employ a Logit model.

26
Therefore the model becomes mathematically

CHCs = f (education, age, income, religion, distance)…………………. (1)

Child healthcare demand is a function of education, age, income, distance and religion.
The explanatory variables used in this model may be quantitative or qualitative in nature.
The quantitative variables are price of health care service, cost of drug, income of the
consumer, mother’s age, mother’s education level, distance to the provider and mother’s
religion.

The logit form of the demand for health care function is written as:

Logit Ch. = In π

π–1
Thus
lnπ= βo+β1 EDUC + β2A + β3Y + β4DIST + D0REL +U

π-1

Ch.– Child healthcare services , Ed – Education, Ag – Age, Y – income, DST – Distance,

R – Religion and U – error term

3.4 DESCRIPTION AND JUSTIFICATION OF VARIBLES

From both the theoretical and empirical literature that has been reviewed in the previous
Chapter, demand for health care services is likely to be influenced by the following
variables namely age of household head (mother), religion, and education, income of
household mother, household income, distance, and access to village health workers and
availability of drugs. The study will use these Variables to test whether they are
significant in influencing health care services.

27
3.4.1 Child healthcare utilization (ch)

In this study the demand for health care (ch) is dichotomous variable. It takes the value
one (ch = 1) with probability if the mother has demand for health care and zero (ch = 0)
with probability 1- if the mother has no demand for health care services for her child.

3.4.2 Mother’s education(Ed)

In this study, education is a categorical variable which is measured by the level of formal
education attained by the mother of the household. The variable is coded as follows: No
education = 0, Primary education = 0, Secondary education = 0 and Tertiary education
=1.One of the justifications of using education as a categorical variable is that the
variables will take into custody the incremental effects of the levels of education ( Kaija
and okwi, 2004). Grossman’s theoretical model on health revealed that education is an
important factor which affects health care seeking behaviour. Education increases
efficiency in health production and thus reduces the price of health investment .The
returns on health are higher for the more educated (Grossman, 1972). Education of a
household head may affect the recognition of symptoms and link them with presence of a
disease. Grossman, (1972); Hjortsberg, (2003) revealed that this will affect the perception
of sickness, its level of sternness and therefore the probability of visiting a health care
provider. Therefore, anticipation is that the more educated the household mother is, the
higher the probability of looking for health care services utilisation by a household child.
Therefore, the study anticipates a positive relationship between demand for child health
care services utilisation and the education of the household mother.

3.4.3Mother’s or household Income (Y)

The level of household income reflects the economic status of the mother of the
household. Both the theoretical and empirical evidence have confirmed a positive
relationship between income and the demand for health care service utilisation. In the
investment variant of Grossman’s model, wage has positive impact on the demand for

28
medical care. Lawson,(2004), Fredrick, (1998) among several studies have found that
household income increases the possibility of searching for treatment from health care
facilities , therefore the study expects the probability of seeking health care services by a
household child to increase with higher household’s mother income. The monthly income
earned will be used as a proxy for household mother income. The income of the mother is
treated as a continuous variable.

3.4.4 Mother’s Age

This is a continuous variable which captures the age of the mother in years. The study
anticipants a higher utilisation of child health care services in households whose mothers
are still young. Since at young age most household heads have low understanding in
managing illness, the prospect of seeking treatment from health care facilities is likely to
be high. Lawson,(2004) revealed that when the mother is very old, looking for child
health care services is expected to decrease as chances of self-treatment will be very high
.Hence, the study uses age squared to capture the effects of old age of a mother on
demand for child health care services.

3.4.5 Distance to the health services\0

Distance to the nearest health facility will be measured in kilometres and the variable will
be continuous, however for the case of this study distance will be categorised as far (6km
– 9km) and near (1km – 5km). Distance to the nearest facility measures physical
accessibility of health services in the community. Muhofa, (2010)reviewed that the
demand for child health care services utilisation tends to decline with distance. The study
expects a negative relationship between demand for child health care services and
distance to the nearest health care facility. There is a higher probability that a household
child will seek health care services during illness if the household is close to a health care
facility.

29
3.4.6 Religion

MoHCW, (2010) revealed that Religion and socio-cultural beliefs play a role in
influencing health care seeking behaviour in rural areas of Zimbabwe for this case
Shamva District. Mothers who believe in traditional healers or in apostolic sect religion
tend to discourage the use conventional medical care, they are liable to prefer self-
treatment as substitute to treatment from hospitals when a household child has fallen ill.
The study expects lower probabilities of seeking health care services in mothers who
believe in traditional healers or in apostolic sect religion. The variable will be categorised
as follows: apostolic sect = 1 or otherwise = 0.

3.5 RESEARCH INSTRUMENTS

This study used primary data which was collected from twenty nine Wards in Shamva
District. The survey was carried out in January and February 2015. All the respondents
that do not belong to any of the main categories will be recorded as other religions. The
questionnaires were given to those Household mothers who were present within three
months prior the survey. The course of action was adopted from the Zimbabwe
Demographic Health Survey (ZDHS, 2010).Data was collected from households using a
pre-coded structured questionnaire. Closed ended questions were used with all possible
answers pre- specified and the respondents made their selections from the answers
provided. The questionnaire was self-administered.

Data collection was done using a face to face individual interview method. This was
important because it helped the respondent to recognize the questions by interpreting
them to fit his or her understanding. Personal interviews have an advantage of making it
possible for the interviewer to set up understanding with the respondents and motivating
them. On data quality control, the questionnaire was pre-tested to determine the reliability
and validity of the instrument. Pretesting was done before the actual survey to test for the
clarity of questions, instructions and to identify any other anomalies in the questionnaire.
The essential adjustments were done after pretesting.

30
3.5 Sample size and sampling procedure

The study area is divided into 29 wards and these wards were taken as clusters. Due to
resources and time constraints, the study used one stage cluster sampling to only select 15
wards for the survey. Out of these 15 wards, 290 households were randomly selected and
interviewed. Although the study is based on illness, in which data for only households
that had experienced illness will be used for regression analysis, this is done following
Mwabuet al., (1993), at least 10 households were interviewed at ward level. Sekeran and
Bougie, 2004) revealed that one of the advantages of using cluster sampling as a
probability sampling design is that the cost of data collection is very low and it offers a
generalisability of the population.

3.5.1 Mathematical tools

The demand for health care model cannot be estimated by using standard regression
techniques due to the dichotomous nature of the dependent variable. Several studies like Kazi
J.A and A.N.K Norman, (2013) following logistic regression technique it is then employed to
estimate the determinants of the demand for child health care service utilisation. Logistic
regression explains the relation between dependent qualitative variable and one or more
qualitative and quantitative explanatory variables. Sarkar, (2004) revealed that Logistic
regression model cannot be estimated by standard ordinary least squares (OLS) method
because the logistic regression model suffers from heteroscedasticity and non-normality
problem in the error term. Therefore maximum likelihood method is used to estimate the
parameters of logistic regression model.

Econometric package
This study uses the Logit model as a technique of estimation. The assumption of the Logit
Model is that the error term follows a logistic distribution. Wooldridge, (2004) revealed
regressions of categorical dependent variables are nonlinear. Even when linearity is
created, for example in the Logit model, marginal effect is not stable, but depends on the
values of the complete vector of explanatory variables. In the nonlinear models, the only

31
standard approach to estimate either the Logit model is to use the Maximum Likelihood
Estimation (MLE) based on the idea of sample information (Wooldridge, 2004). The
Maximum Likelihood Estimation is concerned about selecting parameter estimates that
entail the maximum probability of having it obtained from the observed sample. The
purpose of the MLE is to decide the estimates of the coefficients that formulate the
likelihood of observing an exact outcome of the dependent variable as big as possible.
Studies by Fredrick, (1998) successfully used logit model to find the determinants of
health care demand in Tanzania.

3.5.2 Summary statistics

Summary statistics is going to be used to summaries data in order to communicate the

largest percentage change as simple as possible. The information is given in quick and
simple descriptions such as mean, median, mode, range, and the standard deviation

3.5.3 Diagnostic checking

Before the regression is run the researcher is going to use diagnostic test below. The
adequacy of the model estimated had to be evaluated before the model can be used for
forecasting purposes. The diagnostic tests that to be done include.
 Specification Test
 Unit Root Test (Dickey Fuller)
 Multicollinearity

3.5.4 Unit root tests

With the development of modern time series modelling, unit root test on all variables are
done to determine their time series properties. This is done to avoid the problem of

32
spurious regression when non-stationary series are estimated in their levels in stochastic
models (Badawi, 2003). Determining the order of integration of the macroeconomic
variables thus help in the identification of the model. The Augmented Dickey Fuller
(ADF) tests that take into account the possibility of structural breaks in the time series are
used to analyse the time series properties of these series. To test the order of integration
of variables, standard test for unit root such as the ADF test proposed by Dickey and
Fuller (1979) is to be used.

3.5.5Multicollinearity

This is the presence of linear relationship among the explanatory variables as a result of
the stochastic nature of most regressor’s correlation and interrelationships are bound to
exist among them making multicollinearity inherent in most explanatory variables. It has
the effect of making the normal equation X1X  = X1Y indeterminate, that is it, becomes
impractical to obtain numerical values for parameter  and the Maximum Likelihood
Estimation (MLE) method breaks down since the moment matrix X1X is then singular or
non-invertible. When any two explanatory variables are changing nearly the same way it
becomes extremely difficult to establish the influence of each one regressor, Xi on the
dependent variable Y separately. Hence R-squared is used to test for multicollinearity. If
R2 is in excess of 0.8 the F test the researcher will reject the hypothesis that the partial
slope coefficients are simultaneously equal to zero, but individual t-tests will show that
non-or very few of the partial slope coefficient are statistically different from zero.

3.5.6 F Value and Prob (F)

The researcher is also going to use the "F value'' and "Prob (F)'' which is the statistics test
the overall significance of the regression model. Specifically, they test the null
hypothesis that all of the regression coefficients are equal to zero. This tests the full
model against a model with no variables and with the estimate of the dependent variable

33
being the mean of the values of the dependent variable. The F value is the ratio of the
mean regression sum of squares divided by the mean error sum of squares. Its value will
range from zero to an arbitrarily large number.

The value of Prob (F) is the probability that the null hypothesis for the full model is true
(i.e., that all of the regression coefficients are zero). For example, if Prob (F) has a value
of 0.01000 then there is 1 chance in 100 that all of the regression parameters are zero.
This low a value would imply that at least some of the regression parameters are nonzero
and that the regression equation does have some validity in fitting the data (i.e., the
independent variables are not purely random with respect to the dependent variable).

3.57 Specification Test

The Ramsey Regression Equation Specification Error test (RESET) for misspecification
will be carried out. The test detects omitted variables and incorrect functional form of the
model(Ramsey, 1969). The mechanics of the test is that, if non-linear combinations of the
explanatory variables have any power in explaining the dependent variable, then the
problem of misspecification exists. If the model can be significantly improved by
artificially including powers of the predictions of the model, then the original model must
have been inadequate.

3.6 Data sources and problems

The study uses primary data which was obtained from Shamva District Questionnaire.
The empirical evidence period chosen was long enough to analyse the impact of mother’s
level of education on child healthcare services utilisation taking into account other
variables of health care demand. For further evidence to support analysis data was
obtained from ZIMSTAT. Although the secondary data has got problems in terms of
quality, consistency; accuracy and reliability are very acute in less developed countries

34
like Zimbabwe included. This is due to inadequate monitoring of the Health section
reports and inaccurate reporting and recording of data.

3.7 SUMMARY

The approach adopted from Grossman (1972) and Jacobson (2003) will be used to assess
the impact of mother’s level of education on child healthcare services utilization.
Distance to the healthcare facility and religion are two variables added to the Grossman
Health function.

35
 CHAPTER 1V

DATA ANALYSIS AND RESULTS PRESENTATION

4.0 INTRODUCTION

The purpose of this study is to examine mother level of education influence to child
healthcare services in rural areas of Zimbabwe using household data from Shamva
District. This chapter focuses on the estimation, presentation and interpretation of
empirical findings. The selected data is going to be processed using Stata 12. For
regression analysis, only data for 290 households was used to find the determinants of
demand for Child health care services utilisation.

4.1 SUMMARY STATISTICS

Descriptive statistics

From a sample of 290 households, 290 households had reported having at least one sick
child within the last 3 months before the surveys.

Table 4.1
Ch. Ag Dst Ed R Y

Mean 0.375862 35.55517 0.827586 0.827586 0.620689 298.2759

1 2 2 7

Maximum 1 57 9 1 1 760

36
Minimum 0 20 1 0 0 100

Std. Dev. 0.485181 8.044775 0.378392 0.485181 0.486054 156.2597

9 7 9 2

Observations 290 290 290 290 290 290

The table above shows the summary statistics that include measures of central tendency.
Minimum and maximum are least and greatest elements of a set respectively and there
used to check for outliers. For all variables there are no outliers as shown by the smaller
gaps between the minimum and the maximum values except on income. Income has the
highest standard deviation of 156.2597 which indicates that data spread out over a large
range of values. Child health care services utilisation (Ch.) has the lowest standard
deviation of 0.3758621 wwhich indicates the high degree of its reliability on its
contribution towards being explained in independent variables. The lower the standard
deviation indicates that the data tend to be very close to mean high standard deviation
indicates that the data is spread out over large range of values. Probability ranges from 0
to 1, it deals with the likelihood of occurrence. The number of observation is 290 for each
and every variable.The table shows that on average, the age of the household head was
35.6 years. The average household income was approximately US$298.2759. Average
distance to the nearest health centre was approximately 5.9 kilometres while the average
consultation fee paid at the nearest health facility was US$1.36.

4.2 Diagnostic test

Table 4.2 Correlation matrix

ch ag ed dst y r
ch 1.0000
ag -0.0492 1.0000
ed 0.9853 -0.0315 1.0000
dst 0.0715 -0.0366 0.0715 1.0000

37
 y 0.8468 -0.0492 0.8340 0.1401 1.0000

Multicollinearity exists when explanatory variables are highly correlated. If variables

have multicollinearity below 0.8, it means that there is no strong relationship existing
between variables and deserves no treatment. Multicollinearity from the table above is
approximately within the permissible range. There is therefore no need to correct for
multicollinearity since absolute values are within the range of 0.8.

4.2.1 Heteroskedasticity test

A heteroskedasticity test was performed to check if variances of error terms were

homoskedastic. A chi –square test statistic of 5.94 was obtained with p- value of 0.0148.
At 5% level of significance, we failed to reject the null hypothesis and hence there was no
evidence of heteroskedasticity.

Table 4.2.3Model estimation

The table below summarises the logit results after dropping insignificant variables in the
first estimation. All the variables in the restricted model are significant either at 1%, 5%
or 10% level of significance.

Table 4.3 Model estimation

Ch Coefficient Std. Error Probability

Ag -0.0008809 0.0005473 0.109

Y 0.0002097 0.000052 0.000
Ed 0.8013499 0.0240339 0.000
R -0.141036 0.0214756 0.000
Dst -0.0115355 0.0117837 0.328

38
Cons 0.1405347 0.0328981 0.000

R-squared = 0.9767 F (5, 284) = 2382.72

Adjusted R2 =0.9763 Prob> F = 0.0000

**Significant at 1% level of significance. *Significant at 5% level of significance.

Logit Ch= βo+β1 Ed+ β2Ag + β3Y + β4Dst+ D0R +U

Substituted coefficients

Logit Ch = 0.1405347+0.8013499Ed-0.0008809Ag + 0.0002097Y- 0.0115355Dst+-

0.141036 R

The results presented in the table above are results for the model. As shown, the R-

squared is 0.9767which shows that the model is significant gives information about the

goodness of fit of a model, and a model is not useful unless its R2 is at least 0.5 (Gujarati

2005). Percentage changes in education, age distance,income and religion are significant.

Using the model above Adjusted R2 value is 0.9763 this shows that the variation in the

explanatory variable accounted for over 97.6% in the variation in the dependent variable

Child healthcare services utilisation. This shows that the model has a very high

explanatory power. The model is reliable in estimating Child healthcare services

39
utilisation. The Prob F-statistic is 0.0000which means there is no chance that all of the

regression parameters are zero. This implies that no regression parameters are zeros and

that the regression equation has some validity in fitting the data. The independent

variables are purely random with respect the dependent variable. If there is no variability

in age, education, income, religion distance there is constant C with a coefficient of

0.1405347

4.2.4 Econometrical interpretation

Education of mother

Education has a coefficient of 0.8013499 (0.00) showing a positive relationship with

child healthcare services utilisation. This variable is found to be significant at all levels.
This means that a percentage change in education will result in 80% increase in child
healthcare services utilisation. The sign of the coefficient shows that there is a positive
relationship between the demand for health care services and education of the mother.
The results confirm the predictions of Grossman’s model (1972) in which education is
found to positively influence the demand for health and health care. The results could be
also explained by the evidence from empirical literature that better schooling or education
of the household head has a potential to improve understanding and appreciation of
benefits of seeking health care services. The results of the predicted probabilities show
that highly educated mother, highly seek child health care services utilisation.

Distance to the health facility

Distance has a coefficient of -0.0115355 (0.328) showing a negative relationship with

child healthcare services utilisation. This variable is found to be not significant at all

40
level. This means that a percentage increase in distance will result in a 1.2% decrease in
child healthcare services utilisation.This is supported by theory as well as empirical
evidence which showed a weak positive relationship. The results show a predictable
negative relationship between distance to the nearest health care facility and the demand
for health care services. The relationship illustrates that households who live far away
from the health care facilities have a higher probability of not seeking health care services
than those who stay close to health facilities.

The marginal effects results show that an extra kilometre to the nearest health care facility
reduces the probability of seeking health care services by approximately 0.328. There are
a number of previous studies that have found similar results. For example, a study by
Hutchinson (1999) in Uganda found out that distance negatively reduces the physical
accessibility to health care services. A further distance away from the nearest health care
facility suggests that people have to incur costs to access health care services in the
community. These costs include transportation costs. As a result, poor families may rely
on self-treatment. Distance is viewed as one of major barriers to health care services
utilisation in rural areas like Shamva District.

Household income

Household income has a coefficient of 0.0002097 (0.00) showing a positive relationship

with child healthcare services utilisation. This variable is found to be significant at all
levels. This means that a percentage change in Household income education will result in
0.021%increase in child healthcare services utilisation. The sign of the coefficient shows
that there is a positive relationship between the child healthcare services and education of
the Household income

The results show that monthly household income has a positive relationship with demand
for child health care services utilisation. The study confirmed the positive influence of
income on child health care services consumption as predicted by Grossman (1972).
Grossman (1972) constructed a model where individuals use medical care and their own

41
time to produce health. Individual is assumed to invest in health production until the
marginal cost of health production is equal to the marginal benefits of improved health
status. Health status was assumed to affect utility directly by the value that individual
place on good health per se and indirectly through increasing healthy time and, hence,
labour income. The demand for health increases with the income, because as income
increases the value of the labour time increases. Cropper (1977) assumed that savings are
possible if all income takes the form of earnings and that consumers are risk-neutral in the
sense that their objective is to maximize the expected discounted value of lifetime wealth.
According to Fredrikx (1998), rural households are constrained by low levels of income
to seek treatment from health facilities as access to health care services includes both
direct and indirect costs. Therefore income has an effect on child healthcare services in
Shamva District.

Age of the mother

Age of the mother has a coefficient of -0.0008809 (0.109) showing a negative

relationship with child healthcare services utilisation. This variable is found to be not
significant at all levels. This means that a percentage change in age will result in
-0.0980% change in child healthcare services utilisation. The sign of the coefficient
shows that there is a negative relationship between the demand for health care services
and age of the mother. This means old aged mother does not demand much of child
healthcare services. Several theories like Grossman, (1972) revealed that the demand for
health decreases when people grow old because age reduces payoffs from investment in
health. Older people are less efficient at turning health investment in health stock. If age
increases, the marginal cost of holding an additional unit of health stock also increase.
Thus, age reduces demand for health in both the consumption and investment variants.
However, the model predicts that health stock decreases with age and people are expected
to demand more health care as they grow up. This study showed that as the mother’s age
increases less of healthcare services are demanded hence less of child healthcare services
are demanded by the mother.

42
Religion

Age of the mother has a coefficient of --0.141036 (0.00) showing a negative relationship
with child healthcare services utilisation. This variable is found to be not significant at all
levels. This means that a percentage change in age will result in -14.1%changein child
healthcare services utilisation. The sign of the coefficient shows that there is a negative
relationship between the demand for health care services and age of the mother.
Rajaramet al, (1999) showed that using an ordered logit model to estimate maternal care
demand, the study found out that there as a substantial variation in the likelihood of
utilisation of maternal care services by religion. This reveals that an apostolic mother
does not demand child healthcare services utilisation instead has less utility to derive
from healthcare services utilisation. The apostolic believe in the power of prayer and
believe that illness can be treated by prayer. This was supported by Randolph Byrd
(1988) is often cited as proof of ―the power of prayer.‖

4.6 Conclusion

The econometric results showed that education of mother, household income and religion
of the mother are significant factors that determine the demand for health care services in
rural areas. However, the study has found out that age of mother and distance to the
health facility are insignificant variables. The positive effect and statistical significance of
the household income on the probability of seeking health care services suggests that
people face direct and indirect costs of accessing health care services.

Summary
This chapter looked at results presentation and discussion. The next chapter will conclude
this study by looking at the summary, suggested recommendations and conclusions.

43
 CHAPTER FIVE

CONCLUSIONS AND POLICY RECOMMENDATIONS

5.0 Introduction

This chapter presents a summary and conclusion of the study’s findings. The chapter also
provides policy implications and recommendations derived from the findings.
Limitations of the study and areas for further research will be discussed in the last section
of the chapter.

5.1 Summary of main findings

The study investigated the relationship of mother’s level of education to child healthcare
services utilization the case for Shamva District from 2000 to 2014.The study was
stimulated by the desire to establish factors which are influencing seeking of child health
care services utilisation especially the influence of socio economic and institutional
factors. Utilisation of child health care services in rural areas has remained low despite
efforts by the government to improve utilisation of health care services. The study used
cross sectional household data that was collected in January and February, 2014. Data
was collected from households based on illness. A self-administered questionnaire was
used in the collection of data.

The study used a binary dependent variable thus Child healthcare services utilisation and
a value of one was assigned if a mother sought her child treatment from a clinic or
hospital and zero in the case when mother sought self-treatment in managing child illness.
A set of socio demographic, economic and institutional variables were used as
independent variables. These variables were: age of the mother, education of household
head, religion of the mother, household income and distance to the health care facility. A
logit model was then used to find the determinants of demand for health care services.
The regression results revealed the importance education of mother, household income

44
and religion of the mother to have a significant statistical relationship with the demand for
child health care services utilisation .In the study, household income and education of
mother had a positive influence on the demand for child health care services utilisation.
Religion had an inverse relationship with the demand for child healthcare services
utilisation.

However, the results of the study reveal no significant relationship between age of mother
and demand for child healthcare services utilisation. There was also no significant
relationship between distance to the health care facility and demand for child healthcare
services utilisation .All the hypotheses for insignificant variables were rejected while the
hypotheses for significant variables were accepted.

5.2 Policy Implications and Recommendations

The study has found positive significant relationship between education of household
head and demand for child health care services. In this regard, policies that promote
universal education should be maintained by policy makers in order to improve seeking
of health care services in rural areas. If a household mother is more educated, he or she
understand the benefits and importance of seeking formal child health care services in the
case of illness of household child.

There are several policy insights that can be derived from the empirical findings of this
study. From the results of the study, distance is one of the factors that decrease the
probability of seeking health care services although it was statistically insignificant.
Therefore, the policy implication of this finding is that policies that aim to reduce
distance to the nearest health care facility are likely to increase the probability of seeking
health care services. In light of the above, policy makers should implement policy
interventions that aim to shorten the distance which rural people travel to access health
care services. Such interventions include increasing the number of health care facilities in
the rural areas. This can be done by introducing community based mobile clinics.

45
The positive significant effect of household income on demand for child health care
services utilisation suggests that policies that aim to increase household income are likely
to increase child health care services utilisation and reduce chances of self-treatment and
mortality rate. Introduction of income generating projects should be effected to improve
rural household incomes. Higher household income can improve the chances of seeking
child health care services in rural areas because there are direct and indirect costs of
accessing child health care services.

The study has found negative significant relationship between religion of mother and
demand for child health care services. In this regard, policies that promote educating
apostolic sector the importance of child healthcare services utilisation are essential and
should be maintained by policy makers in order to improve seeking of child health care
services utilisation in rural areas. If a household mother is more educated, he or she
understand the benefits and importance of seeking formal child health care services in the
case of illness of household child.

5.3 Areas for further Research

Owing to time and financial restrictions, the study only used 290 households to appeal
conclusions on the factors which influence utilisation of child health care services. Given
the availability of sufficient resources, data from more households from the Shamva
District can be used in future studies to make rich conclusions. Furthermore, the scope of
study was limited to Shamva District in Mashonaland Central Province. Thus, more
convincing results on the factors which influence demand for child health care services
utilisation in rural Zimbabwe can be found if similar studies are done in other rural
districts of Zimbabwe.

46
REFERENCES
Hjortsberg, C. (2002). Determinants of Household Health care expenditure. The Case
for Zambia.

Hjortsberg, C., and Mwikisa, C.N. (2003).Cost of access to health services in Zambia:
Health Policy Planning

Howlader, S.R., Subrata,R., Hossain, A., Saha,N.C., and Khuda, B. (2000). Demand for
Health care by Rural Households in Bangladesh: Level and Determinants.

Hutchinson, P. (1999), Household Demand for Health Services in Uganda. World

Bank

Kevany, S., Murimo, O., Singh, B., Hlubinka, D., Kulich, M., Morin, S.F., and Sweat, M.
(2012).Socio- economic Status and Health care Utilisation in Rural Zimbabwe:
Findings from Project Accept (HPTN 043).

Lawson, D. (2004). Determinants of Health Seeking Behaviour in Uganda

Lepine, A., and Nestour, A. (2011).Health Care Utilisation in Rural Senegal

Lindelow, M. (2000).Health Care Demand in Rural Mozambique

Litvack, J.I., and Bodart, C. (1993). ―User fees plus quality equals improved access
healthcare

Ministry of Health and Child Welfare, UNICEF, and World Bank. ( 2010).“The
Zimbabwe Health Sector Investment Case (2010)

Mugilwa, L.O. (2005), Health care services in Kenya: Poverty Monitoring,

Measurement and Analysis.

Muhofa, T.(2004).Determinants of Use of Formal Health care services in Butaleja

Sub –country, Butaleja District, Uganda

Munyuki, E., and Shorai, J. (2009), Capital Flows in the Health Sector in Zimbabwe;
Trends and implications for the Health system.

Naaktgeboren, G. (2012). Health Care Utilisation in the Netherlands: Before and

after the Health Insurance Reform.

National Health Strategy for Zimbabwe, (2009-2013). Equity and Quality in Health

Ministry of Health and Child Welfare, (2009) Government of Zimbabwe, Harare.

47
Okwe, A. O.,Okurut, F. N., and Adebua, A. (2006). The determinants of Health care
demand in Uganda:

Osariemen, G. (2011). Theoretical Issues in the Understanding of Maternal Health

Services Utilisation in Lagos, Nigeria.

Ramsey, J.B. (1969). Tests for Specification errors in Classical Linear Least Squares
Regression Analysis.

Sarma, S.(2000). Demand for Outpatient Health Care: Empirical Findings from
rural India.

Sekeran,U., and Bougie, R. (2004).Research Methods for Business

WHO (2000).World Health Report 2000: Health Systems

Wooldridge J. M. (2004). Introductory Econometrics -A modern approach. 2nd

edition.

World Health Organisation (2001). The World Health Report 2001.

Zimbabwe Census Preliminary Report (2012), Zimbabwe National Statistics Agency,

Harare, Zimbabwe.

Zimbabwe Demographic and Health Survey (2010-11). Zimbabwe National Statistical

Agency (ZIMSTAT) Harare, Zimbabwe.

Zimbabwe Health System Assessment (2010), Ministry of Health and Child welfare.

Zimbabwe Vulnerability Assessment Committee (ZIMVAC) (2012), Zimbabwe Rural

Food Security and Vulnerability Assessments.

48
APPENDICES

Appendix a: Correlation matrices

Stata 12 command :Correlates Ag Ed DST y r
(Obs=290)

| ch ag ed dst y r
-------------------------------------------------------------------
ch | 1.0000
ag | -0.0492 1.0000
ed | 0.9853 -0.0315 1.0000
dst | 0.0715 -0.0366 0.0715 1.0000
y | 0.8468 -0.0492 0.8340 0.1401 1.0000
r | -0.9193 0.0452 -0.9047 -0.0934 -0.7889 1.0000

Appendix b : Heteroskedasticity test

hettest

Breusch-Pagan / Cook-Weisberg test for heteroskedasticity

Ho: Constant variance
Variables: fitted values of ch

chi2 (1) = 5.94

Prob>chi2 = 0.0148

Appendix c : regression
Stata 12 command : regress ch ag ed r y dst

Source | SS df MS Number of obs = 290

------------------------------------------- F( 5, 284) = 2382.72
Model | 66.44704645 13.2894093 Prob> F = 0.0000

49
Residual | 1.58398808 284 0.005577423 R-squared = 0.9767
------------------------------------------- Adj R-squared = 0.9763
Total | 68.0310345 289 .235401503 Root MSE = 0 .07468

------------------------------------------------------------------------------
ch | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-----------------------------------------------------------------------------
ag | -.0008809 .0005473 -1.61 0.109 -.0019583 .0001964
ed | .8013499 .0240339 33.34 0.000 .7540428 .8486571
r | -.141036 .0214756 -6.57 0.000 -.1833076 -.0987644
y | .0002097 .000052 4.03 0.000 .0001072 .0003121
dst | -.0115355 .0117837 -0.98 0.328 -.0347299 .011659
_cons | .1405347 .0328981 4.27 0.000 .0757797 .2052897
------------------------------------------------------------------------------

Appendix d: summary statistics

. summarize ch ag r y dst ed

Variable | Obs Mean Std. Dev. Min Max

---------------------------------------------------------------------
ch | 290 .3758621 .4851819 0 1
ag | 290 35.55517 8.044775 20 57
r| 290 .6206897 .4860542 0 1
y| 290 298.2759 156.2597 100 760
dst | 290 .8275862 .3783927 0 1
---------------------------------------------------------------------
ed | 290 .3758621 .4851819 0 1

Appendix e. Marginal effects after regress

Stata command: mfx
Marginal effects after regress

50
y = Fitted values (predict)
= .37586207
------------------------------------------------------------------------------
variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X
---------+--------------------------------------------------------------------
ag | -.0008809 .00055 -1.61 0.107 -.001954 .000192 35.5552
ed*| .8013499 .02403 33.34 0.000 .754244 .848456 .375862
r*| -.141036 .02148 -6.57 0.000 -.183127 -.098945 .62069
y | .0002097 .00005 4.03 0.000 .000108 .000312 298.276
dst*| -.0115355 .01178 -0.98 0.328 -.034631 .01156 .827586
------------------------------------------------------------------------------
(*) dy/dx is for discrete change of dummy variable from 0 to 1

Appendix f: data used

hh ch ed ag dst r y
1 0 0 40 0 1 140
2 0 0 45 0 1 200
3 1 1 33 1 0 450
4 1 1 30 1 0 500
5 1 1 26 1 0 400
6 0 0 34 0 1 150
7 1 1 30 0 0 450
8 1 1 25 1 0 380
9 0 0 20 1 1 190
10 0 0 21 1 1 200
11 1 1 31 0 0 392
12 1 1 29 1 0 350
13 1 1 28 1 0 300
14 0 0 35 1 1 160
15 0 0 25 0 1 135
16 0 0 29 0 1 160

51
17 0 0 41 1 1 150
18 0 0 20 1 1 230
19 0 0 23 1 1 200
20 0 0 34 0 1 210
21 1 1 28 0 0 400
22 1 1 32 0 0 350
23 0 0 26 1 1 100
24 0 0 21 1 1 180
25 1 1 32 1 1 360
26 0 0 36 1 1 200
25 0 0 27 1 1 100
28 0 0 37 1 1 132
29 0 0 34 1 1 150
30 0 0 33 0 0 140
31 1 1 29 0 0 392
32 0 0 24 1 1 180
33 0 0 44 1 1 200
34 1 1 39 0 0 369
35 0 0 37 1 1 210
36 0 0 26 1 1 170
37 0 0 20 1 1 180
38 0 0 22 1 1 170
39 0 0 27 1 1 200
40 1 1 29 0 1 400
41 0 0 33 0 1 190
42 0 0 31 0 1 200
43 1 1 41 1 0 420
44 1 1 29 0 0 430
45 0 0 24 1 1 180
46 0 0 43 0 1 200
47 0 0 28 1 1 150

52
48 0 0 22 0 1 190
49 0 0 24 1 1 210
50 1 1 29 1 0 450
51 0 0 44 1 1 200
52 1 1 41 1 0 390
53 0 0 24 1 1 160
54 0 0 33 1 1 200
55 0 0 34 1 1 210
56 1 1 33 1 0 480
57 0 0 23 1 1 200
58 0 0 27 1 1 180
59 0 0 23 1 1 160
60 0 0 32 1 1 200
61 1 1 32 1 0 350
62 1 1 43 1 0 490
63 0 0 32 1 1 200
64 1 1 43 1 0 205
65 1 1 42 1 0 450
66 1 1 43 1 0 452
67 0 0 42 1 1 188
68 1 1 33 1 0 380
69 1 1 32 1 0 440
70 1 1 33 1 0 450
71 1 1 32 1 0 450
72 0 0 31 1 1 173
73 0 0 34 1 1 180
74 0 0 34 1 1 189
75 0 0 22 1 1 173
76 1 1 32 1 0 470
77 0 0 23 1 0 192
78 0 0 34 1 1 194

53
79 0 0 44 1 1 200
80 0 0 23 1 1 205
81 0 0 32 1 1 195
82 0 0 43 1 1 210
83 1 1 34 1 0 480
84 0 0 24 1 1 213
85 0 0 33 1 1 208
86 1 1 42 1 0 180
87 0 0 34 1 1 198
88 0 0 44 0 1 190
89 1 0 34 1 0 460
90 0 0 34 1 1 200
91 0 0 35 0 1 197
92 0 0 36 0 1 188
93 1 1 34 0 0 350
94 1 1 25 0 0 450
95 1 1 23 1 0 470
96 0 0 23 1 1 200
97 1 1 34 1 0 450
98 0 0 35 1 1 215
99 0 0 33 1 1 200
100 0 0 33 1 1 205
101 1 1 34 1 0 380
102 0 0 34 0 1 155
103 0 0 24 1 0 200
104 0 0 23 1 1 205
105 0 0 44 1 1 200
106 0 0 53 1 1 200
107 0 0 53 0 1 200
108 0 0 54 0 1 135
109 0 0 53 0 1 123

54
110 1 1 32 0 1 450
111 1 1 30 1 0 480
112 0 0 43 0 1 200
113 0 0 43 1 1 210
114 0 0 43 0 1 205
115 0 0 44 1 1 360
116 0 0 44 1 1 150
117 0 0 33 1 1 250
118 0 0 33 0 1 155
119 0 0 32 1 1 520
120 1 1 34 0 0 400
121 0 0 34 0 1 180
122 0 0 34 1 1 140
123 0 0 33 1 1 220
124 0 0 32 1 1 200
125 0 0 32 0 1 200
126 0 0 33 1 1 180
127 0 0 34 1 1 205
128 0 0 42 1 1 200
129 1 1 42 1 0 390
130 0 0 41 1 1 120
131 0 0 43 1 1 160
132 1 1 40 1 0 480
133 0 0 42 1 1 180
134 1 1 33 1 0 480
135 0 0 29 1 1 200
136 0 0 37 1 1 215
137 0 0 34 1 1 180
138 0 0 34 1 1 280
139 0 0 33 1 1 390
140 0 0 23 1 1 320

55
141 1 1 28 1 0 450
142 1 1 43 1 0 500
143 1 1 42 1 0 450
144 0 0 26 1 1 205
145 0 0 54 1 1 200
146 0 0 54 1 1 190
147 0 1 54 1 1 180
148 0 0 54 1 1 155
149 0 0 53 1 1 320
150 0 0 43 0 1 500
151 0 0 43 1 0 232
152 0 0 42 1 1 650
153 0 0 33 1 1 210
154 0 0 34 1 1 200
155 0 0 34 1 1 190
156 0 0 33 1 1 200
157 1 1 33 1 0 520
158 0 0 33 1 1 160
159 0 0 29 1 1 120
160 1 1 43 1 0 550
161 1 1 32 0 0 450
162 0 0 42 1 1 200
163 0 0 33 0 1 200
164 0 0 44 0 1 150
165 0 0 35 0 1 135
166 0 0 45 0 1 123
167 0 0 35 0 1 150
168 0 0 45 1 1 130
169 0 0 35 1 1 170
170 0 0 44 0 1 120
171 0 0 34 0 1 205

56
172 0 0 34 0 1 160
173 0 0 43 1 1 450
174 1 1 40 1 0 560
175 1 1 21 1 0 500
176 1 1 32 0 0 530
177 1 1 33 1 0 500
178 0 0 54 1 1 160
179 0 0 34 1 1 140
180 0 0 45 1 1 180
181 0 0 55 1 1 120
182 0 0 56 1 1 130
183 0 0 45 0 1 130
184 0 0 34 0 1 100
185 0 0 33 1 1 150
186 1 1 33 1 0 440
187 1 1 42 1 0 140
188 1 1 40 1 0 760
189 1 1 42 1 0 600
190 1 1 38 1 0 550
191 1 1 34 1 0 390
192 1 1 45 1 0 750
193 0 0 36 1 1 120
194 1 1 36 1 0 557
195 0 0 47 1 1 135
196 0 0 57 1 1 123
197 0 0 55 0 1 150
198 1 1 54 1 0 490
199 1 1 44 1 0 540
200 1 1 44 1 0 560
201 1 1 43 1 0 600
202 0 0 33 1 1 160

57
203 0 0 33 1 1 450
204 1 1 33 1 0 500
205 1 1 42 1 0 450
206 1 1 42 1 0 430
207 0 0 32 1 1 100
208 0 0 32 1 1 132
209 0 0 34 1 1 326
210 0 0 34 1 1 371
211 1 1 45 1 0 392
212 0 0 55 1 1 113
213 0 0 56 1 1 130
214 0 0 46 1 1 160
215 0 0 45 1 1 200
216 1 1 34 1 0 600
217 0 0 33 1 1 203
218 0 0 33 1 1 215
219 0 0 33 1 0 300
220 1 1 32 1 0 540
221 0 0 42 1 1 198
222 1 1 33 1 0 555
223 1 1 34 1 0 570
224 1 1 35 1 0 480
225 1 1 30 0 0 540
226 1 1 28 1 0 450
227 1 1 31 1 0 500
228 0 0 25 1 1 170
229 0 0 25 1 1 200
230 0 0 26 1 1 113
231 1 1 34 1 0 540
232 1 1 30 1 0 560
233 0 0 43 1 1 160

58
234 1 1 35 1 0 470
235 0 0 34 1 1 130
236 0 0 24 1 1 160
237 1 1 34 1 0 460
238 1 1 36 1 0 590
239 1 1 42 1 0 650
240 0 0 24 1 1 178
241 1 1 36 1 0 497
242 1 1 34 1 0 440
243 0 0 30 1 1 190
244 1 1 31 1 0 450
245 1 1 38 1 0 490
246 1 1 23 1 0 524
247 1 1 32 1 0 490
248 0 0 32 1 1 180
249 0 0 34 1 1 143
250 1 1 30 1 0 560
251 0 0 34 1 1 213
252 0 0 25 1 1 208
253 0 0 36 1 1 600
254 1 1 34 1 0 530
255 0 0 35 1 1 190
256 1 1 33 1 1 500
257 0 0 27 1 1 200
258 1 1 35 1 1 480
259 0 0 45 1 0 188
260 1 1 31 1 0 500
261 0 0 35 1 1 160
262 1 1 32 1 0 450
263 1 1 36 1 0 470
264 1 1 33 1 0 520

59
265 0 0 25 1 1 215
266 1 1 34 1 0 560
267 1 1 44 1 0 550
268 0 0 53 1 1 200
269 0 0 44 1 1 198
270 1 1 33 1 0 550
271 1 1 32 1 0 480
272 1 1 34 1 0 580
273 0 0 45 1 1 200
274 0 0 34 1 1 215
275 1 1 42 1 0 500
276 1 1 33 1 0 490
277 0 0 46 1 1 120
278 0 0 46 1 1 195
279 0 0 33 1 1 210
280 0 0 34 1 1 178
281 1 1 44 1 0 440
282 0 0 43 1 1 208
283 0 0 36 1 1 194
284 1 1 33 1 0 500
285 0 0 34 1 1 200
286 0 0 43 1 1 140
287 0 0 34 1 1 160
288 1 1 45 1 0 400
289 1 1 54 1 0 300
290 1 1 32 1 0 450

60
 BINDURA UNIVERSITY OF SCIENCE EDUCATION

Appendix g: Household Questionnaire

My name is Werengani Simon student number B0823136 and I am a final year in

Bachelor of Science in Economics student at Bindura University. I am carrying out a
research on the topic entitled: The relationship of mother’s levels of education to child
healthcare services utilization in Zimbabwe - the case for Shamva District from 2000
to 2014. The purpose of this study is therefore to collect data concerning the utilisation of
child healthcare care services. The research is for academic purposes and your responses
will be kept strictly confidential. Your cooperation will be highly appreciated.

Questionnaire No _______________
Name of the interviewer_______________ Date of interview__________________
Ward _______________Household number _________ village ____________
Respondent

SECTION A: HOUSEHOLD CHARACTERISTICS

1. Age of mother ____________________ Years

2. How many how many children are currently staying here? ___________
3. What is the highest level of education attained by the mother of the household?
[0] No education at all __________ [0] Secondary education ___________
[0] Primary education ___________ [1] Tertiary education ___________
4. What is the mother’s monthly income ______________________ US $
5. What is religion of the mother?
[0] Traditional _________ [1] Apostolic Sect churches _________

61
SECTION B: ILLNESS AND SEEKING OF HEALTH CARE SERVICES

6. Has any children of the household reported ill within the last three months? [1]YES
[0] NO
7. If YES to Question 7, was treatment sought from a health care facility? [1] YES
[0] NO
8. If YES to Question 7, was the childconfined to bed before being attended to?[1]YES
[0]NO

SECTION C: INSTITUTIONAL FACTORS

9. Approximately, what is the distance to your nearest health facility? ____________

Codes: 1 - 5 Km , low = 1 6 – 9 Km , high = 0
Thank you very much for taking your time to respond to the questions.

62