MARGINALIZATION AND EDUCATION:
INTERNATIONAL PRESENCE, PERFORMANCE, AND POLICY
By
Alisha M. B. Brown

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Educational Policy—Doctor of Philosophy
2015

ABSTRACT
MARGINALIZATION AND EDUCATION:
INTERNATIONAL PRESENCE, PERFORMANCE, AND POLICY
By
Alisha M. B. Brown
This dissertation is a collection of three separate but interrelated essays exploring the
overarching topic of marginalization and education internationally. Across the essays a
combination of qualitative and quantitative techniques are used to analyze marginalization and
education for different student populations in different international contexts. In terms of the
different marginalized populations that are studied, the first and third essays focus on disability
while the second essay examines gender, socioeconomic status, home language, and immigrant
status. Geographically, the first essay has the broadest focus at the global level, while the second
and third essays focus on select countries in sub-Saharan Africa.
The first essay analyzes the presence and representation of disability in mainstream
international education development discourse. Discourse analysis is used to systematically
examine the attention that is paid to disability as a form of marginalization and the way disability
is linguistically situated in recent international development strategy documents. Three
documents produced by agencies that play a major role in international education development
were purposefully sampled and analyzed, namely the World Bank, the United Nations
Educational, Scientific and Cultural Organization (UNESCO), the United Nations Children’s
Fund (UNICEF), and the Brookings Institute.
The second essay analyzes the relationships between multiple categories of
marginalization, feelings of belongingness, and academic achievement across three sub-Saharan
African countries using 8th-grade data from the Trends in International Mathematics and

Science, 2011 (TIMSS, 2011). Specifically, marginalization on the basis of one’s gender,
socioeconomic status, home language, and immigrant status is examined in Botswana, Ghana,
and South Africa. A central component of the analysis involves investigating whether feelings
of belongingness buffer the negative impact that marginalization can have on academic
achievement in mathematics and science.
The third essay explores how the national inclusive education policy reform process for
students with disabilities has been shaped by global and local forces in Tanzania. I investigate
the perspectives and experiences of national policy actors working in the government, nongovernmental organizations and universities by conducting semi-structured interviews in Dar es
Salaam. My analysis focuses on understanding the appropriation of inclusive education policy in
an effort to demystify the policy process and emphasize the agency of global and local actors.
Together, these three essays contribute to literature on marginalization in education at
global and local levels. Individually, each essay offers a unique contribution by having different
foci within the larger topic of marginalization and education internationally. While the first
essay focuses on the presence and representation of a marginalized group within international
education discourse, the second focuses on the academic performance of marginalized groups,
and the third on educational policy for a marginalized group. In addition, the individual
contribution of each essay is further enhanced by the diversity of research approaches that are
used across the essays, including document analysis, quantitative secondary data analysis and
qualitative research methods. The diversity of research approaches and range of subtopics
explored within this dissertation enables each essay to provide a unique perspective on the
presence, performance, and policy of particular marginalized student groups, while at the same
time speaking to the larger issue of marginalization in education internationally.

ACKNOWLEDGEMENTS

This dissertation was a product, in large part, of the generous support I received from
several individuals throughout my doctoral career. Many thanks to my advisor, Dr. Amita
Chudgar, and my dissertation committee members Dr. Peter Youngs, Dr. John Metzler, Dr.
Kristin Phillips, and Dr. Troy Mariage. Given the interdisciplinary nature of my dissertation
bridging educational policy, international and comparative education, and special education, I am
grateful for the opportunity to have worked closely with such tremendous faculty across
disciplines, all of whom have been instrumental to my development as a scholar.
This research was jointly funded by the Michigan State University Graduate School,
College of Education, Education Policy Center, Tanzania Partnership Project, and the Alberta
Scholarships Program. I would also like to extend my appreciation to all of my interview
participants for graciously sharing their time and experiences with me. I am especially grateful
to Veronica Sarungi at Aga Khan University in Dar es Salaam, Tanzania for her valuable
guidance identifying and connecting me with key policy actors in the Ministry and nongovernmental organizations.
Lastly, I would like to thank my family, particularly my parents and my partner Michael
Braun. I am incredibly grateful for your endless support and encouragement.

iv

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................................... ix
LIST OF FIGURES ....................................................................................................................... xi
KEY TO ABBREVIATIONS ....................................................................................................... xii
CHAPTER 1. THE PRESENCE AND REPRESENTATION OF DISABILITY IN SHIFTING
GLOBAL EDUCATION DISCOURSE: A DOCUMENT ANALYSIS ........................................1
Introduction ..................................................................................................................................1
Entering a New Era: International Development Education Strategy Context ........................2
Consequences of the Shifting International Education Agenda on Development ...................4
Consequences for disability .................................................................................................5
Method .........................................................................................................................................7
Research Questions ..............................................................................................................7
Sample..................................................................................................................................7
Analytical Approach ..........................................................................................................10
Results ........................................................................................................................................11
RQ1: Presence of Disability in International Development Education Strategy Discourse
............................................................................................................................................11
RQ2: How Disability is Linguistically Situated in International Development Education
Strategy Discourse ..................................................................................................................13
Representation of disability in the text as a whole ...........................................................13
Disability in the background ........................................................................................14
Disability emerging from the background ...................................................................17
Representation of disability in sentences and words ........................................................19
Discussion ..................................................................................................................................22
Implications and Recommendations.......................................................................................22
Limitations and Future Directions ..........................................................................................26
Conclusion ..................................................................................................................................27
REFERENCES ..........................................................................................................................29
CHAPTER 2. ACADEMIC ACHIEVEMENT OF MARGINALIZED POPULATIONS: THE
BUFFERING EFFECT OF BELONGINGNESS .........................................................................33
Introduction ................................................................................................................................33
Literature Review .......................................................................................................................34
The Need to Belong ................................................................................................................35
Feelings of belongingness and academic achievement ......................................................35
Marginalization in Education .................................................................................................37
Marginalization and academic achievement ......................................................................38
Gender ..........................................................................................................................39
Socioeconomic status ...................................................................................................40
Home language ............................................................................................................40
v

Immigrant status...........................................................................................................41
Why Sub-Saharan Africa? ......................................................................................................42
Low academic achievement ...............................................................................................42
Widespread marginalization ..............................................................................................43
Gender ..........................................................................................................................44
Socioeconomic status ...................................................................................................46
Home language ............................................................................................................47
Immigrant status...........................................................................................................49
Theoretical Framework ..............................................................................................................51
Psychological Theory on the Need to Belong....................................................................51
Labeling Theory .................................................................................................................53
Research Questions and Hypotheses ..........................................................................................55
Data and Method ........................................................................................................................59
Instruments .........................................................................................................................59
Variables ............................................................................................................................60
Academic Achievement ...............................................................................................62
Indicators of marginalization .......................................................................................63
Feelings of Belongingness ...........................................................................................64
Control variables ..........................................................................................................64
Sample................................................................................................................................65
Data Preparation.................................................................................................................68
Data Analysis .....................................................................................................................69
Results ........................................................................................................................................71
RQ1: Feelings of Belongingness and Achievement ...............................................................83
RQ2: Marginalization and Achievement ................................................................................84
RQ2a: Gender and achievement ........................................................................................84
RQ2b: SES and achievement .............................................................................................86
RQ2c: Home language and achievement ...........................................................................87
RQ2d: Immigrant status and achievement .........................................................................88
RQ3: Feelings of Belongingness as a Buffer .........................................................................89
RQ3a: Feelings of belongingness and gender ....................................................................90
RQ3b: Feelings of belongingness and SES .......................................................................90
RQ3c: Feelings of belongingness and home language ......................................................90
RQ3d: Feelings of belongingness and immigrant status....................................................91
RQ4: Cross-National Comparisons ........................................................................................92
Discussion ..................................................................................................................................94
Marginalization.......................................................................................................................94
Gender ................................................................................................................................94
SES.....................................................................................................................................95
Home language ..................................................................................................................97
Immigrant status.................................................................................................................97
Feelings of Belongingness ......................................................................................................98
Additional Limitations and Future Directions ......................................................................101
Conclusion ................................................................................................................................102
APPENDIX ..............................................................................................................................104
REFERENCES .........................................................................................................................106
vi

CHAPTER 3. GLOBAL VERSUS LOCAL INFLUENCES ON THE APPROPRIATION OF
INCLUSIVE EDUCATION POLICY IN TANZANIA .............................................................118
Introduction ..............................................................................................................................118
Relevant Background and Literature ........................................................................................121
Inclusive Education Globally...........................................................................................121
The status of education for children with disabilities globally ...................................121
Global inclusive education policy landscape ..............................................................122
Inclusive Education Locally ............................................................................................123
Geographical, economic, and political context ...........................................................123
Inclusive education policy landscape in Tanzania ......................................................126
The status of education for children with disabilities in Tanzania ..............................127
Barriers to inclusive education policy implementation in Tanzania ...........................131
Theoretical Framework ............................................................................................................133
Research Questions ..........................................................................................................135
Method .....................................................................................................................................136
Participants .......................................................................................................................137
Government officials ...................................................................................................137
NGO staff members.....................................................................................................138
Faculty members .........................................................................................................139
Sampling Procedure .........................................................................................................139
Data Collection Procedure ...............................................................................................140
Researcher Positionality...................................................................................................141
Data Organization and Analytical Approach ...................................................................142
Results ......................................................................................................................................143
RQ1. Local Influence on Inclusive Education Policy Appropriation ..............................143
Stigmatizing cultural attitudes towards disability .......................................................144
Political context ...........................................................................................................148
Stakeholder collaboration ....................................................................................148
Political will .........................................................................................................150
Economic conditions ...................................................................................................152
Government budget allocation .............................................................................152
Familial economic constraints .............................................................................154
RQ1a. Barriers to Policy Implementation Experienced by Local Actors ...................155
Attitudes ...............................................................................................................156
Funding and availability of resources ..................................................................161
Physical barriers ...................................................................................................162
Teachers ...............................................................................................................163
RQ2. Global Influence on Inclusive Education Policy Appropriation ............................165
Funding ........................................................................................................................166
Global doctrine ............................................................................................................167
RQ3. Local and Global Intersection and Divergence ......................................................168
Intersection ..................................................................................................................169
Global documents intersect with local needs .......................................................169
Global funding priority area intersects with local reliance on donor funds .........170
Divergence...................................................................................................................172
RQ3a. Local Resistance, Adaptation and Subterfuge .................................................173
vii

Resistance ............................................................................................................173
Adaptation ............................................................................................................175
Subterfuge ............................................................................................................177
Discussion ................................................................................................................................180
Limitations and Future Directions ...................................................................................183
Implications......................................................................................................................185
Conclusion ................................................................................................................................187
APPENDICES ..........................................................................................................................189
APPENDIX A Interview Protocols ......................................................................................190
APPENDIX B Informed Consent Forms .............................................................................203
REFERENCES .........................................................................................................................208

viii

LIST OF TABLES

Table 1.1 The Education MDGs and EFA Goals ............................................................................3
Table 1.2 Sample .............................................................................................................................8
Table 1.3 Presence of the Word Disability Relative to Total Words in Each Document ..............13
Table 2.1 GPI of Countries Included in Study...............................................................................46
Table 2.2 Languages of Countries Included in Study ....................................................................48
Table 2.3 Hypotheses .....................................................................................................................58
Table 2.4 Variable Details .............................................................................................................61
Table 2.5 Sub-Saharan African Countries Participating in TIMSS 2011 ......................................65
Table 2.6 Country Descriptive Statistics .......................................................................................67
Table 2.7 Correlation Matrix with Botswana Mathematics Data ..................................................72
Table 2.8 Correlation Matrix with Botswana Science Data .........................................................73
Table 2.9 Correlation Matrix with Ghana Mathematics Data .......................................................74
Table 2.10 Correlation Matrix with Ghana Science Data .............................................................75
Table 2.11 Correlation Matrix with South Africa Mathematics Data ..........................................76
Table 2.12 Correlation Matrix with South Africa Science Data ...................................................77
Table 2.13 OLS Estimates for Models Predicting Mathematics and Science Achievement in
Botswana ........................................................................................................................................78
Table 2.14 OLS Estimates for Models Predicting Mathematics and Science Achievement in
Ghana .............................................................................................................................................79
Table 2.15 OLS Estimates for Models Predicting Mathematics and Science Achievement in
South Africa ...................................................................................................................................80
Table 2.16 OLS Estimates for Predictors of Interest in Key Models Predicting Mathematics and
Science Achievement in Each Country .........................................................................................81
Table 2.17 Hypotheses Supported by Country ..............................................................................82
ix

Table 3.1 School Enrollment in Tanzania by Disability Category ..............................................129

x

LIST OF FIGURES

Figure 1.1. Number of times disability is explicitly mentioned in the main text of each document
........................................................................................................................................................12
Figure 1.2. Percentage of time the word disability falls at the beginning, middle and end of a list
of marginalized categories within a sentence ................................................................................20
Figure 2.1. Conceptual model ........................................................................................................55
Figure 2.2. Ordinal interaction between feelings of belongingness and immigrant status in South
Africa .............................................................................................................................................92
Figure 3.1. Map of Tanzania and bordering countries .................................................................124
Figure 3.2. Theoretical framework ..............................................................................................135
Figure 3.3. Participant categorization summary ..........................................................................137
Figure 3.4. National DPO structure in Tanzania..........................................................................139
Figure 3.5. Sample newspaper clippings of individuals with albinism whose limbs had been
chopped off ..................................................................................................................................146
Figure 3.6. Sample inclusive education advocacy materials .......................................................159

xi

KEY TO ABBREVIATIONS

ADHD – Attention Deficit Hyperactivity Disorder
ASD – Autism Spectrum Disorder
BEST – Basic Education Statistics Tanzania
CIA – Central Intelligence Agency
COSTECH – Tanzania Commission for Science and Technology
CRPD – Convention on the Rights of Persons with Disabilities
CSO – civil society organization
DA – discourse analysis
DPO – disabled people’s organization
EFA – Education for All
EMDE – emerging market and developing economy
ESR – Education for Self-Reliance
GDP – gross domestic product
GNI – gross national income
GPI – gender parity index
HER – home educational resources
ICF – International Classification of Functioning, Disability and Health
IDB – International Database
IEA – International Association for the Evaluation of Educational Achievement
IMF – International Monetary Fund
INGO – international non-governmental organization
IRB – Institutional Review Board
xii

IRT – Item Response Theory
LMTF – Learning Metrics Task Force
MDG – Millennium Development Goal
MoEVT – Ministry of Education and Vocational Training
MoLYDS – Ministry of Labour, Youth Development and Sports
NGO – non-governmental organization
NPD – National Policy on Disability
OLS – ordinary least squares
PIRLS – Progress in International Reading Literacy Scale
SACMAQ – Southern and Eastern Africa Consortium for Monitoring Educational Quality
SES – socioeconomic status
SHIVYAWATA – Shirikisho la Vyama Vya Watu Wenye Ulemavu (Tanzania Federation of
Disabled People’s Organizations)
TIE – Tanzania Institute of Education
TIMSS – Trends in International Mathematics and Science
UN – United Nations
UNESCO – United Nations Educational, Scientific and Cultural Organization
UNICEF – United Nations International Children’s Emergency Fund
USAID – United States Agency for International Development
WHO – World Health Organization

xiii

CHAPTER 1. THE PRESENCE AND REPRESENTATION OF DISABILITY IN SHIFTING
GLOBAL EDUCATION DEVELOPMENT DISCOURSE: A DOCUMENT ANALYSIS
Introduction
There is an emerging understanding in global education dialogues that the focus on
access to education is being replaced with a focus on quality of education systems. Worldwide,
children are increasingly enrolling in school, so now there is a growing interest in understanding
what are they learning in school. This focus on excellence may generate new concerns about
equity because marginalized children are likely to be left behind in a system driven solely by
learning outcomes. Children with disabilities have a heightened risk of educational
marginalization and as a result are especially likely to be left behind. The United Nations
Educational, Scientific and Cultural Organization (UNESCO, 2010) considers disability to be
“one of the least visible but most potent factors in educational marginalization” (p. 181). In light
of this, I will explore the presence and representation of disability in these changing international
education development discussions. Specifically, I pose two research questions. First, to what
extent is disability present in international development education strategy discourse? And
second, to what extent is disability linguistically situated in international development education
strategy discourse?
This essay begins by setting the current international development education strategy
context and explaining the consequences of the shifting global education agenda for development
in general and for disability specifically. A close inspection of the presence and representation
of disability in a selection of global education development discourse follows. Next, details on
the sample of international education development documents that are analyzed, the discourse

1

analysis approach that is used, findings and implications are presented. In closing, limitations,
future directions, implications, and recommendations are discussed.
Entering a New Era: International Development Education Strategy Context
We are in the midst of a historical transition as the focus of international development in
education shifts from access to quality. This global shift is reflected in key international
education development events which will be highlighted. After a series of global agenda setting
meetings in the 1990s, at the Millennium Summit in September 2000, 189 members of the
United Nations (UN) adopted a Millennium Declaration, from which eight Millennium
Development Goals (MDGs) were formed to be reached in 15 years by 2015. Two of the eight
goals focused on education: MDG2 achieve universal primary education and MDG3 promote
gender equality and empower women (UN Millennium Project, 2005). A complimentary
education development event occurred in Dakar in March 2000 when the Education for All
(EFA) framework was adopted after reviewing progress of EFA goals set in 1990 by the
international community. The EFA framework articulated six education goals which, like the
MDGs, had a 2015 target date (see Table 1.1 for a complete list of the EFA and MDG education
goals). If we view the international education development arena as a canvas, for the past two
decades the MDGs and EFA goals have narrowed the educational canvas by focusing on
universal primary education and gender parity indicators and targets (Ahmed, 2014).

2

Table 1.1
The Education MDGs and EFA Goals
Goal #

Education target

MDG2

Achieve universal primary education
Target 2A: By 2015, all children, girls and boys, complete a full course of
primary schooling.

MDG3

Promote gender equality and empower women
Target 3A: Eliminate gender disparity in primary and secondary education
preferably by 2005, and at all levels by 2015.

EFA Goal 1

Expanding and improving comprehensive early childhood care and education,
especially for the most vulnerable and disadvantaged children.

EFA Goal 2

Ensuring that by 2015 all children, particularly girls, children in difficult
circumstances and those belonging to ethnic minorities, have access to, and
complete, free and compulsory primary education of good quality.

EFA Goal 3

Ensuring that the learning needs of all young people and adults are met
through equitable access to appropriate learning and life-skills programs.

EFA Goal 4

Achieving a 50% improvement in levels of adult literacy by 2015, especially
for women, and equitable access to basic and continuing education for all
adults.

EFA Goal 5

Eliminating gender disparities in primary and secondary education by 2005,
and achieving gender equality in education by 2015, with a focus on ensuring
girls’ full and equal access to and achievement in basic education of good
quality.

EFA Goal 6

Improving all aspects of the quality of education and ensuring excellence of
all so that recognized and measurable learning outcomes are achieved by all,
especially in literacy, numeracy and essential life skills.

Source. MDGs retrieved from http://www.un.org/millenniumgoals/. EFA goals retrieved from
http://www.unesco.org/new/en/education/themes/leading-the-international-agenda/education-forall/efa-goals/
As we leave the MDG-EFA era, the international emphasis on access to education is
being replaced by an emphasis on quality education. Beginning in 1990, and accelerating to the
2015 deadline, critique and attention has turned to learning quality (McGrath, 2014b). A major
criticism of the MDG-EFA framework is its failure to sufficiently address learning, which is at
the core of the educational experience (Adams, 2012). While efforts have been focused on

3

getting children into school, less attention has been paid to learning in school, or the acquisition
of knowledge and skills that enable children to live productive, healthy, safe lives. For example,
data from the Brookings African Learning Barometer project shows that at the end of 4-5 years
in primary school in 26 African countries, 45% of children fail to meet basic learning outcomes
(Van Fleet, Watkins, & Greubel, 2012). Quality education is a dominant cross-cutting theme in
global discourse concerning the post-2015 development agenda.
Consequences of the Shifting International Education Agenda on Development
The post-2015 global agenda has the potential to bring about a radical change in research
and development funding priorities (McGrath, 2014a). Global discourse, including the
documents that are analyzed in this essay, plays a fundamental role in shaping what this change
will look like. International education development strategy documents are extremely influential
because they reach a global audience of policy and decision makers, educational practitioners,
and international funding institutions. Education sector policy documents produced by
multinational agencies like the World Bank and the UN are unquestionably used as a key
referent in decisions and negotiations by developing countries that rely on foreign aid (Klees,
Samoff, & Stromquist, 2012). The power that these sorts of agencies and the education strategy
documents that they produce have over international education development projects cannot be
understated. To illustrate the breadth of this power, Klees et al. (2012) argue that over the past
three decades the World Bank has situated “itself as the architect, implementer, and enforcer of
global education policy” (p. xv).
Not surprisingly, the money that is tied to fulfilling the goals outlined in international
education strategy documents has a great influence on the types of educational reforms and
projects that are adopted around the world. An important concept to consider with regards to the

4

influence of international aid is ‘donor logic’ or what Steiner-Khamsi (2012) defines as
understanding the nuances behind why and how aid is given to whom. Multinational agencies
like the World Bank and the UN have assumed global leadership on matters of aid and
international development due to their vast monetary and human resources have been
increasingly serving as global policy advisors to national governments (Steiner-Khamsi, 2012).
To receive a grant or loan, recipient governments have to subscribe to the priorities of the aid
strategy of the donor.
Consequences for disability. The shift in international research and funding priorities
from educational access to quality has important implications for children with disabilities
globally. Children with disabilities are among the most marginalized in the world. In many
countries children with disabilities are largely limited to neglect, abandonment, or
institutionalization (UNICEF, 2013). One of the main obstacles facing children with disabilities
is that they are often ‘invisible’ because the number of children living with disabilities globally is
unknown. The most widely used global estimate of childhood disability prevalence is 93 million
children, or 1 in 20 of those aged 14 years or younger (WHO, 2011). However, this global
estimate is essentially speculative. Children with disabilities often do not get registered at birth
and few countries have reliable data on disability prevalence rates (UNICEF, 2013). Since the
shifting international education development strategy focus is on visible children who are already
in school, the invisible children (i.e., children with disabilities) are at risk of being overlooked in
efforts to improve learning outcomes.
In many countries of the world children with disabilities are disproportionately denied the
right to an education. Disability restricts children’s educational opportunity and has been found
to be related to low enrollment, low attendance, and high dropout rates (UNESCO, 2010;

5

USAID, 2011). For example, a WHO (2011) survey in 51 countries found that 51% of boys with
disabilities complete primary school compared to 61% of boys without a disability. In contrast,
the same survey found that 42% of girls with a disability complete primary school compared to
53% of girls without a disability. Disability presents many barriers to educational access such as
insufficient physical access to schools, lack of accessible learning resources and assistive
technology, attitudinal barriers, and limited teacher capacity. To illustrate the severity of the
restricted educational opportunities for children with disabilities, a recent monitoring report of
the Convention on the Rights of the Child acknowledges that “the challenges faced by children
with disabilities in realizing their right to education remain profound” and goes on to recognize
“children with disabilities as one of the most marginalized and excluded groups in respect of
education” (UN, 2011, p. 8).
Global efforts need to continue to overcome such challenges to improve the disparate
access that children with disabilities have to education. Building upon Adams’ (2012)
recognition of the MDGs and EFA goals as important ‘building blocks’ for the betterment of
education globally, it is important not to prematurely abandon efforts focused on educational
access for children with disabilities specifically. As the statistics presented in the previous
paragraph illustrate, the foundational step or building block of getting children with disabilities
into their community’s schools has yet to be realized. As a result, the shifting global agenda
focusing on quality over access is concerning for children with disabilities because domestic
policies and projects in developing countries, and development and financing policies in donor
countries are emphasizing educational quality. Yet, most development initiatives for children
with disabilities target educational access, so as the broader international development focus
moves beyond access to quality the priority of educational access for children with disabilities

6

may receive less global attention and resources. In light of the concerns for children with
disabilities caused by the shifting international education development agenda, select education
strategy documents are carefully analyzed to see if and how disability is represented.
Method
A systematic document analysis of three international development education strategy
documents was conducted. The research questions, sample of international education
development documents, and analytical approach are explained below.
Research Questions
To examine the presence and representation of disability in the mainstream international
education development discourse, my analysis of international education development
documents was guided by the following research questions:
RQ1. To what extent is disability present in international development education
strategy discourse?
RQ2. How is disability linguistically situated in international development education
strategy discourse?
Sample
Table 1.2 lists the three international education development documents that are
analyzed. The title, date, objective, and main agency(ies) involved for each report are specified
in the table. These select documents are limited to recent World Bank, UNESCO, and UNICEF
reports in an effort to capture the dominant discourse on global education development within
the current zeitgeist. Additionally, a recent report from an influential international nongovernmental organization (INGO) is also included in the sample, namely The Brookings
Institute. The Brookings Institute was selected as an influential INGO in international education

7

development due to its joint effort with the UNESCO Institute for Statistics to convene the
Learning Metrics Task Force (LMTF), which will be explained below. Together, all of these
agencies play a major role in education development internationally, so recent key documents
produced by these agencies were purposefully sampled. A more detailed description of each of
the documents that comprise the sample follows.
Table 1.2
Sample
Report Title
Learning for all: Investing in
people’s knowledge and skills
to promote development:
World Bank Group education
strategy 2020

Report
Date
April
2011

Report Objective
Present the Bank’s new 10year strategy to achieve their
learning for all objective

Main Agency(ies)
Involved
The World Bank

Making education a priority in September Summarize main themes and
the post-2015 development
2013
messages that emerged from
agenda: Report of the
global consultations on
thematic consultation on
education to inform the posteducation in the post-2015
2015 education agenda
development agenda

UNESCO;
UNICEF

Toward universal learning:
Recommendations from the
Learning Metrics Task Force

LMTF from the
Brookings
Institution;

September Make recommendations for
2013
the global measurement of
learning outcomes to help
ensure quality education for
all

UNESCO
Institute for
Statistics

The earliest published document included in the sample is the World Bank’s new
Education Strategy 2020 entitled Learning for All: Investing in People’s Knowledge and Skills to
Promote Development (World Bank, 2011). This document presents the World Bank’s 10-year
strategy to achieve their learning for all objective through two strategic goals: reforming
educational systems beyond inputs and building the knowledge base for educational reform. The

8

World Bank education strategy is a highly influential document in the international development
arena that has been widely disseminated to people all over the world working in the field of
education and development (Pitman et al., 2012). The remainder of the documents were selected
to reflect recent discussions on the post-2015 global discourse.
As the deadline to achieve the MDGs and EFA goals is fast approaching at the end of
2015, UN stakeholders are engaged in discussions and consultations for a post-2015
development agenda. Numerous documents have been released on the discussions related to the
post-2015 development agenda, and two such documents are analyzed here that were published
during the time that this study was carried out. In September 2013 a report was released by the
Global Thematic Consultation on Education in the Post-2015 Development Agenda, which was
co-led by UNESCO and UNICEF. The title of the report is Making Education a Priority in the
Post-2015 Development Agenda (UNESCO & UNICEF, 2013). As part of its post-2015
development discussions, the UN is conducting global thematic consultations with academics,
the media, the private sector, employers, civil society, and other stakeholders covering eleven
areas, which include education, health, water, environmental sustainability, energy, inequalities,
governance, growth and employment, population dynamics, food security and nutrition, and
conflict, violence and disaster. The report that is analyzed came out of the education global
thematic consultation held September 2012 to March 2013.
The Brookings Institution is also playing an active role in post-2015 development
discussions. The most recent report launched in September 2013 by the Learning Metrics Task
Force (LMTF) is also analyzed. The LMTF, convened by the UNESCO Institute for Statistics
and the Center for Universal Education at the Brookings Institution, is working to ensure that
learning is a central component of the post-2015 development agenda and makes

9

recommendations for goals to improve learning opportunities and outcomes worldwide (LMTF,
2013). The summary report is tilted Toward Universal Learning: Recommendations from the
Learning Metrics Task Force (LMTF, 2013).
Analytical Approach
Discourse analysis (DA) is used to examine the attention that is paid to disability as a
form of marginalization and the way in which disability is linguistically situated in the sampled
international development education strategy documents. There is no unitary or standardized
analytical approach for DA. Instead, DA is best viewed as a shared perspective that
encompasses a variety of theoretical and methodological approaches (Phillips & Hardy, 2002;
Schiffrin, Tannen & Hamilton, 2001). For my analysis, I draw largely upon the analytical
approach and tools shared by Huckin (1997). Following Huckin’s recommendation, I
approached the text in two stages. First, I read the documents in an uncritical manner. The goal
here was to approach the text as a typical reader who is reading for comprehension. Second, I
then revisit the text in more of a critical and meticulous manner. At the second stage I
approached the text at three different levels, moving from the text as a whole, down to sentences,
and finally down to words.
While focusing on the text as a whole, Huckin (1997) recommends analyzing how the
content of the text is presented and structured, or how the text is framed. Helpful devises to pay
attention to that signify framing include photographs, diagrams, tables, and headings. As part of
my analysis I focused on identifying how disability was framed within the international
education strategy documents. Other elements to consider when analyzing the text as a whole
include foregrounding/backgrounding. These concepts refer to the author’s emphasis or

10

diminishing of certain factors. As part of my analysis I will attempted to see whether disability
was emphasized or deemphasized in the text.
Moving to more minute levels of analysis, I used additional tools recommended by
Huckin (1997) when analyzing text at the sentence and word levels. One such tool is
topicalization. Writers influence readers’ perceptions of importance with their decision of where
to grammatically place the topic of a sentence. As part of my analysis, I made note of where the
topic of disability falls within sentences. Additionally, sentences convey information about
power relations. In sentences that mention disability, I paid attention to who is depicted in power
over whom, which is referred to as agency. Finally, individual words carry with them certain
connotations. When analyzing the word ‘disability’ in the documents I aimed to decipher which
connotations that specific word carries in the text. Together, applying these analytical tools to
my discourse analysis of select international education strategy documents helped me closely
examine the presence and representation of disability to understand the ideological meanings
behind the text.
Results
Discourse analysis results are presented below for each research question. Throughout
the results comparisons are made between the three international education development strategy
documents concerning the presence of disability and the ways in which disability is linguistically
situated.
RQ1: Presence of Disability in International Development Education Strategy Discourse
Disability is explicitly mentioned in all three of the international development education
strategy documents. Figure 1.1 illustrates that the UNESCO and UNICEF document mentions
disability the most (disability word count = 15), followed by the Word Bank document

11

(disability word count = 6), and the LMTF document mentions disability the least (disability
word count = 2). Note that in order to capture all possible stems of the word disability, these
frequency counts include the number of times that the words disability, disabilities, or disabled
are mentioned.
20

# of Times Disability in Text

18
16
14
12
10
8
6
4
2
0
LMTF (2013)
UNESCO & UNICEF (2013)
Word Bank (2011)
International Education Strategy Document

Figure 1.1. Number of times disability is explicitly mentioned in the main text of each document.
Given that the documents differed in length, in addition to examining the raw word count,
it is also important to examine the relative presence of disability within the text as a whole.
Table 1.3 reports the ratio of the number of times that disability is mentioned relative to the total
number of words in each document. When document length is taken into account, the presence
of disability is comparable between the LMTF and World Bank documents with their respective
disability to total words ratios of 0.00018 and 0.00021. The presence of disability is slightly
higher in the UNESCO and UNICEF document, with a ratio of 0.00086.

12

Table 1.3
Presence of the Word Disability Relative to Total Words in Each Document
Disability Word
Count

Total Words

Ratio of Disability to
Total Words

LMTF (2013)

2

10,988

0.00018

UNESCO & UNICEF (2013)

15

17,382

0.00086

Word Bank (2011)

6

28,596

0.00021

Report

Note. Disability word count includes all possible stems of the word disability, notably disability,
disabilities, and disabled. Total word count includes all words in the main body of the text.
Words on the title page, table of contents, abbreviations page, references, and appendices were
not included in the total word count.
Overall, although disability is explicitly mentioned in all of the international education
development strategy documents that were analyzed, its presence is quite limited. This limited
presence is reflected in the small number of times that disability is mentioned in each document
overall and in relation to the total word count. Moving to the next stage of analysis, the ways in
which disability is linguistically situated in the documents when it is present are explained
below.
RQ2: How Disability is Linguistically Situated in International Development Education
Strategy Discourse
Results concerning how disability is linguistically situated in the international education
development strategy documents are divided by the textual level of analysis to illustrate the
nuanced insights that were found at each layer of analysis. First, findings at the macro level are
presented when the three texts were analyzed as a whole. Second, findings at the more minute
levels of analysis (i.e., at sentence and word levels) are presented.
Representation of disability in the text as a whole. When approaching the text as a
whole, I analyzed how disability is framed within the documents and whether disability is
13

emphasized or deemphasized in the text given its textual prominence. I found that disability is
primarily linguistically situated in the background of the text as a whole. This was particularly
true in two of the three documents (those by the LMTF and the World Bank).
Disability in the background. The overarching frame of the LMTF and World Bank
documents was educational quality for all children. Although disability was briefly
acknowledged in both documents as falling under the all umbrella, it was not emphasized. To
explain, the structure of the LMTF document begins by calling attention to the global learning
crisis, then presents seven recommendations to improve the quality of education for children
worldwide, and closes with a call to action. Disability is mentioned in the section that discusses
Recommendation 5: Equity, more fully stated as, “Measurement of learning must include a focus
on equity” (p. 31). Within this section disability is framed as a marginalized group, as a
“sociodemographic” (p. 32) source of educational “disparities within countries” (p. 31) that
should be measured and whose “differences in learning opportunities” (p. 31) should be
addressed. Disability is represented as one sociodemographic dimension among many others
including sex, age, race, urban or rural residence, socioeconomic status, mother tongue,
ethnicity, citizenship, and emergency situations – all of which characterize the “most [emphasis
added] marginalized children and youth” (p. 32). Besides being included in the equity section of
the document, disability does not surface in any of the other six recommendations, any tables or
figures, or any headings or subheadings. The main message of the document is to highlight the
need for improving the learning of children and youth around the world and that global
measurement is an important mechanism to meet this cause. Disability is in the background
under the banner of equity along with numerous other marginalized groups.

14

The structure of the World Bank document follows a similar pattern by first identifying
the importance of improved education for development, and then proceeds to map out their new
education development strategy to improve the learning outcomes of all children. An additional
section is devoted to reviewing lessons learned from past World Bank Group development work
in education, and then the document closes by specifying what they term “implementation
levers” for the new strategy. In contrast to the LMTF document, in the World Bank document
disability appears in more than one section. However, like in the LMTF document, disability is
framed as one of many marginalized “disadvantaged populations” (pp. 10, 78) whose educational
progress is lagging. Disability is most often presented in a long list of other marginalized,
disadvantaged population groups. More specifically, disability is listed as a population group
that faces “barriers” (p. 5) to educational attainment and is listed as a reason for why youth “drop
out of school early” (p. 26) in the body of the text. Structurally, not unlike the LMTF document,
disability is not in any figures, headings or subheadings in the World Bank document. However,
disability is in a table that is included twice in the document (once in the Executive Summary on
page 10 and again in the final section of the document on page 78). The table is titled
Performance, Outcome, and Impact Indicators for the 2020 Education Strategy and disability
occurs once in brackets as an example of a disadvantaged population alongside income groups,
gender, and ethnolinguistic groups in the Impact Indicators column.
On first glance it appears that disability has slightly more textual prominence in the
World Bank document compared to the LMTF document because it is in a table. Huckin (1997)
noted that visual aids such as tables point to framing and textual prominence when analyzing the
text as a whole, so the inclusion of disability in a table in the World Bank document should
theoretically signify its emphasis in the text. However, I argue that disability is still primarily in

15

the background due to its omission in the body of the text that corresponds to the portion of the
table where it is listed. In the context of the table, disability is an example of a disadvantaged
population group whose reduced schooling and learning gaps should be measured as an impact
indicator. Yet, in the larger body of the text disability is entirely absent from all discussions
around educational measurement for knowledge generation, which is a key priority of the
Education Strategy (Priority 2: Building a High-Quality Knowledge Base to Underpin Education
Reforms). For instance, considerable attention is paid to discussing the establishment and
improvement of countries’ education management and information systems (EMIS).
Furthermore, the issue of educational measurement is relevant to the multiyear System
Assessment and Benchmarking for Education Results (SABER) program that is promoted and is
also central to all three knowledge implementation levers: (1) system assessment and
benchmarking tools, (2) learning assessments, and (3) impact evaluations and analytical work.
Thus, educational measurement is a recurring theme in the World Bank text as a whole, but
disability is omitted from this larger dialogue. Huckin (1997) asserts that omission is the
ultimate form of backgrounding, which is relevant to my analysis here because the omission of
disability in the text discussing measurement in essence overshadows any textual prominence
that is afforded by its presence in the table.
On the whole, in both the LMTF and World Bank documents, disability is never
discussed in detail. Instead, it is briefly mentioned as an educationally disadvantaged
sociodemographic category along with other marginalized population groups such as low-income
children, ethnolinguistic minorities, girls, children displaced by conflict, and children living in
rural areas. Because disability is never brought to the forefront within the larger structure and
frame of these documents, it is deemphasized in the texts as a whole.

16

Disability emerging from the background. In contrast to the World Bank and LMTF
documents, disability has more textual prominence in the UNESCO and UNICEF document.
The reason for this notable difference is that inequality is a central theme in the UNESCO and
UNICEF document, as opposed to being on the periphery in the other documents. Disability has
more of a natural fit under this larger inequality frame focused on a human rights-based approach
to education, so it is not surprising that disability is featured more prominently throughout this
document. Disability is mentioned in nearly all sections of the document, and although disability
is not explicitly in any tables, figures, headings or subheadings, it is showcased throughout a
subsection titled Inclusion in Section 5.4 Cross-cutting issues. It is worth noting that the other
two subsections in Section 5.4 are Gender and Emergencies, so girls and children who are
displaced by conflict or natural disasters do get more attention structurally in the document than
children with disabilities. Also, children with disabilities share the Inclusion subsection with
numerous other marginalized and vulnerable groups (those living in remote and rural contexts,
ethnic groups, indigenous peoples, refugees, migrants, internally displaced people, children in
hazardous work or armed forces, those living with HIV, without parental care, and in
institutions).
Another structural detail that signifies more foregrounding of disability in this document
compared to the others is that disability is mentioned in a highlighted textbox in Section 5.3
Quality Education titled Frequently Mentioned Obstacles to Good-Quality Education under the
bullet labeled Inequity. Here, attention is called to “discrimination against marginalized and
socially excluded groups, and failure to include and respond to the needs of children and young
people with disabilities” (p. 23) as a barrier to quality education. Children living in rural areas

17

and areas of conflict, political instability, and disasters are also highlighted in this bullet point,
but nonetheless disability is explicitly highlighted.
A final piece of evidence concerning the greater emphasis on disability in the UNESCO
and UNICEF document is that when the challenge of inequality is introduced in Section 2.0
Progress and Gaps in Education, readers learn that special attention was paid to disability during
the educational consultations for the post-2015 development agenda. This purposeful attention
to disability is highlighted in its own paragraph at the end of the subsection:
In the educational consultations, particular attention was given to children and adults with
disabilities, who have not been systematically included in development planning, policies
or budgeting. Estimates suggest that 30 per cent of all primary age children who are out
of school are children with disabilities. (p. 11)
While it is clear that disability has more textual prominence in the UNESCO and
UNICEF document compared to the LMTF and World Bank documents, disability still remains
somewhat in the background when compared to other disadvantaged groups and due to its
omission in some areas of the body of the text. For example, in a small subsection Linking
Education to Employment, disability is listed alongside other marginalized groups (gender,
location, wealth, ethnicity, migrant status, sexual orientation, and age) as a marker of inequity
that should be measured and disaggregated. However, disability is not included in other sections
of the text that are focused on measurement like Section 4.3 Accountability, the Developing
Education Targets and Indicators subsection, and the Global vs. National Outcome Measures
subsection. Additionally, disability was omitted from the Learning Environment subsection.
Gender, on the other hand, was included by mentioning the need for “separate toilets in sufficient

18

numbers for boys and girls” (p. 28). Disability is relevant to this discussion about toilet
accessibility and safety, but it is absent from the text.
Lastly, although disability is emphasized in the text as a whole more in the UNESCO and
UNICEF document compared to the other two, disability is framed the same way across all three
documents: as a marginalized group faced with educational inequalities. The language
contextualizing disability in the UNESCO and UNICEF document mirrors that of the LMTF and
World Bank documents by couching disability as one “marginalized and vulnerable group” (pp.
16, 30, 38) among many others. The UNESCO and UNICEF document does go one step further
than the others in its conceptualization of marginalization by referring a few times to the
historical and structural nature of inequalities that lead to the marginalization of particular groups
by referring to “exploitive economic/social relations and inequitable social structures” (p. 30)
and “discrimination” and social exclusion (p. 23). The UNESCO and UNICEF document also
provides slightly more detail about disability by mentioning the use of information and
communication technologies (ICT), “such as talking computers – to assist people with
disabilities” (p. 28), and alluding to diversity within disability as a group by specifying “learning
and physical disabilities” (p. 30). However, like in the LMTF and World Bank documents, the
majority of the time disability is mentioned in a very broad sense as an example of a
marginalized population group.
Representation of disability in sentences and words. Turning to the more minute
levels of the text, I analyzed where disability falls within sentences, the agency of people with
disabilities within sentences, and the connotation of the word disability. Figure 1.2 shows that

19

the word disability most commonly falls in the middle of a sentence listing marginalized
categories.1

% of Time Disability Falls at Each Position

100
90
80
70
60
50
40
30

20
10
0
Beginning

Middle
Position of Disability in Sentence

End

Figure 1.2. Percentage of time the word disability falls at the beginning, middle and end of a list
of marginalized categories within a sentence.
Only once was disability listed as the first marginalized category within a sentence,
which occurred in the UNESCO and UNICEF document. Twice disability fell at the end of a list
of marginalized categories within a sentence: once in the UNESCO and UNICEF document and
once in the World Bank document. All other times disability fell in the middle of a list of
marginalized categories within a sentence, which amounted to 71.43% of the time. Thus, across
the documents the topic of disability was rarely emphasized in a list of marginalized categories
within a sentence by being placed at either the beginning or the end of the list to draw readers’

1

Note that the ordering of marginalization categories was checked to confirm that the placement of disability did not
appear to be due to alphabetization.

20

attention to it. Instead, it appeared as if the topic of disability was hidden among other
marginalized categories.
Turing to agency, people with disabilities were presented as passive recipients of aid in
the context of the sentences where disability was mentioned in the documents. In other words,
people with disabilities were not represented in the text as active agents. This finding is drawn
from the way that disability was commonly framed as a marginalized, disadvantaged group
across the international education development strategy documents. In this regard, people with
disabilities are represented as victims of educational inequality, implying a sense of vulnerability
and powerlessness. In fact, the word “vulnerable” was repeatedly used when framing disability
as a marginalized group in the UNESCO and UNICEF document (pp. 11, 16, 28, 30, 39). An
additional element to consider regarding sentence-level power relations is what Huckin (1997)
refers to as “the agent-patient relationship” in the case of these documents. I argue that disability
is the patient and international development agencies are the agents because international
development agencies are represented as initiating actions (and thus exerting power) over the
disabled population in the documents. For example, let us consider the following sentence from
the World Bank document: “This goal will require lowering the barriers that keep girls, people
with disabilities, and ethnolinguistic minorities from attaining as much education as other
population groups” (p. 5). Although the agent who will lower barriers for people with
disabilities and other marginalized groups is not explicitly mentioned in this sentence, because
the text is talking about the World Bank’s 2020 Education Strategy goals, we can infer that the
World Bank and its international development partners serve as the active agents here.
At the word level, disability carried a negative connotation. The words that were
commonly used to frame disability across the documents including “marginalized,” “vulnerable,”

21

and “disadvantaged” are all negative words themselves, and this negative connotation carries
over to the word disability. Throughout the international education development strategy
documents, disability was never presented in a positive light. Instead, the word disability served
as a sociodemographic marker of disadvantage and inequality.
Discussion
Implications and Recommendations
The results of this document analysis have numerous implications for the post-2015
education development agenda. First and foremost, the limited presence of disability across the
documents, the fact that disability largely remains in the background of the text as a whole, and
the way in which disability is hidden among other categories of marginalization within sentences
calls to attention the lack of emphasis that international education development agencies are
placing on disability in their post-2015 vision of education development work. This has clear
implications for children with disabilities, as so much development work and research is based
upon the goals and objectives that are communicated in these types of documents. While it is
encouraging that disability has more textual prominence in the UNESCO and UNICEF document
compared to the LMTF and World Bank documents, this document analysis does by no means
demonstrate that disability is at the absolute forefront of the discourse.
While the UNESCO and UNICEF document brings disability out of the background
slightly more than the other documents, there is still room for improvement. In all documents,
details about the intricate meaning of disability and about the unique learning needs of students
with disability are largely absent. Instead, disability is oversimplified as a sociodemographic
marker of marginalization. The documents begin to scrape the surface by including disability in
varying degrees, with UNESCO and UNICEF doing a better job than the LMTF and the World

22

Bank in terms of the emphasis placed on disability as a marginalized population group.
However, a deeper discussion of what disability means and how disability is difficult to measure
is missing from all of the documents. Although in reality space is limited in these documents to
add details such as this, I feel that some additional explanation is warranted due to the unique
complexity of disability. To explain, the other population groups that disability is commonly
grouped together with are in many ways easier to understand and measure. For example, a
child’s gender is essentially a biological trait and their income level is computed using parental
salary amounts. While I acknowledge that these constructs can be nuanced to some degree as
well, such as challenges classifying the gender of transgender children and difficulties obtaining
accurate parental salary data in developing economies, the construct of disability is arguably
more complex to define and measure and these nuances will be elaborated upon below.
There are multiple reasons that it is difficult to define and measure disability in general
and cross-nationally. First, disability is not monolithic – it is a heterogeneous construct with
multiple forms or categories existing under the general category of disability. In a given
education system a child is rarely designated as disabled or not. Instead, he or she is labeled with
a certain type of disability such as Autism Spectrum Disorder (ASD), Attention Deficit
Hyperactivity Disorder (ADHD), Learning Disability, Visual Impairment, Hearing Impairment,
or Intellectual Disability. The representation of disability as a uniform construct in the
documents analyzed here is an oversimplification of a highly complex category of
marginalization.
An added layer of complexity for international educational development is the challenge
of working across multiple contexts and cultures. Different national educational systems each
have their own policies and procedures for disability identification and classification. What

23

officially “counts” as a disability is different in different countries. To illustrate how such
differences might play out with an example, a child in the U.S. who acts out behaviorally in class
may be classified as Emotionally Disturbed, but in another context that same child may be given
a different disability classification or may not be classified as having a disability at all. Thus,
there are cross-cultural differences in disability classification that are not reflected in the
international education strategy documents. An acknowledgment of these sorts of complexities
is important to provide readers with a more accurate understanding of disability as a
marginalized group.
Another aspect to consider is the increased recognition of the social element of disability.
Since the 1990s there has been a theoretical shift away from the ‘medical model’ where
disability was solely viewed as an abnormal ailment that exists within an individual towards the
‘social model’ where disability is viewed as being caused by discriminatory societal attitudes and
environmental barriers (Lindsay, 2003; Oliver, 1996; Thomas, 1999). While there is no globally
agreed upon definition or measurement of disability, the conceptual framework and classification
system of the World Health Organization International Classification of Functioning, Disability
and Health (ICF) (WHO, 2001) is becoming increasingly recognized as the gold standard. There
is clear overlap between the social model of disability and the WHO ICF conceptualization of
disability, as the WHO ICF recognizes disability as the interaction between a person with a
health condition and environmental contextual factors. However, the applicability of this
classification system to non-Western contexts is uncertain. For example, in her study on the
conceptualization of disability in Ghana, Anthony (2011) asserts that neither the medical model
nor the social model adequately explain the experience of disability in Ghanaian culture. The
cultural relevance of Western ways of conceptualizing and measuring disability should be

24

critically examined when developing international education development strategies that
encompass non-Western contexts.
Finally, given the larger context and purpose of these documents to inform the post-2015
education development agenda, it is not surprising that people with disabilities were presented in
the text as passive recipients of aid or that the word disability carried with it a negative
connotation. Nonetheless, if people with disabilities are not mentioned as a potential resource to
strengthen the economy and communities, then this is akin to the classic search for pathology or
deficit-driven thinking that traditionally characterizes the medical model of disability. The
reductionist deficit-driven mentality of special education is classically defined and critiqued as
spending a considerable amount of time and resources on seeking and remediating deficits
(Poplin, 1984a, 1984b, 1988). Deficit-driven thinking runs counter to the social justice ideology
of inclusion, which seeks profound social change where difference is actually celebrated (Barton,
1997, p. 234). It is problematic if organizations that are strong proponents of inclusion like
UNICEF and UNESCO may actually perpetuate more deficit-driven perspectives of people with
disabilities as opposed to more positive or empowering strength-based perspectives.
There were notable instances of lost opportunities to represent people with disabilities as
having more agency and in a more positive light in the documents. For instance, all three
documents included a discussion about shared responsibility through strategic partnerships
between multiple stakeholders towards achieving quality education for all. To illustrate with an
example, the UNESCO and UNICEF document argues that the involvement of “civil society
organizations, including local NGOs…must be expanded to lobbying and influencing
educational policy and improving inclusive, quality education” (p. 17). The involvement of
disabled people’s organizations (DPOs) could be added to this discussion to show that there is an

25

important role to be played by the people who are experiencing educational inequalities and
marginalization themselves, as opposed to outside advocates only. DPOs are comprised of
people with disabilities, and recommending their involvement in post-2015 education
development initiatives is one way that people with disabilities could be portrayed more as
active, positive agents. While DPOs technically falls under the NGO umbrella term, most
readers would not make this connection, and specifically including DPOs in the text would bring
attention to people with disabilities as active agents. Including people with disabilities as part of
the strategic partnerships that the documents refer to is important to have a shared distribution of
power and agency in the planning and development process.
Limitations and Future Directions
Looking forward, there are numerous extensions of this research that can address some
limitations of this document analysis. The primary limitation of this research is the number of
documents that were analyzed. The three documents analyzed here were selected as a sample
that captures the perspectives of key international education development agencies. As a result,
documents produced by international development agencies outside of the World Bank, LMTF,
UNESCO, and UNICEF were not included. This study provides a valuable starting point and
future work could expand the breadth of the current document analysis by analyzing additional
documents published by other international development agencies.
A related limitation is the timely nature of the documents and wider discourse concerning
the post-2015 development agenda. The engagement of multiple stakeholders in discussions and
plans for their post-2015 development agenda is ongoing, and this document analysis is limited
to the discourse as it stood when the documents were released (April, 2011 and September,
2013). International education development is a dynamic, evolving process. This document

26

analysis provides insight into the early stages of the post-2015 international education
development process, with the hope that my findings might in some small way inform the
process as it continues to advance. More recently published as well as future international
education development strategy documents should be analyzed to continue to monitor and
critique the presence and representation of disability.
Finally, the focus of this essay was limited to analyzing the presence and representation
of disability in the documents. Disability is one of many marginalized groups in education, and
future research should document the presence and representation of other marginalized groups in
post-2015 international education development documents. For example, the rural poor, young
girls from disadvantaged homes, slum dwellers, and pastoralists are identified as the most
marginalized children in the learning crisis in Africa (Watkins, 2013). Similar analyses to the
one done here but focusing on a different marginalized group would provide an interesting point
of comparison. Moreover, it would be valuable to compare and contrast the presence and
representation of multiple marginalized groups in such documents in a relative manner. For
instance, future work should consider if disability is proportionately under or over attended to
compared to other sociodemographic categories of marginalization such as ethnolinguistic
minorities, girls, or children displaced by conflict. Additionally, an alternative way to extend this
work is to analyze how disability (compared to other marginalized categories) is related to
broader terms reflective of marginalization such as poverty, under-represented populations,
disadvantaged learners, or at-risk learners.
Conclusion
Together, this document analysis provides insight into the presence and representation of
disability in international education development discourse. The analysis shows that although

27

disability is explicitly mentioned in all three of the international development education strategy
documents, disability is linguistically situated in the background and is underemphasized for the
most part. Furthermore, the document analysis demonstrates that disability is conceptualized as
a vulnerable marginalized population group in the text as a whole, disability is often hidden in
the middle of a list of other marginalized groups in sentences, people with disabilities are not
depicted as active agents, and the word disability holds a negative connotation. Until disability
more fully moves into the spotlight in the international education development discourse,
children with disabilities will continue to be marginalized in future documents and in turn in
international education development efforts. Additionally, the implications and
recommendations of this analysis suggest the value in recognizing the conceptual and
measurement challenges associated with disability in light of inconsistencies in the definitions of
disability within and across countries, as well as the heterogeneity of the socially and culturally
bound concept of disability.
Overall, this essay highlights that it is not only important to carefully consider the
emphasis that is placed on disability in international education development discourse, but also
the ways in which disability is linguistically situated and represented. Representing disability in
a more nuanced and even empowering way may help increase the likelihood that the unique
needs of children with disabilities are adequately taken into account in future international
education development conversations and initiatives.

28

REFERENCES

29

REFERENCES

Adams, A. M. (2012). The education link: Why learning is central to the post-2015 global
development agenda (Working Paper 8). Washington, DC: The Brookings Institution.
Ahmed, M. (2014). Squaring the circle: EFA in the post-2015 global agenda. International
Journal of Educational Development, 39, 64-69.
Anthony, J. (2011). Conceptualizing disability in Ghana: Implications for EFA and inclusive
education. International Journal of Inclusive Education, 15(10), 1073-1086.
Barton, L. (1994). Inclusive education: Romantic, subversive or realistic? International Journal
of Inclusive Education, 1(3), 231-242.
Huckin, T.N. (1997). Critical discourse analysis. In T. Miller (Ed.), Functional approaches to
written text: Classroom applications (pp. 78-92). Washington, DC: United States
Information Agency.
Klees, S. J., Samoff, J., & Stromquist, N. P. (Eds.). (2012). World Bank and education: Critiques
and alternatives. Rotterdam, NLD: Sense Publishers.
Learning Metrics Task Force. (2013). Toward universal learning: Recommendations from the
Learning Metrics Task Force. Montreal and Washington, DC: UNESCO Institute for
Statistics and Center for Universal Education at the Brookings Institution.
Lindsay, G. 2003. Inclusive education: A critical perspective. British Journal of Special
Education, 30(1), 3–12.
McGrath, S. (2014a). Education and the post-2015 international development agenda.
International Journal of Educational Development, 39, 1-3.
McGrath, S. (2014b). The post-2015 debate and the place of education in development thinking.
International Journal of Educational Development, 39, 4-11.
Oliver, M. 1996. Understanding disability: From theory to practice. Basingstoke: Macmillan.
Phillips, N., & Hardy, C. (2002). Discourse analysis. Thousand Oaks, CA: Sage.
Poplin, M. (1984a). Summary rationalizations, apologies and farewell: What we don’t know
about the learning disabled. Learning Disability Quarterly, 7(2), 130-134.
Poplin, M. (1984b). Toward a holistic view of persons with learning disabilities. Learning
Disability Quarterly, 7(4), 290-294.

30

Poplin, M. S. (1988). The reductionist fallacy in learning disabilities: Replicating the past by
reducing the present. Journal of Learning Disabilities, 21(7), 389-400.
Schiffrin, D., Tannen, D., & Hamilton, H. E. (2001). Introduction. In D. Schiffrin, D. Tannen, &
H. E. Hamilton (Eds.), The handbook of discourse analysis (pp. 1-10). Malden, MA:
Blackwell.
Steiner-Khamsi, G. (2012). For all by all? The World Bank’s global framework for education. In
S. J. Klees, J. Samoff, N. P. & Stromquist (Eds.), World Bank and education: Critiques
and alternatives (pp. 3-20). Rotterdam, NLD: Sense Publishers
Thomas, C. 1999. Female forms: Experiencing and understanding disability. Buckingham: Open
University Press.
United Nations. (2011, August 3). Status of the convention on the rights of the child: Report of
the secretary-general (UN Report No. A/66/230). Retrieved from the UNICEF website
http://www.unicef.org/disabilities/files/Issue48-12-A66150.pdf
United Nations Educational, Scientific and Cultural Organization. (2010). Education for all
global monitoring report 2010: Reaching the marginalized. Paris, France: Author.
United Nations Educational, Scientific and Cultural Organization, & United Nations Children’s
Fund. (2013). Making education a priority in the post-2015 development agenda: Report
of the thematic consultation on education in the post-2015 development agenda. Paris
and New York: Authors.
United Nations Millennium Project. (2005). Investing in development: A practical plan to
achieve the millennium development goals. New York: United Nations Development
Programme.
United States Agency for International Development. (2011). Education opportunity through
learning: USAID education strategy 2011-2015. Washington, DC: Author.
Van Fleet, J. W., Watkins, K., & Greubel, L. (2012). Africa learning barometer. Retrieved
from Center for Universal Education at
Brookings website: http://www.brookings.edu/research/interactives/africa-learningbarometer
Watkins, K. (2013). Narrowing Africa’s education deficit. In the Brookings Africa Growth
Initiative, Foresight Africa: Top priorities for the continent in 2013 (pp. 17-21).
Washington, DC: Brookings.
World Bank. (2011). Learning for all: Investing in people’s knowledge and skills to promote
development: World Bank Group education strategy 2020. Washington, DC: Author.

31

World Health Organization. (2001). International classification of functioning, disability and
health. Author: Geneva, Switzerland.
World Health Organization. (2011). World report on disability 2011. Malta: WHO Press.

32

CHAPTER 2. ACADEMIC ACHIEVEMENT OF MARGINALIZED POPULATIONS IN SUBSAHARAN AFRICA: THE BUFFERING EFFECT OF BELONGINGNESS
Introduction
The attention paid to advancing the academic achievement of marginalized populations
has been increasing globally. International campaigns such as the first Education for All (EFA)
conference, held in Jomtien, Thailand in 1990 have played a large role in building this
momentum. The United Nations Scientific and Cultural Organization (UNESCO) EFA goals
call on all world nations to improve the quality of education so that measurable learning
outcomes are achieved by all children. Furthermore, the EFA goals explicitly highlight the need
to focus on what they refer to as “the most vulnerable and disadvantaged children,” and identify
girls and ethnic minorities as marginalized populations that deserve special attention (UNESCO,
1990).
Yet despite this increased attention to improving the academic achievement of
marginalized student populations, limited progress has been made. The most recent EFA global
monitoring report detailed how progress towards many of the goals has been slowing down and
projected that most goals will remain unmet, particularly in sub-Saharan Africa (UNESCO,
2012). For significant progress to ever become possible in the advancement of the academic
achievement of marginalized groups, it is essential that researchers identify and understand
factors that can be modified or changed to break the cycle of marginalization and poor academic
outcomes. Unfortunately, a child’s membership to a particular marginalized group can often not
be changed (e.g., being a girl). However, their feelings of belongingness can be changed.
Feelings of belongingness in relation to marginalization and academic achievement constitute a

33

gap in existing research that has important implications for policy and practice, particularly in
developing contexts where marginalization and poor academic performance is widespread.
This study addresses this gap by investigating the relationship between feelings of
belongingness, multiple forms of marginalization, and student achievement across three subSaharan African countries using eighth grade data from the Trends in International Mathematics
and Science, 2011 (TIMSS, 2011). Specifically, marginalization on the basis of a child’s gender,
socioeconomic status (SES), home language, and immigrant status is examined in Botswana,
Ghana, and South Africa. Student achievement in both mathematics and science is explored
because competence in mathematics and science is critical in today’s competitive global
innovation economy, particularly for developing economies. In addition to examining the direct
effects of feelings of belongingness and marginalization on academic achievement, the central
component of this analysis investigates whether feelings of belongingness buffers2 the negative
impact of marginalization on academic achievement in mathematics and science.
Literature Review
The following literature review is divided into three major sections. The first section
reviews literature on the need to belong with a focus on the relationship between feelings of
belongingness and academic achievement. The second section reviews literature on
marginalization in education with a focus on the relationship between marginalization and
academic achievement. Literature on the academic achievement of multiple marginalized
student populations is summarized and is divided into specific subsections for marginalization on
the basis of a child’s gender, SES, home language, and immigrant status. Finally, the third

2

Buffering refers to a type of moderation where the effect of the primary independent variable on the dependent
variable is weakened as the moderator increases (West & Aiken, 1991).

34

section reviews literature specific to the sub-Saharan African context which demonstrates why
sub-Saharan Africa is a suitable context to study marginalization.
The Need to Belong
The need to belong has been established by empirical research in the field of psychology.
In their review of empirical research on this topic, Baumeister and Leary (1995) tested what they
referred to as the ‘belongingness hypothesis,’ which posits that the need to belong is a
fundamental human motivation. After reviewing the literature, Baumesiter and Leary concluded
that the need to belong is indeed central to human wellbeing. A large body of literature also
exists on the negative consequences of having this core social need threatened in the form of
social exclusion or ostracism (see reviews by Williams, 2007; Williams & Nida, 2011).
Feelings of belongingness and academic achievement. While the connection between
feelings of belongingness to psychological wellbeing has been well established by research, less
is known about the link between feelings of belongingness and academic achievement. Research
suggests that it is important for schools to consider the effects of belongingness on academic
achievement (Johnson, 2009). However, there are inconsistencies in the literature concerning the
exact nature of the relationship between feelings of belongingness and academic achievement.
To explain, many correlation studies show that children who are preferred by their
teachers and peers tend to be high achievers while those who are rejected tend to be low
achievers (Green, Forehand, Beck, & Vosk, 1980; Katz, 2013; Ladd, 1990; Swift & Spivack,
1969; Taylor, 1989; Wentzel & Asher, 1995). Whether this commonly found positive
association between feelings of belongingness and academic achievement is due to a direct or
indirect effect is less clear, however. Most research assumes an indirect effect, and there is a
substantial amount of corresponding literature demonstrating that feelings of belongingness are

35

related to achievement indirectly through engagement or motivation (see Anderson & Freeman,
2004; Osterman, 2000 for reviews).
Conversely, other studies assume a direct effect and have yielded mixed results regarding
the direction of the relationship. For example, despite hypothesizing a positive direct
relationship, Kennedy and Tuckman (2013) found a significant negative relationship between
college students’ grade point average and their perceived level of belongingness. In contrast,
Anderman (2003) found a positive relationship between feelings of belongingness and
achievement, albeit in the opposite direction. Instead of being positioned as a predictor, the
dependent variable of interest in Anderman’s study was feelings of belongingness, meaning that
in her model grade point average positively predicted school belongingness.
Furthermore, previous literature has shown that different marginalized groups within
schools have different academic and social experiences. Associations between feelings of
belongingness and academic engagement and/or achievement have been suggested in research on
the perceived educational experiences of students of different races, ethnicities, and immigrant
statuses (e.g., Berhanu, 2005; Booker, 2007; Clark, Mercer, Zeigler-Hill, & Dufrene, 2012;
Georgiades, Boyle, & Fife, 2013). In spite of this, the buffering effect of feelings of
belongingness on the relationship between marginalization and academic achievement has yet to
be empirically tested. As will be explained in more detail in the theoretical framework section
later in this essay, this study presupposes that since feelings of belongingness constitute a basic
psychological need (Baumeister & Leary, 1995), these positive feelings have the power to buffer
the negative impact that being a member of a marginalized group can have on a child’s academic
performance. My study explores the moderating effect of feelings of belongingness on the

36

relationship between marginalization and academic achievement, which is a unique contribution
of this study.
Another factor to consider regarding the relationship between feelings of belongingness
and academic outcomes is the element of diversity, that is, diversity in terms of the contexts and
student populations studied. The majority of prior research on this topic has been conducted in
Western countries. A significant contribution of this study is to explore the relationship between
feelings of belongingness and academic achievement in the understudied context of sub-Saharan
Africa.
Overall, the presence of mixed results, inconsistencies in the ways that feelings of
belongingness has been previously studied in relation to academic outcomes, and a general
neglect of diversity in participants and in context in existing literature demonstrates that
additional research is needed to more fully understand the relationship between feelings of
belongingness and academic achievement. The present study contributes to existing literature by
testing the direct effect of feelings of belongingness on academic achievement within three subSaharan African countries, as well as the buffering effect of feelings of belongingness on the
relationship between marginalization and academic achievement. The unique contributions of
this study include exploring these relationships in the understudied context of sub-Saharan
Africa, as well as exploring the moderating effect of feelings of belongingness on the
relationship between marginalization and academic achievement.
Marginalization in Education
This study will adopt the UNESCO (2010) definition of marginalization as, “a form of
acute and persistent disadvantage rooted in underlying social inequalities” (p. 135). The social
inequalities underlying marginalization are created and maintained by ingrained social,

37

economic, and political processes that operate throughout multiple spheres of society. The
education system is one such sphere where marginalization takes place. Social reproduction
theory demonstrates how schooling legitimizes and reproduces social, cultural, and economic
divisions and classes (Anyon, 1981; Aronowitz & Giroux, 1985; Bowles & Gintis, 1976; Lareau,
1987). Furthermore, this study shares the conviction held by Amartya Sen (2009) that
marginalization in education is a “clearly remediable injustice.” The same social, economic, and
political processes that create marginalization and injustice also have the power to reduce or
eliminate it. This viewpoint is consistent with Durkheim’s (1969) theory of social change which
asserts that educational change is only possible if the social structure in which the social
institutions of education are located changes. Thus, marginalization in schools can be changed if
the greater social context is also changed.
Marginalization and academic achievement. Not surprisingly, marginalization
generally has a negative impact on academic achievement. Returning to the definition of
marginalization stated above, if a child experiences acute and persistent disadvantage rooted in
social inequalities, their disadvantage will manifest itself in multiple forms across different social
structures. Within the educational system, academic achievement is the hallmark indicator of
success. So, if you are marginalized, than your ability to become academically successful will be
inherently constrained by the social, economic, and political forces that collectively create and
sustain your disadvantage. Literature illustrating the link between marginalization and academic
disadvantage will be reviewed for multiple marginalized groups below. It should be noted that
existing literature on marginalization in education is vast and diverse, and the literature reviewed
here will be limited to marginalization on the basis of a child’s gender, SES, home language, and
immigrant status because these are the categories of marginalization that were explored in this

38

study. A challenge of research on marginalization in education is that it is difficult to measure
exhaustively because national data are rarely detailed enough to identify all the different types of
marginalized groups (UNESCO, 2010). Fortunately, the TIMSS 2011 data provides multiple
cross-country markers of marginalization, all of which are analyzed in this study. 3
Gender. Although girls are not always an educationally marginalized group and much
progress has been made towards gender parity, in many countries being born a girl still carries
with it a significant education disadvantage. Evidence of this disadvantage is reflected in the fact
that 54% of out-of-school children are girls (UNESCO, 2010). This means that in many
countries girls are less likely than boys to attend school. In terms of differences between primary
and secondary school enrollment, 68 countries have not achieved gender parity in primary
education, while 97 countries have not achieved gender parity in secondary education
(UNESCO, 2012). So, the education disadvantage of girls increases with age. For girls who are
fortunate enough to go to school, girls generally tend to perform as well as, or better than their
male classmates (UNESCO, 2012). Thus at first glance, the issue of access appears to be
characterized by greater inequalities compared to academic achievement when it comes to
gender.
However, a closer examination of achievement differences between boys and girls
reveals significant gender-based disparities by subject. A considerable amount of national and
cross-national research has found that girls tend to be ahead of boys in reading and language but
lag behind in mathematics and science (Johnson, 1996; Stephens et al., 2004). These gender
gaps in subject achievement are found to be established in pre-adolescence and are then
strengthened as students advance through their secondary schooling (Johnson, 1996). In line
3

Other large-scale international student assessment datasets such as the Progress in International Reading Literacy
Study (PIRLS) also provide multiple cross-country markers of marginalization. However, TIMSS 2011 was
selected for this study to maximize the number of sub-Saharan African countries examined.

39

with these findings, this study anticipates that boys will outperform girls in mathematics and
science in the TIMSS 2011 eighth grade data since the participants are at an age where the
gender gap is most striking in these particular subjects.
Socioeconomic status. Of all of the student factors which lead to marginalization in
education, socioeconomic status (SES) is one of the most widely researched. In the literature
SES has a long history as a correlate or predictor of academic outcomes (Coleman, 1966; Cuff,
1933; Holley, 1916; Sirin, 2005; White, 1982). Across such studies, the relationship between
SES and academic outcomes is positive, such that students of high SES have high academic
achievement and students of low SES have low achievement. Alexander and Simmons (1975)
established that the impact of SES on academic achievement is not restricted to the U.S., where
most research up to this point had been done. From the 1970s onwards, further cross-cultural
research has shown that although the extent to which SES influences academic achievement in
low-income countries may not be as large as in high-income countries like the U.S., the
relationship between SES and academic achievement in low-income countries is nonetheless
significant and is in the positive direction (Chiu, 2007; Chudgar & Luschei, 2009; Heyneman,
1976a, 1976b, 1979; Heyneman & Loxley, 1983; Nonoyama-Tarumi & Willms, 2010; Schiller,
Khmelkov, & Wang, 2002). Since all sub-Saharan African countries participating in TIMSS
2011 are designated as low-income countries (International Monetary Fund, 2013), it is expected
that a positive relationship between SES and academic achievement will be found in the present
analysis, consistent with previous literature.
Home language. Language capital, also called mother-tongue, is the set of language
skills that are acquired with minimal effort in childhood and are then strengthened in school
(Chiswick & Miller 1995; Chiswick 1991). Children who do not speak their home language in

40

school are at risk of being marginalized compared to children whose home language is consistent
with the primary language spoken at school. It is difficult for teachers and peers to communicate
with students who are non-native speakers, which can have negative social and academic
repercussions. For instance, Herbert, Hau, and Kong (2002) and Papanastasiou (2000) have
shown that students who receive instruction in or are tested in their second language perform
worse academically compared to native speakers. Additionally, an analysis of the TIMSS 2007
data by Martin, Mullis, and Foy (2008) revealed that students who reported sometimes speaking
the language of the test at home had science test scores 10% lower than students who reported
always speaking the language of the test at home, and the scores of students who reported never
speaking the language of the test at home were 20% lower. Consequently, a similar relationship
between home language and academic achievement is expected in this analysis, such that
students who speak the language of the test at home more frequently will exhibit higher
mathematics and science achievement compared to students who speak the language of the test at
home less frequently.
Immigrant status. Migrant and immigrant children constitute yet another marginalized
category of students in education. When a family immigrates to another country they are likely
to experience one of several distinct paths of adaptation which either can lead to upward or
downward social mobility (Portes & Rumbaut, 2006; Portes & Zhou, 1992; Zhou & Bankston,
1998). The role of education in the segmented assimilation paths can play an important role in
attaining middle class status through educational achievement. However, this is no easy feat and
rarely occurs before the second generation (Portes & Fernandez-Kelly, 2008; Glick & HohmannMarriott, 2007). In studies on the academic performance of immigrant students, there is
evidence of an achievement gap between students who are born in the country (i.e., native

41

students) compared to students who have immigrated to the country (i.e., immigrant students)
(Brown, 2006; Schnepf, 2004). The size of this achievement gap can be quite substantial. For
instance, a study using the Programme for International Student Assessment (PISA) data found a
1.5 year lag in achievement between immigrant students and their native counterparts (OECD,
2007). While these studies have focused on high-income countries, a similar effect of immigrant
status on academic achievement is expected in this study.
Why Sub-Saharan Africa?
Geographically, sub-Saharan Africa generally consists of all African countries that lie
south of the Sahara Desert (UN, 2013). The multifaceted economic, political, and social
challenges that are faced by this region of the world informed my motivation to focus my
analysis on Botswana, Ghana, and South Africa, and will be elaborated on below. Sub-Saharan
Africa is an important context to study the buffering effect of feelings of belongingness on the
relationship between marginalization and academic achievement for two primary reasons that
will be explained below.
Low academic achievement. First, academic achievement is relatively low in this
region compared to the rest of the world. Of all the regions of the world, sub-Saharan Africa
seems to consistently fare less well compared to others in terms of their progress towards EFA
goals. For example, in terms of universal primary education, South and West Asia and subSaharan Africa started from similar positions in 1999 with approximately 40 million primary
aged out-of-school children, and by 2008 this amount dropped by 26 million in South and West
Asia but only by 13 million in sub-Saharan Africa (UNESCO, 2012). In addition to access,
education quality indicators are also generally much lower in sub-Saharan Africa compared to

42

other regions (Dembélé & Lefoka, 2007; Johnson & Beinart, 2008; Van Fleet, Watkins, &
Greubel, 2012; Verspoor, 2008).
The global perception that sub-Saharan Africa is “behind” in terms of its educational
development is not a new one. There are many possible reasons as to why educational
development in sub-Saharan Africa seems to be lagging. One such explanation offered by Abdi,
Puplampu, and Dei (2006) is that the deleterious effect on educational development in subSaharan Africa is due to globalization and neo-colonialism. The lack of a political will and
financial problems have also been suggested as important contributors to the failure of EFA in
developing contexts (Heyneman, 2009). Regardless of the reason, there is clearly an established
need for improvement and a great potential for change in advancing the learning outcomes of
children in sub-Saharan African countries.
Widespread marginalization. Undeniably, sub-Saharan African education systems are
faced with many challenges in the context of economic hardships, conflict and HIV/AIDS,
inadequate numbers of qualified teachers, under resourced classrooms and school buildings, and
the unavailability of learning materials (e.g., Dembélé & Oviawe, 2007; Sifuna, 2007; Verspoor,
2008). These challenges are further compounded by the added challenge of broadening access
and quality learning opportunities to “hard to reach” children such as those living in remote areas
or children of nomadic families (Dembélé & Oviawe, 2007). Due to the many challenges faced
by sub-Saharan African education systems, many children in this context live on the margins of
society and experience educational inequalities.
This leads to the second reason that sub-Saharan Africa is a valuable context to focus on
in this study: the fact that marginalization is extremely prevalent and takes on many different
forms in the region. It is important for research like this study to explore the nuances involved in

43

educational marginalization because it is applicable to the lives of so many children and has the
potential to have a far reaching impact. The widespread nature of marginalization is due to the
extreme and persistent educational inequality that leaves many groups of the society
marginalized in this region of the world. Numerous populations have been identified as
marginalized or disadvantaged in sub-Saharan African education systems, such as girls, children
living in poverty, children living in rural areas, children displaced by conflict, aboriginal
children, immigrant children, children with disabilities, and ethnic, linguistic, and religious
minorities. Watkins (2013) identifies the most educationally marginalized children in Africa as
the rural poor, young girls from disadvantaged homes, slum dwellers, and pastoralists. Literature
specific to the sub-Saharan African context for each of the categories of marginalization included
in this study is briefly reviewed below to contextually situate the analysis.
Gender. In many sub-Saharan African countries, it is not uncommon to hear stories
about women and girls being subjected to abuse, exploitation, and oppression at the hands of
males (e.g., Baker, 2010; Jewkes & Abrahams, 2002; UNICEF, 2011). The economic and social
positions of women in African societies in relation to men are related to gender relations in the
household, school, labor market, and society. This is referred to as the ‘gender stratification
hypothesis’ proposed by Baker and Jones (1993) which maintains that gender differences in
achievement are closely related to the cultural and societal opportunity structures for girls and
women. Else-Quest, Hyde, and Linn (2010) recently found empirical support for this hypothesis
using the TIMSS and PISA 2003 data. Else-Quest et al. found that gender equity in school
enrollment, women’s share of research jobs, and women’s parliamentary representation were the
most powerful predictors of cross-national variability in gender gaps in mathematics.

44

Unique to the African context, Colclough, Al-Samarrai, Rose, and Tembon (2003)
explain that a notable influence on the basis and reproduction of male power is the rural-urban
migration in Africa of predominantly males whilst females remained secluded in home villages
and economically dependent. The authors go on to describe how patriarchy and capitalism also
play a role, as the economic incentives are lower to educate girls due to lower prospective
salaries despite similar qualifications, as well as higher perceived costs by parents due to the lost
labor associated with sending girls to school since girls traditionally play a larger role in
housework. In the context of poverty, girls are needed for the heavy labor required for essential
daily activities such as subsistence farming and transporting water and fuel.
Within this larger social and economic context in relation to gender dynamics, girls are
generally educationally disadvantaged in sub-Saharan Africa. This is evidenced by the fact that
the vast majority of countries where girls face severe educational disadvantage, as measured by a
gender parity index (GPI) of less than 0.90, are in sub-Saharan Africa (UNESCO, 2012). This
pattern holds true at both the primary and secondary school levels, although it is important to
highlight that the incidence of severe gender disparity is higher at the secondary level compared
to the primary level. Given that this study analyzed eighth grade data, recent GPI figures for
lower and upper secondary education for each of the three sub-Saharan African countries are
listed in Table 2.1. Based on these figures it appears that girls in Ghana have the worst access to
secondary school out of the three countries, while girls in Botswana and South Africa fare better.
Some issues related to access for girls include the absence of safe, sanitary toilet facilities as well
as being at an increased risk for sexual abuse and early pregnancy, which impact attendance and
dropout rates (Datta, Phillip, & Verma, 2009).

45

Table 2.1
GPI of Countries Included in Study
Lower Secondary

Upper Secondary

Botswana

1.05

1.09

Ghana

0.93

0.87

South Africa

1.01

1.08

Note. Figures based on 2010 data presented in the UNESCO (2012) Education for All Global
Monitoring Report. GPI = gender parity index; a measure of relative access to education for
males and females.
Regarding the academic performance in the three sub-Saharan African countries that
were explored in this study, previous research suggests that the majority of the countries exhibit
a gender gap in mathematics achievement in favor of boys. For example, the effect sizes for
gender differences in mathematics achievement computed by Etse-Quest et al. (2010) were -0.04
for Botswana, 0.19 for Ghana, and 0.02 for South Africa. Positive values represented higher
scores for boys than girls, meaning that girls seem to be educationally disadvantaged in all
countries except Botswana. However, the most recent data from the Southern and Eastern Africa
Consortium for Monitoring Educational Quality (SACMAQ) shows that boys significantly
outperformed girls in mathematics in Botswana and South Africa after controlling for all other
significant factors (Hungi, 2011). In fact, the gender gap was even larger in Botswana (-2.7)
compared to South Africa (-0.8). Note that SACMEQ does not collect data in Ghana. Due to
inconsistencies in the direction of the gender gap in mathematics achievement, further research is
needed.
Socioeconomic status. Sub-Saharan Africa is the poorest region of the world. All three
sub-Saharan African countries included in this study are classified as low-income countries, or
what the International Monetary Fund (IMF, 2013) calls ‘emerging market and developing
economies’ (EMDEs). Of the 153 countries with the EMDE designation, 45 are in sub-Saharan
46

Africa. This region only accounts for 2.5% of the world’s gross domestic product (GDP) and
exports 2.1% of the world’s goods and services, yet is home to 12.3% of the world’s population
(IMF, 2013). Annual growth in GDP is projected for 2013 for the three sub-Saharan African
countries included in this study at rates of 4.1% for Botswana, 6.9% for Ghana, and 2.8% for
South Africa (IMF, 2013). Even so, poverty continues to be a perpetual issue for this region of
the world and in these countries.
Education is reflective of economic and social inequalities inherent with poverty and
family background. As stated previously, the effect of SES on students’ academic performance
is a relationship that has been well established across the globe (Hanushek & Luque, 2002;
Wöβmann, 2003; Sirin, 2005; Ammermüller, Heijke, & Wöβmann, 2005; Engin-Demir, 2009;
Martins & Veiga, 2010). Likewise, evidence of educational disadvantage on the basis of a
child’s SES has been found in studies focused on the countries included in this study’s analysis,
whereby inequalities in school attainment and academic performance have been associated with
household poverty and parent education (Bouhilia, 2011; Frempong, Reddy, & Kanjee, 2011;
Lloyd & Hewett, 2009; Timaeus, Simelane, & Letsoalo, 2013). Children from disadvantaged
backgrounds in the countries of the present study are not exempt from the risks of being less
successful in school. It is anticipated that this pattern will hold true in the present study as well.
Home language. Due to the colonial history of Africa, the languages of Europe (English,
French, Dutch, Spanish, and Portuguese) have been and continue to be the languages of power
(Brock-Utne & Holomarsdottir, 2004; Alexander, 2000). In modern day Africa, knowing the
language of the conqueror is a marker of privilege for the middle and upper class, meaning that
the language issue is a class issue. Table 2.2 lists the languages of all three countries included in
the present study. The countries’ languages are divided according to those of past colonizers

47

versus vernacular or native languages to illustrate that colonial languages still hold a prominent
position in the multilingual fabric of the countries today.
Table 2.2
Languages of Countries Included in Study

Botswana

Colonizer Language(s)

Vernacular Languages

Englisha

Setswana, Kalanga, Sekgalagadi

a

Ghana

English

Asante, Ewe, Fante, Boron, Dagomba,
Dangme, Dagarte, Akyem, Ga, Akuapem

South Africa

Englisha, Afrikaansa (a Dutch IsiZulua, IsiXhosaa, IsiNdebelea, Sepedia,
dialect)
Sesothoa, Setswanaa, siSwatia, Tshivendaa,
Xitsongaa

Note. Source of data used to create table is “Field Listing: Languages,” by the Central
Intelligence Agency, 2013, retrieved from The World Factbook website:
https://www.cia.gov/library/publications/the-world-factbook/fields/2098.html.
a
Official language(s) of country.
In many African countries it is quite common for children to be taught and tested in
languages they do not speak at home, which has negative implications for learning (Brock-Utne
2001, 2010; Brock-Utne & Holomarsdottir, 2004). The countries examined here are no
exception to this practice. For example, Opoku-Amankwa (2009) found that Ghana’s Englishonly language-in-education policy negatively influenced student learning by creating anxiety and
reducing classroom participation. Similarly, in their investigation of language use and
preferences of children and parents in Botswana, Arua and Magocha (2002) found that Setswana
is the language of the home and English is the language of the school for the vast majority of
students. Furthermore, this pattern of language use was found to be consistent with parent
preferences and educational policy. English also dominates as the language of access and power
in South Africa, as the majority of African students speak their native language in their home and
communities, while having little direct contact with English outside of school (PANSALB, 2000;
Probyn, 2006; Probyn et al., 2002). Research has established that the majority of South African
48

students do not have the necessary English language proficiency to successfully engage with the
curriculum, which leads to marginalization and an inability to communicate knowledge in the
language of the test (Howie, 2001; McKay & Chick, 2001; Posel & Zeller, 2011; Probyn, 2006).
Given these educational issues surrounding language in the African context and the detrimental
effect of marginalization in the form of language on academic outcomes, it is expected that
infrequent speaking of the language of the test at home will have a similar effect on mathematics
and science outcomes in the TIMSS 2011 data.
Immigrant status. Numerous African countries have absorbed millions of children
displaced by conflict, which places additional strain on educational systems (UNESCO, 2010).
In terms of contemporary migration within Africa, the dominant trend is North-South emigration
largely motivated by economic reasons (Appleyard, 1992; Oucho, 1995). Unskilled laborers are
predominantly drawn to the mining areas of South Africa. An unfortunate consequence of
migration is xenophobia due to nationalism and economic factors. In the African context
characterized by limited resources, immigrants are seen as competing with locals for land and
economic opportunities (Kerting, 2009). In Ghana for example, migrants were victimized as
“Zombies” (Nyamjoh, 2006) and mass expulsions of aliens was based on ethnic nationalist
criteria (Whitacker, 2005). Evidence of widespread xenophobia has also been found in
Botswana and these negative attitudes also appear to be influenced by nationalism and economic
factors (Campbell, 2003; Nyamnjoh, 2006; van Dijk, 2002).
Of the three countries included in the present study, there is the greatest amount of
research on immigrants in South Africa, which will be explored in more detail here. Xenophobia
is a large problem in South Africa and concerns have been rising with increased instances of
xenophobic violence since the 1990s (Neocosmos, 2006; Nyamnjoh, 2006). In a nationwide

49

survey, Mattes, Taylor, McDonald, Poore, and Richmond (1999) found that 25% of South
Africans support a total ban on immigration and 45% support strict limits on the number of
immigrants allowed in. This hostility is also seen in relation to the education system as well, as
39% of South Africans opposed granting foreign children equal access to education and only
37% supported it (Mattes et al., 1999). The social context in which immigrants are received is a
major factor related to successful assimilation; a hostile reception by the public and authorities
makes it very difficult for immigrant families to become educated and acquire new occupational
skills (Portes & Fernandez-Kelly, 2008). Thus, the presence of such negative attitudes by South
Africans has the potential to threaten the academic success of immigrant children. While there is
a limited amount of research in this area, a study on the socio-cultural experiences of Black
female immigrant students in South Africa by Vandeyar and Vandeyar (2011) found that the
immigrant students experienced discrimination, harassment, and attrition. This study anticipates
that the educational disadvantage of immigrants will be found in the TIMSS 2011 data in the
form of a direct effect of immigrant status on academic achievement, and that feelings of
belongingness can buffer this effect of marginalization on the basis of immigrant status.
Overall, girls, students of low SES, students who infrequently speak the language of the
test at home, and immigrant students appear to be at risk of being educationally disadvantaged in
the sub-Saharan African countries included in this analysis. Given the contextual circumstances
reviewed above, it is especially important to understand the process of marginalization in relation
to academic achievement and how feelings of belongingness can serve as a buffer to be able to
enact change and enhance the outcomes of disadvantaged children in sub-Saharan African
countries. Additionally, to acknowledge the role of context, similarities and differences in the

50

relationships between feelings of belongingness, marginalization, and academic achievement
will be compared across the three countries to explore cross-national variation.
Theoretical Framework
The theoretical framework of this study draws from the social sciences of psychology and
sociology. To be clear in how theory from each respective discipline is incorporated into the
theoretical conceptualization of my study, this section is divided into two parts. The first
subsection describes how the study connects psychological theory on feelings of belongingness
to the educational context. The second subsection is devoted to applying labeling theory, which
is rooted in sociology. This section closes with a summary paragraph that brings the theories
together and presents my conceptual model.
Psychological Theory on the Need to Belong
The theoretical framework of this study connects psychological theory on feelings of
belongingness to the educational context. Theory concerning feelings of belongingness is rooted
in the field of psychology. Maslow’s hierarchy of needs (originally published in his 1943 paper
and more fully expressed in his 1954 book) serves as the theoretical foundation for this area of
research. In Maslow’s hierarchy of needs, feelings of love and belongingness constitute the third
level of human needs after physiological and safety needs are fulfilled. Maslow argued that all
humans need to feel a sense of belongingness among their social groups, and that this need of
belongingness is especially strong in children.
Relating theory on feelings of belongingness to an educational context, students need to
feel a sense of belongingness in their school for their basic psychological wellbeing. This study
will use the definition of feelings of belongingness provided by Hagerty, Lynch-Sauer, Patusky,
Bouwsema, and Collier (1992) as, “the experience of personal involvement in a system or

51

environment so that persons feel themselves to be an integral part of that system or environment”
(p. 173). Applying this definition to the field of education, the system or environment of interest
is the setting where formal education takes place, which is the school. Students themselves need
to feel that they are an integral part of their school.
One of the many benefits of feeling like you belong at school is improved academic
performance. Osterman (2000) offers a helpful social cognitive theoretical perspective which
posits that students have psychological needs, the satisfaction of those needs impacts their
perceptions and behaviors, and the characteristics of the social context influences how well these
needs are met. Applying this reasoning to the link between feelings of belongingness and
academic achievement, feelings of belongingness constitutes a basic psychological need and
academic achievement can be conceptualized as the result of student perceptions and behaviors.
Thus, it can be theoretically reasoned that student perceptions of their feelings of belongingness
are related to their academic achievement.
Going one step further, the theoretical framework of this study contends that
marginalization is a consequence of the social context that Osterman (2000) stressed as
influencing how well the need to belong is met and the resulting behaviors related to a child’s
enhanced academic performance. As stated previously, the social inequalities underlying
marginalization are created and maintained by social, economic, and political processes that
operate within the education system. Thus, taking into account the role of context like Osterman
recommends encourages us to consider how being marginalized interacts with feelings of
belongingness and the resultant academic performance. Regardless of their individual
background or characteristics, having all students come together and feel that they belong in their

52

school is one of the impetuses behind desegregation and inclusion initiatives (Kunc, 1992),
which has implications for the marginalized groups in this study.
Labeling Theory
In addition to extending psychological theory on feelings of belongingness to the
educational context, an additional component of this study’s theoretical framework is labeling
theory. Major contributors to the development of labeling theory include Becker (1963, 1964),
Lemert (1951, 1972, 1974), Matza (1964, 1969), and Tannenbaum (1938). Labeling theory has
its origins in sociology in the study of social deviance and the self-fulfilling prophecy, and has
extended into the fields of education and disability studies (Fitch, 2002; Rist, 1977).
Tannenbaum (1938) first introduced the idea of ‘tagging’ based on his finding that a negative tag
or label would contribute to further involvement in the delinquent activities of youth. This early
idea concerning the influence of socially imposed labels on behavior has expanded to include the
impact of labels on self-identity. A central tenant of labeling theory is that, “Labels are applied
to individuals which fundamentally shift their definitions of self and which further reinforce the
behavior which had initially prompted the social reaction,” (Rist, 1977, p. 302).
Applying labeling theory to the concept of marginalization, being deemed ‘marginalized’
or a member of a marginalized group (e.g., ‘immigrant,’ ‘disabled’) is in essence a form of
labeling. Labels of marginalization assigned by the larger society to minorities or those who are
viewed as deviating from the norm can carry stigma and negative social meanings. In the case of
marginalization in education, a common metric of deviance from the cultural norm is academic
achievement. As such, a child’s failure to succeed academically can be perceived as a form of
deviance. Labeling theory would argue that if children internalize the negative label and see
themselves as marginalized and at risk of academic failure, this would in turn contribute to poor

53

academic achievement. Alternatively, labels may also have the power to illicit positive
outcomes. For example, in special education disability labels can provide children with access to
individualized services that can support their academic development. Labels are no doubt a
complex social phenomenon that can operate in multiple ways within the school setting.
Whether labels have negative or positive connotations associated with them, it is
important to highlight that the pervasive assumptions which are attached to a label can be
challenged. For example, Goodley (2001) argues for a reconsideration of impairment in relation
to the label ‘learning difficulties.’ In his presentation of inclusive epistemological foundations,
Goodley advocates that we deconstruct impairment by exposing the social nature of diagnostic
criteria and by destabilizing the naturalized notions of ‘learning difficulties.’ Applying
Goodley’s advice, marginalization can be deconstructed by exposing the ways that the system
sets these students up to fail and by destabilizing the notion that marginalized student groups
academically underperform. A child who is a member of a marginalized group can challenge
their label by academically succeeding. I argue that feelings of belongingness can assist in this
process, because if a child feels like they belong at their school this reduces the likelihood that
marginalization becomes a defining part of their self-identify and that they will behave in a way
which reinforces their label. If a child does not perceive himself or herself as being marginalized
since they feel like they belong, this places the child in a positive psychological state that is
conducive to academic success. On the other hand, if the child does not feel like they belong this
can exacerbate the negative impact of their label of marginalization on their academic
performance and self-identity.
Overall, this study presupposes that because feelings of belongingness are such a basic
psychological need (Baumeister & Leary, 1995; Maslow, 1943, 1954), these feelings have the

54

power to buffer the negative impact that being a member of a marginalized group can have on a
child’s academic success. Even if a student is a member of a marginalized group at risk of
academic failure, if the student feels like they belong at their school, than this will help them
perform well academically. Feelings of belongingness can serve to challenge negative feelings
of the self and academic behaviors that are associated with the label of marginalization. Figure
2.1 depicts my conceptual model illustrating the relationships between feelings of belongingness,
marginalization, and academic achievement. Specifically, Figure 2.1 shows the direct effects of
feelings of belongingness and marginalization on academic achievement, as well as the buffering
effect of feelings on belongingness on the relationship between marginalization and academic
achievement.

Figure 2.1. Conceptual model.
Research Questions and Hypotheses
To test the relationships between feelings of belongingness, marginalization, and
academic achievement, the following research questions are asked:
RQ1. How are feelings of belongingness associated with mathematics and science
achievement?

55

RQ2. How are categories of marginalization associated with mathematics and science
achievement?
RQ2a. How is gender associated with mathematics and science achievement?
RQ2b. How is SES associated with mathematics and science achievement?
RQ2c. How is home language associated with mathematics and science
achievement?
RQ2d. How is immigrant status associated with mathematics and science
achievement?
RQ3. Do feelings of belongingness buffer the relationship between categories of
marginalization and mathematics and science achievement?
RQ3a. Do feelings of belongingness buffer the relationship between gender and
mathematics and science achievement?
RQ3b. Do feelings of belongingness buffer the relationship between SES and
mathematics and science achievement?
RQ3c. Do feelings of belongingness buffer the relationship between home
language and mathematics and science achievement?
RQ3d. Do feelings of belongingness buffer the relationship between immigrant
status and mathematics and science achievement?
RQ4. How do the relationships investigated between feelings of belongingness,
categories of marginalization, and mathematics and science achievement differ
across the three sub-Saharan African countries participating in TIMSS 2011?
Table 2.3 lists the hypotheses that correspond to Research Questions 1 through 3. Note that all
hypothesized relationships are anticipated to hold true for both mathematics and science

56

achievement. Research Question 4 is exploratory in nature (i.e., no hypothesis was made for
RQ4) due to the absence of existing literature on the nature of the relationships between feelings
of belongingness, categories of marginalization, and academic achievement in sub-Saharan
African countries.

57

Table 2.3
Hypotheses
Corresponding
RQ

Hypothesis #

Relationship

Direction

RQ1

H1

Feelings of belongingness
will be associated with
variation in achievement

+

RQ2

H2a

Gender will be associated
with variation in
achievement

Being a girl will be
associated with lower
achievement

H2b

SES will be associated
with variation in
achievement

+

H2c

Home language will be
associated with variation
in achievement

Infrequent speaking of the
language of the test at home
will be associated with lower
achievement

H2d

Immigrant status will be
associated with variation
in achievement

Being an immigrant will be
associated with lower
achievement

H3a

Feelings of belongingness
buffers the relationship
between gender and
achievement

Being a girl will be less
strongly associated with
lower achievement if the
student feels like she belongs

H3b

Feelings of belongingness
buffers the relationship
between SES and
achievement

Being of lower SES will be
less strongly associated with
lower achievement if the
student feels like they belong

H3c

Feelings of belongingness
buffers the relationship
between home language
and achievement

Infrequent speaking of the
language of the test at home
will be less strongly
associated with lower
achievement if the student
feels like they belong

H3d

Feelings of belongingness
buffers the relationship
between immigrant status
and achievement

Being an immigrant will be
less strongly associated with
lower achievement if the
student feels like they belong

RQ3

58

Data and Method
The TIMSS 2011 data set was used in the present study. Since 1995, TIMSS has
measured trends in mathematics and science achievement of fourth and eighth grade students in
participating countries all over the world. TIMSS 2011 is the fifth and most recent data
collection organized by the International Association for the Evaluation of Educational
Achievement (IEA) in the TIMSS cycle of studies that have been conducted every four years in
1995, 1999, 2003, 2007, and 2011. The TIMSS database comprises student achievement,
student, home, teacher, school, and curricular background data for 77 participating jurisdictions
(63 countries and 14 benchmarking participants) all over the world (Foy, Arora, & Stanco,
2013). In total, the database contains data from 608,641 students (Foy et al., 2013). These
students were selected using a two-stage random sample design in each participating jurisdiction,
where a sample of schools is drawn at the first stage and one or more intact classroom of students
is sampled at the second stage (Jonas & Foy, 2012).
Instruments
The TIMSS 2011 international database contains data from contextual questionnaires and
student assessments. As part of the TIMSS 2011 assessment, participating fourth and eighth
grade students complete grade-level mathematics and science assessments and a student
questionnaire, their parents complete a home questionnaire, their teachers complete a teacher
questionnaire, their school principals and department heads complete a school questionnaire, and
the TIMSS 2011 national research coordinators complete a curriculum questionnaire.
For this study, data from the student questionnaire, student assessments, teacher
questionnaire, and school questionnaire were used. The student questionnaire asks about various
aspects of students’ home and school lives, including basic demographic information, their home

59

environment, their school learning climate, and their self-perceptions and attitudes toward
learning mathematics and science (Foy et al., 2013). Regarding the student assessment
instruments, to assess a wide range of topics in mathematics and science, a matrix-sampling
booklet design was used where each student is only administered a subset of the entire TIMSS
mathematics and science item pools (Martin & Mullis, 2012). TIMSS 2011 test booklets contain
items which are designed to assess multiple content and cognitive domains in mathematics and
science. The eighth grade content areas for mathematics include numbers, algebra, geometry,
data and chance, and the content areas for science include biology, chemistry, physics, and earth
science. For both mathematics and science, the cognitive domains include knowing, applying,
and reasoning. These content and cognitive domains serve as the foundation of the TIMSS 2011
assessments (Mullis, Martin, Ruddock, O’Sullivan, & Preuschoff, 2009). Lastly, the teacher
questionnaire asks about teachers’ background and their teaching practices, and the school
questionnaire asks about the school environment, programs, and resources.
Variables
Table 2.4 provides a complete list of variables from the TIMSS 2011 dataset that were
used in my analysis. Details for each of the variables are provided in Table 2.4, including the
TIMSS 2011 general/integrated variable name and variable description. Below, each of the
variables is explained.

60

Table 2.4
Variable Details
Variable

TIMSS 2011
Variable Name

Description

Student mathematics
achievement

BSMMAT01B Five plausible values
SMMAT02BS
MMAT03BSM
MAT04BSMM
AT05

Student science
achievement

BSSSCI01
BSSSCI02
BSSSCI03
BSSSCI04
BSSSCI05

Five plausible values

Student age

BSDAGE

Age in years

Student gender

BSBG01

Dichotomous dummy variable indicating boy or
girl

Student time spent on math
homework per week

BSDMWKHW Categorical variable indicating 3 hours or more,
more than 45 minutes but less than 3 hours, or 45
minutes or less*

Student time spent on
science homework per week

BSDSWKHW

Categorical variable indicating 3 hours or more,
more than 45 minutes but less than 3 hours, or 45
minutes or less*

Student home language

BSBG03

Categorical variable indicating always, almost
always, sometimes, or never speak language of
test at home*

Student immigrant status

BSBG09A

Dichotomous dummy variable indicating whether
born in country or not

Student SES (Home
Educational Resources
Scale)

BSBGHER

Derived scale variable ranging from few to some
to many resources

Student likes mathematics
(Students Like Learning
Mathematics Scale)

BSBGSLM

Derived scale variable ranging from do not like to
somewhat like to like learning mathematics

Student likes science
(Students Like Learning
Science Scale)

BSBGSLS

Derived scale variable ranging from do not like to
somewhat like to like learning science

Student values mathematics
(Students Value
Mathematics Scale)

BSBGSVM

Derived scale variable ranging from do not value
to somewhat value to value mathematics

61

Table 2.4 (cont’d)
Variable

TIMSS 2011
Variable Name

Description

Teacher experience

BTBG01

Experience teaching in years

Teacher gender

BTBG02

Dichotomous dummy variable indicating male or
female

Teacher age

BTBG03

Categorical variable indicating under 25, 25-29,
30-39, 40-49, 50-59, or 60 or more years

Teacher education level

BTBG04

Categorical variable indicating did not complete
<ISCED level 3>, finished <ISCED level 3>,
finished <ISCED level 4>, finished <ISCED level
5b>, finished <ISCED level 5b, first degree>, or
finished <ISCED level 5b, second degree> or
higher

School urbanicity

BCBG05A

Categorical variable indicating more than 500,000,
10,001 to 500,000, 50,001 to 100,000, 15,001 to
50,000, 3,001 to 15,000, or 3,000 or fewer people
live in the city, town or area where the school is
located*

Note. BTBG04 variable responses consisted of nationally defined options that varied slightly
across the countries and were recoded for international comparability. To provide an example,
Botswana’s response set for the teacher education level variable is: did not complete senior
secondary, finished senior secondary, finished vocational/technical certificate, finished diploma,
finished first degree, and finished Master’s degree or higher.
SES = socioeconomic status. *Variable was reverse coded.
Academic achievement. The dependent variables in this study are students’
mathematics and science achievement. The BSMMAT01, BSMMAT02, BSMMAT03,
BSMMAT04, and BSMMAT05 variables were used as indicators of overall mathematics
achievement, and the BSSSCI01, BSSSCI02, BSSSCI03, BSSSCI04, BSSSCI05 variables were
used as indicators of overall science achievement. Based on student responses from 14
mathematics and science test booklets, five estimates of each student’s achievement score known
as “plausible values” were computed using Item Response Theory (IRT) scaling methodology
(see Martin & Mullis, 2012 for a detailed description of the TIMSS 2011 scaling approach). The
variables listed above are the five plausible values that equally represent a student’s achievement
62

in mathematics and in science. The TIMSS 2011 user manual states, “The plausible values for
any given scale are the best available measures of student achievement on that scale in the
TIMSS 2011 International Database, and should be used as the outcome measure in any study of
student achievement” (Foy et al., 2013). Accordingly, the five plausible value variables for
mathematics and science achievement were used as the academic outcome dependent variables.
The scale of these plausible achievement values ranges from 0 to 1,000, with a mean score of
500 and a standard deviation of 100.
Indicators of marginalization. Responses from individual items in the student
questionnaire and a derived variable provided by IEA were used as indicators of marginalization
in my analyses. Three individual items were used as independent variables to indicate
marginalization on the basis of a child’s gender (BSBG01), home language (BSBG03), and
immigrant status (BSBG09A). Gender is a categorical variable where students were asked to
report whether they are a girl or a boy. Home language is an ordinal variable where students
were asked to rate the frequency they speak the language of the test at home on a 4-point Likert
scale (always; almost always; sometimes; never). Responses to the home language item were
recoded to create a categorical variable consisting of two groups: (1) infrequent speaking of the
language of the test at home (sometimes and never) and (2) frequent speaking of the language of
the test at home (always and almost always) to aid in the interpretation of my results. Immigrant
status is a categorical variable where students were asked to report whether they were born in the
participant country or not. Students who reported being born in the country were coded as nonimmigrants and students who reported being born outside of the country were coded as
immigrants.

63

A derived index variable was used as an independent variable to indicate the fourth
category of marginalization included in the analysis, which is SES. The derived index variables
provided by IEA is the combination of responses from multiple questionnaire items and are more
reliable than the component questions (Akyüz & Berberoğlu, 2010). The particular derived
variable that will be used in this analysis is called the Home Educational Resources (HER) scale.
The HER scale is a learning-specific metric of SES computed using IRT that was designed to be
internationally comparable. The HER scale is based on student responses concerning the
availability of three home resources, namely parent education, number of books in the home, and
number of study supports (i.e., own room and/or internet connection) (Martin & Mullis, 2012).
The HER scale rates students as having few, some, or many resources. In this study, the HER
scale and will be used as a proxy for SES as a form of marginalization as an independent
variable.
Feelings of belongingness. One item from the student questionnaire was used to indicate
feelings of belongingness (BSBG12C). Feelings of belongingness is an ordinal variable4 where
students were asked to rate the extent to which they agree that they feel like they belong at their
school on a 4-point Likert scale (agree a lot; agree a little; disagree a little; disagree a lot). This
item was reverse coded so that higher scores on the item indicate higher feelings of
belongingness. Feelings of belongingness was entered into the model as an independent variable
to test its direct effect on academic achievement, as well as its buffering effect on the
relationship between categories of marginalization and academic achievement.
Control variables. At the school level, an individual item from the school questionnaire
was used to control for the effect of urbanicity (BCBG05A). At the classroom level, individual
4

Although feelings of belongingness is technically an ordinal variable, in this study it was treated as a continuous
variable because feelings of belongingness is conceptually a continuous phenomenon and in the social sciences
Likert scale data is commonly treated as continuous.

64

items from the teacher questionnaire were used to control for the effects of teacher experience
(BTBG01), gender (BTBG02), age (BTBG03), and education level (BTBG05). At the student
level, individual items from the student questionnaire were used to control for the effects of
student age (BSDAGE) and time spent on homework (BSBMWKHW; BSBSWKHW).
Additionally, derived subject-specific index variables were used to control for students’ liking of
the subject (BSBGSLM; BSBGSLS), value of the subject (BSBGSVM; BSBGSVS), and selfconfidence in the subject (BSBGSCM; BSBGSCS)5.
Sample
This study analyzed the TIMSS 2011 eighth grade student data from the three
participating sub-Saharan African countries, namely Botswana, Ghana, and South Africa. The
rationale for analyzing the eighth grade data as opposed to the fourth grade data was to include
the maximum number of sub-Saharan African countries participating in TIMSS 2011, because
only one of the three countries collected data for fourth grade students (see Table 2.5).
Table 2.5
Sub-Saharan African Countries Participating in TIMSS 2011
Assessment

Botswana

4th grade

8th grade





Ghana



South Africa



Note. Botswana and South Africa administered the fourth and eighth grade assessments to their
sixth and ninth grade students. Adapted from “Countries,” by the National Center for Education
Statistics, 2013, retrieved from the U.S. Department of Education Institute of Education Sciences
website https://nces.ed.gov/timss/countries.asp.

5

Two variables are listed because one refers to mathematics and the other to science.

65

It should be noted that Botswana and South Africa elected to administer the eighth grade
assessments to their ninth grade students because it was expected that the eighth grade students
would find the assessments too difficult (Mullis, Martin, Foy, & Arora, 2012). The average
TIMSS 2011 mathematics and science scores for all sub-Saharan African countries fell below the
TIMSS scale centerpoint (500), mostly falling below the Low (400) International Benchmark
(refer to Table 2.6 for average TIMSS 2011 mathematics and science scores for each of the three
countries). Of the three countries, Ghana had the lowest TIMSS 2011 mathematics and science
average achievement scores and Botswana had the highest.

66

Table 2.6
Country Descriptive Statistics
TIMSS 2011
Average
Achievement
Score*

Number of Students in
Sample

Math

Total

Male

Science

Mean
Age

Mean
SES

8.43
(1.84)

Number of Students
who Frequently
Speak Language of
Test at Home

Number of
Students
who are
Immigrants

Average
Feelings of
Belongingness

625

298

3.73 (.61)

Female

Botswana

397
(2.5)

404
(3.6)

5,400

2,628

2,772

15.84
(.90)

Ghana

331
(4.3)

306
(5.2)

7,323

3,822

3,501

15.74 7.96
(1.51) (1.89)

2,041

530

3.42 (.89)

South
Africa

353
(2.5)

332
(3.7)

11,966

6,078

5,888

15.93 8.91
(1.17) (1.89)

3,986

2,204

3.32 (.91)

Note. Standard deviations appear in parentheses.
*Data from Mullis et al. (2012) and Martin et al. (2012).

67

The total sample size for Botswana was 5,400 students (48.7% male, 51.3% female). Of
these, only 625 students frequently spoke the language of the test at home (11.6%) and 298
students were immigrants (5.5%). The age of students ranged from 10.25 to 19.58 years (M =
15.84, SD = .90). On average, students were of low SES (M = 8.43, SD = 1.43) and reported
high feelings of belongingness at their school (M = 3.73, SD = .61).
The total sample size for Ghana was 7,323 students (52.2% male, 47.8% female). Of
these, only 2,041 students frequently spoke the language of the test at home (27.9%) and 530
students were immigrants (7.2%). The age of students ranged from 9.67 to 18.92 years (M =
15.74, SD = 1.51). On average, students were of low SES (M = 7.96, SD = 1.89) and reported
high feelings of belongingness at their school (M = 3.42, SD = .89).
The total sample size for South Africa was 11,966 students (50.8% male, 49.2% female).
Of these, only 3,986 students frequently spoke the language of the test at home (33.3%) and
2,204 students were immigrants (18.4%). The age of students ranged from 10.00 to 19.92 years
(M = 15.93, SD = 1.17). On average, students were of low SES (M = 8.91, SD = 1.89) and
reported high feelings of belongingness at their school (M = 3.32, SD = .91). Table 2.5 presents
the distribution of the Botswana, Ghana, and South Africa TIMSS 2011 student samples by
gender, SES, home language, and immigrant status.
Data Preparation
To prepare the data for the subsequent analysis, all continuous variables6 were grand
mean centered7 and all categorical variables were dummy coded. Specifically, gender was
6

The predictors of interest that were mean centered include feelings of belongingness and SES. The controls that
were mean centered include school urbanicity, teacher experience, teacher age, teacher education level, student age,
student time spent on homework, student liking of the subject, student value of the subject, and student selfconfidence in the subject.
7
Mean centering continuous variables eliminates non-essential collinearity (i.e., correlation due to scales of the
predictors) between each predictor and the interaction term, resulting in unbiased statistical estimation of all
relationships and significance values (Cohen, Cohen, West, & Aiken, 2003).

68

dummy coded with males as the referent group (i.e., coded as ‘0’), home language was dummy
coded with infrequent speaking of the language of the test at home as the referent group, and
immigrant status was dummy coded with non-immigrant as the referent group.
Eighth grade data for Botswana, Ghana, and South Africa was downloaded from the IEA
International Database (IDB) Analyzer (IEA, 2012). More specifically, the eighth grade student
achievement data files8, the eighth grade student background data files9, the eighth grade studentteacher linkage files10, the eighth grade mathematics teacher background data files11, the eighth
grade science teacher background data files,12 and the eighth grade school background data
files13 for Botswana, Ghana, and South Africa were downloaded and merged using the IEA IDB
merge module.
Data Analysis
All analyses were run using the IDB Analyzer analysis module in conjunction with IBM
SPSS Statistics, Version 21. The IDB Analyzer generates SPSS syntax that takes into account
information from the TIMSS 2011 sampling design in the computation of sampling variance,
generates SPSS code to handle the use of plausible values, and accounts for clustering within the
data (i.e., students nested within classrooms within schools) using jackknifed standard errors
(IEA, 2013). Before running any models, for my preliminary analysis I created correlation
matrices for each country for mathematics and science with all of the variables to begin seeing
how the predictors of interest (feelings of belongingness and the categories of marginalization)
were related to academic achievement.

8

The names of these data files include BSABWAM5, BSAGHAM5, and BSAZAFM5.
The names of these data files include BSGBWAM5, BSGGHAM5, and BSGZAFM5.
10
The names of these data files include BSTBWAM5, BSTGHAM5, and BSTZAFM5.
11
The names of these data files include BTMBWAM5, BTMGHAM5, and BTMZAFM5.
12
The names of these data files include BTSBWAM5, BTSGHAM5, and BTSZAFM5.
13
The names of these data files are BCGBWAM5, BCGGHAM5, and BCGZAFM5.
9

69

A series of iterative ordinary least squares (OLS) multiple regression models were then
used to answer my research questions. All conceptual models are explained below and the
corresponding list of statistical models run can be found in the Appendix. Note that in the
subsequent equations Y is mathematics or science achievement, β0 is the intercept, age is student
age, hmwork is student time spent on mathematics or science homework per week, like is student
likes mathematics or science, value is student values mathematics or science, confident is student
confidence in mathematics or science, tchfemale is teacher gender, tchage is teacher age, tchexp
is teacher experience, tchedu is teacher education level, urbanicity is school urbanicity, belong is
student feelings of belongingness, female is student gender, SES is student socioeconomic status,
hlang is student home language, immig is student immigrant status, and e is the error term.
All control variables were entered into the model first to remove all variance in academic
achievement due to student age, student time spent on homework, student liking of the subject,
student value of the subject, student confidence in the subject, teacher gender, teacher age,
teacher experience, teacher education level, and school urbancity.
Model 1: Y = β0 + βaage + βhhmwork + βllike + βvvalue + βcconfident
+ βmtchfemale + βgtchage + βptchexp + βdtchedu + βuurbanicity + e

(1)

Next, to answer RQ1 and RQ2, the student feelings of belongingness and categories of
marginalization variables were added to the model to determine the main effects of feelings of
belongingness and marginalization on academic achievement.
Model 2: Y = β0 + βaage + βhhmwork + βllike + βvvalue + βcconfident
+ βmtchfemale + βgtchage + βptchexp + βdtchedu + βuurbanicity
+ βbbelong + βffemale + βsSES + βhhlang + βiimmig + e

70

(2)

Building upon these models, to answer RQ3, interaction terms between feelings of belongingness
and each category of marginalization were added to the model to examine the buffering effect of
feelings of belongingness on the relationship between marginalization and achievement.
Model 3: Y = β0 + βaage + βhhmwork + βllike + βvvalue + βcconfident
+ βmtchfemale + βgtchage + βptchexp + βdtchedu + βuurbanicity
+ βffemale + βsSES + βhhlang + βiimmig + βbbelong
+ βfbfemale*belong + βsbSES*belong + βhbhlang*belong
+ βibimmig*belong + e

(3)

Lastly, for RQ4 this process was repeated for each of the three countries for mathematics and
science achievement. The resulting models were compared across the countries to examine
cross-national differences in the estimated relationships. For all analyses, the student sampling
weight TOTWGT was applied and listwise deletions were made for missing data.
Results
Tables 2.7-2.12 present the correlation matrices of all dependent and independent
variables for each sub-Saharan African country. Tables 2.13-2.15 are the complete OLS
regression tables for all 3 models predicting mathematics and science achievement in each of the
three countries. To allow for easier comparison across countries, Table 2.16 below summarizes
OLS estimates for the predictors of interest in the main effect and interaction models (Models 2
and 3). Furthermore, Table 2.17 summarizes which hypotheses were supported in each country
in each subject.

71

Table 2.7
Correlation Matrix with Botswana Mathematics Data
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

1. math plausible value 1

1.00

2. age

-.435**

1.00

3. homework

.065**

-.070**

1.00

4. likes math

.340**

-.100**

.030**

1.00

5. values math

.335**

.143**

.069**

.461**

1.00

6. confident math

.287**

-.047**

-.039**

.543**

.278**

1.00

7. teacher gender

-.010

.000

.011*

-.031**

-.007

-.030**

1.00

8. teacher age

.027**

-.018**

-.047**

-.014**

-.019**

.002

-.151**

1.00

9. teacher experience

.076**

-.040**

-.035**

-.002

-.002

-.001

-.104**

.807**

1.00

10. teacher education

-.002

-.04

-.002

-.048**

-.020**

-.031**

.132**

-.084**

.064**

1.00

11. urbanicity

.199**

-.115**

-.043**

-.018**

.028**

-.048**

.126**

.051**

.133**

.126**

1.00

12. feelings of belongingness

.066**

-.016**

.031**

.157**

.197**

.090**

.015**

.009

.010

-.009

.009

1.00

13. gender

.085**

-.168**

.086**

.017**

.082**

-.072**

.044**

-.017**

-.031**

.008

-.015**

.014**

1.00

14. SES

.080**

-.104**

.005

-.005

.021**

.073**

-.041**

.039**

.040**

.042**

.115**

-.055**

-.042**

1.00

15. home language

.086**

-.059**

.003

-.011*

.001

.067**

.005

.022**

.035**

.062**

.098**

-.017**

.009

.231**

1.00

16. immigrant status

-.137**

.120**

-.015**

-.060**

-.108**

.020**

-.023**

.033**

.028**

-.012*

-.030**

-.033**

-.020**

.065**

.091**

*Correlation is significant at the .05 level. **Correlation is significant at the .01 level.

72

16

1.00

Table 2.8
Correlation Matrix with Botswana Science Data
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

1. science plausible value 1

1.00

2. age

-.429**

1.00

3. homework

-.116**

.051**

1.00

4. likes math

.448**

-.172**

.005

1.00

5. values math

.380**

-.171**

.011

.539**

1.00

6. confident math

.329**

-.113**

.000

.520**

.333**

1.00

7. teacher gender

.061**

-.034*

-.036*

-.038**

.007

-.038**

1.00

8. teacher age

.049**

-.017

-.019

-.002

.009

.003

-.170**

1.00

9. teacher experience

.035*

-.021

-.019

-.007

.008

-.026

-.132**

.713**

1.00

10. teacher education

.002

-.014

-.018

-.048**

-.034*

-.037**

.135**

-.128**

-.094**

1.00

11. urbanicity

.194**

-.107**

-.087**

-.028

.008

.055**

.146**

.051**

-.011

.066**

1.00

12. feelings of belongingness

.076**

-.015

.021

.132**

.164**

.072**

.004

.016

.003

-.005

.002

1.00

13. gender

.060**

-.170**

-.009

.035*

.068**

-.077**

-.007

-.005

-.009

-.018

-.015

.019

1.00

14. SES

.102**

-.101**

.004

.030*

.007

.139**

.008

.049**

.051**

.052**

.108**

-.059**

-.043**

1.00

15. home language

.071**

-.051**

-.021

-.014

-.049**

.094**

.005

.023

.045**

.037**

.087**

-.021

.003

.232**

1.00

16. immigrant status

-.158**

.113**

-.013

-.109**

-.130**

-.023

-.012

-.035*

-.021

.046**

-.026

-.040**

-.023

.071**

.094**

*Correlation is significant at the .05 level. **Correlation is significant at the .01 level.

73

16

1.00

Table 2.9
Correlation Matrix with Ghana Mathematics Data
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

1. math plausible value 1

1.00

2. age

-.271**

1.00

3. homework

.049**

.023*

1.00

4. likes math

.365**

-.146**

-.048**

1.00

5. values math

.238**

-.074**

-.012

.409**

1.00

6. confident math

.308**

-.099**

-.102**

.525**

.341**

1.00

7. teacher gender

.013

.005

-.003

.029*

.054**

-.023

1.00

8. teacher age

.063**

-.070**

.004

.034**

.014

-.016

.033**

1.00

9. teacher experience

.094**

-.068**

-.042**

.009

.005

-.019

.084**

.806**

1.00

10. teacher education

.098**

-.090**

.018

.063**

.029*

.028*

.133**

.460**

.384**

1.00

11. urbanicity

.304**

-.223**

.019

.098**

.045**

.033**

.142**

.209**

.262**

.231**

1.00

12. feelings of belongingness

.059**

-.028*

.027**

.155**

.225**

.105**

-.021

-.019

-.019

.021

-.029*

1.00

13. gender

-.130**

-.077**

.025*

-.083**

-.026*

-.129**

.014

.005

.021

.030**

.022

.002

1.00

14. SES

.100**

-.174**

.063**

.004

-.020

.046**

.009

.120**

.177**

.084**

.248**

-.059**

.007

1.00

15. home language

.011

.024*

.025*

.038**

.037**

.051**

.038**

.013

.045**

-.035**

-.005

.020

-.043**

.129**

1.00

16. immigrant status

-.090**

.017

-.036**

-.056**

-.056**

-.030**

.010

.014

.047**

.020

.008

-.046**

.018

.075**

.037**

*Correlation is significant at the .05 level. **Correlation is significant at the .01 level.

74

16

1.00

Table 2.10
Correlation Matrix with Ghana Science Data
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

1. science plausible value 1

1.00

2. age

-.282**

1.00

3. homework

-.055

.003

1.00

4. likes math

.432**

-.178**

-.061**

1.00

5. values math

.258**

-.083**

-.031**

.425**

1.00

6. confident math

.362**

-.128**

-.095**

.602**

.415**

1.00

7. teacher gender

-.048**

.037**

.011*

-.023*

-.009

-.026*

1.00

8. teacher age

.146**

-.130**

.023

.033**

-.030*

-.005

-.024*

1.00

9. teacher experience

.076**

-.067**

-.035**

.046**

-.011

.004

.103**

.681**

1.00

10. teacher education

.130**

-.094**

.000

.062**

.000

.002

.066**

.413**

.268**

1.00

11. urbanicity

.327**

-.232**

-.022

.140**

.045**

.077**

.089**

.318**

.207**

.306**

1.00

12. feelings of belongingness

.079**

-.029*

.028*

.109**

.198**

.121**

.003

-.037**

-.030*

-.035**

-.028*

1.00

13. gender

-.141**

-.078**

-.004

-.058**

.006

-.095**

.005

.012

.017

.038**

.012

.001

1.00

14. SES

.150**

-.75**

.025*

.023*

-.011

.066**

.056**

.142**

.148**

.081**

.250**

-.059**

.008

1.00

15. home language

.008

.027*

.018

.033**

.042**

.057**

-.010

.000

-.001

-.036**

-.013**

.019

-.041**

.128**

1.00

16. immigrant status

-.106**

.015

-.013

-.074**

-.041**

-.026*

.001

-.001

.000

-.033*

.008

-.046**

.018

.074**

.037**

*Correlation is significant at the .05 level. **Correlation is significant at the .01 level.

75

16

1.00

Table 2.11
Correlation Matrix with South Africa Mathematics Data
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

1. math plausible value 1

1.00

2. age

-.356**

1.00

3. homework

.055**

-.030**

1.00

4. likes math

.117**

-.088**

.052**

1.00

5. values math

.104**

-.080**

.069**

.484**

1.00

6. confident math

.244**

-.073**

.061**

.560**

.349**

1.00

7. teacher gender

.173**

-.080**

.021*

-.006

-.004

.024*

1.00

8. teacher age

.000

.001

-.003

-.007

-.080**

-.025*

.122**

1.00

9. teacher experience

-.033**

-.002

.009

.019*

.030**

-.019

.124**

.786**

1.00

10. teacher education

.073**

.020*

.006

-.009

.011

.015

-.015

-.019

.067**

1.00

11. urbanicity

.377**

-.146**

.011

-.053**

-.001

-.020*

-.055**

.072**

.028**

.129**

1.00

12. feelings of belongingness

.009

-.017

.046**

.209**

.240**

.150**

.012

.013

.028**

.011

-.030**

1.00

13. gender

-.009

-.213**

.039**

-.022*

.001

-.066**

.025**

-.122**

.018

-.011

.008

.023**

1.00

14. SES

.390**

-.162**

.067**

-.030**

.010

.091**

.111**

.021*

.007

.037**

.231**

-.040**

-.022*

1.00

15. home language

.371**

-.160**

.015

-.093**

-.029**

.021*

.096**

.067**

.041**

.032**

.204**

-.047**

.011

.281**

1.00

16. immigrant status

-.339**

.207**

-.067**

-.107**

-.124**

-.021*

-.038**

-.029**

-.014

-.012

-.125**

-.049**

-.049**

-.067**

-.084**

*Correlation is significant at the .05 level. **Correlation is significant at the .01 level.

76

16

1.00

Table 2.12
Correlation Matrix with South Africa Science Data
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

1. science plausible value 1

1.00

2. age

-.362**

1.00

3. homework

.070**

.054**

1.00

4. likes math

.192**

-.111**

.065**

1.00

5. values math

-.023*

.002

.121**

.523**

1.00

6. confident math

.197**

-.108**

.033**

.635**

.426**

1.00

7. teacher gender

.065**

-.008

.021*

-.040**

-.028**

-.023**

1.00

8. teacher age

-.024*

.012

-.003

.006

.022*

.013

-.026**

1.00

9. teacher experience

.049**

-.047**

-.001

.021*

.025*

.031**

-.071**

.791**

1.00

10. teacher education

.125**

.016

.020*

-.047**

-.023*

-.026**

.103**

-.058**

-.067**

1.00

11. urbanicity

.383**

-.146**

-.050**

-.019

-.085**

-.009

.115**

-.031**

.015

.047**

1.00

12. feelings of
belongingness

.004

-.017

.042**

.173**

.197**

.143**

-.009

.004

.000

-.012

-.030**

1.00

13. gender

.006

-.213**

-.007

-.001

-.036**

-.016

.014

-.016

-.014

-.014

.008

.023*

1.00

14. SES

.391**

-.162**

.001

.031**

-.021*

.082**

.058**

-.032**

.011

.105**

.231**

-.040**

-.022*

1.00

15. home language

.430**

-.160**

-.052**

-.031**

-.110**

.062**

.075**

.011

.098**

.114**

.204**

-.047**

.011

.281**

1.00

16. immigrant status

-.363**

.207**

.005

-.151**

-.037**

-.097**

-.034**

-.004

-.030**

-.008

-.125**

-.049**

-.049**

-.067**

-.034**

*Correlation is significant at the .05 level. **Correlation is significant at the .01 level.

77

16

1.00

Table 2.13
OLS Estimates for Models Predicting Mathematics and Science Achievement in Botswana
Predictors
intercept

1

Mathematics Models
2

3

1

Science Models
2

3

400.78 (2.63)

400.03 (3.14)

400.42 (3.13)

402.72 (3.53)

411.18 (3.34)

411.27 (3.29)

-30.02***
(1.61)

-28.19***
(1.70)

-28.12***
(1.70)

-35.99***
(2.04)

-34.79***
(1.92)

-34.76***
(1.91)

homework

5.87** (1.86)

5.72** (1.88)

5.79** (1.87)

-13.06**
(2.27)

-13.07***
(2.34)

-12.98***
(2.34)

likes math

5.91*** (.93)

6.05*** (.90)

6.08*** (.90)

14.92***
(1.10)

14.53***
(1.11)

14.53***
(1.11)

values math

5.61*** (.79)

5.05*** (.75)

5.01*** (.75)

8.18*** (1.26)

8.14*** (1.26)

8.13*** (1.25)

confident math

6.89*** (.78)

7.02*** (.72)

7.04*** (.72)

7.01*** (.94)

6.25*** (.93)

6.29*** (.92)

teacher gender

-1.43 (4.39)

1.66 (4.73)

1.67 (4.73)

10.74* (5.26)

1.69 (5.64)

1.62 (5.61)

.69 (5.00)

.41 (4.58)

.32 (4.56)

4.46 (4.79)

3.13 (3.98)

2.94 (3.95)

.37 (.67)

.41 (.64)

.42 (.64)

.26 (.54)

.24 (.53)

.24 (.52)

11.24 (9.83)

11.72 (9.03)

11.63 (8.99)

4.08 (4.79)

3.28 (4.79)

3.21 (4.79)

11.26***
(2.00)

10.83***
(1.92)

10.79***
(1.90)

12.81***
(2.30)

12.41***
(2.10)

12.42***
(2.09)

FB

-.71 (1.41)

-.09 (1.52)

-.20 (1.38)

.04 (1.89)

gender

2.90 (3.21)

2.93 (3.17)

-11.24* (4.52)

-11.22* (4.51)

.14 (.70)

.13 (.70)

.66 (1.01)

.64 (1.01)

25.38* (10.53)

25.44* (10.55)

7.49 (12.07)

7.81 (12.17)

-32.58**
(10.82)

-28.36*
(11.91)

-34.10*
(14.55)

-34.03*
(14.77)

-4.38 (4.87)

-4.24 (4.88)

14.49** (4.86)

14.47** (4.84)

-40.41**
(15.12)

-40.40**
(15.18)

-1.99 (20.84)

-2.03 (20.66)

immigrant status*tchgender

25.79 (19.93)

19.09 (21.38)

18.69 (24.05)

17.43 (24.39)

gender*home language

-3.24 (10.96)

-3.22 (10.94)

13.27 (14.87)

13.13 (14.94)

gender*immigrant status

-1.64 (13.81)

-5.90 (14.98)

-25.06 (17.03)

-25.48 (16.98)

-29.98 (20.10)

-36.65 (20.57)

-9.36 (34.47)

-7.89 (33.92)

age

teacher age
teacher experience
teacher education
urbanicity

SES
home language
immigrant status
gender*tchgender
home language*tchgender

home language*immigrant
status
gender*FB

-2.12 (2.07)

SES*FB
home language*FB
immigrant status*FB

-.46 (3.21)

.06 (.69)

.66 (.73)

-1.05 (3.71)

-4.27 (4.46)

9.92 (5.78)

8.94 (8.00)

gender*hlang*immig

70.17* (35.30)

82.10* (35.29)

94.23 (55.29)

97.38 (53.27)

gender*hlang*tchgender

39.10* (15.30)

38.87* (15.33)

6.74 (25.37)

6.51 (25.26)

gender*immig*tchgender

-22.05 (30.63)

-13.42 (31.19)

15.33 (37.08)

19.79 (37.28)

hlang*immig*tchgender

54.67 (50.80)

66.96 (46.08)

28.09 (55.21)

25.70 (55.17)

gender*hlang*immig*tchgender
-67.66 (71.33)
-88.64 (63.12)
-64.31 (84.19)
-70.05 (82.55)
Adjusted R2
.33
.35
.35
.37
.38
.38
∆R2
.02
.00
.01
.00
Note. OLS = Ordinary Least Squares; tchgender = teacher gender; FB = feelings of belongingness; SES = socioeconomic status; hlang = home
language; immig = immigrant status; standard errors of regression coefficients are in parentheses. *p < .05, **p < .01, ***p <.001

78

Table 2.14
OLS Estimates for Models Predicting Mathematics and Science Achievement in Ghana
Predictors
intercept

1

Mathematics Models
2

3

1

Science Models
2

3

343.40 (3.89)

355.75 (3.96)

355.92 (4.00)

322.32 (4.69)

338.69 (5.70)

338.70 (5.71)

-10.06***
(1.45)

-10.43***
(1.30)

-10.43***
(1.31)

-11.61***
(2.00)

-12.16***
(1.85)

-12.16***
(1.84)

homework

7.83** (2.47)

8.88*** (2.51)

8.87*** (2.49)

8.90** (3.17)

8.49** (3.16)

8.52** (3.18)

likes math

10.76***
(1.21)

10.41***
(1.28)

10.46***
(1.27)

16.05***
(1.80)

15.66***
(1.88)

15.64***
(1.88)

values math

1.85 (.98)

1.84 (1.08)

1.81 (1.08)

4.66*** (1.37)

4.84** (1.56)

4.82** (1.56)

confident math

7.21*** (.99)

6.69*** (1.05)

6.66*** (1.05)

8.98*** (1.12)

7.92*** (1.31)

7.93*** (1.31)

teacher gender

-11.03 (7.44)

-6.68 (8.68)

-7.23 (8.76)

-15.72 (9.69)

-11.54 (10.30)

-11.40 (10.36)

-2.93 (6.26)

-5.25 (6.03)

-5.28 (6.01)

-.92 (6.01)

-.24 (5.65)

-.28 (5.70)

teacher experience

.73 (.66)

1.07 (.63)

1.07 (.63)

-.01 (.84)

-.06 (.81)

-.06 (.82)

teacher education

.31 (4.25)

1.23 (4.25)

1.31 (4.26)

2.29 (4.49)

2.19 (4.48)

2.20 (4.51)

9.90*** (1.76)

9.72*** (1.81)

9.71*** (1.81)

13.18***
(2.12)

13.31***
(2.13)

13.33***
(2.13)

-1.67 (2.76)

-1.51 (3.42)

-.62 (3.13)

-1.80 (3.98)

gender

-20.95***
(3.56)

-20.89***
(3.57)

-25.81***
(4.85)

-25.83***
(4.81)

SES

-.37 (1.37)

-.39 (1.37)

1.15 (1.63)

1.15 (1.64)

home language

3.58 (5.62)

3.53 (5.28)

4.71 (7.28)

4.90 (7.32)

-28.87*
(11.69)

-29.19*
(11.70)

-25.94**
(10.07)

-25.41*
(10.36)

-5.07 (7.82)

-5.61 (8.05)

-3.96 (12.73)

-4.10 (12.59)

-11.77 (13.51)

-11.14 (13.29)

5.67 (20.85)

5.24 (21.05)

49.61 (34.86)

53.26 (36.59)

3.02 (37.16)

3.05 (37.20)

gender*home language

-4.15 (7.10)

-4.16 (7.08)

-8.21 (9.80)

-8.36 (9.76)

gender*immigrant status

7.22 (13.98)

7.72 (13.69)

-15.17 (15.27)

-15.05 (15.29)

-10.75 (19.44)

-7.88 (19.12)

-17.08 (21.62)

-17.09 (20.80)

age

teacher age

urbanicity
FB

immigrant status
gender*tchgender
home language*tchgender
immigrant status*tchgender

home language*immigrant
status
gender*FB

-.68 (3.58)

2.11 (3.82)

SES*FB

1.26 (1.32)

-.07 (1.78)

home language*FB

-.50 (3.85)

-1.13 (4.88)

immigrant status*FB

5.60 (6.07)

6.12 (6.63)

gender*hlang*immig

12.35 (25.06)

9.59 (24.22)

11.99 (30.66)

11.23 (30.26)

gender*hlang*tchgender

-1.16 (16.72)

-.28 (16.87)

-11.47 (37.20)

-11.34 (37.15)

gender*immig*tchgender

-67.24 (42.47)

-71.20 (43.19)

26.67 (40.60)

27.36 (41.74)

hlang*immig*tchgender

-48.97 (56.31)

-52.20 (57.33)

-25.94 (50.88)

-26.74 (50.30)

gender*hlang*immig*tchgender
71.30 (62.49)
72.91 (63.49)
-67.60 (84.92)
-66.72 (82.17)
Adjusted R2
.24
.26
.27
.29
.31
.31
∆R2
.02
.01
.02
.00
Note. OLS = Ordinary Least Squares; tchgender = teacher gender; FB = feelings of belongingness; SES = socioeconomic status; hlang = home
language; immig = immigrant status; standard errors of regression coefficients are in parentheses. *p < .05, **p < .01, ***p <.001

79

Table 2.15
OLS Estimates for Models Predicting Mathematics and Science Achievement in South Africa
Predictors

1

Mathematics Models
2

intercept

350.54 (4.03)

351.44 (4.25)

351.41 (4.27)

342.21 (7.07)

346.50 (7.54)

346.55 (7.53)

-19.34***
(1.68)

-14.04***
(1.36)

-14.01***
(1.35)

-26.20***
(1.85)

-19.36***
(1.58)

-19.30***
(1.57)

7.09*** (2.06)

3.88* (1.73)

3.94* (1.73)

-2.87 (3.22)

-2.35 (3.14)

-2.25 (3.11)

1.50 (1.08)

1.66 (.85)

1.65 (.85)

12.40***
(1.06)

11.18*** (.86)

11.23*** (.87)

-.08 (.83)

-.81 (.80)

-.86 (.79)

-8.06***
(1.23)

-5.75*** (1.03)

-5.85*** (1.01)

confident math

10.90***
(1.13)

9.92*** (.88)

9.97*** (.88)

8.82*** (1.08)

4.68*** (1.07)

4.73*** (1.07)

teacher gender

33.02***
(6.36)

30.69***
(5.19)

30.68***
(5.22)

8.47 (8.92)

-3.21 (9.23)

-3.21 (9.22)

4.58 (5.11)

4.87 (3.69)

4.80 (3.68)

-16.60**
(6.22)

-9.14* (4.66)

-9.17* (4.64)

-.72 (.63)

-.88 (.47)

-.87 (.47)

1.84* (.77)

.91 (.54)

.90 (.54)

7.11* (3.61)

4.85 (2.97)

4.85 (2.94)

9.88* (4.33)

6.64* (3.06)

6.65* (3.06)

17.18***
(2.49)

11.79***
(2.03)

11.75***
(2.04)

18.02***
(3.16)

11.95***
(2.29)

11.87***
(2.30)

-.12 (1.28)

.03 (2.07)

1.02 (1.76)

-.50 (2.56)

age
homework
likes math
values math

teacher age
teacher experience
teacher education
urbanicity
FB
gender
SES
home language
immigrant status
gender*tchgender

3

1

Science Models
2

3

-1.73 (4.62)

-1.46 (4.62)

-13.58* (6.32)

-13.32* (6.29)

7.27*** (1.02)

7.25*** (1.03)

10.04***
(1.17)

10.01***
(1.19)

58.67***
(7.95)

58.69***
(7.97)

92.13***
(9.22)

92.15***
(9.24)

-37.85***
(6.06)

-37.59***
(6.27)

-61.53***
(11.35)

-61.61***
(11.59)

-8.66 (5.74)

-8.69 (5.79)

7.30 (8.94)

7.23 (8.93)

home language*tchgender

-6.67 (11.98)

-6.64 (11.96)

-15.10 (10.84)

-15.28 (10.87)

immigrant status*tchgender

-12.43 (8.32)

-11.94 (8.45)

6.48 (16.37)

7.69 (16.45)

gender*home language

-13.31 (7.73)

-13.56 (7.82)

-20.92* (9.44)

-21.11* (9.41)

-9.44 (7.64)

-9.63 (7.67)

-8.52 (12.04)

-8.03 (12.13)

-41.99***
(12.07)

-41.67***
(12.23)

-57.11**
(19.80)

-57.54**
(19.99)

gender*immigrant status
home language*immigrant
status
gender*FB

-1.49 (2.14)

SES*FB
home language*FB
immigrant status*FB
gender*hlang*immig

-.67 (2.92)

.27 (.68)

.48 (.91)

-1.07 (2.85)

.14 (3.71)

4.06 (2.90)

8.47* (3.85)

20.64 (16.50)

20.50 (16.81)

25.60 (23.48)

26.55 (24.25)

7.53 (9.18)

7.60 (9.21)

18.06 (13.41)

18.10 (13.33)

gender*immig*tchgender

10.22 (12.32)

9.25 (12.49)

16.52 (18.26)

14.44 (18.10)

hlang*immig*tchgender

5.02 (21.19)

4.04 (21.35)

20.67 (27.18)

22.24 (27.19)

gender*hlang*tchgender

gender*hlang*immig*tchgender
-.85 (26.42)
.39 (26.64)
-13.98 (31.28)
-15.00 (32.03)
Adjusted R2
.31
.46
.46
.29
.47
.47
∆R2
.15**
.00
.18**
.00
Note. OLS = Ordinary Least Squares; tchgender = teacher gender; FB = feelings of belongingness; SES = socioeconomic status; hlang = home
language; immig = immigrant status; standard errors of regression coefficients are in parentheses. *p < .05, **p < .01, ***p <.001

80

Table 2.16
OLS Estimates for Predictors of Interest in Key Models Predicting Mathematics and Science Achievement in Each Country
Botswana
Mathematics
Predictors

Model 2

Model 3

Ghana
Science

Model 2

Model 3

Mathematics
Model 2

Model 3

South Africa
Science

Model 2

Model 3

Mathematics
Model 2

Model 3

Science
Model 2

Model 3

Intercept

400.03
(3.14)

400.42
(3.13)

411.18
(3.34)

411.37
(3.29)

355.73
(3.96)

355.92
(4.00)

338.69
(5.70)

338.70
(5.71)

351.44
(4.25)

351.41
(4.27)

346.50
(7.54)

346.55
(7.53)
-.50
(2.56)

Feelings of
belongingness

-.71
(1.41)

-.09
(1.52)

-.20
(1.38)

.04
(1.89)

-1.67
(2.76)

-1.51
(3.42)

-.62
(3.13)

-1.80
(3.98)

-.12
(1.28)

.03
(2.07)

1.02
(1.76)

Gender

2.90
(3.21)

2.93
(3.17)

-11.24*
(4.52)

-11.22*
(4.51)

-20.95***
(3.56)

-20.89***
(3.57)

-25.81***
(4.85)

-25.83***
(4.81)

-1.73
(4.62)

-1.46
(4.62)

-13.58*
(6.32)

-13.32*
(6.29)

SES

.14
(.70)

.13
(.70)

.66
(1.01)

.64
(1.01)

-.37
(1.37)

-.39
(1.37)

1.15
(1.63)

1.15
(1.64)

7.27***
(1.02)

7.25***
(1.03)

10.04***
(1.17)

10.01***
(1.19)

Home language

25.38*
(10.53)

25.44*
(10.55)

7.49
(12.07)

7.81
(12.17)

3.58
(5.62)

3.53
(5.28)

4.71
(7.28)

4.90
(7.32)

58.67***
(7.95)

58.69***
(7.97)

92.13***
(9.22)

92.15***
(9.24)

Immigrant
status

-32.58**
(10.82)

-28.36*
(11.91)

-34.10*
(14.55)

-34.03*
(14.77)

-28.87*
(11.69)

-29.19*
(11.70)

-25.94**
(10.07)

-25.41*
(10.36)

-37.85***
(6.06)

-37.59***
(6.27)

-61.53***
(11.35)

-61.61***
(11.59)

Feelings of
belongingness*
gender

-2.12
(2.07)

-.46
(3.21)

-.68
(3.58)

2.11
(3.82)

-1.49
(2.14)

-.67
(2.92)

Feelings of
belongingness*
SES

.06
(.69)

.66
(.73)

1.26
(1.32)

-.07
(1.78)

.27
(.68)

.48
(.91)

Feelings of
belongingness*
home language

-1.05
(3.71)

-4.27
(4.46)

-.50
(3.85)

-1.13
(4.88)

-1.07
(2.85)

.14
(3.71)

Feelings of
belongingness*
immigrant
status

9.92
(5.78)

8.94
(8.00)

5.60
(6.07)

6.12
(6.63)

4.06
(2.90)

8.47*
(3.85)

Adjusted R2

.34

.35

.37

.37

.26

.27

Note. SES = socioeconomic status; standard errors of regression coefficients are in parentheses.
*p < .05, **p < .01, ***p <.001

81

.31

.31

.46

.46

.47

.47

Table 2.17
Hypotheses Supported by Country
Country

Subject

Botswana

Math

Ghana
South
Africa

H1

H2a

H2b

H2c

H2d

X

X

Science

X

X

Math

X

X

Science

X

X

Math
Science

X

X

X

X

X

X

X

H3a

H3b

H3c

H3d

X

Note. An X indicates that the designated hypothesis was supported for the respective country in
the respective subject. Remember that main effects for feelings of belongingness (H1), gender
(H2a), SES (H2b), home language (H2c), and immigrant status (H2d) on achievement were
predicted. Interaction effects for feelings of belongingness and gender (H3a), SES (H3b), home
language (H3c), and immigrant status (H3d) on achievement were also predicted.
In terms of overall model fit, when just the controls were entered first, Model 1 explained
33% of the variance (i.e., adjusted R2) in mathematics achievement and 37% of the variance in
science achievement in Botswana. In Ghana, the statistical controls accounted for 24% of the
variance in mathematics and 29% of the variance in science achievement. In South Africa, the
statistical controls accounted for 31% of the variance in mathematics and 29% of the variance in
science achievement.
Next, when the main effects were added, Model 2 explained 35% of the variance in
mathematics and 38% of the variance in science achievement in Botswana. In Ghana, the
controls and main effects accounted for 26% of the variance in mathematics and 31% of the
variance in science achievement. In South Africa, the controls and main effects accounted for
46% of the variance in mathematics and 47% of the variance in science achievement. The
change in R2 between Model 1 and Model 2 was not statistically significant in Botswana for
mathematics (∆R2 = 0.02, p > .05) or science (∆R2 = 0.01, p > .05) or in Ghana for mathematics

82

(∆R2 = 0.02, p > .05) or science (∆R2 = 0.02, p > .05), however it was statistically significant in
South Africa for mathematics (∆R2 = 0.15, p < .01) and science (∆R2 = 0.18, p < .01).
Finally, when the interaction terms were added, Model 3 explained 35% of the variance
in mathematics and 38% of the variance in science achievement in Botswana. In Ghana, the
controls, main effects, and interactions together accounted for 27% of the variance in
mathematics and 31% of the variance in science achievement. In South Africa, the controls,
main effects, and interactions together accounted for 46% of the variance in mathematics and
47% of the variance in science achievement. The change in R2 between Model 2 and Model 3
was not statistically significant in Botswana for mathematics (∆R2 = 0.00, p > .05) or science
(∆R2 = 0.00, p > .05), in Ghana for mathematics (∆R2 = 0.01, p > .05) or science (∆R2 = 0.00, p >
.05), or in South Africa for mathematics (∆R2 = 0.00, p > .05) or science (∆R2 = 0.00, p > .05). A
detailed description of the individual models and relationships used to test each of my research
questions is presented below.
RQ1: Feelings of Belongingness and Achievement
Bivariate analyses suggested that student feelings of belongingness were positively, albeit
weakly, related to achievement in two of the three countries for both subjects. In Botswana,
student feelings of belongingness was positively correlated with mathematics (r = 0.07, p < .01)
and science achievement (r = 0.08, p < .01). Positive correlations between feelings of
belongingness and achievement were also found in Ghana in both mathematics (r = 0.06, p <
.01) and science (r = 0.08, p < .01). However, in South Africa the correlation between feelings
of belongingness and achievement was not significant in mathematics (r = 0.00, p > .05) or
science achievement (r = 0.01, p > .05). These findings suggest that higher feelings of

83

belongingness appear to be associated with higher achievement in mathematics and science in
Botswana and Ghana, but not in South Africa.
Despite the presence of positive associations between feelings of belongingness and
achievement in the bivariate analyses, student feelings of belongingness were not found to be
significantly associated with mathematics and science achievement in the OLS regression
models across all three countries. In the main effects model (Model 2), when student feelings of
belongingness was added as a predictor with the controls and marginalization variables, no
statistically significant main effect was found for feelings of belongingness on achievement in
Botswana for mathematics (b = -0.54, p > .05) and science (b = -0.19, p > .05), in Ghana for
mathematics (b = -1.01, p > .05) and science (b = 0.26, p > .05), and in South Africa in
mathematics (b = -0.62, p > .05) and science (b = 0.66, p > .05). These results show that when
holding constant the controls and marginalization, students’ feelings of belongingness do not
have a direct effect on mathematics and science achievement in Botswana, Ghana, and South
Africa. Thus, H1 was not supported.
RQ2: Marginalization and Achievement
Results examining the association between marginalization and achievement will be
discussed separately for each category of marginalization (gender, SES, home language, and
immigrant status).
RQ2a: Gender and achievement. Bivariate analyses showed mixed results when
examining the point-biserial correlations14 between student gender and achievement across the
three countries. In Botswana, being a girl was positively correlated with mathematics (r = 0.09,
p < .01) and science achievement (r = 0.06, p < .01). In Ghana, being a girl was negatively
14

Point-biserial correlations are used to estimate a linear relationship between a continuous variable and a true
dichotomous variable. They are statistically and inferentially identical to the Pearson product moment correlation
used to estimate linear relationships between two continuous variables.

84

correlated with mathematics (r = -0.13, p < .01) and science achievement (r = -0.14, p < .01). In
South Africa, student gender was not significantly correlated with mathematics (r = -0.01, p >
.05) or science achievement (r = 0.01, p > .05). These findings suggest a positive association
between being a girl and achievement in Botswana, a negative association in Ghana, and no
association in South Africa.
Regression findings were also mixed regarding the main effect of student gender on
mathematics achievement across the three countries, but were statistically significant for science
achievement in all countries. In Botswana, when holding constant the controls, feelings of
belongingness, and other categories of marginalization in Model 2, no statistically significant
main effect was found for gender on mathematics achievement (b = 2.90, p > .05). However,
there was a statistically significant main effect for gender on science achievement in Botswana (b
= -11.24, p < .05). This means that on average Batswana girls score approximately 11 points
lower in science compared to boys.
In Ghana, when holding constant the controls, feelings of belongingness, and other
categories of marginalization in Model 2, a statistically significant main effect was found for
gender on mathematics achievement (b = -20.95, p < .001) and science achievement (b = -22.81,
p < .001). This means that on average Ghanaian girls score approximately 21 points lower in
mathematics compared to boys and 23 points lower in science.
In South Africa, when holding constant the controls, feelings of belongingness, and other
categories of marginalization in Model 2, no statistically significant main effect was found for
gender on mathematics achievement (b = -1.73, p > .05). However, a statistically significant
main effect was found for gender on science achievement in South Africa (b = -13.58, p < .05).

85

This means that on average South African girls score approximately 14 points lower in science
compared to boys after accounting for other relevant differences in their background.
Overall, the OLS regression results show that being a girl was negatively associated with
science achievement in all three countries and was negatively associated with mathematics
achievement in Ghana. Therefore, H2a was supported in all three countries in science and in
Ghana in mathematics, but was not supported in Botswana and South Africa in mathematics.
RQ2b: SES and achievement. Bivariate analyses suggested that student SES was
positively related to achievement in all countries for both subjects. In Botswana SES was
positively correlated with mathematics (r = 0.08, p < .01) and science achievement (r = 0.10, p <
.01), as well as in Ghana in mathematics (r = 0.10, p < .01) and science (r = 0.15, p < .01), and in
South Africa in mathematics (r = 0.39, p < .01) and science (r = 0.39, p < .01). These findings
suggest that higher SES appears to be associated with higher achievement in mathematics and
science in all three countries.
The positive association between SES and achievement only held in one of the three
countries when examining the main effect of student SES on mathematics and science
achievement in the OLS regression models. In Botswana, when holding constant the controls,
feelings of belongingness, and other categories of marginalization in Model 2, no statistically
significant main effect was found for SES on mathematics (b = 0.14, p > .05) or science
achievement (b = 0.66, p > .05). The same non-significant results were found in Ghana for
mathematics (b = -0.37, p > .05) and science achievement (b = 1.15, p > .05).
In contrast, in South Africa, when holding constant the controls, feelings of
belongingness, and other categories of marginalization in Model 2, a statistically significant main
effect was found for SES on mathematics (b = 7.27, p < .001) and science achievement (b =

86

10.04, p < .001). This means that on average a 1 unit increase in SES is associated with a 7 point
increase in mathematics achievement and a 10 point increase in science achievement in South
Africa.
Overall, South Africa was the only country where student SES was positively associated
with mathematics and science achievement. Therefore, H2b was supported in South Africa in
mathematics and science, but it was not supported in Botswana and Ghana in either subject.
RQ2c: Home language and achievement. Bivariate analyses suggested that frequently
speaking the language of the test at home was positively related to achievement in two of the
three countries for both subjects. In Botswana frequent speaking of the language of the test at
home was positively correlated with mathematics (r = 0.09, p < .01) and science achievement (r
= 0.07, p < .01). In Ghana student home language was not significantly correlated with
mathematics (r = 0.01, p > .05) or science achievement (r = 0.01, p > .05). In South Africa
frequent speaking of the language of the test at home was positively correlated with mathematics
(r = 0.37, p < .01) and science achievement (r = 0.43, p < .01). These findings suggest that
frequent speaking of the language of the test at home appears to be associated with higher
achievement in mathematics and science in Botswana and South Africa, but not in Ghana.
The positive associations between student home language and achievement found in the
bivariate analyses held fairly well when examining the main effect of student home language on
mathematics and science achievement across the three countries. In Botswana, when holding
constant the controls, feelings of belongingness, and other categories of marginalization in
Model 2, a statistically significant main effect was found for home language on mathematics (b =
25.38, p < .05) but not science achievement (b = 7.49, p > .05). This means that Batswana
students who frequently speak the language of the test at home score approximately 25 points

87

higher in mathematics compared to students who do not frequently speak the language of the test
at home.
In Ghana, when holding constant the controls, feelings of belongingness, and other
categories of marginalization in Model 2, no statistically significant main effect was found for
home language on mathematics (b = 3.58, p > .05) or science achievement (b = 4.71, p > .05).
In South Africa, when holding constant the controls, feelings of belongingness, and other
categories of marginalization in Model 2 for South Africa, a statistically significant main effect
was found for home language on mathematics (b = 58.67, p < .001) and science achievement (b
= 92.13, p < .001). This means that on average South African students who frequently speak the
language of the test at home score approximately 59 points higher in mathematics and 92 points
higher in science compared to students who do not frequently speak the language of the test at
home.
Overall, being a student who frequently speaks the language of the test at home was
positively associated with mathematics achievement in Botswana and South Africa, as well as
science achievement in South Africa. Therefore, H2c was supported in Botswana for
mathematics and in South Africa for mathematics and science, but was not supported in
Botswana for science or in Ghana for mathematics or science.
RQ2d: Immigrant status and achievement. Bivariate analyses suggested that being an
immigrant was negatively related to achievement in all three countries for both subjects. In
Botswana being an immigrant was negatively correlated with mathematics (r = -0.14, p < .01)
and science achievement (r = -0.16, p < .01), as well as in Ghana in mathematics (r = -0.09, p <
.01) and science (r = -0.11, p < .01), and in South Africa in mathematics (r = -0.34, p < .01) and

88

science (r = -0.36, p < .01). These findings suggest that being an immigrant is associated with
lower achievement in mathematics and science in all three countries.
The OLS regression results are consistent with the bivariate results, as student immigrant
status was found to be significantly associated with mathematics and science achievement across
all three countries when modeling main effects. In Botswana, when holding constant the
controls, feelings of belongingness, and other categories of marginalization in Model 2, a
statistically significant main effect was found for immigrant status on mathematics (b = -32.58, p
< .01) and science achievement (b = -34.10, p < .05). This means that on average Batswana
immigrant students score approximately 33 points lower in mathematics compared to nonimmigrants and 34 points lower in science. The same pattern of results was found in Ghana in
mathematics (b = -28.07, p < .05) and science (b = -25.94, p < .01), as well as in South Africa in
mathematics (b = -37.85, p < .001) and science (b = -61.53, p < .001). This means that on
average Ghanaian immigrant students score approximately 28 points lower in mathematics and
26 points lower in science compared to non-immigrants, while South African immigrant students
score 38 points lower in mathematics and 62 points lower in science. Overall, these findings
demonstrate that being an immigrant is associated with lower mathematics and science
achievement in Botswana, Ghana, and South Africa. Therefore, H2d was supported in all
countries in both subjects.
RQ3: Feelings of Belongingness as a Buffer
In the vast majority of the interaction models, student feelings of belongingness was not
found to significantly buffer the negative relationship between marginalization and academic
achievement. Details and the one exception to this pattern of results are explained below

89

separately for the buffering effect of feelings of belongingness on the relationship between each
category of marginalization and achievement.
RQ3a: Feelings of belongingness and gender. Student feelings of belongingness did
not significantly buffer the relationship between student gender and achievement for any country
in either subject. In Botswana, when holding the controls, marginalization, and other
interactions constant in Model 3, no statistically significant interaction was found for feelings of
belongingness and gender on mathematics (b = -2.12, p > .05) or science achievement (b = -0.46,
p > .05). The same pattern of non-significant results was found in Ghana for the interaction
between feelings of belongingness and gender on mathematics (b = -0.68, p > .05) and science
achievement (b = 2.11, p > .05), as well as in South Africa in mathematics (b = -1.49, p > .05)
and science (b = -0.67, p > .05). Thus, H3a was not supported in any country in either subject.
RQ3b: Feelings of belongingness and SES. Student feelings of belongingness did not
significantly buffer the relationship between student SES and achievement for any country in
either subject. In Botswana, when holding the controls, marginalization, and other interactions
constant in Model 3, no statistically significant interaction was found for feelings of
belongingness and SES on mathematics (b = 0.06, p > .05) or science achievement (b = 0.66, p >
.05). The same pattern of non-significant results was found in Ghana for the interaction between
feelings of belongingness and SES on mathematics (b = 1.26, p > .05) and science achievement
(b = -0.07, p > .05), as well as in South Africa in mathematics (b = 0.27, p > .05) and science (b
= 0.48, p > .05). Thus, H3b was not supported in any country in either subject.
RQ3c: Feelings of belongingness and home language. Student feelings of
belongingness did not significantly buffer the relationship between student home language and
achievement for any country in either subject. In Botswana, when holding the controls,

90

marginalization, and other interactions constant in Model 3, no statistically significant interaction
was found for feelings of belongingness and home language on mathematics (b = -1.05, p > .05)
or science achievement (b = -4.27, p > .05). The same pattern of non-significant results was
found in Ghana for the interaction between feelings of belongingness and home language on
mathematics (b = -0.50, p > .05) and science achievement (b = -1.13, p > .05), as well as in South
Africa in mathematics (b = -1.07, p > .05) and science (b = 0.14, p > .05). Thus, H3c was not
supported in any country in either subject.
RQ3d: Feelings of belongingness and immigrant status. Student feelings of
belongingness significantly buffered the relationship between student immigrant status and
achievement in one of the three countries in one subject. In Botswana, when holding the
controls, marginalization, and other interactions constant in Model 3, no statistically significant
interaction was found for feelings of belongingness and immigrant status on mathematics (b =
9.92, p > .05) or science achievement (b = 8.94, p > .05). The same pattern of non-significant
results was found in Ghana for the interaction between feelings of belongingness and immigrant
status on mathematics (b = 5.60, p > .05) and science achievement (b = 6.12, p > .05), as well as
in South Africa in mathematics (b = 4.06, p > .05).
There was, however, one exception to this pattern of non-significant results. In South
Africa, when holding the controls, marginalization, and other interactions constant in Model 3, a
statistically significant interaction was found between feelings of belongingness and immigrant
status on science achievement (b = 8.47, p < .05). As depicted in Figure 2.2, the interaction
between feelings of belongingness and immigrant status is an ordinal interaction, whereby the
regression lines do not cross within the observed range of data (Cohen, Cohen, West, & Aiken,
2003). This means that although higher feelings of belongingness significantly increase science

91

achievement for immigrant students, their achievement never surpasses that of their nonimmigrant peers. Overall, H3d was supported in South Africa in science only. H3d was not
supported in Botswana and Ghana in either subject or in South Africa in mathematics.
400
390

Science Achievement

380
370
360
350

Non-Immigrant

340

Immigrant

330
320
310
300
Low Feelings of
Belongingness

High Feelings of
Belongingness

Figure 2.2. Ordinal interaction between feelings of belongingness and immigrant status in South
Africa.
RQ4: Cross-National Comparisons
Comparisons in the relationships analyzed between feelings of belongingness,
marginalization, and academic achievement between Botswana, Ghana, and South Africa were
made throughout the RQ1-3 subsections above and are summarized in Table 2.16. Notable
similarities and differences in the results across the three sub-Saharan African countries will be
highlighted here.
Examining the results across the countries, the findings for RQ1 and RQ3 had the
greatest overlap. One striking similarity is that H1 was not supported in any country in either
92

subject. Student feelings of belongingness were found to have no significant main effect on
mathematics or science achievement across Botswana, Ghana, and South Africa. Another
similarity is that H3a, H3b, and H3c were not supported in any country in either subject. The
interactions between student feelings of belongingness and gender, SES, and home language on
mathematics and science achievement were not significant across Botswana, Ghana, and South
Africa.
Similarities across countries diverge considerably when comparing the results for RQ2.
A notable difference between the three countries is that South Africa is the only country where
the main effect of marginalization on science achievement was significant for all four categories
of marginalization (gender, SES, home language, and immigrant status). In contrast, the main
effect of marginalization on science achievement was only significant for two of the four
categories of marginalization (gender and immigrant status) in Botswana and Ghana. More
cross-national differences emerge when comparing the main effect of marginalization on
mathematics achievement. Similar to the results for science, in South Africa the main effect of
marginalization on mathematics achievement was significant for more categories of
marginalization than Botswana and Ghana. In South Africa the main effects of three categories
of marginalization (SES, home language, and immigrant status) on mathematics achievement
were significant, whereas significant main effects were only found for two categories of
marginalization in the other countries (SES and immigrant status in Botswana and gender and
immigrant status in Ghana).
Finally, perhaps the most salient cross-country difference is that South Africa was the
only country where a significant interaction between feelings of belongingness and
marginalization was found. The results show that student feelings of belongingness significantly

93

buffer the relationship between immigrant status and science achievement in South Africa, but
all other interaction terms between feelings of belongingness and marginalization were nonsignificant.
Discussion
The findings of this study as illustrated above provide some insight into the relationships
between marginalization, feelings of belongingness, and academic achievement for eighth grade
students in sub-Saharan Africa. This section connects my findings to previous literature,
speculates why some hypotheses were not supported, posits limitations of the study, and makes
recommendations for future research.
Marginalization
The findings of this study produced mixed results regarding the relationships between
categories of marginalization and academic achievement. Some categories of marginalization
operated as expected in relation to mathematics and science achievement, such as immigrant
status, but others operated in ways that were not predicted, such as SES. The relationship
between each category of marginalization and mathematics and science achievement will be
discussed in more detail below.
Gender. The presence of a main effect of gender on academic achievement in Ghana for
both subjects is consistent with previous research demonstrating that girls are most academically
at risk in Ghana compared to Botswana and South Africa (Etse-Quest et al., 2010; UNESCO,
2012). The presence of a main effect of gender on science achievement in Botswana and South
Africa is also consistent with previous research that demonstrates an achievement gender gap in
these two sub-Saharan African countries as well (Etse-Quest et al., 2010; Hungi, 2011), however
the absence of a significant main effect of gender on mathematics achievement for these

94

countries was surprising. The direction of the effect of gender on achievement was as expected
in Ghana and South Africa, yet in Botswana there has been mixed results with some studies
showing girls outperforming boys (e.g., Etse-Quest et al., 2010) and others with boys
outperforming girls (e.g., Hungi, 2011). The present study supports the former, finding that
female students scored significantly lower than males in all three countries in science.
Further research should continue to investigate the relationship between gender and
mathematics achievement in the three countries studied here. Examining why no main effect of
gender was found in Botswana and South Africa in mathematics would be especially valuable. If
the gender gap in these countries is indeed closing and being a girl is no longer a disadvantage in
mathematics, than this study shows promising evidence for the education systems of Botswana
and South Africa. Further research investigating why the gender gap in these contexts is closing
could aid other countries in achieving greater gender equality in secondary schools.
SES. It is surprising that South Africa was the only country where a main effect of SES
on achievement was found given the well-established relationship between SES and academic
outcomes in low-income countries (Chiu, 2007; Chudgar & Luschei, 2009; Heyneman, 1976a,
1976b, 1979; Heyneman & Loxley, 1983; Nonoyama-Tarumi & Willms, 2010; Schiller,
Khmelkov, & Wang, 2002). Although all three countries are considered to be developing
nations characterized by low GDPs (IMF, 2013), South Africa has the greatest within-country
inequality in the distribution of family income, as it has the second highest GINI index in the
world (CIA, 2013). South Africa also has one of the largest score distributions from the 5th to
95th percentile for mathematics and science of all participating TIMSS 2011 countries (Mullis et
al., 2012; Martin et al., 2012), signaling large disparities in performance across students.
Together, this wide variation in SES and in achievement may explain why the main effect of

95

SES on achievement was only statistically significant for South Africa due to the large amount of
variance in both variables. Although Botswana has the third highest GINI index in the world,
Ghana is ranked much lower as sixty-third (CIA, 2013). Additionally, neither Botswana nor
Ghana had as much within-country inequality in TIMSS scores as South Africa (Mullis et al.,
2012; Martin et al., 2012). This means that compared to South Africa, Botswana and Ghana
have less variance in these variables to be able to find a significant effect.
The lack of a main effect of SES on achievement in Botswana and Ghana could also be
due to measurement issues. The issue of measuring SES is complex because there is a lack of
consensus or best practices on how to account for SES in cross-national educational research,
and this variability in approaches leads to variability in results and interpretation (Chudgar,
Luscheli, & Fagioli, 2014). Furthermore, measuring SES in developing countries poses
additional challenges because many indicators are unreliable and insensitive in the context of
developing world economies (Doocy & Burnham, 2006). In this study, the HER scale was used
as a proxy for SES and was based on the number of books in the home, parent education, and
home study supports. Additional components of SES that were not included in the proxy used in
this study might include family income, parent occupation, and additional home possessions
(e.g., computer, Internet, refrigerator, television, radio, bicycle, livestock, and land). Further
research should determine if other measures of SES are related differently to achievement in the
TIMSS 2011 data for Botswana and Ghana, similar to the approach taken by Chudgar et al.
(2014) in their cross-national comparison of the relative importance of variables commonly used
to measure SES models using TIMSS 2007 eighth grade data.
On the other hand, if the finding here of no main effect of SES on achievement is not a
variance or measurement issue this could mean that SES is indeed unrelated to mathematics and

96

science achievement in Botswana and Ghana. Similar to the lack of gender effects discussed
above, this could be evidence of progress towards more equitable academic achievement in these
countries. Further research into the relationship between SES and achievement in Botswana and
Ghana would be helpful to confirm or refute this finding.
Home language. The main effect of home language on mathematics in Botswana and on
mathematics and science in South Africa is consistent with previous research demonstrating
higher academic performance when a student’s home language is consistent with the language of
instruction and testing (Herbert et al., 2002; Martin et al., 2008; Papanastasiou, 2000). It is
surprising that this main effect was not found in Ghana as well, given that Opoku-Amankwa
(2009) found detrimental effects of English-only instruction on learning in Ghana. However, of
the three countries, the vast majority of previous research on home language and achievement
has been conducted in South Africa (Howie, 2001; McKay & Chick, 2001; PANSALB, 2000;
Posel & Zeller, 2011; Probyn, 2006; Probyn et al., 2002), so perhaps more work is needed in the
Ghanian context to better understand the nuanced relationship between home language and
academic achievement. Additionally, students who speak a language other than the language of
instruction and testing are more likely to fail and drop out of school (Brock-Utne, 2001), so
perhaps the effect that Opoku-Amankwa (2009) found at the primary level was not found here at
the secondary level because students who were struggling in school due to language challenges
had already dropped out and were not represented in the eighth grade sample. Additional
research would be helpful to inform our understanding of the relationship between home
language and achievement in secondary education, most notably in Ghana.
Immigrant status. The main effect of immigrant status on both mathematics and science
achievement found in all three countries is consistent with previous research demonstrating an

97

achievement gap between immigrant and non-immigrant students (Brown, 2006; OECD, 2007;
Schnepf, 2004). Furthermore, this finding suggests that the documented presence of xenophobic
attitudes in Botswana, Ghana, and South Africa (Campbell, 2003; Nyamjoh, 2006; Kerting,
2009; van Dijk, 2002; Whitacker, 2005) can create an educational disadvantage for immigrant
students. This is an important issue for future research and educational policy to address to
ensure that all students can succeed in school, regardless of their immigrant status.
Feelings of Belongingness
Despite the theoretical framework and conceptual model developed above suggesting that
feelings of belongingness have a direct and buffering effect on academic achievement, H1 and
H3 were largely not supported by my analysis. The absence of a main effect of feelings of
belongingness on mathematics and science achievement means that student feelings of
belongingness do not seem to play an important role relative to the other independent variables
in the model (the controls and marginalization variables). This finding is inconsistent with
previous literature that found a direct effect of feelings of belongingness on academic
achievement (Anderman, 2003; Kennedy & Tuckman, 2013). Furthermore, the absence of
significant interaction terms between feelings of belongingness and marginalization on
achievement suggests that student feelings of belongingness do not appear to buffer the negative
relationships between gender, SES, and home language in any of the countries analyzed here.
Limitations associated with the measurement of student feelings of belongingness in this
study may help explain the absence of significant main and interaction effects. Although TIMSS
2011 data was desirable for my study because it afforded me the opportunity to analyze and
compare multiple relationships incorporating multiple student-level, teacher-level, and schoollevel variables across multiple sub-Saharan African countries, its measurement of feelings of

98

belongingness is limited. Due to the realities of what was available in the TIMSS 2011 data, the
student feelings of belongingness variable was based on student responses to one item asking the
extent to which they feel like they belong at their school. While this one item is the closest
variable that the TIMSS 2011 data has to the psychological construct of feelings of
belongingness, a scale designed to capture the nuanced construct of feelings of belongingness
would have been more ideal. It is more customary in psychological literature on student feelings
of belongingness to use a scale to measure this construct, such as the 18-item Psychological
Sense of School Membership Scale (Goodenow, 1993). Future research examining the
relationships between feelings of belongingness, marginalization and achievement might
consider using a more detailed measure of feelings of belongingness than the one used here.
Cross-cultural generalizability is a related issue to the limitations of my measure of
feelings of belongingness. The psychological theory that informs my conceptualization of the
construct of feelings of belongingness comes from Western scholars, and in this study I am
applying it to the global South by examining sub-Saharan African countries. The application of
Western psychological theories to African contexts can be problematic and in some cases is
inappropriate (e.g., Cole & Bruner, 1971; Cole & Scribner, 1977). Future research may wish to
explore the conceptualization and measurement of the construct of belongingness from the
perspective of Batswana, Ghanaian, and South African students to understand how student
feelings of belongingness may or may not relate to academic achievement for marginalized
students in those particular contexts.
An additional factor to consider in light of the non-significant results is the way that the
direct and indirect influence of feelings of belongingness was modeled in this study. Due to the
lack of consensus in existing literature on whether feelings of belongingness are directly or

99

indirectly related to student achievement, this study aimed to provide some clarity by modeling
both a direct effect and an indirect effect through an interaction with marginalization. The lack
of significant findings here may not mean that there is no relationship at all between feelings of
belongingness, marginalization, and academic achievement. Instead, the significant positive
bivariate correlations between feelings of belongingness and mathematics and science
achievement would suggest that there is a relationship, but the nuanced relationships between
these variables may be incorrectly specified by the models. Previous research has pointed to the
roles that student engagement and motivation might play as moderators of the relationship
between feelings of belongingness and achievement (Anderson & Freeman, 2004; Osterman,
2000). As such, future research should consider how student engagement and motivation impact
the relationship between feelings of belongingness, marginalization, and achievement.
Extending the present study by incorporating students engagement and/or motivation could
further our understanding of the psychological process of how feelings of belongingness may
indirectly impact the academic achievement of marginalized students.
The one exception to the pattern of non-significant findings concerning student feelings
of belongingness was H3d regarding the buffering effect of feelings of belongingness on the
relationship between immigrant status and academic achievement, which was supported in South
Africa for science achievement. This finding suggests that even though immigrant students in
South Africa are at risk of lower science achievement, positive feelings of belongingness help
buffer the negative effect of immigrant status on achievement. One possible implication of this
finding is that enhancing immigrant students’ feelings of belongingness may have the potential
to improve their academic outcomes. However, because the relationship was an ordinal
interaction, it appears that feelings of belongingness can never completely make up for the

100

educational disadvantage that immigrant students have. For this reason, further research and
educational policy initiatives should work to identify additional ways to improve the academic
achievement of marginalized students in sub-Saharan African countries.
Additional Limitations and Future Directions
In addition to the limitations and future directions already discussed, there are some
broader ways to extend this study. First, while the present study was an attempt to analyze all
cross-country markers of marginalization provided by the TIMSS 2011 data, the categories of
marginalization that were included was by no means exhaustive. This is a common challenge of
research on marginalization in education because national data are rarely detailed enough to
identify all the different types of marginalized groups (UNESCO, 2010). Due to the
measurement limitations of the TIMSS data, additional categories of marginalization such as
disability status, race, ethnicity, religious affiliation, and sexual orientation were not able to be
included in this analysis. Future research using other data sources should explore how student
feelings of belongingness might impact the academic experiences and corresponding
achievement of students from these and other marginalized groups to examine if there are any
similarities and differences across marginalized groups.
A related extension of this study is to explore the nuanced and overlapping nature of
marginalization in relation to feelings of belongingness and academic achievement. Similar to
the previous discussion about the potential cross-cultural differences in the understanding and
measurement of student feelings of belongingness, the definition of particular categories of
marginalization may not necessarily be consistent across different contexts and this is a
limitation of using large cross-national datasets. Instead of taking presupposed categories of
marginalization as was done here, an alternative approach to studying the relationships between

101

feelings of belongingness, marginalization, and academic achievement would be to see what
categories of marginalization emerge within given educational contexts through classroom
observation research for example. In addition to examining the cultural relevance of categories
of marginalization in school, future research could also explore the overlapping nature of
marginalization in relation to feelings of belongingness by examining for instance the
educational experiences of female immigrant students. Having multiple markers of
marginalization can create greater educational disadvantage and social isolation by making
students experience what Purdie-Vaughns and Eibach (2008) term as “intersectional invisibility”
(i.e., possessing multiple markers of marginalization renders a person invisible relative to those
belonging to a single category of marginalization). Perhaps the buffering effect of feelings of
belongingness on the relationship between marginalization and academic achievement is stronger
for students experiencing intersectional invisibility.
A final recommended extension of this study is geographically to include other countries
beyond the three examined here and other world regions beyond sub-Saharan Africa. Given the
concentration of previous research on student feelings of belongingness and academic
achievement in the U.S., cross-national and regional comparisons would be interesting to see if
student feelings of belongingness operate differently in the global North versus South. Or
perhaps different results would be found in contexts outside of sub-Saharan Africa where
marginalization is less widespread.
Conclusion
Overall, this study challenges us to consider how student feelings of belongingness and
marginalization operate independently and together to impact academic achievement in subSaharan Africa. This study is among the first to empirically test the buffering effect of feelings

102

of belongingness on the relationship between marginalization and academic achievement. While
student feelings of belongingness were only found to buffer the relationship between one of the
four categories of marginalization and achievement in one of the two subjects in one of the three
countries (immigrant status and mathematics in South Africa), this study can serve as a launch
point for future work continuing to explore the nuanced relationship between student feelings of
belongingness, marginalization, and academic achievement in ways that address the study’s
limitations. Future research in developing and developed contexts can inform the potential of
feelings of belongingness to advance the academic achievement of marginalized students.

103

APPENDIX

104

Statistical Equations
The following statistical equations specify the iterative ordinary least squares multiple
regressions that were conducted using all five plausible values as the dependent variable. The
models were estimated separately for mathematics and science in each of the three countries.
Note that the interaction terms in grey represent predictors that were added to the conceptual
equations (see Data Analysis section) when actually running the models in order to correctly
specify the model statistically when multiple categorical independent variables were present.
Specifically, these interaction terms are needed in the models because dummy variables are
needed to represent all cross-categories otherwise the regression results are biased (Cohen et al.,
2003). They are listed in grey because they are not the primary variables of interest to answer
my research questions.
Model 1: Y = β0 + βaage + βhhmwork + βllike + βvvalue + βcconfident + βmtchfemale + βgtchage
+ βptchexp + βdtchedu + βuurbanicity
Model 2: Y = β0 + βaage + βhhmwork + βllike + βvvalue + βcconfident + βmtchfemale + βgtchage
+ βptchexp + βdtchedu + βuurbanicity + βffemale + βsSES + βhhlang + βiimmig +
βbbelong + βfhfemale*hlang + βfifemale*immig + βhihlang*immig +
βfmfemale*tchfemale + βhmhlang*tchfemale + βimimmig* tchfemale + βfhifemale*
hlang* immig + βfhmfemale*hlang*tchfemale + βfimfemale*immig*tchfemale +
βhimhlang* immig*tchfemale + βfhimfemale* hlang* immig*tchfemale
Model 3: Y = β0 + βaage + βhhmwork + βllike + βvvalue + βcconfident + βmtchfemale + βgtchage
+ βptchexp + βdtchedu + βuurbanicity + βffemale + βsSES + βhhlang + βiimmig +
βbbelong + βfbfemale* belong + βsbSES* belong + βhbhlang*belong + βibimmig*
belong + βfhfemale*hlang + βfifemale*immig + βhihlang*immig +
βfmfemale*tchfemale + βhmhlang*tchfemale + βimimmig* tchfemale + βfhifemale*
hlang* immig + βfhmfemale*hlang*tchfemale + βfimfemale*immig*tchfemale +
βhimhlang* immig*tchfemale + βfhimfemale* hlang* immig*tchfemale
Where Y is mathematics or science achievement, β0 is the intercept, age is student age, hmwork is
student time spent on mathematics or science homework per week, like is student likes
mathematics or science, value is student values mathematics or science, confident is student
confidence in mathematics or science, tchfemale is teacher gender, tchage is teacher age, tchexp
is teacher experience, tchedu is teacher education level, urbanicity is school urbanicity, belong is
student feelings of belongingness, female is student gender, SES is student socioeconomic status,
hlang is student home language, immig is student immigrant status, and e is the error term.

105

REFERENCES

106

REFERENCES

Abdi, A. A., Puplampu, K. P., & Dei, G. J. S. (Eds.). (2006). African education and
globalization: Critical perspectives. Lanham, MD: Lexington Books.
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions.
Newbury Park, NJ: Sage.
Akyüz, G., & Berberoğlu, G. (2010). Teacher and classroom characteristics and their relation to
mathematics achievement of the students in the TIMSS. New Horizons in Education,
58(1), 77-95.
Alexander, L., & Simmons, J. (1975). The determinants of school achievement in developing
countries: The educational production function (Staff Working Paper 201). Washington,
DC: World Bank.
Alexander, N. (2000). English unassailable but unattainable: The dilemma of language policy in
South African education (PRAESA occasional papers no. 3). Cape Town, South Africa:
PRAESA.
Ammermüller, A., Heijke, H., & Wöβmann, L. (2005). Schooling quality in Eastern Europe:
Educational production during transition. Economics of Education Review, 24, 579-599.
Anderman, L. H. (2003). Academic and social perceptions as predictors of change in middle
school students’ sense of school belonging. The Journal of Experimental Education,
72(1), 5-22.
Anderman, L. H., & Freeman, T. M. (2004). Students’ sense of belonging in school. In M. Maehr
& P. Pintrich (Eds.), Advances in motivation and achievement: Motivating students,
improving schools: The legacy of Carol Midgley (Vol. 13, pp. 27-63). Greenwhich, CT:
JAI Press.
Anyon, J. (1981). Social class and school knowledge. Curriculum Inquiry, 11 , 3-42.
Appleyard, R. (1992). International migration and development – An unresolved relationship.
International Migration, 30(3/4), 251-266.
Aronowitz, S., & Giroux, H. (1985). Reproduction and resistance in radical theories of
schooling. In S. Aronowitz & H. Giroux (Eds.), Education under siege: The conservative,
liberal, and radical debate over Schooling (pp. 69-114). Boston, MA: Bergen & Garvey.
Arua, A. E., & Magocha, K. (2002). Patterns of language use and language preference of some
children and their parents in Botswana. Journal of Multilingual and Multicultural
Development, 23(6), 449-461.
107

Baker, B. (2010). Security in post-conflict Africa: The role of nonstate policing. Boca Raton, FL:
CRC Press, Taylor & Francis Group.
Baker, D. P., & Jones, D. P. (1993). Creating gender equality: Crossnational gender stratification
and mathematical performance. Sociology of Education, 66, 91–103.
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal
attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497529.
Becker, H. S. (1963). Outsiders. New York, NY: The Free Press.
Becker, H. S. (1964). The other side. New York, NY: The Free Press.
Berhanu, G. (2007). Normality, deviance, identity, cultural tracking and school achievement:
The case of Ethiopian Jews in Isreal. Scandinavian Journal of Educational Research,
49(1), 51-82.
Booker, K. C. (2007). Perceptions of classroom belongingness among African American college
students. College Student Journal, 41(1), 178-186.
Bouhilia, D. S. (2011). The quality of secondary education in the Middle East and North Africa:
What can we learn from the TIMSS results? Compare, 41(3), 327-352.
Bowles, S., & Gintis, H. (1976). Education, inequality, and the meritocracy. In S. Bowles & H.
Gintis (Eds.), Schooling in capitalist America: Educational reform and the contradictions
of economic life (pp. 102-24). New York: Basic Book.
Brock-Utne, B. (2001). Education for all – in whose language? Oxford Review of Education,
27(1), 115-134.
Brock-Utne, B. (2010). Research and policy on the language of instruction issue in Africa.
International Journal of Educational Development, 30, 636-465.
Brock-Utne, B., & Holmarsdottir, H. B. (2004). Language policies and practices in Tanzania and
South Africa: Problems and challenges. International Journal of Educational
Development, 24, 67-83.
Campbell, E. K. (2003). Attitudes of Botswana citizens toward immigrants: Signs of
xenophobia? International Migration, 41(4), 71-111.
Central Intelligence Agency. (2013). Field listings: Languages. Retrieved from The World
Factbook website: https://www.cia.gov/library/publications/the-worldfactbook/fields/2098.html

108

Central Intelligence Agency. (2013). The world factbook: Country comparison: Distribution of
family income – Gini index. Retrieved from https://www.cia.gov/library/publications/theworld-factbook/rankorder/2172rank.html
Chiswick, B. R. (1991). Speaking, reading and earnings among low-skilled immigrants. Journal
of Labor Economics, 9, 149–70.
Chiswick, B. R., & Miller, P. W. (1995). The endogeneity between language and earnings:
International analysis. Journal of Labor Economics, 13, 246–88.
Chiu, M. M. (2007). Families, economies, cultures and science achievement in 41 countries:
Country, school, and student level analyses. Journal of Family Psychology, 21, 510-519.
Chudgar, A., & Luschei, T. F. (2009). National income, income inequality, and the importance
of schools: A hierarchical cross-national comparison. American Educational Research
Journal, 46, 626-658.
Chudgar, A., Luschei, T. F., & Fagioli, L. (2014). A call for consensus in the use of student
socioeconomic status measures in cross-national research using the Trends in
International Mathematics and Science Study (TIMSS) (Research Brief ID# 17564).
Teachers College Record, June 16, 2014. Retrieved from http://www.tcrecord.org
Clark, C. R., Mercer, S. H., Zeigler-Hill, V., & Dufrene, B. A. (2012). Barriers to the success of
ethnic minority students in school psychology graduate programs. School Psychology
Review, 41(2), 176-192.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation
analysis for the behavioral sciences (3rd ed.). Mahwah, NJ: Lawrence Erlbaum
Associates.
Colclough, C., Al-Samarrai, S., Rose, P., & Tembon, M. (2003). Achieving schooling for all in
Africa: Costs, commitment, and gender. Burlington, VA: Ashgate.
Cole, M., & Bruner, J. S. (1971). Cultural differences and inferences about psychological
processes. American Psychologist, 26, 867-76.
Cole, M. & Scribner, S. (1977). Developmental theories applied to cross-cultural cognition
research. Annals New York Academy of Sciences, 285, 366-373.
Coleman, J. S., Campbell, E. Q., Hobson, C. J., McPartland, F., Mood, A. M., Weinfeld, F. D., et
al. (1966). Equality of educational opportunity. Washington, DC: U.S. Government
Printing Office.
Cuff, N. B. (1933). Relationship of socioeconomic status to intelligence and achievement.
Peabody Journal of Education, 11, 106-110.

109

Datta, D., Phillip, S., & Verma, P. K. (2009). Capacity building of a district education system:
Insights from Kenya. Education 3-13, 37(3), 203-217.
Dembélé, M., & Lefoka, P. (2007). Pedagogical renewal for quality universal primary education:
Overview of trends in sub-Saharan Africa. International Review of Education, 53(5/6),
531-553.
Dembélé, M., & Oviawe, J. (2007). Quality education in Africa: International commitments,
local challenges and responses. International Review of Education, 53(5/6), 473-483.
Doocy, S., & Burnham, G. (2006). Assessment of socio-economic status in the context of food
insecurity: Implications for field research. World Health & Population, 8(3), 32-42.
Durkheim, E. (1969). L’évolution pédagogique en France. Paris: Presses universitaires de
France.
Else-Quest, N. M., Hyde, J. S., & Linn, M. C. (2010). Cross-national patterns of gender
differences in mathematics: A meta analysis. Psychological Bulletin, 136(1), 103-127.
Engin-Demir, C. (2009). Factors influencing the academic achievement of the Turkish urban
poor. International Journal of Educational Development, 29, 17-29.
Fitch, E. F. (2002). Disability and inclusion: From labeling deviance to social valuing.
Foy, P., Arora, A., & Stanco, G. M. (Eds.). (2013). TIMSS 2011 user guide for the international
database. Chestnut hill, MA: TIMSS & PIRLS International Study Center, Boston
College.
Frempong, G., Reddy, V., & Kanjee, A. (2011). Exploring equity and quality education in South
Africa using multilevel models. Compare, 41(6), 819-835.
Georgiades, K., Boyle, M. H., & Fife, K. A. (2013). Emotional and behavioral problems among
adolescent students: The role of immigrant, racial/ethnic congruence and belongingness
in schools. Journal of Youth Adolescence, 42, 1473-1492.
Glick, J. E., & Hohmann-Marriott, B. (2007). Academic performance of young children in
immigrant families: The significance of race, ethnicity, and national origins.
International Migration Review, 41(2), 371-402.
Goodenow, C. (1993). The psychological sense of school membership among adolescents: Scale
development and educational correlates. Psychology in the Schools, 30(1), 79-90.
Goodley, D. (2001). ‘Learning difficulties’, the social model of disability and impairment:
Challenging epistemologies. Disability and Society, 16(2), 207-231.

110

Green, K. D., Forehand, R., Beck, S. J., & Vosk, B. (1980). An assessment of the relationship
among measures of children’s social competence and children’s academic achievement.
Child Development, 51, 1149-1156.
Hagerty, B. M. K., Lynch-Sauer, J., Patusky, K. L., Bouwsema, M., & Collier, P. (1992). Sense
of belonging: A vital mental health concept. Archives of Psychiatric Nursing, 6(3), 172177.
Hanushek, E., & Luque, J. A. (2002). Efficiency and equity in schools around the world (NBER
Working Paper 8949). Cambridge, MA: National Bureau of Economic Research.
Herbert, W. M., Hau, T. K., & Kong, K. C. (2002). Multilevel causal ordering of academic selfconcept and achievement: Influence of language of instruction (English compared with
Chinese) for Hong Kong students. American Educational Research Journal, 39, 727–63.
Heyneman, S.P. (1976a). A brief note on the relationship between socio-economic status and test
performance among Ugandan primary school children. Comparative Education Review,
20(1), 42-47.
Heyneman, S. P. (1976b). Influences on academic achievement: A comparison of results from
Uganda and more industrialized settings. Sociology of Education, 49(3), 200-211.
Heyneman, S.P. (1979). Why impoverished children do well in Ugandan schools. Comparative
Education, 15(2), 175-185.
Heyneman, S. P. (2009). The failure of education for all as a political strategy. Prospects, 39, 510.
Heyneman, S.P., & Loxley, W. (1983). The effect of primary-school quality on academic
achievement across twenty-nine high- and low-income countries. American Journal of
Sociology, 88, 1162-1194.
Holley, C. E. (1916). The relationship between persistence in school and home conditions. In G.
M. Whipple (Ed.), The 15th yearbook of the National Society for the Study of Education
(Pt. 2). Bloomington, IL: Public School Publishing.
Howie, S.J. (2001) Mathematics and Science Performance in Grade 8 in South Africa
1998/1999. Pretoria, South Africa: Human Sciences Research Council.
Hungi, N. (2011). Accounting for variations in the quality of primary school education (Working
Paper 7). Paris, France: SACMEQ.
International Association for the Evaluation of Educational Achievement. (2012). International
database analyzer (version 3.0). Hamburg, Germany: IEA Data Processing and Research
Center.

111

International Association for the Evaluation of Educational Achievement. (2013). Help Manual
for the IDB Analyzer (Version 3.1). Hamburg, Germany: Author.
International Monetary Fund. (2013). World economic and financial surveys: World economic
outlook April 2013: Hopes, realities, risks. Washington, DC: Author.
Jewkes, R., & Abrahams, N. (2002). The epidemiology of rape and sexual coercion in South
Africa: An overview. Social Science and Medicine: 55(7), 1231-1244.
Johnson, D., & Beinart, W. (2008). The changing landscape of education in Africa: Quality,
equality and democracy. In D. Johnson (Ed.), The changing landscape of education in
Africa: Quality, equality, and democracy (pp. 7-12). Oxford: Symposium Books.
Johnson, L. S. (2009). School contexts and student belonging: A mixed methods study of an
innovative high school. The School Community Journal, 19(1), 99-118.
Johnson, S. (1996). The contribution of large-scale assessment programmes to research on
gender differences. Educational Research and Evaluation, 2, 25-49.
Jonas, M. & Foy, P. (2012). Sample design in TIMSS and PIRLS. In M. O. Martin, & I. V. S.
Mullis (Eds.), Methods and procedures in TIMSS and PIRLS 2011 (pp. 1-21). Chestnut
Hill, MA: TIMSS & PIRLS International Study Center, Boston College.
Katz, J. (2013). The three block model of universal design for learning (UDL): Engaging
students in inclusive education. Canadian Journal of Education, 36(1), 153-194.
Kennedy, G. J., & Tuckman, B. W. (2013). An exploration of the influence of academic and
social values, procrastination, and perceived school belongingness on academic
performance. Social Psychology of Education, 16, 435-470.
Kersting, N. (2009). New nationalism and xenophobia in Africa – A new inclination? Africa
Spectrum, 44(1), 7-18.
Kreft, I., & De Leeuw, J. (1998). Introducing multilevel modeling. Thousand Oaks, CA: Sage.
Kunc, N. (1992). The need to belong: Rediscovering Maslow’s hierarchy of needs. In R. Villa,
J. Thousand, W. Stainback, & S. Stainback (Eds.), Restructuring for caring and effective
education. Baltimore, MD: Paul Brookes.
Ladd, G. W. (1990). Having friends, keeping friends, making friends, and being liked by peers in
the classroom: Predictors of children's early school adjustment? Child Development, 61,
1081-1100.
Lareau, A. (1987). Social class differences in family school relationships: The importance of
cultural capital. Sociology of Education, 60, 73-85.

112

Lemert, E. (1951). Social pathology. New York, NY: McGraw-Hill.
Lemert, E. (1972). Human deviance, social problems and social control. Englewood Cliffs, NJ:
Prentice-Hall.
Lemert, E. (1974). Beyond Mead: The societal reaction to deviance. Social Problems, 21, 457468.
Lloyd, C., & Hewett, P. (2009). Educational inequalities in the midst of persistent poverty:
Diversity across Africa in educational outcomes. Journal of International Development,
21, 1137-1151.
Martin, M. O., Mullis, I. V. S., & Foy, P. (2008). TIMSS 2007 international science report:
Findings from IEA’s trends in mathematics and science study at the eighth and fourth
grades. Chestnut Hill, MA: Boston College.
Martin, M. O., & Mullis, I. V. S. (Eds.). (2012). Methods and procedures in TIMSS and PIRLS
2011.Chestnut Hill, MA: TIMSS & PIRLS International Study Center, Boston College.
Martins, L.,& Veiga, P. (2010). Do inequalities in parents’ education play an important role in
PISA students’ mathematics achievement test score disparities? Economics of Education
Review, 29(6), 1016-1033.
Maslow, A. H. (1943). A theory of human motivation. Psychological Review, 50(4), 370-396.
Maslow, A. H. (1954). Motivation and personality. New York, NY: Harper.
Mattes, R., Taylor, D. M., McDonald, D. A., Poore, A., & Richmond, W. (1999). Still waiting
for the barbarians: South Africa attitudes towards immigrants and immigration
(Migration Policy Series No. 14). Cape Town, South Africa: Southern Africa Migration
Project.
Matza, D. (1964). Delinquency and drift. New York, NY: Wiley.
Matza, D. (1969). Becoming deviant. Englewood Cliffs, NJ: Prentice-Hall.
McKay, S., & Chick, K. (2001). Positioning learners in post-Apartheid South African schools: A
case study of selected multilingual Durban schools. Linguistics and Education, 12(4),
393-408.
Mullis, I. V. S., Martin, M. O., Foy, P., & Arora, A. (2012). TIMSS 2011 international results in
mathematics. Chestnut hill, MA: TIMSS & PIRLS International Study Center, Boston
College.

113

Mullis, I. V. S., Martin, M. O., Ruddock, G. J., O’Sullivan, C. Y., & Preuschoff, C. (2009).
TIMSS 2011 assessment frameworks. Chestnut hill, MA: TIMSS & PIRLS International
Study Center, Boston College.
National Center for Education Statistics. (2013). Countries. Retrieved from the U.S. Department
of Education Institute of Education Sciences website
https://nces.ed.gov/timss/countries.asp.
Neocosmos, M. (2006). From ‘foreign natives’ to ‘native foreigners’: Explaining xenophobia in
post-Apartheid South Africa. Dakar, Senegal: Council for the Development of Social
Science Research in Africa.
Nonoyama-Tarumi, Y., & Willms, J. D. (2010). The relative and absolute risks of disadvantaged
family background and low levels of school resources on student literacy. Economics of
Education Review, 29, 214-224.
Nyamnjoh, F. B. (2006). Insiders and outsiders: Citizenship and xenophobia in contemporary
Southern Africa. New York, NY: Zed Books.
Opoku-Amankwa, K. (2009). English-only language-in-education policy in multilingual
classrooms in Ghana. Language, Culture and Curriculum, 22(2), 121-135.
Organisation for Economic Co-Operation and Development. (2007). PISA 2006: Science
Competencies for Tomorrow's World. Paris: Author and Programme for International
Student Assessment.
Osterman, K. F. (2000). Students’ need to belong in the school community. Review of
Educational Research, 70(3), 323-367.
Oucho, J. O. (1995). International migration and sustainable human development in Eastern and
Southern Africa. International Migration, 33(1), 31-53.
PANSALB. (2000). Language Use and Language Interaction in South Africa: A Sociolinguistic
Survey. Pretoria: PANSALB.
Papanastasiou, C. (2000). Internal and external factors affecting achievement in mathematics:
Some findings from TIMSS. Studies in Educational Evaluation, 26, 1-7.
Portes, A., & Fernandez-Kelly, P. (2008). No margin for error: Educational and occupational
achievement among disadvantaged children of immigrants. The ANNALS of the American
Academy of Political and Social Science, 620, 12-36.
Portes, A., & Rumbaut, R. G. (2006). Immigrant America: A portrait (3rd ed.). Berkeley, CA:
University of California Press.

114

Portes, A., & Zhou, M. (1992). Gaining the upper hand: Economic mobility among immigrant
and domestic minorities. Ethnic and Racial Studies, 15, 491-522.
Posel, D., & Zeller, J. (2011). Home language and English language ability in South Africa:
Insights from new data. Southern African Linguistics and Applied Language Studies,
29(2), 115-126.
Probyn, M. J. (2006). Language and learning science in South Africa. Language and Education,
20(5), 391-414.
Probyn, M. J., Murray, S., Botha, L., Botya, P., Brooks, M. and Westphal, V. (2002) Minding the
gaps – an investigation into language policy and practice in four Eastern Cape districts.
Perspectives in Education 20 (1), 29-46.
Purdie-Vaughns, V., & Eibach, R. P. (2008). Intersectional invisibility: The distinctive
advantages and disadvantages of multiple subordinate-group identities. Sex Roles, 59(56), 377-391.
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data
analysis methods (2nd ed). Thousand Oaks, CA: Sage.
Rist, R. C. (1977). On understanding the process of schooling: The contributions of labeling
theory. In J. Karabel & A. H. Halsey (Eds.), Power and ideology in education (pp. 292305). New York, NY: Oxford University Press.
Schiller, K. S., Khmelkov, V. T., & Wang, X. Q. (2002). Economic development and the effects
of family characteristics on mathematics achievement. Journal of Marriage and Family,
64, 730-742.
Schnepf, S. V. (2004). How different are immigrants? A cross-country and cross-survey analysis
of educational achievement (IZA Discussion Paper). Southampton, UK: University of
Southampton, Southampton Statistical Sciences.
Sen, A. (2009). The idea of justice. Cambridge, MA: Belknap Press of Harvard University Press.
Sifuna, D. N. (2007). The challenges of increasing access and improving quality: An analysis of
universal primary education interventions in Kenya and Tanzania since the 1970s.
International Review of Education, 58(5/6), 687-699.
Sirin, S. R. (2005). Socioeconomic status and academic achievement: A meta-analytic review of
research. Review of Educational Research, 75(3), 417-453.

115

Stephens, M., Krainer, K., Jungwirth, H., Stadler, H., May, S.,Whitney, L., Spee, A. A. J.,
Wester, A., & Hodgkinson, D. (2004). Gender differences in achievement: Overview and
examples of data use in five countries. In J. H. Moskowitz & M. Stephens (Eds.),
Comparing learning outcomes: International assessment and education policy (pp. 123145). London: Routledge Falmer.
Swift, M. S., & Spivack, G. (1969). Achievement related classroom behavior of secondary
school normal and disturbed students. Exceptional Children, 35, 677-684.
Tannenbaum, F. (1938). Crime and community. New York, NY: Columbia University Press.
Taylor, A. F. (1989). Predictors of peer rejection in early elementary grades: Roles of problem
behavior, academic achievement, and teacher preference. Journal of Clinical Child
Psychology, 18(4), 360-365.
Timaeus, I. M., Simelane, S., & Letsoalo, T. (2013). Poverty, race, and children’s progress at
school in South Africa. Journal of Development Studies, 49(2), 270-284.
Tripp, A. M., Casimiro, I., Kwesiga, J., & Mungwa, A. (2009). African women’s movements:
Changing political landscapes. New York: Cambridge University Press.
United Nations. (2013). Composition of macro geographical (continental) regions, geographical
sub-regions, and selected economic and other groupings. Retrieved from United Nations
Statistics Division website: http://unstats.un.org/unsd/methods/m49/m49regin.htm
United Nations Children’s Fund. (2011). Violence against children in Tanzania: Findings from a
national survey 2009. Dar es Salaam: United Republic of Tanzania.
United Nations Educational, Scientific and Cultural Organization. (1990). The world declaration
of education for All. Paris, France: Author.
United Nations Educational, Scientific and Cultural Organization. (2010). Education for all
global monitoring report 2010: Reaching the marginalized. Paris, France: Author.
United Nations Educational, Scientific and Cultural Organization. (2012). Education for all
global monitoring report 2012: Youth and skills: Putting education to work. Paris,
France: Author.
Vandeyar, S., & Vandeyar, T. (2011). Articulating cultures: Socio-cultural experiences of Black
female immigrant students in South African schools. Gender and Behavior, 9(2), 41614188.
Van Dijk, R. (2002). Localizing anxieties: Ghanaian and Malawian immigrants, rising
xenophobia, and social capital in Botswana (ASC working paper no. 49). Leiden, the
Netherlands: Afrika-Studiecentrum.

116

Van Fleet, J. W., Watkins, K., & Greubel, L. (2012). Africa learning barometer. Retrieved from
Center for Universal Education at Brookings website:
http://www.brookings.edu/research/interactives/africa-learning-barometer
Verspoor, A. M. (2008). The challenge of learning: Improving the quality of basic education in
sub-Saharan Africa. In D. Johnson (Ed.), The changing landscape of education in Africa:
Quality, equality, and democracy (pp. 13-43). Oxford: Symposium Books.
Wentzel, K. B., & Asher, S. R. (1995). The academic lives of neglected, rejected, popular, and
controversial children. Child Development, 66, 754-763.
Whitaker, B. (2005). Citizens and foreigners: Democratization and the politics of exclusion in
Africa. African Studies Review, 48(1), 109-126.
White, K. R. (1982). The relation between socioeconomic status and academic achievement.
Psychological Bulletin, 91(3), 461-481.
Williams, K. D. (2007). Ostracism. Annual Review of Psychology, 58, 425-452.
Williams, K. D., & Nida, S. A. (2011). Ostracism: Consequences and coping. Current Directions
in Psychological Science, 20(2), 71-75.
Wöβmann, L. (2003). Schooling resources, educational institutions and student performance:
The international evidence. Oxford Bulletin of Economics and Statistics, 65(2), 117-170.
World Health Organization. (2011). World report on disability 2011. Malta: WHO Press.
Zhou, M., & Bankston, C. (1998). Growing up American: How Vietnamese immigrants adapt to
life in the United States. New York: Russell Sage.

117

CHAPTER 3. GLOBAL VERSUS LOCAL INFLUENCES ON THE APPROPRIATION OF
INCLUSIVE EDUCATION POLICY IN TANZANIA
Introduction
As globalization continues to spread, educational policy is becoming increasingly
influenced by global forces beyond the national borders of individual countries. For developing
contexts, the global diffusion of educational ideals arguably mostly flow in a unidirectional
fashion from the global North/West to the global South/East (Anderson-Levitt & Alimasi, 2001;
Steiner-Khamsi, 2010). Contributing factors include the desire on the part of developing
countries to be perceived as “modern” nation-states (Ramirez & Boli, 1987b; Ramirez &
Ventresca, 1992), as well as the increased presence and influence of non-governmental
organizations and multilateral agencies since the 1990s (Kamens & McNeely, 2009; King, 2007;
Martens, 2005). Policies are transferred within the context of asymmetrical power relations, as
international organizations directly or indirectly influence policy adoption in developing
countries with external funds (Anderson-Levitt & Alimasi, 2001; Ginsburg, Cooper, Raghu, &
Zegarra, 1991; Steiner-Khamsi, 2010). While recipient countries certainly modify or recontextualize borrowed policies (Philips & Ochs, 2003; Schriewer, 2000; Steiner-Khamsi, 2004;
Steiner-Khamsi & Stolpe, 2006), their ability to mediate global policy pressure is limited by the
asymmetrical power relations that permeate the policy transfer process.
Similar to other developing nations, instrumental global influences are also evident in
Tanzania’s policy initiatives. Upon independence, President Nyerere developed a Three Year
Development Plan for 1961-1964 to strengthen the nation’s economy. In her discussion of this
plan, Vavrus (2003) cautions, “Although Tanzanian in name, the plan was based primarily upon
the views of the World Bank and the United States Agency for International Development

118

consultants who worked on it” (p. 54). With the exception of the period of African socialism and
self-reliance from 1967 through the 1980s when Tanzania’s policy was uniquely Tanzanian, this
trend has persisted over time as the Tanzanian government has continued to seek and has needed
help from national and international organizations (Vavrus, 2003). The complex interplay
between global and local forces in Tanzania is summarized nicely by Samoff (1994):
Education policy-making in Tanzania is not a solely Tanzanian activity. Like their
colleagues elsewhere in Africa, Tanzanian policy-makers look to the North Atlantic for
models, analyses and diagnoses, and approval. Often subtle, this deference to external
authority conditions policies–from specifying what is problematic to designing
intervention strategies to evaluating outcomes. Even more important, most new projects
in education and even a portion of recurrent expenditures rely on externally provided
funding. (p. 143)
The present study explores the interplay of global and local forces on Tanzania’s policy process
in the case of inclusive education reform. The primary goal of my study is to examine how
global and local forces have shaped the inclusive education policy process, notably policy
formation and implementation, from the perspective of key policy actors in the education sector.
Inclusive education serves as an informative and relevant case through which to analyze
global and local influences in Tanzania for three main reasons. First, the global movement
towards inclusion has been gaining momentum, yet the inclusive ideology has been hotly
contested and no consensus has been reached as to how best to define and implement it (Kavale
& Forness, 2000; Slee, 2004). The lack of a global consensus on inclusion complicates local
implementation, unlike other global policies that have clear messages and signals. SteinerKhamsi (2010) explains that policy makers only adopt rhetoric during the phase of explosive

119

growth during the global transfer of policies, and inclusive education as a global reform
movement is at that phase. Inclusive education reform is thus at risk of being limited to rhetoric
due to such inconsistencies in understanding, coupled with the fact that the inclusion movement
is a relatively recent initiative. Second, Tanzania’s socio-cultural, historical, economic, and
political context adds a unique element to such an analysis. As will be elaborated upon below,
Tanzania’s history with African socialism has potential implications for inclusive education that
may prove that there is a local influence on policy appropriation. Finally, as my literature review
will demonstrate, students with disabilities are exceptionally educationally disadvantaged in
Tanzania. To improve the state of education for students with disabilities in Tanzania,
sustainable reform is needed. It is imperative that we understand the policy process behind
inclusive education reform to ensure that the policy is not limited to rhetoric, but that it will lead
to implementation and action.
To summarize, Tanzania like several other African nations has been influenced by and
responded to Western (donor) pressures and rhetoric in the education development arena.
Locally, the educational marginalization of children with disabilities in Tanzania is a serious
issue requiring urgent attention. Globally, while inclusive education conversations have become
more prominent, several disagreements prevail about how inclusion should be defined and
implemented. In light of this, my study analyzes the influence of global and local forces on the
process of inclusive education policy formation and implementation in Tanzania. National
policy actors are interviewed to understand the appropriation of inclusive education policy from
their perspective. To inform sustainable reform for this disadvantaged student population that
moves beyond rhetoric, it is especially important to determine the extent local policy actors have
mediated and adapted the global ideal of inclusion to make it appropriate for the local context.

120

Relevant Background and Literature
Inclusive Education Globally
Global discourses surrounding the inclusive education movement reflect a vision of a
utopia where all students with disabilities are educated alongside their non-disabled peers
without discrimination. Inclusive education is characterized by a “feel-good rhetoric” with a
radical humanistic philosophical premise (Armstrong, Armstrong, & Spandagou, 2010, p. 4).
The interest and enthusiasm around inclusive education has been steadily growing around the
world, despite inconsistencies in its definition and implementation (Slee, 2009). Such variation
makes inclusion an especially interesting construct to study internationally to see how different
local contexts interpret this seemingly ambiguous concept that is becoming increasingly
prevalent within the global education discourse. Although inclusive education theory
conceptualizes the term quite broadly by incorporating multiple forms of difference, the bulk of
research and policy initiatives in this area have focused on students with disabilities (Kozleski,
Artiles, & Waitoller, 2011). Consistent with this dominant discourse, this paper will focus on the
inclusion of students with disabilities as well. An overview of the status of education for
children with disabilities globally, as well as the key international policies which are recognized
as being fundamental to the global inclusive education movement follows.
The status of education for children with disabilities globally. Disability is a major
exclusionary factor when considering a child’s access to a quality education, particularly in lowincome countries. Of the 600 million people with disabilities in the world, 80% live in
developing countries (UNESCO, 2006). One third of the estimated 77 million children who are
not attending school are excluded on the basis of their disability (UNESCO, 2006). In Africa, it
has been estimated that fewer than 10 percent of children with disabilities who are of school-

121

going age attend school, and of those who do, the majority do so in a segregated setting (World
Vision, 2007).
Global inclusive education policy landscape. Many policies have shaped the global
inclusive education movement. The imperative of education for all was first recognized in the
Universal Declaration of Human Rights (UN, 1948) and is stated as:
Everyone has the right to education. Education shall be free, at least in the elementary
and fundamental stages. Elementary education shall be compulsory... Education shall be
directed to the full development of the human personality and to the strengthening of
respect for human rights and fundamental freedoms. It shall promote understanding,
tolerance and friendship among all nations, racial or religious groups, and shall further
the activities of the United Nations for the maintenance of peace. (UN, 1948, Article 26)
At the first Education for All (EFA) conference, held in Jomtien, Thailand in 1990, every child’s
right to an education was further promoted globally.
Education for children with disabilities was incorporated into the United Nations
Educational, Scientific and Cultural Organization (UNESCO) World Declaration on Education
for All in Article 9: “The learning needs of the disabled demands special attention. Steps need to
be taken to provide equal access to education to every category of disabled persons as an integral
part of the education system” (UNESCO, 1990). In 1994, the Salamanca Statement and
Framework for Action on Special Needs Education further increased global attention to the
education of children with disabilities. It is important to highlight that the Salamanca Statement
explicitly identified inclusive education as a key strategy to reach EFA goals:
Regular schools with this inclusive orientation are the most effective means of combating
discriminatory attitudes, creating welcoming communities, building an inclusive society

122

and achieving education for all; moreover, they provide an effective education to the
majority of children and improve the efficiency and ultimately the cost-effectiveness of
the entire education system. (UNESCO, 1994, Article 2)
In 2007, the global movement towards inclusive education for children with disabilities was
reiterated in the UN Convention on the Rights of Persons with Disabilities (CRPD), which states:
“Parties should recognize the right of persons with disabilities to education. With a view to
realizing this right without discrimination and on the basis of equal opportunity, States Parties
shall ensure an inclusive education system at all levels (UN, 2007, Article 24). To date, 119
countries have signed and ratified the convention, including Tanzania15 (UN, 2013).
Inclusive Education Locally
Moving from the global to the local, this section will examine the status of education for
children with disabilities, review national level policies on inclusive education, and discuss
barriers to inclusive education policy implementation in Tanzania. First, a brief overview of the
Tanzanian context will be provided to contextualize the discussion on the education of students
with disabilities and inclusive education. Particular attention will be paid to Tanzania’s history
with African socialism due to its potential implications for inclusive education.
Geographical, economic, and political context. As depicted in Figure 3.1, the United
Republic of Tanzania is a country in East Africa that is bordered by Kenya and Uganda to the
north; Rwanda, Burundi, and the Democratic Republic of the Condo to the west; Zambia,
Malawi, and Mozambique to the south; and the Indian Ocean to the east. According to the most
recent population and housing census conducted in 2012, Tanzania has a population of 44.9
million (National Bureau of Statistics, 2013).

15

Tanzania signed the CRPD on March 30, 2007 and ratified the convention on October 11, 2009.

123

Democratic
Republic of
the
Congo

Figure 3.1. Map of Tanzania and bordering countries. Adapted from “Map of Tanzania,
Mozambique and neighboring countries,” by WDGH Educational Foundation, 2011, retrieved
from the Public Broadcasting Service website
http://www.pbs.org/frontlineworld/stories/tanzania605/profile.html.
Despite aspirations of becoming a ‘middle income country’ by 2025 (United Republic of
Tanzania, 2011), Tanzania remains one of the poorest countries of the world. According to
World Bank classifications based on gross national income (GNI) per capita, Tanzania is
presently classified as a ‘low-income economy’ with a GNI of $1,035 or less (The World Bank,
2013). Economically, Tanzania has an emerging position as a leading exporter of gold, other
minerals, and gas; however the majority of the population is reliant on sustenance agriculture
(Coulson, 2013). Severe food insecurity is a problem faced by many Tanzanians, particularly
those living in rural areas (Phillips, 2013; World Food Programme, 2013).
124

Tanzania’s ujamaa and equality African socialist ideals of the past are important to
highlight when considering the inclusive education of students with disabilities. Tungaraza
(1994) suggests that the plight of individuals with disabilities in Tanzania may very well have
been viewed differently due to the nation’s African socialism ideology. After independence in
1961 from colonial rule by the British and Germans, Tanzania’s founding President Julius
Nyerere released the Arusha Declaration or ‘blueprint for socialist development’ in February of
1967. The Arusha Declaration marked a radical turning point in Tanzanian politics (Nyerere,
1968). Among its numerous components was the establishment of ujamaa, which Nyerere
himself translated as ‘familyhood’ (Coulson, 2013). In March of 1967, a subsequent policy
statement entitled Education for Self-Reliance (ESR) was put forth by Nyerere. ESR served as
the educational component of Nyerere’s ujamaa program, and was designed to reduce regional,
ethnic, and class inequalities throughout the education system (Vavrus, 2003). Although there
was no direct link between disability and ujamaa in the ESR, there is some evidence of a
connection between Nyerere’s African socialist ideals and his approach to education. When
laying a stone for a school of the deaf in 1974, Nyerere was quoted as stating that it is the
responsibility of the people of Tanzania to “ensure that these children have the opportunity to
enter regular schools and to make sure we help them in their problems” (as cited in Tungaraza,
1994). This statement reflects the collective responsibility of all Tanzanians to include and
support children with disabilities, which aligns with the focus of ESR on addressing educational
inequalities. Nyerere retired as President in 1985, but his influence on Tanzanian politics
continued through his friends and contacts in the government up until his death in 1999
(Coulson, 2013). By analyzing global and local influences on inclusive education reform, this

125

study will be able to determine whether the ujamaa political spirit of the past influenced the
formation and implementation of inclusive education reforms of present day Tanzania.
Inclusive education policy landscape in Tanzania. Tanzania is a signatory to several
international conventions that promote education for all, such as the United Nations Universal
Declaration of Human Rights (UN, 1948) and the Convention on the Rights of the Child (UN,
1989). In terms of national Tanzanian policy, in 2004 the Ministry of Labour, Youth
Development and Sports (MoLYDS) released the National Policy on Disability (NPD). The
NPD touches on multiple policy issues for individuals with disabilities, such as health, early
intervention, HIV/AIDS, education, employment, accessibility, and care. The NPD defines the
policy issue of education as:
Education is key to the development of children with disabilities’ potential. The
education policy emphasizes the availability of early learning and basic education to all
children aged 7 years and that children with disabilities will be given a priority. Despite
this commitment the educational system is inaccessible to children with disabilities.
(MoLYDS, 2004, Article 3.8)
To address this policy problem, the NPD provides the following policy statement, “The
government in collaboration with stakeholders shall provide a conducive environment for
inclusive education that takes care of special needs of disabled children” (MoLYDS, 2004,
Article 3.8).
In 2009, the Ministry of Education and Vocational Training (MoEVT) published the
National Strategy on Inclusive Education 2009-2017, whose stated overall goal is that, “All
children, youth and adults in Tanzania will have equitable access to quality education in

126

inclusive settings” (p. 3). The National Strategy on Inclusive Education defines inclusive
education as:
Inclusive education is a system of education in which all children, youths and adults are
enrolled, actively participate and achieve in regular schools and other educational
programmes regardless of their diverse backgrounds and abilities, without discrimination,
through minimization of barriers and maximization of resources. (MoEVT, 2009, p. 2).
The policy document then goes on to specify multiple objectives and strategic areas of action
including having education policies and programs informed by inclusive values and practices,
teaching and learning which is responsive to the diverse needs of learners, educational support
for all learners, strengthened professional capabilities, and enhanced community partnership and
participation in inclusive education. It should be noted that the language of these local
documents is quite broad, and the specific means of how to attain the stated objectives and
strategic areas of action are not clearly detailed.
In summary, Tanzania has signed numerous international conventions related to
education for all and to inclusive education. Additionally, similar to international policies on
disability, Tanzania’s national education policies clearly state inclusive education as the desired
approach to educating students with disabilities.
The status of education for children with disabilities in Tanzania. There appears to
be a disconnect between the policy rhetoric around inclusion with the lived educational
experiences of children with disabilities in Tanzania. Unfortunately in spite of the spirit of
ujamaa, compared to other African countries issues of the education of students with disabilities
are under addressed in Tanzania (Chataika, Mckenzie, Swart & Lyner-Cleophas, 2012). It is
estimated that 2.4 million people in Tanzania have a disability and that the prevalence of

127

disability in Tanzania is approximately 8% (United Republic of Tanzania, 2009; World Health
Organization, 2004). However, it is important to note that these statistics should be interpreted
with caution because it is generally agreed that information and statistics on disability in
developing countries is unreliable and incomplete (International Labour Office, 2004; UNESCO,
2010; UNICEF, 2008).
In Tanzania, having a disability doubles the probability of children never attending
school (United Republic of Tanzania, 2009). It is estimated that less than 40% of children with
disabilities are enrolled in primary school and less than 2% are enrolled in special education
programs (Njelesani, Couto, & Cameron, 2011). This alarmingly high proportion of out-ofschool children with disabilities appears to be relatively consistent across studies. For instance,
in a study of children with mental disabilities in Tanzania by Mbwilo, Smide, and Aarts (2010),
the majority of their sample (58%) never attended school.
According to the most recent United Republic of Tanzania Ministry of Education and
Vocational Training (MoEVT) Basic Education Statistics Tanzania (BEST) data from 2010,
there were a total of 36,252 children with disabilities enrolled in primary school16 (MoEVT,
2011). As illustrated in Table 3.1, the majority of these students were physically impaired
(38%), followed by mentally impaired (22%), and deaf/mute (13%). Given that primary schools
fees were abolished in 2000 in Tanzania (Coulson, 2013) but secondary schools still charge
school tuition fees, it is not surprising that the statistics for secondary school enrollment are
significantly lower than that of primary. According to the BEST data, only 5,265 children with
disabilities were enrolled in secondary school in 2010. Table 3.1 shows similar disability

16

The structure of formal schooling in Tanzania constitutes 7 years of primary school (Standards 1-7 for children
aged 7-13 years), 4 years of junior secondary or O (ordinary) level (Forms 1-4 for children aged 14-17 years), 2
years of senior secondary or A (advanced) level (Forms 5-6 for children aged 18-19 years), and 1-3 years of tertiary
education (for adults aged 20+ years).

128

category enrollment patterns at the primary and secondary levels, with the majority of secondary
students with disabilities being physically impaired (54%), followed by other (12%), visually
impaired (10%), and deaf/mute (10%).
Table 3.1
School Enrollment in Tanzania by Disability Category
Primary
Disability Category

Secondary

Raw Count

%

Raw Count

%

Albino

2,381

6.6

331

6.3

Autism

557

1.5

33

0.6

Deaf/Mute

4,716

13.0

520

9.9

Deaf/Blind

1,345

3.7

155

2.9

Mentally Impaired

7,936

21.9

170

3.2

665

1.8

61

1.2

Others

3,304

9.1

631

12.0

Physically Impaired

13,782

38.0

2,825

53.7

Visually Impaired

1,566

4.3

539

10.2

Total disabled

36,252

100

5,265

100

Multi-Impaired

Total non-disabled

8,419,305

1,638,699

Note. Table created using Basic Education Statistics Tanzania (BEST) online Data Center 2010
data (MoVET, 2011).
Moreover, Table 3.1 provides insight into the disproportionate amount of students with
disabilities compared to students without disabilities that are attending school in Tanzania. The
figures in Table 3.1 show that there are 232 times as many non-disabled pupils as there are
disabled pupils enrolled in primary school (36,252 disabled compared to 8,419,305 nondisabled). This disparity increases at the secondary level, as there are 311 times as many nondisabled pupils as there are disabled pupils enrolled in secondary school (5,265 disabled
compared to 1,638,699 non-disabled). Presenting these statistics another way, a mere 0.4% of
129

the total primary school student population has disabilities, and only 0.3% of the total secondary
student population has disabilities. Comparing these percentages with the estimated 8%
disability prevalence rate in Tanzania, a large discrepancy appears to exist, suggesting that many
students with disabilities in Tanzania are not in school.
For children with disabilities that do attend school, Tanzania has adopted three primary
approaches to educating students with disabilities: special schools, integrated schools and
inclusive schools (Mkumbo, 2008). Special schools are the most segregated setting, as they only
house students with identified disabilities. Special schools cater specifically to students with
disabilities and focus on serving those with a particular type of disability, such as blind or deaf
learners. Due to their history, special schools in Tanzania are closely tied to religion, since
special schools evolved out of the European Christian missionary movement of the 19th century
(Kiyaga & Moores, 2003; Tungaraza, 1994). Integrated schools are slightly less segregated than
special schools, as they educate students with disabilities in mainstream schools but in separate
special classrooms. Akin to special schools, physical division still exists in integrated schools
because students with disabilities are separated from their mainstream peers, albeit the degree of
separation is at the classroom level as opposed to the school level. Finally, in contrast to special
schools and integrated schools, inclusive schools are not segregated at all. Inclusive schools
educate students with disabilities in the same educational environment as their non-disabled
peers. The vast majority of children with disabilities in Tanzania attend either special schools or
integrated schools (Mkumbo, 2008). Additional evidence on school provision for students with
disabilities is provided by Mamdani, Rajani, Leach, Tumbo-Masabo, and Omondi (2009) who
reported that little inclusive education is practiced in the vast majority of schools and only a
handful of special education schools exist. Together, these findings suggest that children with

130

disabilities have limited access to inclusive schools in Tanzania, although more current data is
needed.
A final factor to consider regarding the education of students with disabilities in Tanzania
is the location of schools that serve students with disabilities. Although the vast majority of
children with disabilities in Tanzania live in rural areas, special education facilities are
concentrated in urban areas (Kisanji, 1993; Marshall, 1997a; Njelesani et al., 2011). An absence
of special education facilities in rural areas has led to what Kisanji (1998) and Nambira (1994)
refer to as the ‘casual integration’ of children with special needs. These authors critique casual
integration because students with disabilities do not receive any learning support, in large part
because teachers of regular schools are not adequately trained to teach students with disabilities.
Furthermore, children who are casually integrated have no ‘official’ existence (Miles, 1989).
Casually integrated children have somewhat of an invisible status in the education system
because there is no official education identification or documentation of these students and their
unique learning needs. In sum, existing evidence suggests that children with disabilities are
educationally disadvantaged and experience multiple barriers to accessing quality education in
Tanzania. Such realities are inconsistent with Tanzania’s policy rhetoric which promotes an
inclusive quality education for all students, including those with disabilities.
Barriers to inclusive education policy implementation in Tanzania. As noted in the
previous sections despite having such progressive policies, the educational condition of children
with disabilities leaves a lot to be desired in Tanzania. Evidence suggests that while many
developing countries have developed educational policies that recognize inclusion as a desirable
form of education for students with disabilities, inclusion is not being adequately implemented
(Eleweke & Rodda, 2002). Studies on the topic of inclusive education in sub-Saharan Africa

131

more broadly and in Tanzania specifically identify multiple barriers to inclusive education policy
implementation. Common barriers that are reported in the literature include physical barriers
such as inaccessible school infrastructure and transportation issues; attitudinal barriers such as
insensitivity and discrimination by teachers, principals, peers, and society at large and ashamed
parents hiding children with disabilities at home instead of enrolling them in school; a lack of
teaching and learning materials and resources needed to facilitate the learning of students with
disabilities; and a lack of teacher training in special education and inclusive education practices
which is related to poor identification and recognition of disability and special needs (Chataika et
al., 2012; Eleweke & Rodda, 2002; Marshall, 1997a, 1997b; Mbwilo et al., 2010; Miles, 2011;
Mkumbo, 2008; Mnyani, 2009; Woods, 2008).
In Africa, one of the major obstacles to inclusion has been a lack of support on the part of
governments in terms of planning and resource provision (Abosi, 1996). This is consistent with
the findings of the most comprehensive assessment of inclusive education done in Tanzania to
date by Mkumbo (2008) in conjunction with HakiElimu, a non-governmental organization
(NGO) focused on issues of equity and education. In their study of six districts in Tanzania,
Mkumbo identified the biggest barrier as the lack of a concerted institutionalized effort.
Mkumbo argued that policymakers lacked political will and criticized the government for failing
to mention disability in key educational policy documents, as well as NGOs concerned with
education for leaving disability out of their statutes.
Overall, despite the promotion of increased access to education for children with
disabilities propagated by Tanzania’s national inclusive education policy framework, many
students with disabilities remain out of school and face numerous barriers to accessing an
inclusive quality education where their academic and social development can be fostered. With

132

the goal of informing sustainable inclusive education reform in Tanzania, my study investigates
how global and local forces have shaped the formation and implementation of inclusive
education policy from the perspective of key education sector policy actors working in the
government and non-governmental organizations (NGOs). It is important to determine the
extent to which local policy actors have mediated and adapted the global ideal of inclusion to
make it appropriate for the local context to determine whether inclusive education reform efforts
can advance beyond rhetoric.
Theoretical Framework
Global and local forces are a central component of this study’s theoretical framework. A
dichotomy appears to exist within the field of comparative education, where some scholars apply
a global model of education such as sociology’s neo-institutionalism or world culture theory
(e.g., Boli, 2005; Ramirez & Boli, 1987a), while others from an anthropological or
externalization persuasion focus on local variation and agency at the national, district, and
classroom levels (e.g., Schriewer & Martinez, 2004; Takayama, 2010). Instead of focusing only
on the global or the local level, Anderson-Levitt (2003) encourages us to unite both approaches
in her comment, “If we take seriously both local variability and world culture theory, we
recognize that each perspective on its own misses something crucial” (p. 18). Similarly, Arnove
(1999) stresses the importance of understanding the interactive process between the global and
the local, as well as its tensions and contradictions. In alignment with the sentiments expressed
by Anderson-Levitt and Arnove, my study attempts to understand how global and the local
forces intersect through the appropriation of inclusive education policy reform in Tanzania.
Another component of my study’s theoretical framework is the application of a
sociocultural approach to educational policy analysis, which emphasizes power, process, and

133

practice (Ball, 1994; Levinson & Sutton, 2001). Levinson and Sutton (2001) propose a
sociocultural approach to policy analysis and conceptualize policy as “a complex social practice,
an ongoing process of normative cultural production constituted by diverse actors across diverse
social and institutional contexts” (p. 1). The present study will apply a sociocultural approach to
the analysis of inclusive education policy in Tanzania in order to understand how the policy is
appropriated by national policy actors, analyzing the influence of global and local forces on the
process of policy formation and implementation.
Finally, this study adopts an interpretivist epistemological approach. The central purpose
for research within an interpretivist paradigm is to understand (Glesne, 2011). This purpose fits
with the primary goal of this study to understand how inclusive education policy has been
appropriated in Tanzania from the perspective of key policy actors in the education sector.
Furthermore, this study is aligned with the interpretivist approach because it meets the
approach’s two assumptions that reality is socially constructed and that variables are complex,
interwoven and difficult to measure (Glesne, 2011). Regarding the latter assumption, this study
assumes that there are many global and local socio-cultural, historical, economic and political
variables that influence these policies and are complex, interwoven, difficult to measure, and at
times contradictory. A visual representation of these global and local variables in relation to the
process of inclusive education policy appropriation in Tanzania is depicted in Figure 3.2. As
depicted by the gradient arrow labeled “focus of study” on the left side of Figure 3.2, this study
primarily focuses on the upper portions of the figure by interviewing national policy actors.

134

Figure 3.2. Theoretical framework.
Research Questions
The purpose of this study is to understand the appropriation – commonly referred to as
the formation and implementation (Koyama, 2009) – of inclusive education policy in Tanzania in
relation to global and local forces. To achieve this understanding, I ask the following research
questions:
135

RQ1. How do local factors shape the formation and implementation of inclusive
education policy in Tanzania?
RQ1a. What barriers to inclusive education policy implementation do local actors
experience and how are they related to local contextual factors?
RQ2. How do global factors shape the formation and implementation of inclusive
education policy in Tanzania?
RQ3. How do these local and global forces intersect and diverge in their influence on the
formation and implementation of inclusive education policy in Tanzania?
RQ3a. At the local level, is there any evidence of resistance, adaptation, and
subterfuge to global inclusive education forces?
The first question corresponds to understanding local influences on the appropriation of inclusive
education policy, and the second to understanding global influences. The third question brings
the two levels together and explores the constructed and contested relationship between them.
Method
Two qualitative methods were used to investigate my research questions. The primary
method was semi-structured interviews and the supplemental method was document analysis.
Semi-structured interviews were conducted to collect in-depth information about the experiences
and perspectives of national policy actors about the appropriation of inclusive education policy.
Additionally, documents collected from participants provided background information on
inclusive education policy in Tanzania and implementation efforts. More details about the
participants and method are provided below.

136

Participants
A total of 20 policy actors operating at the national level in Dar es Salaam, Tanzania
participated in semi-structured interviews with the researcher. Although Dodoma is the official
capital city of Tanzania, Dar es Salaam was selected as the location for this study because it
serves as Tanzania’s main commercial city and many government offices are located there. To
be included in the study, all participants were required to have knowledge about or personal
experience with inclusive education policy formation and/or implementation in Tanzania. In
addition, participants had to have a current or past formal appointment with the government or an
NGO that has participated in inclusive education reform efforts. Individuals who did not meet
these inclusion criteria were excluded from participation.
The sample consisted of 3 national government education officials, 15 staff members
from education and disability NGOs, and 2 faculty members. The categorization of all
participants is described below and is summarized in Figure 3.3.

Figure 3.3. Participant categorization summary.
Government officials. Of the three government officials, two were from the Untied
Republic of Tanzania’s Ministry of Education and Vocational Training (MoEVT) and one was
from the Tanzania Institute of Education (TIE). The TIE is the MoEVT agency responsible for
137

curriculum development. The TIE participant worked in the area of special education curriculum
development. The organization structure of the MoEVT is divided into 21 divisions and units17.
One participant was from the ministry’s Special Needs Education Unit and the other was from
the Adult and Non-Formal Education Division.
NGO staff members. Twelve of the NGO staff members were from national NGOs and
three were from multinational NGOs that operated in one or more countries in addition to
Tanzania. The majority of NGO staff whom I interviewed held high-ranking leadership
positions within the organization, such as Secretary General, Coordinator, and Program Officer.
An important distinction among disability NGOs that I learned once I was in the field is what is
referred to as Disabled People’s Organizations (DPOs). DPOs are registered non-governmental
organizations that are founded and operated by people who have disabilities themselves, as
opposed to organizations that are founded and operated by people without disabilities. The latter
was referred to as Organizations for People with Disabilities, indicating that the organization is
largely run by people without disabilities (although this does not necessarily mean that
absolutely no people with disabilities are on staff) to advocate on behalf of and serve those with
disabilities. In Tanzania there is a formal hierarchical structure of ten national DPOs that are
organized under the Tanzania Federation of Disabled People’s Organizations or Shirikisho la
Vyama Vya Watu Wenye Ulemavu Tanzania (SHIVYAWATA). The national Tanzanian DPO
structure is depicted in Figure 3.4. I was able to interview representatives from six DPOs.

17

The 21 MoEVT divisions and units include the Higher Education Division, Technical and Vocation Education
Training Division, Adult and Non-Formal Education Division, Teachers Education Division, School Inspectorate
Division, Special Needs Education Unit, Diversity Unit, Education by Media Unit, School Registration Unit, Pre
Primary and Primary Education Unit, Secondary Education Unit, Policy Planning Division, Administration and
Human Resources Management Division, Procurement Management Unit, Finance and Accounts Unit, Internal
Audit Unit, Government Communication Unit, Legal Services Unit, Information and Communication Technology
Unit, National Commission for UNESCO, and Zonal and District Offices (MoVET, 2013).

138

Figure 3.4. National DPO structure in Tanzania.
Faculty members. The two faculty participants were from the education department in
large, national universities in Dar es Salaam. Faculty interviews provided the researcher with
helpful background and contextual information.
Sampling Procedure
A combination of purposive and snowball sampling was used to identify and recruit
participants. First, purposive sampling was used to identify and recruit participants who were
most likely to have sufficient knowledge of and/or experience with inclusive education policy
formation and implementation initiatives. Selection of eligible participants was done in
consultation with faculty member contacts in Tanzania with local knowledge about inclusive
education reform in Tanzania, as well as the nation’s ministerial and NGO structure.
Additionally, with the goal of incorporating diverse perspectives across multiple disability types
in my sample, I purposefully recruited from all DPOs. Eligible participants were contacted via
email, mobile phone, and in person to ask if they were willing to participate in my study.
At the conclusion of each interview, I asked each participant whether they had any
recommendations for additional government officials or NGO staff whom I should interview that
would be able to talk to me about inclusive education policy formation and implementation (see
final question in interview protocols in Appendix A). For this snowball portion of my sampling

139

procedure, I would then contact, recruit, and interview the informants that my participants
recommend until my sample became saturated (i.e., no new informant names were recommended
and I had a wide representation of government and NGO perspectives represented in my data).
My sample reached saturation at 20 participants.
Data Collection Procedure
This study employed two methods of data collection: semi-structured interviews and
documents. Using the attached interview protocols (see Appendix A), I interviewed national
policy actors about the formation of inclusive education policy in Tanzania and their
organization’s strategies for implementation. In addition, as part of the interview I asked
participants for paper or electronic copies of relevant inclusive education policy documents.
Each interview was approximately one hour in duration. All interviews were audio recorded
with written participant consent (see informed consent forms in Appendix B), so that the
interviews could be later transcribed. Interviews took place in a location of the participant’s
choosing, most often in the participant’s office. The majority of the interviews were one-on-one
with a participant and I, but there were a few instances where I interviewed two to three
individuals from the same organization together. In the latter instances, I counted each group
interview as one interview.
Tanzania’s two official languages are English and Kiswahili. All interviews were
conducted in English because English is the language of commerce and it was anticipated that all
participants would be proficient in spoken English due to their occupations as high-ranking
policy officials and NGO staff members. Copies of the informed consent forms were provided to
all participants in both languages (see Appendix B) in case they preferred to read Kiswahili
instead of English, but the majority of participants selected the English form. Additionally,

140

participants were informed that if they preferred to be interviewed in Kiswahili instead of
English, the interview could be conducted through a hired translator. However, all participants
chose to be interviewed in English. My data collection procedures were approved by the
Michigan State University Institutional Review Board (IRB# x13-992e; i044630) and the
Tanzania Commission for Science and Technology (COSTECH).
Researcher Positionality
The purpose of my interpretive approach was to gain an understanding of how inclusive
education policy is being appropriated in Tanzania from the perspective of key policy actors in
the education sector. As a white woman visiting Tanzania from the global North/West, I actively
attempted to resist falling into the stereotype of a privileged international expert coming with all
of the answers. Instead, I tried to position myself as a researcher whose goal was to share
Tanzanian voices and experiences. To do so, I feel that my social status as a female and as a
student helped me convey my genuine search for meaning and understanding. Although my
language ability was limited, I did my best to practice customary greetings in Kiswahili and
express my gratitude for my participants’ time and valuable contribution to my research. To help
build rapport, I also strove to connect with each of my participants through our shared moral
commitment to children with disabilities, trying to position myself as a comrade or fellow
advocate. Participants from disability NGOs especially seemed to appreciate my interest in
including their respective disability group in my research and in learning about their efforts in
Tanzania so that I can share it with a wider global audience.
Admittedly, my ‘outsider’ status amplified by the color of my skin and hair, my dress,
and my language limited the extent to which I could connect with my participants. While I feel
that my demeanor articulated above enabled me to attain insightful responses to my interview

141

questions, I also feel that I had to frequently navigate the challenge of peeling away the layers of
politically correct responses to uncover true perspectives. Some participants, particularly
government officials, were brief with some of their answers and I had to consistently probe
further throughout the interviews. Even so, I feel that many of my participants were very open
with me and I believe that the data I was able to obtain provides interesting insights into multiple
Tanzanian perspectives and experiences of inclusive education policy appropriation.
Data Organization and Analytical Approach
During my time in the field, I recorded field notes and themes in a personal notebook. I
also developed weekly summary reports of my activities, emerging themes and findings, and
upcoming data collection plans and strategies. These weekly summaries helped me focus my
data collection efforts on emerging themes and identify areas of confusion to ask future
participants about to help clarify my understanding. For example, when interviewing my first
government official I quickly realized that he and I had different understandings of what
constitutes a policy; I perceived the National Strategy on Inclusive Education 2009-2017 as a
policy but he did not. Instead, he asserted that it was a strategy, which seemed to hold different
meaning to him beyond a subtle difference in nomenclature. In addition to probing him further
to try and understand the definitional disconnect, in future interviews I asked participants to
clarify for me the difference between a government policy compared to a strategy.
Audio recordings of interviews were transcribed and subsequently coded to identify
themes related to the formation and implementation of inclusive education policy in Tanzania. I
analyzed my data with an eye towards patterns, relationships, and divergent cases as Erickson
(1986) recommends. To explain, I looked for patterns in how participants reported inclusive
education policy being formed and implemented across time and space in Tanzania. Applying

142

my global versus local theoretical framework to my data analysis, I looked for relationships and
interactions between global and local influences on Tanzania’s inclusive education policy
appropriation. When exploring this global versus local relationship with inclusive education
policy appropriation, I looked for instances where their respective influence and actions
intersected and diverged. Furthermore, as part of my analysis of local influences, I looked for
relationships between policy appropriation and Tanzania’s contextual socio-cultural, historical,
economic and political factors. To help me answer my research questions, first I applied what
Hatch (2002) refers to as a post positivist perspective and typological approach to analysis by
using the following a priori themes: global, local, socio-cultural, historical, economic, and
political factors. After I coded my data using these a priori themes, I then looked at my data to
see what additional themes emerged so that I did not unnecessarily limit the scope of my
interpretations.
Results
All identified themes and supporting evidence are reported in relation to each of my
research questions.
RQ1. Local Influence on Inclusive Education Policy Appropriation
From the perspectives of the policy actors I interviewed, the most influential local factors
impacting inclusive education policy appropriation in Tanzania fall under three main categories:
socio-cultural beliefs, political context, and economic conditions. I developed these three main
categories from my a priori themes of local socio-cultural, economic, and political factors.
Within these larger categories the following subcategories emerged from the data: cultural
attitudes towards disability, stakeholder collaboration, political will, government budget, and
familial economic constraints.

143

Stigmatizing cultural attitudes towards disability. Socio-cultural beliefs, specifically
cultural attitudes towards disability, were commonly reported as playing a vital role in the
appropriation of inclusive education policy. Participants explained that the vast majority of
Tanzanians have negative attitudes towards people with disabilities and there is a great deal of
stigma and discrimination associated with disability. To illustrate the general cultural attitude
towards disability, an NGO representative explained:
The general attitude of people with disability is more of, you know, it’s, it’s put them at a
distance. It’s not real inclusive if you want to use the word. Umm it’s them and us, it’s
still like that. And so there a lot of attitudes that are need to be changed.
In some cases there were religious or spiritual beliefs associated with comments about
cultural attitudes towards disability. For example, a handful of individuals used the word
“cursed” to describe traditional cultural beliefs and understandings about disability. Also, this
comment from an NGO staff member explicitly mentions God as playing a role in creating
disability: “Social context that many parents they feel that having a disabled child it’s not a
blessing from God, and they regret.”
Additionally, the following two quotations from different NGO representatives provide
further insight into the social shame of bearing a child with a disability: “Most of the time
someone having a child with disability, generally speaking it’s a shame. They don’t come in
front of people and say, ‘This is my child!’ It is a shame, you see?” and, “You know the habit
was once a baby is born with albinism in the family, the father is the first to run away saying,
‘That this is not my kid. Where did you find this boy?’” Likewise, this statement from another
NGO representative shows the gravity of social stigma around disability in Tanzania: “What you

144

should know is that the level of stigma is very high here. And every service, every service to any
person, to any person with disability revolve around stigma. The main barrier is stigma.”
These types of negative attitudes towards disability appear to span all levels of society,
including the government. The widespread nature of these beliefs about disability can
potentially limit inclusive education efforts, which is implied in the following quotation from an
NGO representative:
Most of disabilities are associated with suppression, troubles, are a lot of troubles which
are very difficult to dismantle. So what you will know is that stigma is very high. And
when you want to get any information how far is the progress of inclusive education in
Tanzania you should know that stigma is still in the level of the government.
In addition to negative attitudes towards disability in general, multiple participants
discussed cultural beliefs about albinism specifically. Albinism is one type of disability that is
socially and politically recognized in Tanzania which has created a lot of public controversy, so
it is important to highlight these perspectives for this particular cultural context. In Tanzania
people with albinism seem to be especially at risk of social discrimination, stigma, and physical
harm. I noticed this even when looking through a local newspaper during my time in the field.
News stories and corresponding photos like the ones pictured in Figure 3.5 were quite
commonplace.

145

Figure 3.5. Sample newspaper clippings of individuals with albinism whose limbs had been
chopped off. Source: The Citizen, May 25, 2014, p. 28 (left), May 21, 2014, p. 15 (right).
Insight into the socio-cultural beliefs about individuals with albinism can be seen from the
following childhood memory shared by an NGO representative:
When we were born I saw one albino and I asked my mother, “Who is this?” and she told
me, “It is an albino and they call him a Zeruzeru.” It is an albino and she said, “These
people do not die.” And I asked, “Why?” She said “No you can’t see they just
disappear.” It is because albinos was born to be buried, people will kill and hide. So
they were believing that albino just evaporate (laughing)! Because the, I came to learn
that it was because people were killing them! And they were hiding them somewhere,
secretly, secretly hide them in secret.
With the increased media coverage on albino mutilation and killings, there is more public
scrutiny about this behavior. However, multiple people I spoke with told me that albino killings
are still taking place secretly in present day Tanzania, so it is still a pertinent issue from their
perspective.
Like the general cultural attitudes towards disability, my data suggests that the cultural
beliefs about albinism may also transcend into the Tanzanian government. A primary reason for
146

this appears to be the cultural beliefs that justify the actions. During our discussion about albino
killings, an NGO representative who was an albino himself explained his understanding of the
rationale behind the behavior and provided multiple examples:
Even the member of the government are doing. Because there is belief that if you have a
blood, or finger or any part of a person with albinism, when treated with maybe rituals
they can bring victory to elections maybe. Uh you may win in economic, and maybe in
the Lake Zone you may net a lot of fish! You may do better in the business, and so forth
and so forth. So it is work of the witchdoctor who say if you want to become say a
member of parliament, if you want to become a president, if you want to become rich
bring me ABC of a person with albinism. So you find people looking for person with
albinism for their body parts.
This particular individual continued to express his concern that albino killings by government
officials were actually going to increase over the next two years during the local government
elections and the general elections. He explained that these types of cases have historically been
much more rampant during election cycles due to beliefs about the prosperity that albino body
parts can bring.
Overall, in Tanzania there seems to be widespread stigma, discrimination, and negative
attitudes towards disability in local communities and perhaps even upwards into the government.
There is a clear sense that disability is an undesirable form of difference and people with
disabilities are often considered social outcasts. The case of albinism as a type of disability
appears to be especially extreme in how these individuals are almost seen as subhuman and
worthy of death for the purpose of individual gain and success. These types of socio-cultural
beliefs about disability conceivably pose a great challenge for moving inclusive education

147

implementation forward because the lack of social acceptance of disability directly opposes the
ideals of inclusive education. Furthermore, to the extent such beliefs may be present within the
government, if community and national leaders do not value or support disability then it is
difficult to convince the greater society of its importance.
Political context. In addition to socio-cultural beliefs about disability, the political
context is another key factor that impacts inclusive education policy appropriation in multiple
ways. Political context appears to be most important in terms of stakeholder collaboration in
policy formation and the political will of the government to implement inclusive education
policy. Both of these political elements will be discussed below.
Stakeholder collaboration. When I asked about policy formation, participants
consistently explained that the policy process in Tanzania is very structured and involves
multiple meetings with stakeholders before a given policy is officially approved by all of the
necessary levels of government. Representatives from nearly all of the NGOs that I conducted
interviews with were formally invited by the Ministry of Education and Vocational Training
(MoEVT) to one to two meetings designed to provide them with opportunities to give feedback
on drafts of the National Strategy on Inclusive Education 2009-2017. The perceived
participatory and collaborative nature of inclusive education policy formation is illustrated by the
following quotation from an NGO representative:
The whole policy development here, you learn from others, it is participatory. It is
participatory. It involves quite a number of stakeholders meeting before it is even go to
the government to restructure. Back and forth with stakeholder reading the document,
stakeholders. It’s not the issue of the government; writing the policy basically is people
writing the policy, it’s so collaborative.

148

Participant perspectives varied considerably concerning the degree to which the
structured stakeholder collaboration element of the inclusive education policy formation process
was truly collaborative. While the previous quotation reflects a very positive perspective, other
viewpoints were more negative. For instance, some NGO staff members who attended these
meetings shared that they did not feel that their recommendations were adequately taken into
account by the government. From the perspective of one NGO representative, the government
appointed a small group of people to write the policy independently without the active
contribution of the disabled community:
And this is also a problem managing from the inclusion issue during the formation of the
act. Yeah, [the government] find some people who sings that they can do it and hide
themselves under the shade do the work without involving people with disability
themselves.
Additionally, there appeared to be some suspicion among participants that the
government exerted its bureaucratic control to promote its own interests by purposefully
selecting individuals with similar political leanings to participate in the inclusive education
policy formation process. Specifically, an NGO representative disclosed:
If you don’t belong to a ruling party say, it’s very hard for you to be appointed into the
team to work on the policy paper of inclusion. So apart from being involving non-state
actors, the non-state actors have to be loyal to the ruling party. So this is the problem
although they were included but only those who were loyal to the government. The
impact is that they are not going to be very critical to the policy… giving opportunity to
speak, opportunity to act which is very much limited in the country. So whichever party
comes into power will have its own controlling mechanism.

149

Overall, there was a clear consensus that there was a formal structured collaborative political
process during the formation of Tanzania’s National Strategy on Inclusive Education 2009-2017.
This collaborative element seemed to be a valued characteristic of policy formation in Tanzania.
Perspectives about how the collaborative process unfolded differed among participants, with
expressing positive viewpoints and others were more negative. So while the formal political
process is important, the nature of collaboration between stakeholders remains unclear.
Political will. “Political will” is a common term that I heard consistently throughout my
interviews with NGO representatives and even government officials themselves. Most often lack
of political will was discussed as a major barrier to inclusive education policy implementation.
The one exception to this was an NGO staff member who interpreted the existence of a national
inclusive education policy as evidence of the political will of the government to support inclusive
education. But even that particular individual acknowledged that there has been limited
implementation by the government, as can be seen in his comment, “The government has made a
commitment, a very positive commitment in this area. So what remains is to actualize it, you
know?”
The vast majority of participants felt that there was a lack of political will on the part of
the Tanzanian government to actualize inclusive education. The blatancy of this issue is evident
in the following response of an NGO representative when asked what impedes inclusive
education in Tanzania:
What I can say? Lack of political will (laughing). In terms of, because inclusive
education requires a dramatic change of the policy system so that it made the element of
the inclusiveness…I know it is a process and it takes time, but not, so far I think we

150

haven’t done much. Not much has been done…so to me I see, like, that is really from the
decision makers and that is the government.
This individual identifies the government and its lack of political will as the primary reason that
limited progress has been made towards inclusive education policy implementation.
Likewise, when asked if inclusive education is a priority of the government, a participant
from another NGO replied, “It is not a priority. No, no! (laughing).” My conversations with
participants about the relative prioritization of inclusive education by the government compared
to other areas reveal some nuances to consider in relation to political will. Expanding secondary
education was cited as a major education initiative that the government has been focusing on
over the past decade. The prioritization of secondary education over inclusive education can be
seen in the following comment by an NGO representative, “Secondary schools – here in the
country now that is education. But when we are talking about inclusive education, it’s
something that we wish to happen, but we have to make sure that some other general concerns
are addressed first.”
In addition to secondary education, multiple participants identified teacher absenteeism
as another important issue that the government is working to address before it can focus on
inclusive education. An NGO representative described, “Take the current situation: the
government is getting big headache about teachers…teachers are striking, they are not going to
schools.” She then went on to explain that addressing teacher absenteeism is a more pressing
need than inclusive education and stated that inclusive education is not a priority because, “there
are different priorities according to existing problems.”
In sum, time and time again, participants brought up political will as an important factor
to consider in relation to inclusive education policy implementation. The majority clearly felt

151

that the government was not committed to inclusive education. The issue of political will
becomes more nuanced when considering inclusive education relative to other educational
challenges that the government is confronted with. Currently, the government seems to be
prioritizing other areas above inclusive education.
Economic conditions. Participants consistently discussed economic conditions in
relation to the implementation of Tanzania’s national inclusive education policy. This occurred
at two levels. One is the level of the national government in terms of the amount of the budget
that is allocated to inclusive education implementation efforts and the other is the familial level
in terms of how economic constraints of families with children with disabilities influence their
decisions about which children to send to school. Both will be explained below.
Government budget allocation. Displeasure about the allocation of government funds to
inclusive education efforts was frequently mentioned during interviews. For example, an NGO
representative told me quite simply, “This inclusive education is not a government funded
project.” Further insight into the implications of government budget allocation for inclusive
education is provided by the following quotation from a different NGO representative in
response to my question asking if inclusive education is a priority of the government:
It’s not, it’s really isn’t. I would love to say it is because I know that there is a lot of
people in government who care about inclusive education, and who wants to see, to see
us have schools that are really equipping the population as diverse as it is, as diverse as
the need of the learners are. But the reality is, uh we are not putting our money towards
these things. And so we will continue to talk but if we don’t change the way we think
about how we spend our money then it will just be talk. And it will just be a nice
declaration on paper that is not being practically implemented.

152

From her point of view, the National Strategy on Inclusive Education 2009-2017 is limited to
policy rhetoric because the government is not allocating funds towards implementation.
It is important to highlight that participant comments about the government budget
allocation to inclusive education seemed to be closely tied to their perceptions about political
will. This connection can be seen in the following comment by a government official: “When
there is no political will it’s clear that inclusive education cannot be implemented successfully
because inclusive education it’s expensive! Now restructuring, which means needs resources
and the resources depend on the political will.” In terms of political will this particular
government official housed in the special education unit seemed to be referring to the political
will of other sections of the MoEVT outside of special education, thereby attributing limited
implementation to the more general lack of political will and the resultant lack of government
funding. This may speak to the hierarchical position of special education in the MoEVT by
implying that special education officials feel beholden to other ministerial units or departments,
particularly in relation to funding. Further insight into the hierarchical nature of the lack of
political will and funding at upper government levels is provided by the following comment the
government official made later in our conversation:
Even the National Assembly…Some of the members of parliament, some they are not
aware. When you talk about people with deaf-blindness, they don’t know what is deafblindness. They haven’t seen them. When you tell them, maybe okay we need enough
funds to buy machine for visual impairment, some of them they don’t see the importance
of allocating enough funds for that…because we always say in order for inclusive
education to be effective we need to a political will. It’s very important because they are
those who are dealing with allocation of funds and whatever. But without political will

153

nothing will be implemented. Because resources will be not or enough funds will be not
allocated for that.
An additional example linking government budget allocation to the lack of political will
or commitment is provided by this quotation from an NGO representative:
If I have told the government would have paid enough resources to invest in the real
inclusive education is where now we can say that we are now doing the inclusive
education system. Apart from that we will be having good policies, strategies, but no
effective implementation and now there are only three years remained for the strategy to
expire. Yeah (laughing) it means we will do the vision but looking in background for the
nine years not something have been committed. So the commitment to government
inclusive system, it is very little. Because that commitment with allocating resources, that
commitment they will change the system of preparing the teachers, that commitment they
will raise awareness to the community, for that commitment they will have to pay enough
resources to buy the learning assistive devices. That’s what I can say.
Overall, participants from NGOs and the government itself seem displeased by the lack
of funding allocated to inclusive education by the Tanzanian government. Perspectives about
government budget allocation appeared to be closely linked with those about political will.
Specifically, the lack of budget allocation to inclusive education implementation by the
government was seen as evidence of the government’s lack of political will towards inclusive
education.
Familial economic constraints. Although less common than comments about the lack of
government funding for inclusive education, some participants mentioned familial economic
constraints as another local economic factor that is related to inclusive education

154

implementation. This factor appeared during discussions about parents of children with
disabilities. More specifically, how their familial economic constraints influenced their
decisions about sending their child with a disability to school. One NGO representative
explained:
In a family that is not well off, well off to do, and when they have a choice who to go to
school because of the financial constraints, who do they choose? They wouldn’t choose
that the one who is disabled. They will choose the normal one to go to school; the
disabled one can stay at home. Yeah, you see? And then even when they are children
disability they need the support to go to school. Like wheelchairs to support, because the
parent will carry the child to school every day. But he has, she has also to go to farm
and come back collect this child from the school is tiring. I am sure not many parents
will volunteer for that for the whole period of schooling of the child.
Likewise, another NGO representative expressed his opinion that, “Most of them [parents]
actually they see if you are spending money to educate your child with disability you lose, you
are not investing, you are just wasting your time and resources.”
These examples suggest that parents may not see adequate economic value in educating
their child with a disability. That is, the rate of return for educating a child with a disability may
not be worth the cost of sending the child to school (e.g., school fees, school uniform expenses)
and the lost cost of time working in the fields spent transporting the child to school. The
economic constraints that many parents face in the developing context of Tanzania likely have a
negative impact on inclusive education policy implementation.
RQ1a. Barriers to Policy Implementation Experienced by Local Actors. Local actors
reported a long list of barriers that hinder inclusive education policy implementation in Tanzania.

155

The most recurrent barriers identified by participants are attitudes, funding and availability of
resources, physical barriers, and teachers. These barriers focus more at the school level
compared to the broader socio-cultural, political, and economic factors discussed above.
However, notable connections to the larger metanarrative of local factors exist and will be
highlighted in the subsequent discussion of each local barrier to policy implementation.
Attitudes. In line with the general stigmatizing socio-cultural attitudes towards disability
discussed above, negative attitudes towards educating children with disabilities were frequently
mentioned as a primary barrier to inclusive education implementation. Participants shared with
me that the majority of community members, parents, school administrators, teachers, and
students believe that children with disabilities are just “useless kids” who do not need an
education. Commonly unquestioned assumptions or myths regarding the dissociation between
disability and education were stated in multiple ways across participants such as, “A disabled
child cannot get education” or “A disabled child cannot learn” or “People with disability have
nothing to do with education.”
These types of unhelpful attitudes seem to be related to a lack of understanding or
knowledge about disability and education. The following excerpt from an NGO representative
about parent and community perceptions reflects such nuances:
They see their children cannot see, they cannot hear correctly and then they imagine
okay, how is he going to get the knowledge? They are not allowed to access, when you
talk of education the community have barrier completely they feel they can’t. That
thinking that they can’t is what, is the biggest challenge.

156

Furthermore, numerous individuals told me that although people do not like to openly
talk about it, it is common for parents to hide their children with disabilities at home instead of
bringing them to school. An NGO representative explained:
There are many, there many children who don’t attend to school because parents or
guardians sometime they don’t have capacity to enroll their children. Sometime parents
they don’t have full knowledge. Sometime parents they don’t know contacts of where
they can enroll their children for getting education. Yeah, sometime social problems
when a man marry and have a baby who is disability a dad can leave, without providing
any support to his children! The society at large can discriminate this woman. Yeah
relatives also can discriminate these woman. Maybe also you can find a woman cannot
allow that child to go outside playing with others. Because the public also does not
accept that kind of disability. That’s why you can find difficult time when you visit those
homes you will find children staying and playing at home, not allowed to go outside.
That’s why if you go to schools you cannot find disabled children, you will find nondisabled only. That’s why, because their parents hide their children. They are fearing
people can laugh, the people can discriminate them.
Once again, this quotation reflects the nuances involved regarding cultural attitudes and beliefs
about disability and education. Lack of information or knowledge about how to enroll a child
with a disability in school, lack of social support, public shame and discrimination all seem to
contribute to the act of keeping children at home instead of enrolling them in school.
In regards to attitudes about inclusive education specifically, complimentary to the beliefs
that children with disabilities should not attend school and should be hidden at home is the
feeling that children with disabilities should not “mix” or be included with those without

157

disabilities in a local community school. This is reflected by the rationale that an NGO
representative described as: “a regular teacher is meant for regular people.” In other words,
children with disabilities are discriminated against as outsiders who do not belong in “regular”
schools with “regular” teachers and “regular” students. Such discrimination is evident in the
story another NGO representative shared with me that, “If you take someone [with a disability]
and bringing to the class all the student will run away. Even the teachers will think you are
cursed, ah!”
Again, lack of community awareness and information likely plays a role in perpetuating
these types of attitudes. An NGO staff member provides additional details concerning attitudes
about inclusion and mixing students with and without disabilities:
The problem is that the community do not know that there is such a thing [inclusion], I
can say. So they don’t bring their kids to these schools. They still have the mind that
these children, student with disability should have their own schools, should have their
own classes, should have their own teachers. These mindset are both at the government
level and the society level. So when you ask someone from the society that you can mix
together the persons with disability with normal students in the same classes, first of all
they wouldn’t believe you. So community awareness is very much needed.
While attitudes undoubtedly pose a significant barrier to inclusive education from the
perspective of the policy actors I interviewed, fortunately I also learned about ongoing advocacy
efforts that many NGOs whom I interfaced with were pursuing. Community awareness and
education were a central advocacy focus. For instance, one NGO representative told me about a
recent public rally they held:

158

Go to Tanga to sensitize communities of the importance of educating children with
disability, and it worked! So we had, we had rallies, we had posters like that one. And
then flags and t-shirts and, you know, things like that, that have also advocacy massages.
We have banners, small ones, big ones advocating for inclusive education and right of
education to children with disabilities.
A sample flag and poster from this rally are pictured below in Figure 3.6.

Figure 3.6. Sample inclusive education advocacy materials.
The Kiswahili on the flag translates to English as, “Equal rights. Equal opportunities. Education
for children with disabilities.” The translations of the captions on the poster in the boxes from
left to right, top to bottom are: “You have chosen to be a prefect head!” and “Yes, mother,
students with disabilities have capabilities, take us to school” in box 1, “Together with others this
one is disabled. I want him to study in a common school with others who are not disabled” and
159

“Okay” in box 2, “Talents of students with disabilities because even we can” in box 3, “Let’s
ensure this budget makes mention of students with disabilities” in box 4, “I like to study because
our school has a friendly environment” in box 5, “I wish to go to school but there are no toilets
for the disabled” in box 6, and “Children with disabilities need better teachers” in box 7.
I was told numerous success stories that were spurred by these particular advocacy
efforts. For example, a school that was visited during the awareness campaign started a sign
language club. The club enabled the peers of a student who had a hearing impairment to learn
sign language themselves so that they could help their colleague understand classroom lessons.
Another example is a community member who volunteered to pay for the school fees of a blind
form four student who could not continue his education because his parents were unable to pay
the last installment of his school fees. The student’s story was publicized by the media during
the campaign and the NGO representative explained, “Because of our advocacy…somebody just
volunteered, ‘I will pay for his fees. Let him go and get his results!’” A final example is of a
local cement business who donated beds and mattresses to a school for children with disabilities
that had poorly equipped dormitories. Together, these success stories illustrate the positive
impact that local advocacy efforts can have on community attitudes and support for the education
of students with disabilities. Increased awareness and knowledge can produce powerful changes.
In sum, people seem to have uninformed and unhelpful attitudes about why and how
students with disabilities should be educated instead of being hidden at home, as well as why and
how they should be mixed or included with students without disabilities. Community awareness
campaigns like the one highlighted here are working to change the types of attitudes that are
serving as barriers to inclusive education implementation in Tanzania.

160

Funding and availability of resources. Similar to the more general economic conditions
discussed above, a major constraint experienced by local actors is serious shortages of resources.
Participants frequently remarked how expensive inclusive education is and that schools do not
have the necessary resources to support the learning of students with disabilities. Special learning
materials mentioned by participants included eye glasses, magnifying glasses, books with large
print, braille machines and books, wheelchairs, and sunscreen to protect the skin of children with
albinism.
The negative impact that such lack of resources can have on learning is shown in the
following scenario provided by a government official:
Now that is where we get the point of resources. Sometime you can send, you can send
for example the visually impaired pupils, you send them to school, but there is no
resources. There is no braille machine for them, which means you are dumping them but
they are not learning. Send them to school and achieve - that is the target.
So, the lack of resources seems to create the unfortunate situation where children with
disabilities are physically “dumped” in schools but they do not have the necessary materials or
supports to facilitate learning. In discussions with other participants I learned that these children
frequently drop out or fail out of school, suggesting that inclusive education is not sustainable
without the necessary supports in place in the classroom.
An important factor in the Tanzanian context when discussing resources is not only the
lack of funds to purchase learning materials (which in and of itself is indeed a major barrier), but
also the limited availability of these materials in the local market. I was told stories about how
different materials such as sunscreen are simply not sold in stores due to a lack of local demand,
even in the urban center of Dar es Salaam. The NGOs focused on albinism reported that they

161

depend on sunscreen donations from Europe and North America because the vast majority of
Tanzanians have dark skin and do not use sunscreen. Another example from an NGO
representative illustrates this supply and demand relationship for materials for students with
visual impairment:
For lens, or magnifiers or monoculars, like that, so there was no demand. Means if there
is no demand there is no supply for that equipments. So you not find them to be sold in
the shop. If you give me money and say okay you can have these to purchase one
hundred piece of magnifiers I will struggle to get them.
Thus, the issue of inadequate resources for inclusion is more nuanced than merely having the
necessary funds to purchase special learning materials and equipment. The unavailability of
these specialized materials in the local market is a significant challenge to overcome to promote
the learning of children with disabilities.
Physical barriers. Another barrier to inclusive education that was consistently mentioned
by local actors was that the school environment is not “friendly” to children with disabilities.
Across interviews, I came to learn that participants used the term “unfriendly environment” to
refer to the fact that school infrastructure was physically inaccessible to many learners,
especially to those with physical disabilities. Environmental barriers mentioned by participants
included: doors, passageways, stairs, and toilets. The following quotation from a government
official explains how stairs act as a physical barrier for learners with disabilities:
You can see that we have a lot of upstairs in schools. You can imagine people with
wheelchairs, how can he manage to use upstairs? Even those with visual [impairments],
although they have that stick [white cane] eee! But if the environment are friendly to all,

162

students will be willingly to go to school. Even parents will be willing to send their
children to school.
Another example about the unavailability of accessible toilets is provided by a faculty member.
Here, she is telling a story about a conversation that she had with a girl with a physical disability
who used a wheelchair:
I talked with her and she said, “When I leave home in the morning I don’t take
breakfast.” I said, “Why don’t you take breakfast?” She said, “I don’t take breakfast
because if feel like going to the toilet I can’t go to the school toilet. So when I go home
it’s when I start eating in this afternoon.” So she remains hungry the whole day because
of the toilet. Yeah, there are a lot of sad stories.
Together, these examples illustrate how school infrastructure serves as a physical barrier to
including students with disabilities in their local community schools.
Similar to the preceding barrier of funding and availability of resources, physical barriers
are also connected to the broader theme of economic conditions. Funding is required to modify
the physical infrastructure of existing schools to make the environment friendlier to students with
disabilities, as well as build new schools that are designed to be physically accessible.
Teachers. Finally, teachers' abilities and attitudes were also frequently mentioned as
barriers to inclusive education in Tanzania. As members of the greater community, the negative
cultural attitudes towards disability and inclusive education explained above are often held by
teachers as well. Similar to the earlier discussion of negative cultural attitudes towards disability
and inclusive education, negative attitudes and a lack of knowledge seem to go hand in hand. I
was told that teachers feel they do not have the sufficient knowledge or training to support
students with disabilities in their classrooms. This can be seen in the following statement where

163

an NGO representative sympathized with teachers, saying that, “It is difficult for teachers. Even
teachers themselves would not feel that they are able to handle a class with mixed kind of
learners.”
The lack of teacher training and the consequences for the inclusion of students with
disabilities is further explained by another NGO representative:
Teachers which are really competent in education, I mean in special education, there are
very, very few. And you find sometime in one school: one, two teachers and sometime
zero. And we say this school is now taking, opting inclusive education. To us we think
some of the children, especially people with disabilities, the deaf, they just visiting the
school. At the end of the day they find they are gaining nothing.
This notion of children with disabilities just visiting school but not learning due to inadequate
teacher preparation came up in many interviews. Another NGO representative describes the
problem as, “Many of the schools they lack special education teachers. So you find children
many are there as routine, that they are there but they are not benefiting really.”
To adequately prepare teachers to support the learning of students with disabilities,
participants alluded to the fact that teacher training for inclusive education must be designed to
meet the unique instructional needs associated with different types of disabilities. For example,
an NGO representative highlighted the following knowledge gaps pertaining to students who are
hearing, visually, and cognitively impaired, respectively: “Teachers should also know about the
hearing aid, should know about the white cane, they should also know the things that will enable
a child who is mentally disturbed.” Likewise, to address the communication needs of students
with visual and hearing impairments, the importance of training teachers in Braille and sign
language was also consistently expressed. The recalled dialogue between a government official

164

whom I interviewed and the head of a secondary school in Njombe artfully illustrates the need to
train teachers to communicate with students with disabilities to foster their learning:
I remember when I met the head of secondary school he said that, “Those students they
don’t want to study, those with hearing impairment.” I asked, “Why?” [The head of the
school said] “Because what we are trying to give them, to teach them, they don’t
understand.” I asked them, “Do you have any teacher within your school who is capable
or expert of sign language?” He said, “No, no.” So how can those hearing impairment
learn when there is no expert of sign language? It becomes a problem because no one
can sign for them…of course they are not learning, they are dumped. The most
[important] issue is to make sure that those students who are sent to school they are able
to achieve.
In terms of connections to the larger metanarrative, teacher training overlaps with the
broader economic and political themes. The development and maintenance of teacher training
programs carry a considerable expense, thereby requiring a political commitment and investment
in human capital and infrastructure. Through my field experience I learned that there is currently
only one special education teacher training college in Tanzania, which is not enough to meet the
needs of inclusive education. Inadequate teacher training is clearly a barrier that needs to be
overcome for inclusive education implementation in Tanzania.
RQ2. Global Influence on Inclusive Education Policy Appropriation
Two key themes emerged in my data with respect to my second research question
focused on global influences on inclusive education policy appropriation in Tanzania. These
include global funding and doctrine, each of which will be elaborated upon below.

165

Funding. Not surprisingly, when I asked participants about global influences on the
creation and implementation of the National Strategy on Inclusive Education 2009-2017, donor
funding was prominently featured in our discussions. The degree to which global donors seemed
to influence both policy formation and policy implementation was quite marked. In terms of
inclusive education policy formation, I was told by government and NGO representatives alike
that Nordic countries funded the entire process. For instance, one NGO staff member said
boldly, “I can say with wide mouth that the question of inclusive education is coming from
outside. I think was it Sweden or Denmark which have funded the project of inclusive education
in the Ministry of Education.” Another NGO representative corroborated, “It’s Finland that
funded the process.”
A similar story was told when discussing the role of international donors in inclusive
education policy implementation. An NGO representative openly shared that, “All the project
that we have been implementing they have been funded by external donors.” Large development
agencies like UNICEF and UNESCO were frequently mentioned as sources of funding for
various inclusive education projects and initiatives that have been carried out by the government
and NGOs. For example, when describing recent and ongoing intervention efforts, a government
official told me:
We have been cooperating with UNESCO, UNICEF on issues of establishing assessment
centers…And then for example last year we inaugurate the guide for people with
albinism, we get the funds from UNICEF and then we prepare the guides for that.
It is important to draw attention to the tone of these types of comments. Across my
interviews, discussions about funding felt very matter-of-fact and seemed to reflect a sense of
fatalism concerning external funding. For instance, the blatancy of the seemingly inevitable

166

reality that inclusive education implementation is funded from abroad is conveyed by the next
quotation. When asked about the funding sources of the projects a particular NGO was pursuing,
the representative responded, “From Washington again (laughing)”. Based on the tone and
brevity of responses such as this, I got the impression that the answers to my questions about
funding sources seemed obvious to my participants; there was almost a sense of confusion as to
why I was asking such an outrageously self-evident question. Furthermore, the preceding
comment alludes to the fact that Tanzania has a historical pattern of relying on donor funds for
development, and inclusive education appears to be no exception. In sum, the link between
global funding and inclusive education policy appropriation in Tanzania was extremely
pervasive.
Global doctrine. In addition to funding, another global influence on inclusive education
policy appropriation in Tanzania is what I am calling global doctrine, or the global policy
documents that promote the rights of children with disabilities and inclusive education. The
main documents were reviewed earlier when presenting the global inclusive education policy
landscape to contextualize this study. Numerous participants directly referred to many of these
global documents when discussing what triggered inclusive education in Tanzania, most notably
the Salamanca Statement, Education for All, and the UN Convention on the Rights of Persons
with Disabilities (CRPD). These three documents and their collective influence on inclusive
education in Tanzania are reflected in the following quotation from an NGO representative:
The international [community] is speaking of the implementation of Salamanca
whatever…education for all…Our country having signed that one, having ratified those
kind of documents, the UN documents…even the optional protocol was also signed. So
that one becomes legal kind of document, it has now to make sure that the country itself

167

has now to make sure that the different provisions have to be putted in action and one of
that is to make sure that inclusive education is there. I think it has a direct impact
because we have agreed as UN member state. So you cannot deny something which you
have already ratified.
The global doctrine of inclusive education appeared to carry a lot of weight in the eyes of
the individuals whom I spoke with, including members of the government. Further evidence of
the influence of global doctrine can be seen in the subsequent quotation from a government
official. In this part of our conversation he is openly telling me that portions of text in the
National Strategy on Inclusive Education 2009-2017 were directly taken from global documents
like the UN CRPD and the Salamanca Statement:
Some have been taken direct as it is from those documents so that it can be implemented
in Tanzania. Yeah, some of them we decided to adopt direct from those are because
Tanzania is one of the signatory of those documents.
From the perspective of the individuals I interviewed, and as can be seen in both of the previous
quotations, there seems to be a shared perspective that Tanzania’s signature on these documents
holds the nation accountable to fulfill its contractual agreement to inclusive education. In
addition to the pressure that global doctrine creates for Tanzania as a signatory to uphold their
commitment to inclusive education, it also seems to significantly shape the dialogue of inclusive
education in Tanzania because the doctrine’s direct influence is manifested in the nation’s
official policy rhetoric.
RQ3. Local and Global Intersection and Divergence
The various local and global factors explored thus far do not operate in isolation. In an
attempt to point out specific local and global influences, divisions were made above, but in

168

reality all of these factors interact with one another in complex ways throughout the policy
process. This section highlights some key instances where the local and the global intersect and
diverge.
Intersection. Two points of intersection between global and local factors appear to play
a major role in the appropriation of inclusive education in Tanzania. These include the
intersection between global documents and local needs and the intersection between global
funding priorities and local reliance on donor funds.
Global documents intersect with local needs. Building upon the previous discussion
about the influence of global doctrine on inclusive education taking hold in Tanzania, most
participants perceived a meaningful overlap between local needs and the global documents. In
other words, global documents were perceived as complimenting local needs concerning
disability that have long existed in Tanzania. I got the sense that some individuals were quite
adamant that I understand this connection and do not misrepresent Tanzania’s adoption of global
documents as a unidirectional transaction. For example, one NGO representative stressed that:
It doesn’t start from somewhere outside; please now help me to establish disability
policy, no! The needs start from inside - within you, your community, your people, your
country realize that there is a need for it. If yes, then you plan for it. If there no
realization of the need it will not happen. The UN declarations help start with knowledge
which opens up the mind. So the biggest push I would say starts inside. But then it’s
supported by those international instruments. So you could see it is the need, it is the link
between the two layers.
He went on to explain that children with disabilities need a lot of support to go to school in
Tanzania and this has been a challenge for many years. This example serves to illustrate the link

169

between the needs of Tanzanian children and the responsiveness of the global documents to
those local needs.
It is important to point out that this perspective of global documents complimenting local
needs differs somewhat from the argument advocated above that global doctrine as an outside
force impacts local inclusive education policy appropriation. In this instance, more agency is
afforded to local policy actors compared to the former positioning of local policy actors as
somewhat submissive signatories that bear the responsibility to uphold their commitment to
inclusive education. Such agency is evident in how the global documents are used as a tool by
local policy actors to promote their local needs. Global inclusive education documents were
frequently referred to by participants as “tools” or “instruments,” as can be seen in the above
quotation. Additionally, another NGO representative explains how his organization uses a
Kiswahili version of the UN CRPD when advocating for the rights of children with disabilities to
the government:
When we argue, when we talk with government, when we send our reports, when we
demand something we refer to that document. And so long it is the document acceptable
within the country they agree. They agree, so to say it has some impact.
Examples such as these illustrate how local policy actors are skillfully using global documents as
an advocacy tool in an effort to meet their local needs. Such actions are made possible due to the
complimentary nature of the global documents to local needs concerning disability.
Global funding priority area intersects with local reliance on donor funds. A second
intersection between the global and the local in relation to inclusive education policy
appropriation in Tanzania is the connection between inclusive education as a priority funding
area for global donors and Tanzania’s dependence on donor funds for national education

170

development. Participants recognize that the presence of inclusive education on the international
development agenda has financial implications for global investment and aid in developing
countries like Tanzania. Given the economic realities of the Tanzanian context, local policy
actors are quite reliant on donor funding and the case of inclusive education is no exception.
Individuals from the government and from NGOs frequently spoke about the willingness of
global donors to support or fund inclusive education in a very positive light. For example, an
NGO representative explained how global attention to and support of inclusive education has
helped their initiatives:
And the, the, I mean the inclusion of person disability in the development process being
one of the international agenda. Yeah, has helped a lot for us to say okay, Tanzania is
not an island playing our own song – we have to play to the music of the world. So that
also to our intervention but also availability of donors who are ready and, you know,
willing to support, you know, our initiatives of promoting uh inclusive education.
Similarly, a government official explained how international aid from organizations like
UNICEF and UNESCO helped address their need for resources:
UNICEF and UNESCO in most cases they help us to implement our plans that is due to
that, sometime we don’t have enough resources. And when we are stacked we tend to ask
them for assistance, especially for resources. For example even these documents was
printed by UNESCO themselves…Even the fund for training was from UNESCO.
Global financial support has been instrumental in advancing inclusive education in Tanzania.
The influence of international funding on inclusive education would not be as great if there was
not such a great need for resources in Tanzania and if inclusive education was not a priority

171

funding area on the international development agenda; the global and local forces intersect and
work in tandem.
Divergence. When examining the interplay between global and local forces in the
appropriation of inclusive education in Tanzania, there were not only points of intersection like
those discussed above, but there were also points of divergence. The most notable instance
where the global and the local diverged was in terms of attitudes about disability and inclusive
education. The negative cultural attitudes towards disability discussed previously for Research
Question 1 can be juxtaposed against broader global beliefs about the rights of people with
disabilities. A government official eloquently describes the stark contrast between global and
local perspectives from her point of view:
You see like, like here they had one myth that a person with disability is a poor person. A
person with disability cannot learn. A person with disability cannot marry. A person
with disability have no right to vote. And if you kill him or her you are going to be rich –
like those are local ideas. Like, or if it’s others they say if you will educate him God will
bless you, because they base on myth. Faith, you know. If I educate this person with
disability maybe God will bless me. But it’s right, it’s her right or his right to take.
But global, global means it’s international. Things are clear, are open, and are in
scientific way. They see person with disability is like other person. And they need to get
all rights to have a sex, to get married, to study, to acquire education, everything.
Local is local my dear! (laughing)
This participant associates local perspectives about disability with myth, religion, and tradition,
while she associates global perspectives with science. Interpreting this scientific versus nonscientific distinction one step further, this may suggest an implicit association between the global

172

with modernity and the local with traditionalism. Additionally, global perspectives are
described as clear and open, implying that local beliefs are more complicated and perhaps
obscured. Such divergent perspectives about disability and in turn inclusive education are
important to consider when diving deeper into specific instances of local resistance, adaptation,
and subterfuge in the next subsection.
RQ3a. Local Resistance, Adaptation and Subterfuge. Instances of local resistance,
adaptation, and subterfuge emerged from my data when looking more broadly for cases of
divergence between global and local forces. In addition to divergence in a broad sense between
global and local attitudes towards disability and inclusion aforementioned, the ways that local
policy actors have actively resisted inclusive education, have adapted inclusive education to fit
the Tanzanian context, and have skillfully navigated the inclusive education policy process to
reach their own objectives presented themselves as additional examples of divergence in more a
more subtle manner. Examples of local resistance, adaptation, and subterfuge are presented
below.
Resistance. On the surface it appeared that all participants supported the global notion to
include all students with disabilities, but when probing deeper I discovered evidence of some
dissenting voices that are resisting inclusive education. Local disagreement about inclusive
education as the most appropriate means to educate children with disabilities seemed to be a
point of contention. For example, an NGO representative disclosed that Tanzanians, “Are
accepting, some are like resistance some, some, but, you know, the government people so is not
like hundred percent.” Another NGO representative gave the following example of dissenting
voices when discussing areas of local conflict concerning inclusive education: “I think, for
people with disability, they say, ‘No inclusive education is not good.’”

173

Differing opinions on the matter seemed to be especially pronounced among various
disability groups. Multiple participants from the government and from NGOs told me about the
general trend that a subset of the disabled community feels that children with severe disabilities
should be educated separately in special schools or units, while children with mild disabilities
can be included in regular schools. For example, a blind individual told me that the blind
community feels that they need special schools or units to be taught braille, which is not feasible
in an inclusive classroom. Consequently, the position advocated by the blind community
according to the NGO representative I spoke with is that inclusive education may be possible in
upper educational levels, but not in lower primary:
Especially at early level like primary one or two…So with lower classless, we do not
believe in inclusive education. Because of the braille language we think if the children
does not get the required support for them to learn braille, good braille, have command
in braille, and then that is going to deny them opportunity to participate fully in the
process of acquiring education in higher levels.
I learned that these types of concerns have been voiced during the collaborative
stakeholder policy formation meetings explained above when discussing the political context.
Using platforms such as these meetings, some disability groups seem to have the reputation
among the DPO community of trying to dissuade the government from adopting the more
holistic global vision of inclusive education. Putting restrictions on who can be included and at
what point in their education they can be included runs counter to global doctrine which
advocates for the inclusion of all children at all stages of their education. Such disagreement or
lack of consensus about inclusive education can create challenges for implementation, as
reflected by the following quotation from an NGO representative:

174

The other barrier, the other barrier is about to these other NGOs, disabled peoples
organizations of which they do not support inclusive education totally. Now for them if
they continue to preach that inclusive education is not favorable, if they continue to do so
and therefore the government will have a lot of challenges because it will continue to
preach the good of inclusive education and then there are people who are going to
oppose that one. So, to me I think that is a challenge because the government and the
civil society have to speak the same thing for the changes to happen.
Overall, partial support for inclusion and active resistance by some disability groups illustrate
divergence from the global; the local understanding and promotion of inclusive education for
some runs counter to the global philosophy of inclusive education for all.
Adaptation. When discussing the relationship between global and local forces,
participants seemed to recognize the need to adapt or localize inclusive education to the
Tanzanian context. For instance, a faculty member explained that to make inclusive education
more locally relevant:
We can say, okay, our education is responsive to all. When we are saying responsive to
all we have to define which type of society are we having? There are the disconnecting,
things that are disconnecting, is actually differences between countries: differences of
economic status, differences in terms of culture, also cultural differences, differences in
terms of connectivity when you are talking of electricity, differences about social
responsibilities. These are the differences that come to make inclusive education to all.
All of these differences are important to consider when developing and implementing an
inclusive education policy in a given context.

175

While there were many instances similar to the preceding quotation where my
participants and I discussed the importance of adapting global policy to the local context in a
broad or abstract sense, it was difficult to attain concrete examples of local adaptations. For
instance, one NGO representative viewed the National Strategy on Inclusive Education 20092017 document itself as evidence of local adaptation:
It was a pressure from outside and we said is worth having it and we, and try to localize
it so it fits to our environment because now we have developed our own national strategy.
So we are not implementing international one but our own national inclusive education
strategy with a lot of ideas borrowing from the international strategy.
When asked about specific examples of how the national strategy diverges from global policy
documents, the conversation continued to stay at the abstract level as he discussed local
“approaches” in a broad sense. Likewise, another NGO representative cited the inclusion of incountry budget details in the National Strategy on Inclusive Education 2009-2017 as an example
of local adaptation.
A final example comes from an interview with a government official where we discussed
his work developing new inclusive education training booklets for teachers in collaboration with
UNESCO. He explained that the documents were originally developed for Asian and Pacific
countries but changes were made to contextualize it to the Tanzanian context. When I asked
about specific changes that were made, he mentioned cultural and language issues in case
studies:
Uh we tried to check with case studies which we have been using in the, which have been
using in the former document. We changed it, we changed those cases, those case studies
so that it can suit within our environment. And even those issues we thought that it is not

176

applicable in Tanzania we decided to change them, to delete them so that we
accommodate all context that we see that it suitable for Tanzanian environments…the
language for the Islamic language which have been used in that book we decided to skip
them. And also there some, also traditions issues for Asia or some cultural issues of
Asian we decided to abandon them.
Based on these types of comments, there appears to be limited local adaptation in Tanzania’s
inclusive education policy documents overall.
Subterfuge. A final way that divergence between the global and the local was seen is
through subterfuge. In my assessment of the data, to an extent I feel that inclusive education was
adopted in Tanzania to meet the complimentary yet distinct objectives of getting international aid
and investing in education generally. Each of these objectives is discussed below.
To a degree, inclusive education is perceived as a point of access to donor funds. As
explained previously, the local actors I interviewed recognized that inclusive education is on the
international agenda and as a result is tied to money from external sources. In the context of the
nation’s low economic status and dependency on international aid, it seems that one motivation
behind Tanzania’s embracement of inclusive education is financial in nature. To explain, it
seems that Tanzania adopted inclusive education almost out of fear that donor funds would be
withdrawn if they resisted inclusive education as a policy initiative. For example when
discussing global donors in relation to inclusive education, a government official openly
admitted, “We need to do what they say…we do not want to make them angry.”
Additional evidence of this fear and its rationale is illustrated in the following excerpts
from an interview with an NGO representative. When talking about Tanzania’s adoption of
inclusive education, I was told:

177

In Tanzania creating polices and ratify is very simple. We are good at that, very simple.
But now, you know, Tanzania is very much worried with anything that will impact from
the international arena. Most of the things which come from the international arena are
being ratified very quickly: protocols, conventions, you know.
When I asked why, he explained:
Most of our budget comes from outside, from outside organizations. So they are worried
to create any, any bad relationship with the international because we are mostly
depending on them, investors from outside. More of our budget, almost half of our
budget are coming from European countries, American countries, so and so on.
Tanzania is very much worried about anything that will defend their status, that are the
international arena. So most of the lobbying from outside will work in Tanzania. When
they are told, “Do this” from outside they will quickly do it.
Based on this portion of our discussion, I got the sense that Tanzania adopted inclusive education
from a policy standpoint quite quickly in large part due to a fear of creating poor international
relations which would in turn impact external funding.
The power of donor funds to shape inclusive education policy in Tanzania is further
illustrated by the following comment from an NGO representative, “Donor, the donors,
development partners, you know? They have come with that song so we have to sing it
(laughing)”. This suggests that Tanzania is singing along or going along with the priorities of
the donors to ensure that financial aid continues. Currently, inclusive education is the song that
policy actors in Tanzania feel that they need to sing along with. Put another way, a different
NGO representative sees the financial utility of joining the global inclusive education movement
by referring to inclusive education as a “good machinist for the sake of the donors.” Here,

178

inclusive education is viewed as a machine operating within the larger apparatus of educational
development that facilitates the transfer of funds. These comments bring into question the
possibility that if the donors come with a new song or a new machine (i.e., a new educational
development priority), then local policy actors will shift gears so to speak and transition to the
next funding priority.
In a related vein, inclusive education was perceived as a way to support educational
development in general. This distinction is nuanced and can be seen in an NGO representative’s
response to what triggered inclusive education in Tanzania: “One is for demand, demand for
inclusive, no, demand for education. Not for inclusive.” I found it interesting that this
participant corrected himself. His self-correction conveys the subtle distinction that the local
demand for education on the whole is greater than the demand for inclusive education
specifically.
This point became clearer after interviewing additional participants who expressed the
need for their nation to improve its education system overall as a central concern. To illustrate,
an NGO representative explains the local demand for education as:
For us we think because we have hunger with education. We do not have enough
opportunity for everybody to get education, in the right way, at the right time. Now there
are people who come to say they have the way how to make it happen.
This participant perceives inclusive education as one potential solution from outside people to
meet Tanzania’s hunger for education. Similarly, another NGO representative disclosed:
Of course it started with, we borrowed this idea from foreigners, especially from Europe,
Westerners, Europe. Whenever any partner would accept funding us, the thing that will

179

ask is that we are funding education. But how do you think of this idea of inclusive
education? So there was some sort of influence from outside.
The two preceding quotations suggest that the idea of inclusive education has been
suggested or imposed by global forces as a way to improve education in Tanzania. Put another
way in a metaphorical sense, inclusive education can be viewed as an international prescription
to remedy Tanzania’s educational ills. Local actors want to improve education on the whole and
outside global actors are prescribing inclusive education as the solution. Local actors are eager
for assistance with educational development, so accept the course of treatment.
When considering subterfuge on the whole, inclusive education appears to be a means to
two ends. The first is as a means to securing international funding. The second is as a means to
promote educational development. There are undoubtedly multiple motivations in play. Aside
from the more obvious goal of providing children with disabilities with an education, these are
two additional objectives that also seem to be underpinning inclusive education policy
appropriation in Tanzania.
Discussion
The findings of this study suggest that the case of inclusive education is somewhat
consistent with Tanzania’s general historical trend of adopting policies from the global
North/West and relying on external funding (Samoff, 1994; Vavrus, 2003). The asymmetrical
power relations that characterize global policy transference between developed and developing
countries (Anderson-Levitt & Alimasi, 2001; Ginsburg et al., 1991; Steiner-Khamsi, 2010) are
no doubt evident in the present study, as global funding and global doctrine seemed to directly
influence inclusive education policy appropriation. However, the findings concerning the
influence of multiple local factors and their intersection and divergence with global factors

180

complicates the notion that the global diffusion of educational ideals flows in a unidirectional
fashion from the global North/West to the global South/East (Anderson-Levitt & Alimasi, 2001;
Steiner-Khamsi, 2010).
In terms of the multiple local socio-cultural, political, and economic factors that emerged
from the data, Tanzania’s low-income economy status (World Bank, 2013) definitely seemed to
play a role in hampering inclusive education policy appropriation through government budget
and familial economic constraints. Socio-cultural beliefs about disability and their ties to social
stigma and discrimination also featured prominently in interviews, particularly concerning
albinism. It is worth noting that although it did not come up in my interviews, other
anthropological literature points to additional sociocultural, historical and economic factors that
impact what was communicated to me as discriminatory attitudes and practices associated with
albino killings. For example, rapid economic and social change fostered by structural adjustment
and economic liberalization has been connected to the efforts of miners and fisherman in
Northwest Tanzania to secure lucky albino charms from healers for finding minerals or fish
(Bryceson, Jønsson, & Sherrington, 2010; Schühle, 2013; Tanner, 2010), thereby illustrating the
multifaceted nature of these beliefs and practices. I want to assert that my findings do not
support a discourse of primitivism, but attempt to reflect participant perspectives that are
temporally and contextually bound.
In terms of political factors, the government’s lack of political will was a primary focus
in discussions about political influence, which is consistent with previous research on inclusive
education in Tanzania (Mkumbo, 2008). Similarly, all barriers to inclusive education that are
commonly reported in the literature including physical infrastructure, attitudes, teacher training,
and lack of learning materials and resources (Chataika et al., 2012; Eleweke & Rodda, 2002;

181

Marshall, 1997a, 1997b; Mbwilo et al., 2010; Miles, 2011; Mkumbo, 2008; Mnyani, 2009;
Woods, 2008) were found in this study. Together, all of these local influences suggest that
inclusive education policy appropriation in Tanzania is not only driven by external global forces.
Beyond documenting the presence of these local factors in Tanzania’s inclusive education policy
process, this study has also generated many nuances in relation to local influences. For example,
with regards to the perceived lack of political will on the part of the government to promote
inclusive education, the data also revealed a degree of recognition that the government has to
address more pressing educational issues like secondary education and teacher absenteeism,
which impacts the low prioritization of inclusive education. Additionally, when considering
attitudes, the findings of this study suggest that it is important to consider the influence of limited
knowledge about disability because negative attitudes and a lack of information seemed to go
hand in hand.
Lastly, this study further complicates the notion of global policy transference by
illustrating the complex interplay between global and local forces in inclusive education policy
appropriation in Tanzania. Points of intersection between the global and the local were present
in the data, including the intersection between global documents and local needs and between
global funding priorities and local reliance on donor funds. These points of intersection show
how global and local forces can work together in tandem to move inclusive education forward.
Conversely, points of divergence between the global and the local were also found in the data
such as divergent attitudes about disability and inclusive education between global and local
levels as well as between different disability groups within Tanzania. Evidence of subterfuge
was also found, suggesting that inclusive education may have been strategically adopted as a
way to access donor funds or reluctantly accepted as a global prescription to treat Tanzania’s

182

educational ills. These points of resistance and subterfuge demonstrate that global and local
forces can clash and the local adoption of inclusive education may have some strategic motives
beyond supporting the education of students with disabilities.
Limitations and Future Directions
Although this study extends what is known about global and local influences on the
inclusive education policy process in Tanzania, it is important to note a few limitations of this
work that should be addressed by future research. First, the perspectives represented by this
qualitative study are limited to the sample of individuals that I interviewed. My combined
methods of purposive and snowball sampling enabled me to capture a variety of perspectives
across national policy actors working in the education sector, but I did not interview government
officials from all divisions and units of the ministry nor did I interview representatives from all
national and multinational organizations operating in Tanzania. The opinions of national policy
actors not included in my sample may differ from those of my participants, and future research
should extend my work to continue to give voice to additional perspectives. Additionally,
further research exploring the perspectives of the disabled community should consider budgeting
for and providing participants with transportation and interpreter services to accommodate travel
and communication needs. I was unable to interview a representative from one DPO who
requested financial compensation for transportation and an interpreter, and this unfortunate
constraint on participation could be prevented in future research.
Related to my sample characteristics is the limitation of the level of my analysis of
inclusive education policy appropriation in Tanzania. Returning to the theoretical framework
depicted in Figure 3.2, this study was largely limited to the upper sections of the diagram
because national policy actors were interviewed. While some insight into the lower levels of the

183

diagram (i.e., implementation at the regional, district, ward, village, and school levels) was
discussed to a degree, it is important to recognize that it was from the perspective of high-level
policymakers who in most cases are fairly removed from these local contexts. Consequently,
future work should more closely analyze policy appropriation at the more micro levels depicted
in Figure 3.2 from the perspectives of local policy actors at regional, district, ward, village, and
school levels. Doing so may yield additional instances of local adaptation beyond the limited
amount that were found here. Research of increasingly more micro levels where practitioners on
the ground are confronted with local realities and attitudes deeply imbedded in local contexts
might tell a different story.
Another direction for future research is to more closely explore the relationship between
ujamaa and inclusive education in Tanzania. Interestingly, Tanzania’s political history with
African Socialism and ujamaa did not emerge during any of my interviews. This may have been
because I did not explicitly ask about ujamaa. Instead, when asking about local factors I
intentionally phrased my questions broadly to avoid biasing or guiding participant responses in a
given direction. Future research may want to explicitly ask about the connection, or lack thereof,
between ujamaa and inclusive education. Alternatively, it is possible that ujamaa was not
mentioned by participants because from their perspective this African socialist ideal is not as
complimentary to the ideology of inclusive education as one might expect. While previous
literature suggests that ujamaa is aligned with inclusive education ideals (Tungaraza, 1994),
perhaps the cultural and political spirit of ujamaa creates social division because disability is a
social marker of difference (i.e., people with disabilities may not fall within the collective notion
of familyhood). Further investigation into this possibility could be helpful for inclusive

184

education implementation efforts to determine whether ujamaa might help or hinder cultural
understanding of inclusive education policy.
Implications
Despite its limitations, the results of this study are valuable for policy actors in Tanzania
and internationally to consider. First, participants listed numerous local barriers that should be
addressed as inclusive education implementation in Tanzania continues. Continued advocacy
efforts like the public rally discussed above will be important to change negative cultural
attitudes towards disability and inclusive education across all levels of society and encourage
parents to send their children with disabilities to school. Given the apparent connection between
negative attitudes and a lack of understanding, educating the citizenry about disability and
providing information about inclusive education can help address the knowledge gap and give
parents of children with disabilities more capacity to enroll their children. Additionally, the
perceived lack of political will should be addressed through improved government budget
allocation to inclusive education by targeting areas such as teacher training, the physical
accessibility of school infrastructure, and the availability of teaching and learning resources.
Continued and improved collaborating between the government and community stakeholders
will also help build a unified inclusive education movement in Tanzania.
Global and local partners should also work together to acknowledge and understand the
reasons behind areas of resistance and subterfuge. Failing to do so may threaten the
sustainability of current and future inclusive education policy development and implementation.
Tanzania, like many other developing nations, has been influenced by and responded to global
pressure to adopt inclusive education. To carry inclusive education forward beyond rhetoric,
local commitment and investment are invaluable. While seemingly positive advancements in

185

policy, advocacy, and implementation have been made in Tanzania, it is too early to tell if the
current direction that inclusive education policy appropriation is taking will be successful long
term. The presence of local resistance may threaten the realization of inclusive education, and
perhaps further local adaptation would be useful to reach a more mutually agreeable education
development solution. Additionally, while the instances of subterfuge (i.e., using inclusive
education as a means to secure international funding or as a means to promote educational
development) do not appear to wholly undermine the inclusive agenda at this point, they may
threaten the sustained advancement of inclusive education. The longevity of inclusive education
reform would be tested for example when global funding priorities shift or the global
prescription of educational development changes. Continued advancements in local adaptations
could be a promising area of opportunity for global and local partnerships to focus on to prevent
the potentially harmful effects of local resistance and subterfuge, thereby promoting sustained
inclusive education policy appropriation in Tanzania.
In addition to implications for Tanzania, the findings from this study raise important
questions regarding the global spread of inclusive education policy and practice. As inclusive
education continues to proliferate across the globe, it will be increasingly important to consider
how global and local forces will intersect and diverge as additional developing countries adopt
inclusive education. It is vital for international development actors to be attentive to local needs
and interests that are shaped by socio-cultural, historical, economic, and political factors to
support sustainable inclusive education efforts. Joint efforts between global and local partners
should be targeted at addressing local barriers, resistance, and subterfuge. To do so, meaningful
local adaptation of the global ideal of inclusion is essential to advance inclusive education reform
beyond rhetoric and create long term change.

186

Conclusion
The primary goal of this study was to examine how global and local forces have shaped
inclusive education policy appropriation in Tanzania from the perspectives of national policy
actors in the education sector. Semi-structured interviews with 20 policy actors in Dar es Salaam
working in the government, non-governmental organizations, and universities revealed how
multiple local and global factors have influenced the creation and implementation of the National
strategy on Inclusive Education 2009-2017. Key local factors identified by participants fell
under the broad themes of socio-cultural beliefs, political context, and economic conditions and
included stigmatizing cultural attitudes towards disability, stakeholder collaboration, political
will, government budget, and familial economic constraints. Additionally, participants felt that
attitudes, funding and availability of resources, physical barriers, and teachers were common
barriers that local actors face during inclusive education implementation. Key global factors
identified by participants were external funding and global doctrine for inclusive education.
Global and local factors were found to intersect and diverge in different ways that served
to facilitate or inhibit inclusive education efforts. Points of intersection that facilitated inclusive
education efforts were global documents intersecting with local needs and the global funding
priority area intersecting with local reliance on donor funds. A point of divergence that inhibited
inclusive education efforts was conflicting global and local attitudes about disability and
inclusive education. Additionally, participants shared instances of local resistance, adaptation
(though minimal), and subterfuge.
Together, these findings illustrate the complex and dynamic relationship between local
and global forces in the case of inclusive education policy appropriation. Tanzania’s national
inclusive policy formation and implementation process is clearly an interactive process between

187

the global and the local that is built upon spaces of uneven and common ground. As inclusive
education continues to expand within Tanzania and globally, it is important to consider the
interplay between global and local forces to address barriers and create sustainable educational
initiatives that meet their intended purpose of supporting the education of children with
disabilities.

188

APPENDICES

189

APPENDIX A

Interview Protocols

190

Note: These interviews will be semi-structured. The protocols are subject to conversational
variation and use of probes as indicated. Below are the main questions that I asked my
participants.
Interview Protocol: Ministry
Before we begin do you have any questions for me that I can answer or clarify? Shall we begin?
Let us begin with some background questions.
1. Can you please tell me briefly about your current responsibilities and how long you have
been working at [insert name of Ministry department]?
2. What does inclusive education mean to you? (probe for source of definition, for scope of
definition - if mention disability probe for definition of disability, for example of what
inclusive education looks like)
Now let’s move on to some questions about your views on the formation of inclusive education
policy in Tanzania to help me understand the process of policy formation in this context.
3. In your opinion, what triggered the creation of an inclusive education policy in Tanzania?
(probe for most important historical moments, for international versus grassroots
pressures, for who was motivated to create an inclusive education policy and why)
4. Who do you feel were the key policy actors and organizations that were involved in the
formation of the policy? Who had the greatest influence and why?
5. Compared to other areas, do you feel that inclusive education is a priority policy area for
Tanzania? Why or why not?
6. Is there anything else that you would like to share that you think is important for me to
understand how Tanzania’s inclusive education policy was formed?
Now I would like to ask you some questions about global influences on inclusive education
policy formation and implementation in Tanzania.
7. Do you think that the ideas represented in Tanzania’s inclusive education policy are more
representative of global ideas, local ideas, or both?
8. Are you familiar with global initiatives for inclusive education such as Education for All,
Universal Primary Education as expressed in the Millennium Development Goals, and
the Salamanca Statement and Framework for Action on Special Needs Education?
a. (If respond no, explain): These types of global initiatives say that all children
should have access to quality education in inclusive settings, regardless of their
ability or background. What do you think about these types of global initiatives?

191

b. (If respond yes, probe further): What do you think about these types of global
initiatives? Are there any additional global initiatives I did not mention that you
feel inform Tanzania’s inclusive education policy?
i. How have these global initiatives influenced Tanzania’s inclusive
education policy formation and implementation? (probe for direct and
indirect influences)
ii. Has the reception in Tanzania been warm, lukewarm, or cold? Why do
you think that may be the case? (probe for examples)
Next I would like to learn about the activities of [insert name of Ministry department] in the
formation and implementation of Tanzania’s inclusive education policy.
9. Has your department been involved in the formation of Tanzania’s inclusive education
policy?
a. (If respond no): *Skip to question 10
b. (If respond yes):
i. Can you tell me about how your department was involved in policy
formation? (probe for departmental and personal roles and responsibilities,
for relevant documents/pamphlets)
ii. What led your department to do this? (probe for decision making process
and structure, for local/global policy models or guidelines)
10. Has your department been involved in the implementation of Tanzania’s inclusive
education policy?
a. (If respond no): Who do you feel is responsible for implementing the policy and
why? Can you tell me about some of the other areas your department focuses on?
*Skip to question 11
b. (If respond yes):
i. Can you please describe the inclusive education implementation strategies
and programs that your department is doing? (probe for examples, for
implementation locations, for relevant documents/pamphlets)
ii. Who is responsible for carrying out these implementation efforts? (probe
for layers of accountability, for flow of implementation from national to
local level)
iii. How are these implementation strategies funded?
iv. What led your department to implement these strategies? (probe for
decision making process and structure, for local/global implementation
models or guidelines)
v. Are these implementation strategies working? Why or why not?

192

1. What sorts of factors would you say influence your implementing
strategies? (e.g., economic conditions, political context, culture,
religion)
2. Are some of these factors particularly helpful to your
implementation efforts? What is the greatest hindrance?
vi. Has your department partnered or consulted with international
governments or global organizations in your efforts to implement
inclusive education policy?
1. (If respond no): Are there any particular reasons why [insert name
of Ministry department] decided not to partner with global
organizations?
2. (If respond yes, probe for power differentials, for resistance,
adaptation and subterfuge):
a. What are their roles and responsibilities? What are yours?
(probe for frequency of partnerships)
b. Was there any conflict of ideas or approaches between your
department and the global partner? (If respond yes): How
did you reach an agreement? (probe for positive versus
negative experiences)
11. Does your department have any inclusive education implementation plans for the future?
Describe.
Let’s close with a question about the future of inclusive education in Tanzania.
12. Do you personally believe that inclusive education has been achieved in Tanzania? Can it
be achieved? Why or why not? What needs to happen to make this possible? (probe for
changes in policy and implementation, for change at different levels, for timeline)
Lastly,
13. Can you recommend any other people that I should talk to who would be able to help me
understand inclusive education policy formation and implementation in Tanzania? (probe
for names and contact information)
That ends our interview. Nime shukuru.

193

Interview Protocol: NGOs
Before we begin do you have any questions for me that I can answer or clarify? Shall we begin?
Let us begin with some background questions.
1. Can you please tell me briefly about your current responsibilities and how long you have
been working at [insert name of NGO]?
2. What does inclusive education mean to you? (probe for source of definition, for scope of
definition - if mention disability probe for definition of disability, for example of what
inclusive education looks like)
Next I would like to learn about the activities of [insert name of NGO] in the formation and
implementation of Tanzania’s inclusive education policy.
3. Has your organization been involved in the formation of Tanzania’s inclusive education
policy?
b. (If respond no): *Skip to question 4
c. (If respond yes):
i. Can you tell me about how your organization was involved in policy
formation? (probe for departmental and personal roles and responsibilities,
for relevant documents/pamphlets)
ii. What led your organization to do this? (probe for decision making process
and structure, for local/global policy models or guidelines)
4. Has your organization been involved in the implementation of Tanzania’s inclusive
education policy?
d. (If respond no): Who do you feel is responsible for implementing the policy and
why? Can you tell me about some of the other areas your organization focuses
on? *Skip to question 5
e. (If respond yes):
i. Can you please describe the inclusive education implementation strategies
and projects that your organization is doing? (probe for examples, for
implementation locations, for relevant documents/pamphlets)
ii. Who is responsible for carrying out these implementation efforts? (probe
for layers of accountability, for flow of implementation from national to
local level)
iii. How are these implementation strategies funded?
iv. What led your organization to do this project? (probe for decision making
process and structure, for local/global implementation models or
guidelines)
v. Are these implementation strategies working? Why or why not?
194

1. What sorts of factors would you say influence your implementing
strategies? (e.g., economic conditions, political context, culture,
religion)
2. Are some of these factors particularly helpful to your
implementation efforts? What is the greatest hindrance?
vi. Has your organization partnered or consulted with international
governments or global organizations in your efforts to implement
inclusive education policy?
1. (If respond no): Are there any particular reasons why your
organization decided not to partner with global organizations?
2. (If respond yes, probe for power differentials, for resistance,
adaptation and subterfuge):
a. What are their roles and responsibilities? What are yours?
(probe for frequency of partnerships)
b. Was there any conflict of ideas or approaches between your
department and the global partner? (If respond yes): How
did you reach an agreement? (probe for positive versus
negative experiences)
5. Does your organization have any inclusive education implementation plans for the
future? Describe.
Now let’s move on to some questions about your views on the formation of inclusive education
policy in Tanzania to help me understand the process of policy formation in this context.
6. In your opinion, what triggered the creation of an inclusive education policy in Tanzania?
(probe for most important historical moments, for international versus grassroots
pressures, for who was motivated to create an inclusive education policy and why)
7. Who do you feel were the key policy actors and organizations that were involved in the
formation of the policy? Who had the greatest influence and why?
8. Compared to other areas, do you feel that inclusive education is a priority policy area for
Tanzania? Why or why not?
9. Is there anything else that you would like to share that you think is important for me to
understand how Tanzania’s inclusive education policy was formed?
Now I would like to ask you some questions about global influences on inclusive education
policy formation and implementation in Tanzania.
10. Do you think that the ideas represented in Tanzania’s inclusive education policy are more
representative of global ideas, local ideas, or both?

195

11. Are you familiar with global initiatives for inclusive education such as Education for All,
Universal Primary Education as expressed in the Millennium Development Goals, and
the Salamanca Statement and Framework for Action on Special Needs Education?
a. (If respond no, explain): These types of global initiatives say that all children
should have access to quality education in inclusive settings, regardless of their
ability or background. What do you think about these types of global initiatives?
b. (If respond yes, probe further): What do you think about these types of global
initiatives? Are there any additional global initiatives I did not mention that you
feel inform Tanzania’s inclusive education policy?
i. How have these global initiatives influenced Tanzania’s inclusive
education policy formation and implementation? (probe for direct and
indirect influences)
ii. Has the reception in Tanzania been warm, lukewarm, or cold? Why do
you think that may be the case? (probe for examples)
Let’s close with a question about the future of inclusive education in Tanzania.
12. Do you personally believe that inclusive education has been achieved in Tanzania? Can it
be achieved? Why or why not? What needs to happen to make this possible? (probe for
changes in policy and implementation, for change at different levels, for timeline)
Lastly,
13. Can you recommend any other people that I should talk to who would be able to help me
understand inclusive education policy formation and implementation in Tanzania? (probe
for names and contact information)
That ends our interview. Nime shukuru.

196

Interview Protocol: INGOs
Before we begin do you have any questions for me that I can answer or clarify? Shall we begin?
Let us begin with some background questions.
1. Can you please tell me briefly about your current responsibilities and how long you have
been working at [insert name of INGO]?
2. What does inclusive education mean to you? (probe for source of definition, for scope of
definition - if mention disability probe for definition of disability, for example of what
inclusive education looks like)
Next I would like to learn about the activities of [insert name of INGO] in the formation and
implementation of Tanzania’s inclusive education policy.
3. Has your organization been involved in the formation of Tanzania’s inclusive education
policy?
f. (If respond no): *Skip to question 4
g. (If respond yes):
i. Can you tell me about how your organization was involved in policy
formation? (probe for departmental and personal roles and responsibilities,
for relevant documents/pamphlets)
ii. What led you/your organization to do this? (probe for decision making
process and structure, for local/global policy models or guidelines)
4. Has your organization been involved in the implementation of Tanzania’s inclusive
education policy?
h. (If respond no): Who do you feel is responsible for implementing the policy and
why? Can you tell me about some of the other areas your organization focuses
on? *Skip to question 5
i. (If respond yes):
i. Can you please describe the inclusive education implementation strategies
and projects that your organization is doing? (probe for examples, for
implementation locations, for relevant documents/pamphlets)
ii. Who is responsible for carrying out these implementation efforts? (probe
for layers of accountability, for flow of implementation from national to
local level)
iii. How are these implementation strategies funded?
iv. What led your organization to do this project? (probe for decision making
process and structure, for local/global implementation models or
guidelines)
v. Are these implementation strategies working? Why or why not?
197

1. What sorts of factors would you say influence your implementing
strategies? (e.g., economic conditions, political context, culture,
religion)
2. Are some of these factors particularly helpful to your
implementation efforts? What is the greatest hindrance?
vi. Has your organization partnered or consulted with the Ministry or
Tanzanian civil society organizations in your efforts to implement
inclusive education policy?
1. (If respond no): Are there any particular reasons why [insert name
of INGO] decided not to partner with local organizations?
2. (If respond yes, probe for power differentials, for resistance,
adaptation and subterfuge):
a. What are their roles and responsibilities? What are yours?
(probe for frequency of partnerships)
b. Was there any conflict of ideas or approaches between your
organization and your local partner? (If respond yes): How
did you reach an agreement? (probe for positive versus
negative experiences)
5. Does your organization have any inclusive education implementation plans for the
future? Describe.
Now let’s move on to some questions about your views on the formation of inclusive education
policy in Tanzania to help me understand the process of policy formation in this context.
6. In your opinion, what triggered the creation of an inclusive education policy in Tanzania?
(probe for most important historical moments, for international versus grassroots
pressures, for who was motivated to create an inclusive education policy and why)
7. Who do you feel were the key policy actors and organizations that were involved in the
formation of the policy? Who had the greatest influence and why?
8. Compared to other areas, do you feel that inclusive education is a priority policy area for
Tanzania? Why or why not?
9. Is there anything else that you would like to share that you think is important for me to
understand how Tanzania’s inclusive education policy was formed?
Now I would like to ask you some questions about global influences on inclusive education
policy formation and implementation in Tanzania.
10. Do you think that the ideas represented in Tanzania’s inclusive education policy are more
representative of global ideas, local ideas, or both?

198

11. Are you familiar with global initiatives for inclusive education such as Education for All,
Universal Primary Education as expressed in the Millennium Development Goals, and
the Salamanca Statement and Framework for Action on Special Needs Education?
a. (If respond no, explain): These types of global initiatives say that all children
should have access to quality education in inclusive settings, regardless of their
ability or background. What do you think about these types of global initiatives?
b. (If respond yes, probe further): What do you think about these types of global
initiatives? Are there any additional global initiatives I did not mention that you
feel inform Tanzania’s inclusive education policy?
i. How have these global initiatives influenced Tanzania’s inclusive
education policy formation and implementation? (probe for direct and
indirect influences)
ii. Has the reception in Tanzania been warm, lukewarm, or cold? Why do
you think that may be the case? (probe for examples)
Let’s close with a question about the future of inclusive education in Tanzania.
12. Do you personally believe that inclusive education has been achieved in Tanzania? Can it
be achieved? Why or why not? What needs to happen to make this possible? (probe for
changes in policy and implementation, for change at different levels, for timeline)
Lastly,
13. Can you recommend any other people that I should talk to who would be able to help me
understand inclusive education policy formation and implementation in Tanzania? (probe
for names and contact information)
That ends our interview. Nime shukuru.

199

Interview Protocol: Faculty
Before we begin do you have any questions for me that I can answer or clarify? Shall we begin?
Let’s begin with some background questions.
1. Can you please tell me briefly about your current responsibilities and how long you have
been working at [insert name of university]?
2. What does inclusive education mean to you? (probe for source of definition, for scope of
definition - if mention disability probe for definition of disability, for example of what
inclusive education looks like)
Now let’s move on to some questions about your views on the formation of inclusive education
policy in Tanzania to help me understand the process of policy formation in this context.
3. In your opinion, what triggered the creation of an inclusive education policy in Tanzania?
(probe for most important historical moments, for international versus grassroots
pressures, for who was motivated to create an inclusive education policy and why)
4. Who do you feel were the key policy actors and organizations that were involved in the
formation of the policy? Who had the greatest influence and why?
5. Compared to other areas, do you feel that inclusive education is a priority policy area for
Tanzania? Why or why not?
6. Is there anything else that you would like to share that you think is important for me to
understand how Tanzania’s inclusive education policy was formed?
Now I would like to ask you some questions about global influences on inclusive education
policy formation and implementation in Tanzania.
7. Do you think that the ideas represented in Tanzania’s inclusive education policy are more
representative of global ideas, local ideas, or both?
8. Are you familiar with global initiatives for inclusive education such as Education for All,
Universal Primary Education as expressed in the Millennium Development Goals, and
the Salamanca Statement and Framework for Action on Special Needs Education?
a. (If respond no, explain): These types of global initiatives say that all children
should have access to quality education in inclusive settings, regardless of their
ability or background. What do you think about these types of global initiatives?
b. (If respond yes, probe further): What do you think about these types of global
initiatives? Are there any additional global initiatives I did not mention that you
feel inform Tanzania’s inclusive education policy?

200

i. How have these global initiatives influenced Tanzania’s inclusive
education policy formation and implementation? (probe for direct and
indirect influences)
ii. Has the reception in Tanzania been warm, lukewarm, or cold? Why do
you think that may be the case? (probe for examples)
Next I would like to discuss inclusive education policy implementation and the role of the
government and international and local organizations.
9. Who do you feel is responsible for implementing the policy and why?
10. Which international organizations, government departments, and local civil society
organizations are taking the greatest action to implement the policy? Which are taking
the least action?
a. Can you describe the types of implementation strategies that [insert name of
identified organization] is taking? (probe for detailed examples, for funding
sources, for flow of implementation from national to local school level, for what
leads organizations to do this work, for future implementation plans) *Repeat
question for each organization identified
11. Are these implementation strategies working? Why or why not?
a. What sorts of factors would you say influence implementing strategies in
Tanzania? (e.g., economic conditions, political context, culture, religion)
b. Which of these factors are particularly helpful to implementation efforts? What is
the greatest hindrance?
12. Does the government or local organizations partner or consult with international
governments or organizations in their implementation efforts?
a. How are roles and responsibilities typically divided between global and local
partners? (probe for frequency of partnerships)
b. Do you feel that there are any conflicts of ideas or approaches between global and
local partners? (If respond yes): How are conflicts commonly resolved?
a. Can you think of any reasons why the government or local organizations might
decide not want to partner with global organizations?
Let’s close with a question about the future of inclusive education in Tanzania.
13. Do you personally believe that inclusive education has been achieved in Tanzania? Can it
be achieved? Why or why not? What needs to happen to make this possible? (probe for
changes in policy and implementation, for change at different levels, for timeline)
Lastly,

201

14. Can you recommend any other people that I should talk to who can help me understand
inclusive education policy formation and implementation in Tanzania? (probe for names
and contact information)
That ends our interview. Nime shukuru.

202

APPENDIX B

Informed Consent Forms

203

Research Project: Inclusive Education in Tanzania
Alisha Brown, M.A., doctoral candidate in Education Policy, Michigan State University
Contact Information: brown248@msu.edu
INTERVIEW CONSENT FORM
You are invited to participate in a research project about inclusive education in Tanzania. You
have been selected to participate in this study because you have participated in inclusive
education policy formation or implementation or you have witnessed its impact in a particular
community in Tanzania. The purpose of this research project is to understand the appropriation
of inclusive education policy in Tanzania.
With your permission, I would like to ask you some questions about your participation and
experiences with inclusive education. Your participation in this project is voluntary. You may
refuse to participate or withdraw from this study at any time without penalty. I estimate that the
interview will take about one hour – but you are free to stop it at any time. With your permission
the interview will be audio recorded. However, you may elect to participate in this research
project without being recorded electronically. The audiotapes and transcriptions, as well as all
other notes and written data collected, will be kept in a locked cabinet in my home and your
privacy will be protected to the maximum extent allowable by law. All information disclosed in
the interview will be kept under a pseudonym. Data collected from the interview will be held in
confidence.
You will have the opportunity to ask questions or express concerns about your participation in
this research at any time. Every effort will be made to provide you an opportunity to ask
questions and express concerns throughout the interview process. You will be given the
opportunity to review information gathered in relation to your participation in the study, which
may be submitted for publication in one or more professional journals pertaining to education.
In any and all cases, your anonymity will be maintained to the fullest extent of the law.
You will not directly benefit from your participation in this study. However, your participation in
this study may contribute to the understanding of inclusive education in Tanzania.
There are no foreseeable risks associated with participation in this study. This research is not
being used for tangible personal benefits by the persons responsible for designing it nor do they
have finanical interests in the service being examined. The information gathered in the course of
this research project is part of my studies as a graduate student in Educational Policy at Michigan
State University. My goal is to obtain a better understanding of the formation and
implementation of inclusive education policy in Tanzania so that I may use this research for my
doctoral dissertation.
If there are any questions or concerns over your rights please feel free to contact the principal
investigator (Dr. Amita Chudgar, Faculty, College of Education, Michigan State University, 408
Erickson Hall, East Lansing MI 48824-1034 USA, PHONE (517) 353-5342, amitac@msu.edu).
If you have questions or concerns about your role and rights as a research participant, would like
to obtain information or offer input, or would like to register a complaint about this study, you
may contact, anonymously if you wish, the Michigan State University's Human Research

204

Protection Program at (517) 355-2180, Fax (517) 432-4503, or e-mail irb@msu.edu or regular
mail at 408 W. Circle Dr. Rm 207 Olds Hall, MSU, East Lansing, MI 48824
Your signature below indicates your voluntary agreement to participate in this study.
_______________________________ Date: _____________
□ yes/no to audio recording
You will be given a copy of this form to keep.

205

Jina la mradi: Elimu Jumuishi katika nchini ya Tanzania
Alisha Brown, M.A., Mwanafunzi wa Shahada ya Uzamivu, Masomo ya Sera za Elimu,
Chuo Kikuu cha Michigan State, East Lansing, USA.
Anwani ya barua pepe: brown248@msu.edu
FOMU YA RIDHAA YA MAHOJIANO
Unaombwa kushiriki katika utafiti kuhusu elimu jumuishi katika Tanzania. Umechaguliwa
kushiriki katika utafiti huu kwa sababu ama umewahi kushiriki katika sera za elimu jumuishi au
umeshuhudia manufaa ya programu za elimu jumuishi katika jamii fulani hapa nchini Tanzania.
Nia ya utafiti huu ni kuelewa sera hizi nchini Tanzania.
Ukikubali, ningependa kukuuliza maswali machache kuhusu kushiriki kwako na uzoefu wako
katika elimu jumuishi. Kushiriki kwako katika mradi huu ni kwa hiari yako tu. Unaweza kukataa
kushiriki au kujiondoa kutoka kwenye utafiti huu wakati wo wote bila adhabu. Ninakadiria
kwamba mahojiano yatachukua muda wa saa moja na nusu hivi—lakini unaweza kuacha
kushiriki wakati wo wote. Ukikubali, mahojiano yatarekodiwa. Hata hivyo, unaweza kuamua
kushiriki katika utafiti huu bila kurekodiwa. Nitaweka kanda na unukuzi pamoja na
kumbukumbu zingine zote katika kabati lililofungwa nyumbani kwangu na jina lako pamoja na
kushiriki kwako katika mradi huu zitahifadhiwa kulingana na sheria. Habari zote utakazozitoa
katika mahojiano hazitatumia majina yako halisi na zitahifadhiwa kwa siri.
Utapewa fursa ya kuuliza maswali au kueleza wasiwasi wako juu ya utafiti huu wakati wo wote.
Tutajitahidi kukupa fursa ya kuuliza maswali na kueleza wasiwasi wako wakati wo wote
mahojiano yanapoendelea. Utapewa nafasi ya kupitia taarifa zitakazokusanywa kuhusu kushiriki
kwako katika utafiti huu, ambao unaweza kuwasilishwa ili kuchapishwa katika majarida ya
elimu. Kulingana na sheria, hakuna wakati jina lako litatajwa. Hutafaidika binafsi kutokana na
kushiriki kwako katika utafiti huu. Lakini, kushiriki kwako katika utafiti huu kutachangia katika
kuelewa vile shule zinavyotekeleza sera ya elimu jumuishi katika Tanzania na matokeo yake.
Hakuna madhara yanayotarajiwa kutokea kutokana na kushiriki kwako. Kushiriki kwako
kunaweza kuchangia kuelewa elimu jumuishi katika nchi ya Tanzania.
Hakuna hatari itakayotokea kutokana na kukishiriki kwako katika mradi huu. Utafiti huu
hutumiwi kuwafaidi watu walioupanga mradi huu au kuwawezesha kupata malipo kutokana na
huduma zinazochunguzwa. Hawakusudii kunufaika kutoka programu au sera za elimu jumuishi.
Habari zote zitakazokusanywa katika utafiti huu zitakuwa sehemu ya masomo yangu kama
mwanafunzi wa shahada ya uzamivu katika masomo ya sera ya elimu katika Chuo Kikuu cha
Michigan State, nchini Marekani. Lengo langu ni kuelewa vizuri matayarisho na utekelezaji wa
elimu jumuishi katika Tanzania ili niweze kutumia matokeo ya utafiti huu katika tasnifu yangu
ya shahada ya uzamivu.
Iwapo una maswali au wasiwasi juu ya haki zako katika utafiti huu, tafadhali wasiliana na mtafiti
mkuu (Daktari Amita Chudgar, Faculty, College of Education, Michigan State University, 408
Erickson Hall, East Lansing MI 48824-1034 USA, PHONE 517-353-5342, amitac@msu.edu).
Iwapo una maswali au wasiwasi juu ya haki zako kama mshiriki katika mradi huu au kama
huridhiki wakati wo wote juu ya jambo lo lote katika utafiti huu wasiliana, kwa siri kama
unataka, kwa kupiga simu au kuandika: Michigan State University's Human Research Protection
206

Program at 517-355-2180, Fax 517-432-4503, au tumia barua pepe: irb@msu.edu au barua ya
kawaida kwa anwani hii: 408 W. Circle Dr. Rm 202 Olds Hall, MSU, East Lansing, MI 48824
Saini yako hapa chini inaonyesha umekubali kwa hiari kushiriki katika mradi huu.
Saini:_______________________________ Tarehe: _______________________
□ Ndio/hapana kurekodiwa kama sehemu ya mradi huu.
Utapewa nakala ya formu hii ubaki nayo.

207

REFERENCES

208

REFERENCES

Abosi, C. O. (1996). Early childhood education of children with disabilities in Botswana. African
Journal of Special Needs Education, 1, 33-40.
Anderson-Levitt, K. (Ed.). (2003). Local meanings, global schooling: Anthropology and world
culture theory. New York, NY: Palgrave Macmillan.
Anderson-Levitt, K., & Alimasi, N. (2001). Are pedagogical ideals embraced or imposed? The
case of reading instruction in the Republic of Guinea. In B. A. U. Levinson & M. Sutton
(Eds.), Policy as practice: Toward a comparative sociocultural analysis of educational
policy (pp. 33-58). Westport, CT: Ablex.
Armstrong, A. C., Armstrong, D., & Spandagou, I. (2010). Inclusive Education: International
policy and practice. London: Sage Publications.
Arnove, R. F. (1999). Introduction: Reframing comparative education: The dialectic of the
global and the local. In R. F. Arnove & C. A. Torres (Eds.), Comparative Education: The
dialectic of the global and the local (pp. 1-23). Boulder, CO: Rowman and Littlefield.
Boli, J. (2005). Contemporary developments in world culture. International Journal of
Comparative Sociology, 46, 383-404.
Bryceson, D., F., Jønsson, J., B., & Sherrington, R. (2010). Miners’ magic: Artisanal mining, the
albino fetish and murder in Tanzania. Journal of Modern African Studies, 48(3), 353-382.
Chataika, T., Mckenzie, J. A., Swart, E., & Lyner-Cleophas, M. (2012). Access to education in
Africa: Responding to the United Nations Convention on the Rights of Persons with
Disabilities. Disability & Society, 27(3), 385-398.
Coulson, A. (2013). Tanzania: A political economy (2nd ed.). Oxford, United Kingdom: Oxford
University Press.
Eleweke, C. J., & Rodda, M. (2002). The challenge of enhancing inclusive education in
developing countries. International Journal of Inclusive Education, 6(2), 113-126.
Erickson, F. D. (1986). Qualitative research in education. In M. C. Wittrock (Ed.), Handbook of
research on teaching (3rd ed., pp. 119-161). Washington, DC: American Educational
Research Association.
Ginsburg, M. B., Cooper, S., Raghu, R., & Zegarra, H. (1991). Educational reform: Social
struggle, the state, and the world economic system. In M. B. Ginsburg (Ed.),
Understanding educational reform in global context: Economy, ideology, and the state
(pp. 3-47). New York: Garland.
209

Glesne, C. (2011). Becoming qualitative researchers: An introduction (4th ed.). Boston: Pearson.
Hatch, J. A. (2002). Doing qualitative research in educational settings. SUNY Press.
International Labour Office. (2004). United Republic of Tanzania country profile: Employment
of people with disabilities: The impact of legislation (East Africa). Geneva, Switzerland:
Author.
Kamens, D. H., & McNeely, C. L. (2009). Globalization and the growth of international
educational testing and national assessment. Comparative Education Review, 54(1), 5-25.
Kavale, K. A., & Forness, S. R. (2000). History, rhetoric, and reality: Analysis of the inclusion
debate. Remedial and Special Education, 34(6), 279-296.
King, K. (2007). Multilateral agencies in the construction of the global agenda on education.
Comparative Education, 40(2), 377-391.
Kisanji, J. (1993). Special education in Africa. In P. Mittler, R. Brouillette, & D. Harris (Eds.),
Special Needs Education (pp. 158-172). London: Kogan Page.
Kisanji, J. (1995). Interface between culture and disability in the Tanzanian context: Part II.
International Journal of Disability, Development and Education, 42(2), 109-124.
Kisanji, J. (1998). The march towards inclusive education in non-Western countries: Retracing
the steps. International Journal of Inclusive Education, 2(1), 55-72.
Kiyaga, N. B., & Moores, D. F. (2003). Deafness in sub-Saharan Africa. American Annals of the
Deaf, 148(1), 18-24.
Koyama, J. P. (2009). Localizing No Child Left Behind: Supplemental Education Services (SES)
in New York City. In F. Vavrus & L. Bartlett (Eds.), Critical approaches to comparative
education: Vertical case studies from Africa, Europe, the Middle East, and the Americas
(pp. 21-37). New York: Palgrave Macmillan.
Kozleski, E. B., Artiles, A. J., & Waitoller, F. R. (2011). Introduction: Equity in inclusive
education: Historical trajectories and theoretical commitments. In A. J., Artiles, E. B.
Kozleski, and F. R. Waitoller (Eds.), Inclusive education: Examining equity on five
continents (pp. 1-14). Cambridge, MA: Harvard University Press.
Levinson, B. A. U., & Sutton, M. (2001). Introduction: Policy as/in practice – A sociocultural
approach to the study of educational policy. In B. A. U. Levinson & M. Sutton (Eds.),
Policy as practice: Toward a comparative sociocultural analysis of educational policy
(pp. 1-22). Westport, CT: Ablex.

210

Mamdani, M., Rajani, R., Leach, V., Tumbo-Masabo, Z., & Omondi, F. (2009). Influencing
policy for children in Tanzania: Lessons from education, legislation and social protection
(Special paper 09.30). Dar es Salaam, Tanzania: Research on Poverty Alleviation.
Marshall, J. (1997a). Planning services for Tanzanian children with speech and language
difficulties. International Journal of Inclusive Education, 1(4), 357-372.
Marshall, J. (1997b). Provision for children with speech and language difficulties in Tanzania.
International Journal of Disability Development and Education, 44, 341-365.
Martens, K. (2005). NGOs and the United Nations: Institutionalization, professionalization and
adaptation. New York: Palgrave Macmillan.
Mbwilo, G. S. K., Smide, B. & Aarts, C. (2010). Family perceptions in caring for children and
adolescents with mental disabilities: A qualitative study from Tanzania. Tanzania
Journal of Health Research, 12(2), 1-12.
Miles, M. (1989). The role of special education in information based rehabilitation. International
Journal of Special Education, 4(2), 111-118.
Miles, S. (2011). Exploring understandings of inclusion in schools in Zambia and Tanzania using
reflective writing and photography. International Journal of Inclusive Education, 15(10),
1087-1102.
Ministry of Education and Vocational Training. (2009). National strategy on inclusive education
2009-2017. Dar es Salaam, Tanzania: Author.
Ministry of Education and Vocational Training. (2011). Education statistics data center.
Retrieved from http://educationstatistics.moe.go.tz/moe/#
Ministry of Education and Vocational Training. (2013). Organization structure. Retrieved from
http://www.moe.go.tz/index.php?option=com_content&view=article&id=1578&Itemid=
630
Ministry of Labour, Youth Development and Sports. (2004). National policy on disability. Dar es
Salaam, Tanzania: Author.
Mkumbo, K. (2008). Do children with disabilities have equal access to education? A research
report on accessibility to education for children with disabilities in Tanzanian schools.
Dar es Salaam, Tanzania: HakiElimu.
Mnyani, C. B. F. (2009). Developing teachers’ work for improving teaching and learning of
children with visual impairment accommodated in ordinary primary schools. European
Educational Research Journal, 8(2), 336-351.

211

Nambira, G. (1994). Educating children with hearing impairment in Tanzania: Comparing
regular and special schools. Stockholm: Institute of International Education, Stockholm
University.
National Bureau of Statistics. (2013). 2012 population and housing census: Population
distribution by administrative units: Key findings. Dar es Salaam, Tanzania: National
Bureau of Statistics and Office of Chief Government Statistician.
Njelesani, J., Couto, S., & Cameron, D. (2011). Disability and rehabilitation in Tanzania: A
review of the literature. Disability and Rehabilitation, 33(23-24), 2196-2207.
Nyerere, J. K. (1968). Freedom and socialism: A selection of writings and speeches 1965-1967.
Oxford, UK: Oxford University Press.
Phillips, D., & Ochs, K. (2003). Processes of policy borrowing in education: Some explanatory
and analytical devices. Comparative Education, 39(4), 451–461.
Phillips, K. D. (2013). Dividing the labor of development: Education and participation in rural
Tanzania. Comparative Education Review, 57(4), 637-661.
Ramirez, F. O., & Boli, J. (1987a). Global patterns of educational institutionalization. In G.M.
Thomas (Ed.), Institutional structure: Constituting state, society, and the individual (pp.
150-172). Newbury Park, CA: Sage.
Ramirez, F. O., & Boli, J. (1987b). The political construction of mass schooling: European
origins and worldwide institutionalization. Sociology of Education, 60, 2-17.
Ramirez, F. O., & Ventresca, M. J. (1992). Building the institution of mass schooling:
Isomorphism in the modern world. In B. Fuller & R. Rubinson (Eds.), The political
construction of education: The state, school expansion, and economic change (pp. 4759). New York: Praeger.
Samoff, J. (Ed.). (1994). Coping with crisis: Austerity, adjustment and human resources. Paris,
France: UNESCO.
Schriewer, J. (2000). Comparative education methodology in transition: Towards a science of
complexity? In J. Schriewer (Ed.), Discourse formation in comparative education (pp. 352). Berlin: Peter Lang.
Schriewer, J., & Martinez, C. (2004). Constructions of internationality in education. In G.
Steiner-Khamsi (Ed.), The global political of educational borrowing and lending (pp. 2953). New York, NY: Teachers College Press.

212

Schühle, J. (2013). Medicine murder of people with albinism in Tanzania – How casino
capitalism creates rumorscapes and occult economies (Working Paper No. 2/2013).
Retrieved from the Center for Area Studies, Freie Universität Berlin website:
http://edocs.fuberlin.de/docs/servlets/MCRFileNodeServlet/FUDOCS_derivate_000000002907/xCASWorking-2-13_Schuehle.pdf?hosts=
Steiner-Khamsi, G. (2004). The global politics of educational borrowing and lending. New
York, NY: Teachers College Press.
Steiner-Khamsi, G. (2010). The politics and economics of comparison. Comparative Education
Review, 54(3), 323-342.
Steiner-Khamsi, G., & Stolpe, I. (2006). Educational import: Local encounters with global
forces in Mongolia. New York, NY: Palgrave Macmillan.
Slee, R. (2009). The inclusion paradox: The cultural politics of difference. In M. W. Apple, W.
Au, and L. A. Gandin (Eds.), The Routledge international handbook of critical education
(pp. 177-189). New York, NY: Routledge.
Takayama, K. (2010). Politics of externalization in reflexive times: Reinventing Japanese
education reform discourses through ‘Finnish PISA success.’ Comparative Education
Review, 54, 51-75.
Tanner, R. (2010). Ideology and the killing of albinos in Tanzania: A study in cultural relatives.
Anthropologist, 12(4), 229-236.
Tungaraza, F. D. (1994). The development and history of special education in Tanzania.
International Journal of Disability, Development and Education, 41(3), 213-222.
Tungaraza, F. D. (2012). Who sinned? Parents’ knowledge of the causes of disability in
Tanzania. International Journal of Special Education, 27(2), 216-224.
United Nations. (1948). The universal declaration of human rights. New York, NY: Author.
United Nations. (1989). Convention on the rights of the child. New York, NY: Author.
United Nations. (2007). Convention on the rights of persons with disabilities. New York, NY:
Author.
United Nations. (2013). Convention and optional protocol signatures and ratifications.
Retrieved from UN enable website:
http://www.un.org/disabilities/countries.asp?navid=12&pid=166
United Nations Children’s Fund. (2008). Monitoring child disability in developing countries:
Results from multiple indicator cluster surveys. New York, NY: Author.
213

United Nations Educational, Scientific and Cultural Organization. (1990). The world declaration
of education for All. Paris, France: Author.
United Nations Educational, Scientific and Cultural Organization. (1994). The Salamanca
statement and framework for action on special needs education. Paris, France: Author.
United Nations Education and Science Cultural Organization. (2006). EFA global monitoring
report 2007: Strong foundations: Early childhood care and education. Paris, France:
Author.
United Nations Educational, Scientific and Cultural Organization. (2010). Education for all
global monitoring report 2010: Reaching the marginalized. Paris, France: Author.
United Republic of Tanzania. (2009). 2008 Tanzania Disability Survey. Dar es Salaam,
Tanzania: United Republic of Tanzania National Bureau of Statistics.
United Republic of Tanzania. (2011). The Tanzanian five year development plan 2011/20122015/2016: Unleashing Tanzania’s latent growth potentials. Dar es Salaam, Tanzania:
President’s Office, Planning Commission.
Vavrus, F. (2003). Desire and decline: Schooling amid crisis in Tanzania. New York, NY: Peter
Lang.
Woods, E. (2008). Progress towards EFA in Tanzania. Prospects, 38, 425-430.
World Bank. (2013). Data: Country and lending groups. Retrieved from
http://data.worldbank.org/about/country-classifications/country-and-lendinggroups#Low_income
World Food Programme. (2013). Comprehensive food security and vulnerability analysis:
Tanzania 2012. Rome, Italy: United Nations World Food Programme.
World Health Organization. (2004). Disability and rehabilitation status: Review of disability
issues and rehabilitation services in 29 African countries. Geneva, Switzerland: Author.
World Vision. (2007). Education’s missing millions: Including disabled children in education
through EFA FTI processes and national sector plans. UK: World Vision.

214