Women Leaders in Kenya
Negotiating Leadership Space
Master’s Thesis: International Development
University of Amsterdam
Olwen Enright, FCMA, BA
Student ID 10500928
University of Amsterdam
Department of International Studies
Supervisor: Dr. Jacobijn Olthoff
Second reader: Dr. Hulya Kosar-Altinyelken
Publication: 5 December 2014
Source: SEED magazine 14(11), cover.
Kwa Nafasi Zote Tunataka Usawa na Haki We want Equality and Justice in All Spaces
Acknowledgements
The woman on the front cover of my thesis is wearing a T-shirt with a message in Kiswahili. It says: Kwa Nafasi Zote Tunataka Usawa na Haki. This means “We want Equality and Justice in all Spaces”. I would like to thank the many women leaders that I have encountered in the last two years, both the Women Leaders of Community Based Organisations in Kenya, and the Women Leaders at the
University of Amsterdam who are teaching and researching in the field of International
Development. Just like this woman, all of you want equality and justice in all spaces, and all of you are doing something about it.
I would especially like to thank my supervisor, Dr. Jacobijn Olthoff, who has encouraged and challenged me throughout this process. Her ability to look through and beyond to the positive qualities of each of her students is a remarkable gift.
I also wish to thank my second reader, Dr. Hulya Kosar-Altinyelken, for her encouragement when I shared my apprehensions with her about embarking on this Master’s degree as a mature student. I find it very fitting that she should be involved now that my degree is drawing to its close.
And finally, I wish to thank my husband Floris for the loving and practical support he has given me while I have been pursuing my Master’s degree. And I thank my daughters Isabel and Marguerite, and my sons Ethan and Duncan, for their understanding. This has been an important transitional phase in my life, and it has not been easy for all of you to have me locked away writing assignments and being away from home during two field trips to Kenya. And to my children: I hope that in years to come you will look back as I have done with my own mother, and say, “Yes, my mother did that. I can do it too.” However, I hope even more, that just like this Kenyan woman, you will teach your children to say: We want Equality and Justice in all Spaces, and that you will show them how to do something about it!
Table of Contents
Acknowledgements ... 2
1.0
Introduction ... 5
1.1 Origins and Scope of this Research ... 5
1.2 Bungoma and Kenyan Context ... 7
1.3. Women’s Leadership relevance to Social Justice and Development ... 8
1.4. The problem and the research questions ... 10
1.5 Thesis Structure ... 11
2.0
Problematizing and Engendering Leadership Theory ... 13
2.1 Critical Biases of Leadership Theories ... 13
2.2 Summary of Critical Biases ... 16
2.3 Consequences of analysis: central tenets of my research ... 17
2.4 Conclusion ... 18
3.0
Theoretical Framework: Tenets, Paradigms and Methodologies ... 19
3.1 Axiology ... 19
3.2 Ontology & Epistemology: Critical Realism ... 20
3.3 Grounded Theory methodology informed by Critical Realism ontology ... 22
3.4 Consequence of Choosing Critical Realism-Grounded Theory ... 23
3.5 Conclusion ... 25
4.0
Methodology ... 26
4.1 Respondents and Sampling... 26
4.2 Mixed Methods – ‘All is Data’ ... 28
4.3 Integrated Data Analysis ... 32
4.4 Quality and Validity: Resonance with Local Reality ... 33
4.5 Conclusion ... 34
5.0
Women Leaders: Conceptualisation as Source and Constraint ... 35
5.1 Unpacking the Concept of Woman Leader ... 35
5.2 Conceptualisation of Women: Introducing the Good Woman ... 36
5.3 The Dominant Concept of a Leader: Ideal versus Actual ... 38
5.4 Interrelations Between Concepts: The Triple Tensions of Women Leaders (arising from the Good Woman frame)... 41
5.5 Re-conceptualizing Women’s Leadership: Six Conceptualisations ... 45
5.6 Conclusion ... 50
6.0
Contexts: Influence of Political and Church Contexts ... 52
6.1 The Political Context ... 52
6.2 The Church Context ... 54
6.3 Conclusion ... 57
7.0
Enactment: Negotiating (Re-)Conceptualisations and Spaces ... 60
7.1 Negotiating Maria’s right to travel to the bank: Francis visits her Financial Officer's husband in his home context ... 60 7.2 Negotiating the freedom to attend meetings: Lourine took her meetings to the men’s space . 62
7.3 Reconceptualisation through new skills: Angela the Public Speaker ... 63
7.4 Reconceptualisation through co-analysis and critical thinking: Rachel got married ... 65
7.5 Conclusion ... 67
8.0
Conclusion and Discussion: Women’s Leadership – from partial resolution of
tensions to transformative potential... 70
8.1 Contextual Interpellation: Conceptualisations of leadership and womanhood and the Triple Tensions facing women leaders ... 70
8.2 Partial reconciliation or exaggeration of tensions ... 72
8.3 Looking through and beyond - critical thinking and strategic action ... 74
8.4 Theoretical Evolution and Methodological Reflections ... 77
8.5 Policy Recommendations... 83
8.6 Recommendations for further research ... 84
REFERENCES ... 86
APPENDICES ... 89
Appendix 1: Example of working with Critical Realism-Grounded Theory ... 89
WOMEN LEADERS IN KENYA 5
1.0 Introduction
This research explores the lived reality of women leaders of Community Based Organisations (CBOs) in Bungoma in western Kenya. Their CBOs focus on community improvement, and include rural action, self-help groups, savings groups, entrepreneurial skills development and providing assistance to marginalised groups in the community. This research uncovers how the women leaders
conceptualise, enact and experience their leadership in connection to their womanhood, in the contexts of their local society and prevalent conceptualisations of womanhood and leadership. This research used a generative methodology to focus on leadership. My approach is founded on Critical Realist ontology and used Grounded Theory methodology to uncover the principal leadership conceptualisations prevailing in local society, and how they do or do not interpellate women as leaders. Influenced by the Strategic Relational Approach, I will argue that, although embedded in limiting social structures, these women leaders can be strategic agents, able to analyse and act to negotiate more space for their leadership. In so doing, I believe women at this community level are contributing to a process of transformation in relation to gender equality, social justice and
development objectives.
In this chapter I explain the origins of the project and the focus on women leaders, provide some relevant background on Bungoma and the Kenyan context and present my arguments for why women’s leadership at the community level in Kenya matters in the international context of social justice and development. Finally, I define the problem my research addresses and my research questions.
1.1 Origins and Scope of this Research
My first fieldwork phase began with a research project for the Libre Foundation on the local relevance and effectiveness of its Leadership Development Programmes (LDPs) in Bungoma, in western Kenya. My theoretical review done at that time confirmed that leadership theory tends to be biased towards northern hemisphere, business and military models, with relatively little focus or value placed on leadership models in civil society or developing countries, especially at the small community scale.
My research group in that first phase was composed of Bungoma CBO leaders who were either in, going to or had already completed an LDP. I wanted to contextualise the lives of local community leaders and I developed a model to become aware of the domains of action and influence leadership is situated and understand where tensions and synergies existed.
Whilst the first phase of this research focused on both men and women community leaders, I
became increasingly aware that the women faced specific and more complex challenges compared to the men. And, I also observed that the LDPs were not consciously addressing their issues. For
example, all leaders were limited in their enactment by the congestion they faced from multiple demands (donors, relatives, supporting their family, etc), but the women were also expected to be based at home, carrying out the vast majority of domestic tasks. They therefore experienced
6 WOMEN LEADERS IN KENYA
difficulties in leaving the home in pursuit of their leadership activities. I also realised that the curriculum focused more on problem solving training than on critical education.
Phase 2 therefore moved on from the more positivist evaluation of the leadership training delivered by Libre in favour of seeking to understand the experience of women’s leadership through an increasingly co-constructive and qualitative research process. As a result, the research was able to perceive different conceptualisations and strategies operating at multiple levels, which limited or enlarged leadership space for women depending on the how they played out in different contexts. I preferred this co-constructive methodology because I believe it creates a more equal and respectful relationship and recognises that the women leaders hold much of the knowledge themselves, consciously or unconsciously, and can, with some facilitation, analyse or critique the data
themselves, adding depth and value to my own analysis. Such an approach shares the learning and ownership and is thus essentially transformative.
This thesis draws together Phase 1 and 2 and seeks to show how the conceptual models evolved. Concepts have not been predefined or operationalised in the traditional manner. Furthermore, I decided not to summarise the extensive work published on leadership theory, but to focus on critiquing the dominance of male, business, and Global North leadership models, both to avoid reinforcing this hegemony and bias, and to create space for alternative models which may be of more relevance in the context of women’s leadership of CBOs in western Kenya, and similar settings. My approach is in line with Jackie Ford, Professor of Leadership and Organisation Studies, who maintains that the notion of leaders and theories of leadership need to be deconstructed in order to understand their exclusionary effects on women (Ford 2005: 243).
I am aware that multi-scalar influences from local, national and international actors and structures are at work in the conceptualisation and interpellation of women’s leadership in Kenya. This research focuses on local actors and structures. It is therefore useful to have a general understanding of the local context in Bungoma as well as the wider context of women’s leadership in Kenya, in terms of social justice and development. I will therefore provide a brief overview of both in the next two sections.
WOMEN LEADERS IN KENYA 7
1.2 Bungoma and Kenyan Context
Source: Bungomaproject.com.au
Bungoma is a largely rural area, with a population of nearly 1.4 million, in the Western Province of Kenya, near the border with Uganda. Roughly two thirds of people are educated to primary school level, while 11% attend secondary education. Around 300,000 people live in towns, where 14% are connected to electricity and 21% to piped water. In rural areas these figures are 1.5% and 7.2% respectively (Kenya Open Data 2009). On the Corruption Perception Index, published by Transparency International (Transparency International 2014), Kenya is ranked 145th out of 175 countries. On the Ease of Doing Business Index (World Bank Group 2014) Kenya is ranked 137 th. Women’s life expectancy is 62 years and men’s 59. And 49% of the population is under 15 years old (Kenya National Bureau of Statistics 2013). There is periodic violence for political and other reasons (Standard Media 2013).
The table below compares Kenya with other countries in the region. Given Kenya’s relatively strong economic position, it might be expected that a social indicator such as number of national
parliamentary seats for women would be higher. However, social development does not necessarily follow from economic development, as the table below shows. Kenya may be able to claim it is just 2 per cent under the worldwide average of 21% of women to men parliamentarians, but it ranks 80th worldwide. Neighbouring Rwanda ranks 1st worldwide with 63.8%, and Uganda ranks 24th with 35%. ‘In many respects, Kenya likes to see itself as the most advanced country in East Africa. With regard to gender equality in politics it surely is not; instead it trails behind all others.’ (Kamau 2008: vi)
8 WOMEN LEADERS IN KENYA
Table 1.2: Economic position of Kenya in the region.
GDP per capita (2012)1 HDI (2012)2 Women Parliamentarians Worldwide Ranking
Kenya $943 0.531 80th
Rwanda $620 0.502 1st
Uganda $598 0.480 24th
International conferences have helped to create pressure on governments to create greater equality and participation of women since the 1970s (Odhiambo & Mitullah 2010: 14). Kabira and Kimani’s (2012) report details the progress that has been made, such as: “the establishment of national and institutional gender policies, the National Gender Commission, a ministry in charge of Women’s Affairs, Children and Social Services….”
However, strong societal constraints make progress difficult. My observations in Bungoma support the findings of Nkomo & Ngambi’s 2009 review of academic research and literature concerning African women leaders, which showed consensus among scholars regarding the factors that impede women from accessing leadership positions. These factors are: ‘early socialisation, limited
educational achievements, multiple roles, gender stereotyping, subtle discrimination, and organisational policies and procedures’ (Nkomo & Ngambi 2009: 55).
1.3. Women’s Leadership relevance to Social Justice and Development
I believe it is useful to have a general understanding of women’s leadership in terms of its relevance to social justice and development agendas. I will therefore give a brief overview, to help the reader see how identifying ways to enhance women’s leadership at the community level in western Kenya can contribute to progress not only in Kenya but in similar contexts where the approach I
recommend in the conclusions could be replicated.
The role of women leaders in community development and in peace building is attracting increased attention at the international level, yet there is still limited recognition of their right to
representation and of the contribution that women make to leadership.
1
Gross Domestic Product in current dollars, 2012
(http://data.un.org/Data.aspx?q=GDP+per+capita&d=SNAAMA&f=grID%3a101%3bcurrID%3aUSD%3bpcFlag%3a1)
2
WOMEN LEADERS IN KENYA 9
Source: Accessed 2 December 2014, https://www.cordaid.org/en/topics/womens-leadership/
Source: Accessed 2 December 2014, http://www.oxfamnovib.nl/Our-work.html
According to Nancy Fraser, the principles of social justice are redistribution, recognition and representation (Fraser 1997 cited in Blackmore 2013; cited in Sweetman 2008). This thesis argues that the underrepresentation of women leaders is indicative of wider gender inequality (Blackmore 2013; Kabeer 2005) and is thus a matter of social justice. Furthermore, this thesis argues that this injustice is reflected in the paucity of attention to women leaders in academic research and publications. According to Professor Nyokabi Kamau, Director of the Centre of the Parliamentary Studies and Training in Kenya, more academic research is needed to counter dominant stereotypes which sustain beliefs that women leaders are not as effective as men. As she writes ‘very little if any of what women leaders do get to the public domain’, even though her research shows that ‘when they get into politics, women’s priority was guided by their vision to see a world where men and women and all other people who may be vulnerable get equal opportunities’ (Kamau 2008: 32). Millennium Development Goal (MDG) 3 addresses gender equality and women’s empowerment as an intrinsic rather than instrumental goal. Representation of women in leadership is an end of itself, and not simply because women are a means to carry development improvement to others (Kabeer
1
0
WOMEN LEADERS IN KENYA
2005: 13). Sweetman argues that women need ‘to be visible politically as women and be empowered to act in that capacity, because … they … have needs and attitudes on vital issues which differ from those of men’ (Sweetman 2008: 437). Professor Kamau’s research reinforces this view, showing that women politicians are largely preoccupied with social justice projects: ‘Without doubt, (the) majority of the women focus on human development issues e.g. education, health, children, nutrition, women’s income, water and other injustices that affect women and girls differently’ (Kamau 2008: 33). The inequality of women in leadership therefore adversely affects the development process (Odhiambo & Mitullah 2010). Skill shortages are widespread in many African countries, and so the ‘challenges of sustainable socioeconomic development mandate that all leadership talent be developed and employed’ (Nkomo & Ngambi 2009).
Kabeer argues for ‘transformative forms of agency that do not simply address immediate inequalities but are used to initiate longer-term processes of change in the structures of patriarchy’ (Kabeer 2005: 23). I argue that women’s leadership at the community level is (potentially) such a
transformative agency for longer-term societal level. In my view, capacities built at the local level can have a cumulative effect on women’s confidence as well as on their skills. As Kabeer notes, such capacities can spill over into the political sphere, and in the abilities of women to interact with officials, and challenge employers, and to participate in protests (Kabeer 2005: 23). Therefore, local level development of women leaders has the potential to effect LR transformative changes in society – in redistribution, recognition and representation. ‘Building this collective capacity of women in all spheres of life to participate and to hold authorities accountable is thus the only basis on which the world’s policymakers can keep the promises that they have made on the issue of gender equality.’ (Kabeer 2005: 23)
In line with Kabeer’s work, Kabir and Kimani have charted the historical journey of women’s leadership in Kenya and point to the pivotal influence of Maendelo ya Wanawake (MYWO) a
grassroots women’s self-help groups which expanded to focus on “enhancing women’s participation in leadership in appointive an elective bodies”. Although some progress has been made, in general the social context remains highly challenging.
1.4. The problem and the research questions
As argued in the previous section I believe my research can contribute to addressing gender
inequality and development challenges in Kenya. I see enlarging the space for women’s leadership at the community level as having value as an intrinsic goal, but also in terms of the contribution it can make to wider social change by helping to advancing gender equality and the development agenda. I seek to address a two-fold problem: women’s underrepresentation in leadership theory and research (especially women from the Global South and at the community level), and the difficulties women face in becoming and staying community leaders. I will show that women community leaders face tensions between the local notions of womanhood and those of male centred leadership that restrict their space for enactment. However, there are tactical strategies and actions that some women leaders are using to expand leadership space and change conceptualisations, and I have formulated these into a series of conceptual frameworks in my conclusion.
WOMEN LEADERS IN KENYA 1
1
I will also show how critical analysis of mainstream leadership theory biases combined with emerging data led me to place women at the centre of my second fieldwork phase, and to reflect and read more about the gendered nature of leadership. My axiological position was reinforced by this critique, and as I result I developed five tenets around which I would formulate my research design and methodologies. These were: Be situated in local social and cultural context: Anchor the research in a local understanding of leadership
Place women leaders at the centre
Move away from positivist methodologies towards more qualitative methods
Make space for a range of leadership models – be generative
Look through and beyond the data to deeper realities: Be rigorous
Ideally, it would be situated in the historical, political and economic contexts also. However the limits of this research required me to focus more narrowly. Ideally, it would also be anchored in a local understanding of followership. Again, the limits of this research meant I could not pursue this aspect. Although I was not fully aware of it in my first fieldwork phase, these tenets reveal that my research and study increasingly have reflected the Transformative Paradigm.
The main research question is:
How is leadership, in connection to womanhood, being conceptualised and enacted by women leaders of community-based organisations in Bungoma in western Kenya?
This question was broken down into three sub-questions:
How are leadership and womanhood being conceptualised and what interrelations exist between these concepts? (Chapter 5)
Which contexts do (not) privilege women leaders, and which factors contribute to this? (Chapter 6)
How are women leaders enacting their leadership? Is their enactment influenced by their own and/or by societal conceptualisations of leadership and womanhood? (Chapter 7)
1.5 Thesis Structure
This first chapter has introduced my research topic and motive, described its evolution, and provided a description of the research context. The main research question and three sub-questions were presented along with the limitations of scope. Chapter 2 problematizes and addresses the
engendering of leadership theory by analysing the critical biases of existing leadership theories and identifying their inadequacies for the study of women leaders in Kenya. Based on this assessment, Chapter 2 concludes by summarising the tenets of this research programme. Chapter 3 outlines the paradigmatic framework for this research, by relating these tenets to my axiology, ontology and
1
2
WOMEN LEADERS IN KENYA
epistemology. This provides the rationale for my methodology choice: combining a Critical Realism ontology with a Grounded Theory methodology. The influence of the Strategic Relational Approach (SRA) and the concept of Interpellation are also acknowledged. Chapter 4 presents the research design in detail, including respondents, mix of methods and modes of analysis. Each design element is related to my tenets. Chapter 5 presents the findings on conceptualisations of leadership and womanhood which Chapter 6 shows to be more favourable for women’s leadership in church contexts than in political contexts. Chapter 7 outlines how women are enacting their leadership, which barriers they experience and how women leaders respond both strategically and tactically. These findings are drawn together in Chapter 8, with an overall discussion which answers the research questions and presents my final results in two conceptual models. As the combination of Critical Realism and Grounded Theory methodology is relatively new in leadership research (Kempster & Parry 2011), it is valuable to reflect on this combination. Therefore, Chapter 8 also provides a reflection on the research process and the evolution of my conceptual framework. Finally, recommendations are made for policy and future research.
WOMEN LEADERS IN KENYA 1
3
2.0 Problematizing and Engendering Leadership Theory
This thesis argues that the social justice issue of underrepresentation of women leaders in public life is reflected in their underrepresentation in academic research and literature. This thesis further maintains that there is an urgent academic and social need for more research concerning African women leaders, and that Western leadership theory is an inadequate lens for viewing the status and experiences of Kenyan women leaders, especially at the community level. I am supported in this view by Nkomo & Ngambi’s research into academic publications on African women leaders which found ‘a paucity of systematic research focusing directly on African women leaders’ (Nkomo & Ngambi 2009: 53).To avoid reinforcing Western hegemony and bias, this chapter does not present leadership theories themselves but rather a critical analysis of their biases. This critique is the rationale for the five research tenets I developed to underpin my research into Kenyan women’s leadership at the community level. Chapters 3 and 4 will show how these tenets are central to my conceptual framework and research design.
2.1 Critical Biases of Leadership Theories
My review of the literature on mainstream leadership theories makes me conclude that the core assumptions of hegemonic leadership concepts and theories suffer from multiple biases, which make them unsuitable for application to women’s leadership in civil society in the Global South.
These biases include:
a) Geographical Bias towards Global North
Both academic scholars and practitioner authors argue that leadership theories suffer from geographical bias because they originate in the Global North and are dominated by our worldviews and leadership practices (Blackmore 2013; Nkomo & Ngambi 2009; Ford 2005; Fowler et al 2002; Hailey 2006). The majority of research also relies on Western theoretical frameworks and discourses, and there is scarce reference to African contexts (Nkomo and Ngambi 2009: 60).
b) Sectoral Bias towards Business and Military
James et al (2005) agree with the geographical bias critique and add that leadership theory was derived originally from the business and military sectors. This analysis is supported by Hailey (2006). Leadership theory and development programs reflect the leadership styles, priorities and biases of their military and business origins. There is scant evidence of research into civil society leadership. Sectoral biases must be taken into account when considering the extent to which leadership theories are applicable at the community level, especially in non-Western cultures.
c) Gender Bias towards masculine models – compounded by Western gender concepts
Critical scholars (such as Blackmore 2013; Ford 2005; Nkomo & Ngambi 2009; and Odhiambo & Mitullah 2010) in their critiques all highlight the underlying assumption that
1
4
WOMEN LEADERS IN KENYA
leaders are male. Leadership literature and practice tends to fail to question or reflect upon its gendered nature and the masculine models which dominate Leadership Theory tend to privilege men. As Ford writes, it is ‘the masculine voice that governs discourse’ and she argues that it is this discourse which sustains the dominance of masculine leadership models, to the exclusion of other models: ‘macho, individualistic, assertive and dominant behaviours continue to take precedence over the more feminine qualities such as empathy, the capacity for listening, relational skills etc.’ (Ford 2005: 245). The lack of relevance to women’s leadership in Africa is compounded by relying on Global North concepts of gender and gender relations, which exhibit minimal
understanding of gender rooted in African perspectives (Nkomo and Ngambi 2009). In their report on gender and leadership specifically in Kenya, Odhirambho and Mitullah argue that despite scientific claims, Western academia has exported Western gender models to non-Western societies. These models, ranging from extreme patriarchalism to egalitarianism, impose: ‘mythical assumptions about women’s and men’s capabilities, the sexual division of labour in early human history, and, as a result, women’s place in today’s society’ (Odhiambho and Mitullah 2010: 12-13). In other words, what the academic community had assumed to be universal experiences, suitable for universal generalisation of concepts, theories and practise, was actually an account of male experiences and ideas, predominantly by and for male actors situated in a Global North context. As argued in f) below, realist ontology and positivist epistemologies have served to institutionalize these models, to the extent that their masculine foundations become invisible.
Adopting a critical approach to hegemonic leadership theory, and specifically including feminist critique, allows alternative perspectives and more diverse leadership models to emerge, especially for women and men who do not practise the leadership traits popularly decreed as masculine. Rationality and assertivity have come to be seen as masculine prerogatives for example. Yet most feminist perspectives have moved away from reductionist perspectives. As Blackmore observed, reductionist binaries ‘… between mind/body, rationality/emotionality, public/private, objective/subjective (…) reduce to essentialist understandings of man/woman’ (Blackmore 2013: 146). In Chapter 5, this binary reductionism will be revealed as a dominant influence in conceptualisations of women.
d) Univocal bias over multi-vocal
Leadership theory and development programmes often focus on leadership being invested in one individual rather than in a collective, multi-vocal entity (Malunga 2006). I argue that the tendency to focus on the results achieved by the leader supports univocal leadership over multifocal forms of cooperative leadership/followership, to the cost of shared relational and experiential aspects. I agree with Blackmore that when leadership is limited to measurable output of an organisation, it can get isolated from wider social environments (Blackmore 2013). To my mind, this can culminate in societal disjuncts. Leadership theory also exhibits tendencies either to high-level universal generalities or to individualist biographies, with less room for other models in between. Or as Fowler et al
WOMEN LEADERS IN KENYA 1
5
(2000: 2) describe it: ‘prevalent leadership theory (…) either overly relies on reductionism or is excessively personalized and particularistic.’ Therefore there is little appreciation of the realities of leadership at smaller scales, such as at the scale of the community. The lack of room both for multiple leadership models and evaluation criteria devalues or renders invisible alternatives to the dominant paradigm. Ford describes this self-reinforcing aspect: ‘Perpetuating single models of univocal and patriarchal leadership behaviours... perpetuates a model that is exclusionary’ (Ford 2005). She argues that, once engrained, such a concept is hard to change: ‘Through a “process of reification”, the concept of leadership takes on an objective existence, which seems to make it beyond challenge.’ (Ford 2005: 243).e) Scalar Bias
There is little appreciation of the realities of leadership at smaller scales such as local community level. As we will later see in Chapter 5, there are multiple models already at work in civil society in Kenya.
f) Bias in leadership research due to the dominance of positivism over qualitative approaches
According to several academic authors (Kempster & Parry 2011; Bryman 2004;
Blackmore 2013; and Ford 2005),mainstream leadership research tends to be based on a realist ontology and a positivist epistemology. Positivist research tends towards pre-determined hypotheses, and to focus on visible data. This contributes to the view that there is a single definable concept of leadership. Therefore, positivist research is inadequate for capturing nuanced, multi-level interactions. ‘Many studies of leadership have focused on quantitative empirical research and whilst there is some evidence of a move towards methodological diversity (Bryman 2004), there is a dearth of qualitative studies, whether hypothesis driven or of an interpretive nature. The favoured
approaches adopted within the literature are laboratory-based experiments or
questionnaires, to the exclusion of contextually specific, qualitative studies’(Ford 2005: 241).
Blackmore (2013: 139) argues that feminist research should move away from numerical representation, to the social relations of gender and power. She contends this generates a wider field of possibilities. ’Feminist understanding provides substantive and normative alternatives to how we theorise and practise leadership’ (Blackmore 2013: 139). Finally, of the small number of academic publication on African women leaders that were found by Nkomo and Ngambi (2009), the majority were based on quantitative methods. Nkomo and Ngambi expressed surprise at the ‘predilection towards quantitative methods, given the arguments feminist scholars have made for the potency of qualitative methods for understanding the experiences of women’ (Phendla 2004 cited in Nkomo and Ngambi 2009: 54). ‘The domination of positivist approaches to knowledge production … tends to devalue and misrepresent African knowledge systems (e.g. Odora-Hoppers 2002)’ (Nkomo and Ngambi 2009: 54).
1
6
WOMEN LEADERS IN KENYA
2.2 Summary of Critical Biases
I have created the diagram below to provide the reader with a summary of these multiple biases and also provide this visual as a tool to increase our awareness that, due to our world views, we tend to carry theses biases with us when we engage with LDPs in Global South countries and that frame our interaction with local leaders (James 2005, Hailey 2006). We do not usually pay attention to the ontological and epistemological foundations of these theories, but they do affect the manner in which we define and research leadership (Kempster & Parry 2010; Ford 2005).
Diagram 2.2: Analysis of Hegemonic Discourses of Leadership.
Key message: The goals, experiences and challenges of a white male leader in a multinational Global
North business, are far distant from those of a woman leader in a community-based organisation orating within civil society in the west of Kenya.
WOMEN LEADERS IN KENYA 1
7
2.3 Consequences of analysis: central tenets of my research
As a result of this analysis and informed by early field experiences, I developed the following five tenets to help me avoid reproducing similar biases. These have been central to my research design:
Tenet 1: Be situated in local social and cultural context: Anchor the research in a local understanding of leadership
Given the geographic bias identified, this research needed to be based in local context. I would design my research in a way that would allow the local meaning of leader to emerge from the respondents themselves, rather than begin by defining leadership in advance, and testing local meanings against a prior definition.
Tenet 2: Place women leaders at the centre
As gender emerged as a key issue, it was important to place women at the centre of my research, to focus on their voices and experiences, but without excluding men. My rationale was that women leaders live and work with men and women, and so these interactions are an important part of their experiences, both positive and negative.
Tenet 3: Move away from positivist methodologies towards more qualitative methods
As the use of positivist methodologies in mainstream leadership research has been critiqued for contributing to the biases towards Global North leadership models, it is desirable to focus research on qualitative methodologies.
Tenet 4: Make space for a range of leadership models – be generative
Defining leadership is itself an exclusionary practice, which I attribute to the tendency towards reductionist binaries. I wanted to avoid a research product that was unwittingly reductionist, contributing to another exclusionary leadership product. I therefore concluded that my research design needed to be generative if it was to make space for a range of leadership models.
This meant that my research design should utilise a generative approach, to assist me to look through and beyond the data. I realised that I needed to be adventurous and take the risks involved in making theoretical abstractions from my data analysis and interpretations.
Tenet 5: Look through and beyond the data to deeper realities: Be rigorous
Due to the research question addressing agency and structure (via conceptualisation, contexts and enactment), my research design needed to be especially rigorous. As Sengue warns, we tend to see data that confirms our preconceptions. Therefore I needed to discipline myself to rigorously collect and analyse data in different ways and at different levels, paying attention to outliers if they arose, rather than smooth them out as positivist research tends to do. Their signals could help make new perspectives visible.
1
8
WOMEN LEADERS IN KENYA
2.4 Conclusion
As I have shown, the dominant paradigm is not suited to the purpose of this research. The biases are geographical, sectoral, gendered, scalar, univocal, and positivist. As Nkomo and Ngambi say, it fails to ‘interrogate the cultural, historical, political, and economical context, influencing the study of African women’ in leadership (Nkomo & Ngambi 2009: 64).
This analysis reinforced my belief that there is a need for research and theory development centred on women leaders in Global South civil society and that my research needed first to deconstruct the meanings and practices attributed to leadership and then to construct a contextual theory of women leaders. The tenets were designed to help me implement and maintain this focus. In the next chapter I explain how my ontological and epistemological choice made it likely that I would develop such tenets and Chapter 4 will show how they fundamentally shaped my research design.
WOMEN LEADERS IN KENYA 1
9
3.0 Theoretical Framework: Tenets, Paradigms and
Methodologies
This chapter compares the research tenets derived from my critique of leadership theory with my own axiological, ontological and epistemological positions. Finding them to be consistent, this chapter then explains the choice of methodology and its consequences.
The aim for my research design was to arrive at an ontological knowingness of a layered reality, and I needed to use epistemologically rigorous and hopefully creative methodology which would enable me to look beyond the empirical, to real and actual realities and causal factors. I wanted this ontological knowingness to be recognisable and debatable between respondents and myself, communicable by some of my respondents to each other, academically communicable to an academic community back at UvA, and communicable to organisations involved in Leadership Development Programmes in developing countries.
I began my studies with an ethical standpoint consistent with my personal history and values. During my studies I became exposed to influences and experiences which enriched my starting point. This led me to critique mainstream leadership theory as well as to select Critical Realism and Grounded Theory as my theoretical and methodological framework. The tenets crystallised out of this
evolution. For example the need to focus specifically on women only became ringingly clear to me at the end of Phase one, leading me to place women at the centre of fieldwork phase two, and to reflect and read more about the gendered nature of leadership.
As the academic reader will no doubt have recognised, these tenets reflect a transformative paradigm. While I certainly would not claim that my research fulfilled the ideals of transformative research, it has rigorously pursued a number of the criteria which are central to transformative paradigm.
Research paradigms are a set of beliefs that reflects our worldview, and hence guide our actions as researchers. A research paradigm is comprised of four philosophical components: Axiology,
Ontology, Epistemology and Methodology (Denzin and Lincoln 2005 cited in Mertens 2010). An integrated research design requires us to examine all four. The following section outlines my research paradigm using these components.
3.1 Axiology
There are many levels at which axiological considerations influenced the formulation of my research questions and design. This section describes the key points of intersection, which noticeably reflect key points of transformative axiology, from the macro to the micro level.
Social justice focus on the marginalised
At the macro level my ethical position went further than simply cultural and social respect, because I intentionally focused on people who tend to be marginalised. I focused on women leaders because they are underrepresented in leadership and tend to encounter more obstacles than men leaders.
2
0
WOMEN LEADERS IN KENYA
This intentional focus on marginalised groups is a hallmark of transformative paradigm, because it is concerned with social justice. I plan to publish the results of my research, and to present it to academics and practitioners, with the aim (and hope) that this will contribute to improvements in leadership development training, e.g. being more targeted to local needs, e.g. emphasising the benefits of critical education over purely instrumental leadership training.
Concern for respondents
The transformative paradigm is also concerned with respondents learning during research process. Their well-being was important to me, so I needed to be self-critical about how my actions could affect them in order to mitigate any complications that might arise, either physical inconveniences or from my being personally intrusive in my questioning. I did engage in several co-constructive
research encounters, however I was careful that I did not provoke the respondent to engage in actions after our meetings, which they might subsequently regret. The respondents’ actions are their own, and while I would be travelling away again, the respondents continue to live in their own community.
Positionality
This meant that I needed to be aware of potential biases in my background and my values, in the methods I used, indeed biased created simply by my presence. I understood that I was not a neutral agent in this research, because of my upbringing in the global North, because of my socio-economic and national affiliation, and because of my gender. I recognized that my previous working
experiences in Africa were potential influences on my research, and could impact my data both positively and negatively. I prepared and reflected carefully before and after interactions. I needed to listen value free to the narratives of the respondents and not to put my own concept of leadership or womanhood between my understanding of what their narratives are.
Quality and Respondent Validation
Adopting a transformative approach means respecting its role in advocacy for social justice. This meant that my research had to be based on a very rigorous research design. Transformative research is not passively descriptive; its results are intended to facilitate change, through interventions and policies that will increase social justice (Sweetman et al. 2010). Thus one must be very sure of the links between findings and recommendations, and that it is participants who are central in validating research results.
3.2 Ontology & Epistemology: Critical Realism
As demonstrated in Chapter 2, mainstream leadership research has tended to originate from positivist ontologies and epistemologies, which has contributed to the hegemonic leadership theories and practises critiqued in Chapter 2.
Ontology concerns how it is that we know reality. Paraphrasing Mertens (2007: 215): How can I decide what kind of evidence I can accept in order to know whether a woman is a leader, or is she is enacting her leadership? A realist might accept a leadership title as evidence, while a Critical Realist would not. I myself recognise that there are multiple realities in any endeavour, especially in one as
WOMEN LEADERS IN KENYA 2
1
multi-layered as leadership, and I position myself with critical realists who propose a ‘stratified’ rather than a ‘flat’ ontology (Sayer 2000: 12).Critical Realism is a philosophy of science which distinguishes between three layers of reality commonly referred to as Empirical, Actual and Real (Sayer 2000: 10; Harrit 2011; Fairclough 2005). My preference is to use the terminology of Kempster and Parry (2011: 110) who categorise the three layers of reality as Empirical, Events, and what they term Deep. Empirical data is that which is
observable and are the outcomes of Events that exist in time and place, caused by Deep level of reality. Deeper powers are often unobservable but causally efficacious. These ‘may not be capable of being observed through events, but rather are interpreted and explored through an understanding of the interplay between agency and structure’ (Archer, M., cited in Kempster and Parry 2011: 110).
Table 3.2: Layers of reality in Critical Realism
Kempster & Parry Sayer
Empirical Empirical
Events Actual
Deep Real
Critical Realism recognises the relational nature of the real, actual and empirical. In other words, depending upon the conditions, properties of the Real may not be activated (Jessop 2005). In this research, we can understand that Real capabilities of women leaders may (not) become activated depending upon the norms of their society.
Critical Realism has a ‘distinctive view of causation’ because it recognises that the capacities and susceptibilities of actors or structures to behave in certain ways depends on many conditions, and that structures and contexts have ‘their own causal powers and liabilities, which may trigger, block or modify its action’ (Sayer 2000: 15). Hence, empirical factors such as the congestion of women leaders’ day-to-day lives are visible, but are caused by social structures based on Deep beliefs. The division of labour according to gender roles is a deeper structure that results in the highly congested nature of the women leaders’ lives, as their domestic duties constrain their ability to be involved in their leadership goals. Thus, we cannot understand leadership based on conventional research that aims to prove causation just by seeking repetitions and regularities. Only part of this reality is visible to us, and we are challenged to conceptualise and interpret the other levels, keeping in mind that, depending on conditions, the same causal powers may result in different outcomes.
Taking Critical Realism further, Kempster and Parry (2011) again cite Margaret Archer’s work, who describes the difference between corporate agents and primary agents. Corporate agents are aware of and sustain societal structures, while primary agents are unaware of being influenced by these structures. In this research, women were observed as both active and passive actors. In Chapter 7, women leaders will be seen as active actors, employing strategies that influenced the Events taking place, and changed the visible outcomes.
2
2
WOMEN LEADERS IN KENYA
However, Critical Realism also recognises that the world has ‘emergent properties’ (Sayer 2000: 10), meaning that although these causal powers exist, their influence is not predictable. This makes it possible to understand how we could be or become many things which currently we are not’ (Sayer 2000: 12). This ‘becoming’ relates to women’s leadership. For example, in Chapter 7, we will observe that despite the constraints on women accessing and enacting their leadership, women are actively enacting their leadership at multiple levels.
Acknowledging invisible realities presents epistemological challenges: How can we know about these realities? How can we access knowledge about these realities? In Critical Realist epistemology, knowledge can be produced by explanation instead of prediction, and by retroduction instead of induction and deduction (Jessop 2005). In Critical Realism we observe events occurring, and are challenged to uncover explanations of the causal mechanism behind these events, perceiving and interpreting the power relations behind what we observe. Retroduction means that we ask ourselves, “What must reality be like for this to be happening?”
In summary, my Critical Realism position required me to try to distinguish the real, actual and empirical levels of reality pertaining to the complex reality of the daily lives of women leaders, to explore the influence of leadership conceptualisations upon the women; and to understand how women go about enacting their leadership.
3.3 Grounded Theory methodology informed by Critical Realism
ontology
In view of my axiology, and with critical realism as my ontological and epistemological base, I needed to make methodological choices. Tenets 1-5 guided my decision to combine my critical realist
ontology with a Grounded Theory methodology, because this can enhance our understanding of leadership, if we look ‘through and beyond’ the visible data to underlying levels of reality (Kempster and Parry 2011). This choice was inspired by Kempster and Parry’s challenge to combine Critical Realism and Grounded Theory, and by Barry Glaser’s determined advocacy that rigorously pursuing Grounded Theory methodological process in its entirety will provide us with fundamental insights, and thereby unleash its generative potential.
In this framework, qualitative research suits the nature of leadership, as it is a rich and contextual topic. In the last 15 years, various forms of qualitative research are increasingly being used, with Grounded Theory in particular being in ‘relative infancy as a research instrument as far as leadership research is concerned’ (Kempster and Parry 2011: 106). Kempster and Parry demonstrate in their article that ‘applying a critical realist frame of reference to grounded theory research can augment the strengths of the Grounded Theory method’ (Kempster and Parry 2011: 117).
Grounded Theory is concerned with the respondents themselves. It does not start with a pre-conceived hypothesis. Instead, the Grounded Theory researcher must constantly ask: “What is the main concern being faced by the participants”, and “What accounts for the continual resolving on this concern?” (Glaser 2004: 11).
Barney Glaser argues that Grounded Theory stands alone as a methodology for generating conceptual theory and he describes its goal as a ‘conceptual theory abstract of time, place and
WOMEN LEADERS IN KENYA 2
3
people’ (Glaser 2011: 10-11). The opportunity of Grounded Theory is to develop concepts andtheoretical relations through a continuous cycle of collection, coding and conceptual analysis (Glaser 2011; Bryman 2012; Strauss and Corbin 1990).
The challenge is to remain in an open coding process, until such time as the core variable has been discovered and then to rigorously analyse, contrast, challenge it until it feels dense. The emphasis is on analysing incidents, and not on descriptions, and this emphasis is crucial to accessing the core variable and theoretical concepts: ‘the GT problem and core variable must emerge and it will’ (Glaser & Holton 2004: 7). This core variable will have high explanatory power, because it has been
constantly examined from different angles and has been verified through saturation, relevance and workability. Above all, it is recognised as being highly relevant by the respondents (Glaser 2004). Thus, the product of good Grounded Theory research is a theory which ‘explains the preponderance of behaviour in a substantive area with the prime mover of this behaviour surfacing as the main concern of the primary participants’ (Glaser 2004: 10).
3.4 Consequence of Choosing Critical Realism-Grounded Theory
The key consequence of my choice for CR-GT is that theory is incorporated later in the research process. Glaser has argued emphatically against prior literature review, due to the risks of losing focus on the main concern of respondents in their local context. In the context of critique of the Western bias of leadership theory and with the tenets of my research, this made sense, but was difficult to do. Not only am I very interested in theory, but the UvA process emphasises the importance of the literature review process.
Despite the gendered nature of leadership, I have not attempted to engage with gender theory, because of the critique by African academics referred to in Chapter 2 that there is little
understanding of African gender perspectives. I have however engaged with the Strategic Relational Approach (Jessop 2005; Hay 2002) both as part of my conceptual frame and as an instrument of analysis to help perceive layers. I also encountered and integrated Althusser’s concept of
interpellation. Both of these helped to visualize multiple levels of reality. As will be seen in Chapter 7, physical and abstract space emerged as significant theoretical factors, which were not originally recognised, but have proved to be significant. The theory boxes below indicate my main influences.
2
4
WOMEN LEADERS IN KENYA
Theory Box – Interpellation into Leadership: Societal Steering Practice
Interpellation is a philosophical concept from the works of Althusser, who in 1971 proposed that the extent to which we recognise ourselves in public discourse affects whether we feel that we are ‘part of a social group that shares a particular orientation toward specific publicly circulating cultural forms (Louis Althusser 1971, cited in Charles Briggs 2007: 555).
In this research, interpellation is used to mean that some actors feel recognised by society as suitable to become leaders. In other words, they recognise themselves feeling “called to leadership”. I argue that the dominant conceptualisation of leaders in the local community interpellates (steers) some actors to leadership, while excluding others. Leadership
conceptualisations can therefore be exclusionary.
I found Foucault’s theory of “Governmentality”* helpful in understanding the effects of invisible societal steering practices on different actors, and in understanding how it is that some actors become privileged and more easily interpolated into leadership. We can view societal
interpellation as a good example of societal steering practises. Applied to this research, I argue that the extent to which women feel themselves interpellated as leaders influences whether they reach for leadership, and how they enact it. In the case of women leaders, the world over and certainly in Kenya, they are also subject to interpellations concerning society’s conceptualisation of womanhood. The concept of interpellation was very helpful in the evolution of my research, as it helped me realise that interpellation worked differently in different contexts. Dominant societal conceptualisations influences the way in which community leaders perceive their own identity, and this impacts their perception of the influence they can exercise in different contexts. _________________________
* Governance refers to all societal steering and regulatory practices that exist both in law and in the norms at work in diverse societal hierarchies and structures. These steering practices can be perceived in the norms, values, rules, and other group governance processes both written and unwritten, and they have a tendency to include and exclude different actors, and to privilege certain actors over others (van Dijk 2014).
WOMEN LEADERS IN KENYA 2
5
3.5 Conclusion
This chapter has compared the research tenets derived from my critique of leadership theory with my own axiological, ontological and epistemological positions and explained the rationale for my choice of Critical Realism ontology with a Grounded Theory methodology. The influence of the Strategic Relational Approach (SRA) and my use of the concept of Interpellation have also been acknowledged. In the following chapter I show how the research methods are derived from the theoretical framework and reflect the central tenets introduced at the end of chapter 2.
Theory Box – Strategic Relational Approach
The agency of women leaders is embedded in their local political, social, cultural and economic structures. Analysing their agency therefore requires a conceptual framework that acknowledges their embeddedness. I chose for the Strategic Relational Approach (SRA), because it recognizes and facilitates analysis of the separateness and the interaction of structure and agency. The SRA is deserving of an in-depth explanation, but for current purposes it is sufficient to understand a few key points:
SRA examines action and structure in relation to one another, recognizing their constant interaction. It also recognizes that specific structures can privilege certain actors, identities, strategies, spatial and temporal horizons, and certain actions over others. This structural
selectivity means that certain structures constrain certain identities and strategies and reinforce
others (Jessop 2005). However, SRA also frames actors as being capable of analysing their context and formulating strategies to influence their context. Their strategic actions are ‘intentional conduct … to realize certain outcomes and objectives which motivate action’ (Hay 2002: 129). This implies reflection, motivation, and the ability to learn from the past, and the idea that one can exercise choices (Hay 2002, Jessop 2005).
2
6
WOMEN LEADERS IN KENYA
4.0 Methodology
In the previous chapter, I described how my central tenets, which reflect a transformative paradigm, were consistent with my choice of a Grounded Theory methodology within a Critical Realist
ontological frame. This chapter will show how the elements of my research design have been integrated at several levels to fit with its ontological and epistemological foundations. I will
demonstrate that the findings are based on a highly integrated research design, which interweaves the requirements of its tenets and paradigm with the methodology for respondent sampling, data analysis and validation to ensure it matches the core problem I aimed to address. I will also describe how I have carried out regular critical review to ensure the research design remained consistent and integrated.3
4.1 Respondents and Sampling
I selected my respondents in line with the guiding principles of tenets 1 and 2 that respondents and their local contexts and understanding should be central. The common denominator of the
respondents was that they were all community leaders participating or having participated in leadership training.
My convictions were reinforced by Mertens’ and Torrance’s arguments that the views, experiences and ideas of our respondents should be central in our designs (Glaser 2004; Torrance 2012; Mertens 2007). I am convinced by Glaser that uncovering the ‘main concern’ of our respondents and the way that it is continually resolved should be the central focus of my research.
The leaders were running organisations and initiatives focussed on improving their local community; women farmers’ knowledge sharing, savings merry-go-rounds, women’s self-help business initiatives such as brickmaking, chair rental, textile design and sale, education provision (early childhood, kindergarten, primary school, secondary school, young adults), widows and orphan support groups, farming initiatives, health, community affairs, pan-African training, church, rural development and youth coaching. They ranged in age from 25 to 68 years old. The table below summarises the number of respondent interviews, expert interviews and focus groups across the two research phases.
3
I wish to acknowledge the influence of the books Thinking with Theory in Qualitative Research by Jackson and Mazzei (2012), and Becoming a Researcher by Dunne et al (2005).
WOMEN LEADERS IN KENYA 2
7
Table 4.1: Respondents
Fieldwork Phase 1 Fieldwork Phase 2 Total Respondent Interviews: Women Men 15 15 18 4 33 19 52 Expert Interviewers 5 6 11 63 Focus Groups 6 3 9
The respondents were organised into nine focus groups, of which five were all female, two were all male and two were of mixed gender. In total approximately 60 people participated.
My sampling was designed in line with Tenet 5, to look through and beyond data to deeper realities. In Fieldwork Phase 1, I built a database of the one hundred and nineteen CBO leaders that had taken part in Libre training courses in Bungoma. The information was not readily available, but was finally built up from correspondence between the Libre trainer and local contact person, old course attendance lists, and some personal notes and papers of the local contact person.
The database provided me with information on the identities of all the trainees to help me see if there were any patterns in the identities of those chosen to participate in the courses, and to chose a fair sample population (Bryman 2012: 168 and 183-185).
While very time-consuming, these statistical data provided a clear description of the research population and on the nature of the CBOs’ activities.
2
8
WOMEN LEADERS IN KENYA
Diagram 4.1: From wider view to narrow focus
4.2 Mixed Methods – ‘All is Data’
Rationale
Tenet 5 also underpinned my choice for mixed methods which I interpreted to mean using a wide mix of methods.
My data collection was guided by the Grounded Theory maxim that “all is data”(Glaser 2004: 12). I enjoyed the Grounded Theory challenge of constant comparison - seeking, analysing and interpreting multiple forms of data, while observing and interacting with my respondents in different locations. I interacted with most of my respondents in several ways.
Creatively extending the ways in which information can be accessed tends to provide rich data, which can yield multiple insights at multiple levels (Creswell Piano Clark 2011 Chapter 3; Bryman 2012, Table 3.2; Bryman 2006). For me this approach is consistent with my position as a critical realist, striving to look ‘through and beyond’ the data, trying to ‘come ever closer to reality’ (Kempster and Parry, 2011) to the empirical, the actual and the real that a critical realist tries to reach. In addition, pursing “all” kinds of data in MM research allows me greater opportunity to notice conflicts in my data or in my interpretations. After all, as described in Chapter 3, my Grounded Theory aim is not only to become aware of my respondents ‘main concern’, and to discover how it is continually resolved, but also to generate a theory which resonates with local reality. As a researcher, I can be much more sure that my findings resonate with local reality if I have stretched my evolving theory through ‘’constant comparison’’ in different settings, with data collected in different ways, and with conflicting data.
WOMEN LEADERS IN KENYA 2
9
Methods Overview
Guided by tenet 4 to be generative, qualitative methods were dominant in my mixed method research design.
Qualitative methods included individual interviews, individual questionnaires, focus groups, analysis of diverse media including local internet reports, newspapers, magazines, homilies, hymns and training manuals of leadership training institutions. Respondents were agreeable to being recorded on video or voice recordings, and allowed me to take photographs, and provided their consent for me to do so. On my second visit I was able to acknowledge the collaboration of all of my phase one respondents by bringing each one a gift of their own framed photograph or their own video
recording on CD ROM.
Multiple Methods per Respondent:
Tenets 1,2,3, and 5 guided me into using several methods per person. Doing so provided better opportunities to understand each respondent’s perspectives from many angles, and gave me the opportunity to compare and interpret findings. So, for example, M completed a survey and an interview both in Phase 1 and Phase 2, and she made a personal video summary; E did the same as M and also took part in a focus group, and provided me with a drawing indicating the congested nature of her daily life.
I interviewed in diverse locations and visited a few leaders at work, observing them with their beneficiaries and their colleagues. Locations included community meetings and events, churches and villages, towns and cities.
Focus groups: I wished to situate the concept of leadership in its local context, and was cognisant of its contested nature. I held nine focus groups with approximately 60 CBO leaders. One group was composed of women in a rural area, another of women in the town centre; a third group was of men in the same rural area, and the fourth for men in the town centre. The fifth and six groups were mixed gender groups held within and around training courses. The seventh, eighth and ninth groups were women-only groups. The four single sex groups discussed ‘Leadership and Gender’, and the mixed sex groups, ‘Leadership in Bungoma’ and ‘African Leadership’.
Participant Observation: In addition to the above, I was able to gather data and gain insights by being a Participant Observer at three leadership development courses provided by a Dutch foundation and a Kenyan based pan-African training organisation. Observing the often intense interactions between men and women leaders in the training environment was especially helpful to make comparisons with the data and emerging concepts from the focus groups and interviews. This made GT
methodology more rigorous. There were plenty of very personal examples of the challenges these leaders faced. Having previously worked with Kenyans and other Africans I was able to partly overcome some of the drawbacks of my gender and ethnic origin.
Quantitative methods: As explained above and in Chapter 2, my research design favoured qualitative methods, as being more suitable to the nature of the research problem. Quantitative methods were used as a secondary support, in three ways:
3
0
WOMEN LEADERS IN KENYA
A survey was used to establish the identity and characteristics of the total research population: Names, ages, gender, tribe; the name of their community organisation and its focus of activities; the history of LDP attendance.
To understand patterns in the perceptions of phase one respondents, concerning changes experienced in their leadership style or skills (if any), which they related to attending leadership development course
When analysing phase one focus groups data in my initial iterations. While focus groups are a qualitative method, I used Ryan et al sorting and x analysis to understand patterns in descriptions and debates concerning their leadership challenges (Ryan and Bernard 2010). The table below provides an overview of research methods employed, indicating their relevance to research questions.
WOMEN LEADERS IN KENYA 3
1
Table 4.2: Comparison of research questions with designated mix of methods
Method SQ1 Conceptual-isations SQ2 Contexts SQ3 Enactment RQ Overall research question
Statistical database: CBO database
creation/analysis - 119 CBO leaders
Characteristics of population and nature of CBOs
Individual interviews √ √ √ √ Questionnaire 1 (Fieldwork 1) - leadership challenges, goals, MSCs √ √ Questionnaire 2 (Fieldwork 2) - definition of woman and leader, enactment challenges √ √ √ Focus groups √ √ √ √ Participant observation at training courses √ √ √ √ Analysis of training manuals √ √ √ √ Drawings of - Circles of influence/ congestion - SRA drawings √ √ √ √ Analysis of - Newspapers - Magazines - Internet news - TV news reports - Homilies - Hymns √ √ √ √ Observation at - rural villages - town market - church locations √ √ √ √
3
2
WOMEN LEADERS IN KENYA
4.3 Integrated Data Analysis
I believe it is important to reflect on how data is handled. It helps to become more conscious of how we often make automatic subconscious choices and what the consequences can be. We can also become more aware of how we are handling and analysing data, be more self critical, and improve our analysing abilities. I agree with Bryman that there are risks of unintentionally mishandling data if we have collected it in a different mode of analysis (Bryman 2006). I developed the table below to help me evaluate how I was integrating my methods and modes of analysis. Table 4.3 lists the methods I used along the left hand column, an along the top it shows a continuum from Qualitative on the left to Quantitative on the right. Each method is indicated along this continuum, with the nature of the method indicated in light green, and the mode in which the data was analysed indicated in dark green.
Table 4.3: Comparing and Integrating Methods and Modes of Analysis
List of methods
<---Qualitative---Quantitative--->
Statistical database Method Analysis
Individual interviews Phase 1 Method Analysis
Individual interviews Phase 2 Method Analysis
Questionnaire 1 Analysis Method Analysis
Questionnaire 2 Method Analysis
Focus groups Method Analysis Analysis
Participant observation at training courses Method Analysis
Analysis of training manuals Method Analysis
Drawings of:
- Circles of influence Method Analysis
- SRA drawings Method Analysis
Analysis of:
- Newspapers Method Analysis
- Magazines Method Analysis
- Internet news Method Analysis
- TV news reports Method Analysis
- Homilies Method Analysis
- Hymns Method Analysis
Observation at:
- Rural villages Method Analysis
- Town market Method Analysis
- Church locations Method Analysis Analysis
