Copyright is owned by the Author of the thesis. Permission is given for a copy to be downloaded by an individual for the purpose of research and private study only. The thesis may not be reproduced elsewhere without the permission of the Author. i “It Makes Me Feel Proud Of Who I Am”: Developing Functional Thinking Through Culturally Located Tasks A thesis presented in partial fulfilment of the requirements for the degree of Master of Education in Mathematics Education at Massey University, Manawatū, New Zealand Bronwyn Elizabeth Gibbs 2020 ii Abstract Success in algebra plays a major role in equity and lifelong opportunities well beyond the mathematics classroom. Nationally, and internationally, high failure rates in algebra see many non-dominant students excluded from equitable higher education, career, and economic opportunities. There appears to be limited research focused on non-dominant students and the development of algebraic thinking through culturally located tasks. This study examines the representations Māori and Pāsifika students use when engaging with contextual functional tasks, and the ways Māori and Pāsifika students generalise culturally located tasks involving functions. A design based research intervention and qualitative research methods, drawing on a Pāsifika research methodology, were selected as most appropriate for the study. Twelve 10-12 year old Māori and Pāsifika students from a low socio-economic, urban school in New Zealand participated. Students engaged in an intervention of eight lessons focused on developing functional thinking with growing patterns drawn from Māori and Pāsifika cultures. A range of data were collected and analysed, including interviews, field notes, video recorded classroom observations, and photographs of student work. Findings revealed that when Māori and Pāsifika students were given opportunities to draw on their cultures to make sense of functional relationships, they constructed increasingly sophisticated and abstract representations to identify, communicate, and justify generalisations. There was significant growth in their conceptual understanding of both contextualised and decontextualised growing patterns. Additionally, aligning tasks with non- dominant students’ traditions and experiences strengthened students’ mathematical and cultural identities. This study offers a contribution to the literature regarding how culturally contextualised tasks support non-dominant students to engage in early algebra, in particular, the representation and generalisation of functions. To address disparities and structural inequities in mathematics education, educators must acknowledge that students bring their own cultural knowledge and strengths to the classroom, and provide opportunities for all students to learn mathematics in ways they see as relevant to their cultural identities and communities. Recognising that mathematics is inherently cultural is a key lever for equity. iii Acknowledgements I would like to acknowledge the support I have received from a number of people throughout my studies. First of all, I want to sincerely thank my two supervisors, Doctor Jodie Hunter and Professor Roberta Hunter. I would like to thank Jodie and Bobbie for so willingly sharing their extensive knowledge, invaluable advice, and guidance throughout this research project. Over the last five years they have consistently encouraged me to work towards a vision for equitable mathematics education, to take up opportunities to extend my practice and understanding, and supported me to take on further study. It has been a privilege to work through this journey with you. Meitaki ma'ata. I would like to thank the student participants for being so open in sharing their experiences, perspectives, and enthusiasm for learning mathematics throughout the study. I learnt a great deal from all of you, and hope that the learning you engage with treasures your world views and expects nothing less than for you to reach your fullest potential. I would like to express my gratitude to the teacher participant for opening up her classroom and teaching practice, and committing her valuable time and resources to this study. Thank you for your belief in this project and in your students. Thank you to the principal and board of trustees for allowing me to undertake the research within your school setting, and to whanau who gave consent for their children to participate in the study. Finally, I would like to thank my friends and family. This study would not have been possible without you encouraging me to achieve my goals. Thank you for believing in me and my ability to complete this research. I am especially grateful to my daughters Poppy and Ruby for your interest, encouragement and motivation. E kore e mutu ngā mihi aroha ki a koutou. iv Table of Contents Abstract ii Acknowledgements iii Table of Contents iv List of Tables and Figures vii Chapter One: Introduction 1 1.1 Introduction 1 1.2 Background to the study 1 1.3 Rationale 2 1.4 Objectives 4 1.5 Overview 4 Chapter Two: Literature Review 6 2.1 Introduction 6 2.2 Sociocultural framework 6 2.2.1 Concept of culture 7 2.2.2 Māori and Pāsifika cultures 7 2.2.3 Culture and mathematics 8 2.3 Contextualisation of tasks 9 2.4 Functional thinking 11 2.4.1 Pathways of student thinking related to functions 12 2.4.2 Layers of generalisation 14 2.4.3 Representations to support reasoning and generalising 15 2.4.4 Forms of representation 16 2.4.5 Difficulties with functions: from patterns to generalisation 19 2.5 Growing patterns and indigenous cultures 21 2.6 Summary 22 Chapter Three: Methodology 24 3.1 Introduction 24 3.2 Justification for methodology 24 3.2.1 Pāsifika framework 25 3.3 Researcher Role 27 3.4 Data Collection 28 v 3.4.1 Interviews 28 3.4.2 Observations 29 3.4.3 Video recording of lessons 30 3.4.4 Student work 31 3.5 Participants and research setting 31 3.6 The research study schedule 32 3.7 Data Analysis 35 3.7.1 Coding and developing themes 36 3.7.2 Validity and reliability 37 3.8 Ethical considerations 38 3.9 Summary 39 Chapter Four: Findings and Discussion 40 4.1 Introduction 40 4.2 Students initial understanding of functional relationships 40 4.3 Representing and generalising functional relationships 42 4.3.1 Using T-charts to support functional thinking 42 4.3.2 Visual representations to support generalisation 49 4.3.3 Natural language to express generalisations 54 4.3.4 Symbolic representation 55 4.3.5 Multiple representations 59 4.3.6 Challenges with functions 62 4.4 Culturally located learners engaging with contextual tasks 63 4.5 Phase three: students final understanding of functional relationships 68 4.6 Summary 69 Chapter Five: Conclusion 71 5.1 Introduction 71 5.2 Summary of research questions 71 5.2.1 What representations do Māori and Pāsifika students use when engaging with contextual functional tasks? 71 5.2.2 How do Māori and Pāsifika students generalise culturally located tasks involving functions? 72 5.3 Key findings, implications and recommendations 73 5.4 Limitations of the study 75 5.5 Suggested areas for further research 75 5.6 Final thoughts 76 vi References 77 Appendices 90 Appendix A1: Semi-structured initial and final interview 90 Appendix A2: Task based interview 91 Appendix B: Record sheet used for classroom observations 92 Appendix C: Assessment tasks 93 Appendix D1: Task one: sāsā 94 Appendix D2: Task two: ngatu 95 Appendix D3: Task three: vaka 96 Appendix D4: Task four: titi 97 Appendix D5: Task five: tukutuku panel 98 Appendix D6: Task six: tivaevae 99 Appendix D7: Task seven: fala 100 Appendix D8: Task eight: kapa haka 101 Appendix E1: Thematic analysis table used to group codes into themes 102 Appendix E2: Thematic analysis 103 Appendix F1: Principal consent form 104 Appendix F2: Board of Trustees consent form 105 Appendix F3: Student participant consent form 106 Appendix F4: Parents of student participants consent form 107 Appendix F5: Teacher participant consent form 108 Appendix G1: Student and parent research information sheet 109 Appendix G2: Board of Trustees research information sheet 111 Appendix G3: Teacher research information sheet 113 vii List of Tables and Figures Summary of Tables Table 1 Cultural context, pattern and function type of tasks used 33 in the lesson sequence Table 2 Summary of research activities and data gathering strategies 35 implemented during each phase of the current study Table 3 Participants functional thinking and representations pre-teaching 41 intervention Table 4 Participants functional thinking and representations post-teaching 68 intervention Summary of Figures Figure 1 A covariation and correspondence approach for a growing pattern 14 (as cited in Wilkie, 2012, p. 4) Figure 2 Pattern generalisation as objectification, showing layers of 15 generality and different types of thinking (as cited in Wilkie, 2016, p. 338) Figure 3 A typical growing pattern and recursive approach to finding the 20 next position (as cited in Beatty, 2014, p.1) Figure 4 The Ula Model for collaborative engagement (as cited in Sauni, 26 2011, p. 57) Figure 5 T-chart showing recursive relationship in the dependent variable 43 Figure 6 The three columned t-chart 46 Figure 7 T-chart representing the quadratic pattern 49 Figure 8 Visual representation of the multiplicative structure of the 50 growing pattern Figure 9 Using the visual representation to justify the function 52 𝛾 = 6 𝑥 + 3 Figure 10 Visual representation of triangular numbers to explore the 53 function y = 𝑥(𝑥+1) 2 viii Figure 11 Variable notation characterising the rule as an equation 58 Figure 12 Multiple representations and generalisation of the ngatu pattern 60 Figure 13 Multiple representations in group workbooks 61 1 Chapter One: Introduction 1.1 Introduction This chapter provides the background to the study and establishes the context of the research. Section 1.2 highlights mathematical achievement challenges for Māori and Pāsifika students in New Zealand schools, and the need for educators to recognise mathematics as a cultural endeavour if equitable outcomes are to be achieved. Section 1.3 explains the rationale of the study, and section 1.4 presents the specific research questions this study investigates. Finally, an overview of the chapters is provided in section 1.5. 1.2 Background to the study New Zealand communities, like those around the world, are becoming increasingly culturally diverse. Within New Zealand, Māori and Pāsifika peoples are two of the fastest growing groups compared to the overall population (Statistics New Zealand, 2018). Current projections indicate that Māori and Pāsifika children will make up the majority of New Zealand primary school students by 2040. The implications for the education system are far- reaching, as these changing demographics require a greater capacity to respond to diverse cultures, and optimise mathematical learning opportunities for all students (Alton-Lee, 2003; Hunter & Hunter, 2018; Seah & Andersson, 2015). In New Zealand, national and international data (e.g. Educational Assessment Research Unit and New Zealand Council for Educational Research, 2018; Ministry of Education, n.d.) show that Māori and Pāsifika students are underachieving in mathematics compared to all other cultural groups. For a number of years, New Zealand governments have sought to reform and improve educational outcomes for Māori and Pāsifika students. Education policies and Ministry of Education initiatives, such as the Ka Hikitea Māori education strategy (Ministry of Education, 2013a) and Pāsifika Education Plan (Ministry of Education, 2013b), promote an equitable education system, focused on raising academic success and reducing disparities in outcomes. However, this vision is yet to become a reality. Evidence from the research (e.g. Hunter & Hunter, 2018; Louie & Adiredja, 2020; Rubel, 2017) suggests that the persistent underachievement of non-dominant students within mathematics education can be attributed to three interrelated factors: deficit theorising; 2 prevailing beliefs that mathematics is culture free; and structural inequities within the education system. A common thread between the three components, is that they all problematise non-dominant student’s cultures. Deficit theories assign blame for disparities in educational outcomes to students themselves, and perceived deficiencies in their family, community, and cultural backgrounds (Hunter & Hunter, 2018; Louie & Adiredja, 2020; Rubel, 2017). Related to deficit theories are pervasive Eurocentric beliefs that mathematics is culture and value free, and involves universally accepted facts, rules and procedures (Hunter & Hunter, 2018; Louie, 2017; Nasir, 2016). Rather than questioning assumptions about the education system non-dominant students are positioned within – for example, which students and their language, values, knowledge and ways of being are privileged while other groups of students are disadvantaged - the structures of the education system are framed as neutral and normal, and students and their cultures as the problem that needs to be fixed (Bills & Hunter, 2015; Davis, 2018; Louie & Adiredja, 2020). In contrast to focusing on gaps in achievement, a significant body of research (e.g. Bills & Hunter, 2015; Hunter & Hunter, 2018; Louie, 2017; Wager, 2012) questions the gaps in opportunities for all students to learn mathematics in ways they see as relevant to their cultural practices and cultural identities. These research studies show that privileging non- dominant students’ cultural capital in the mathematics classroom has a powerful influence on equitable educational outcomes. When learning is viewed as inherently cultural, non- dominant students’ values, languages, traditions, and worldviews, are not deficits to be overcome but assets which can propel students to academic success (Si’ilata et al., 2018). 1.3 Rationale Over the past two decades the teaching and learning of algebra has been a prominent research area in mathematics education (e.g. Blanton et al, 2015; Kaput, 2017; Rivera & Becker, 2011). Algebra is considered critical to students’ mathematical development, and has been labelled as a gateway to academic and professional success (e.g. Blanton et al., 2018; Knuth et al., 2016; Morton & Riegle-Crumb, 2019). However, for many non-dominant students algebra is not a gateway to opportunity, it is a gatekeeper. Nationally, and internationally, high failure rates in algebra exclude non-dominant students from equitable economic, citizenship, higher education and career opportunities - particularly those related to the twenty first century STEM subjects of science, technology and engineering (Hunter & Miller, 3 2018; Knuth et al., 2016; Morton & Riegle-Crumb, 2019). Research into the teaching and learning of algebraic concepts is important, because success in algebra plays a major role in equity and lifelong opportunities well beyond the mathematics classroom for non-dominant students. A key aspect of learning to think algebraically is exploring functional thinking, including generalising the relationship between two variables, and representing, justifying, and reasoning with these generalisations (e.g. Blanton et al., 2018; Wilkie, 2016). Teachers commonly use visual growth patterns to develop functional thinking, because growing patterns provide a means for students to focus on the underlying mathematical structure of a pattern and explore generalisation (Beatty, 2014; Markworth, 2012; Miller & Warren, 2012; Wilkie, 2016). However, the types of growing patterns presented and discussed in the mathematics classroom make a significant difference to the ways in which all students are provided with opportunities to develop functional understandings (Beatty, 2014; Markworth, 2012). The majority of studies into the development of students’ functional thinking through growing patterns are with students from majority backgrounds, or, if they feature non- dominant students, patterns are generally decontextualized rather than culturally located (Hunter & Miller, 2018). There appear to be few international studies which have considered how young non-dominant students’ engage in functional reasoning, and the role of culture in this process. Within the New Zealand context both Māori and Pāsifika cultures have a long and rich mathematical heritage with a strong emphasis on geometric patterning used, for example, within craft design and architecture (Finau & Stillman, 1995; Meaney et al., 2013). There is some evidence that drawing on the mathematics embedded within Pāsifika patterning provides a powerful means of supporting young culturally diverse students to develop understanding of growing patterns and engage in functional thinking (Hunter & Miller, 2018). However little is known about Māori and Pāsifika students’ representations of growing patterns, and how they move through the stages of mathematical generalisation. The current study aims to extend upon the limited research available regarding how culturally contextualised patterning tasks support non-dominant students to develop their understanding 4 of growing patterns, represent their functional thinking, and engage in the generalisation process. 1.4 Objectives The purpose of the study is to explore how 10-12 year old Māori and Pāsifika students’ draw upon culturally embedded mathematics to develop algebraic understandings and make sense of functional relationships. In particular, the study addresses the following research questions: 1) What representations do Māori and Pāsifika students use when engaging with contextual functional tasks? 2) How do Māori and Pāsifika students generalise culturally located tasks involving functions? 1.5 Overview Chapter Two summarises national and international literature that is focussed on the role of culture in the teaching and learning of mathematics. It also defines the role of functional thinking in constructing early algebraic understanding, looking particularly at the research on representing and generalising functions. The literature review describes how culturally contextualised tasks provide opportunities to improve non-dominant students outcomes in mathematics education, as well as strengthening cultural knowledge and understanding. Chapter Three presents the research design and methods used in the study, and explains the data collection and analysis. The participants and research setting are introduced, and the timeframe for the study is outlined. Considerations regarding ethics, the role of the researcher, and the validity and reliability of the research are also addressed. Chapter Four integrates the findings, and discussion of the findings, to present the results of the study. The chapter examines the shifts that occurred in how non-dominant students were able to articulate and symbolise generalisations, and use multiple representational forms to support generalisation. The key developments in student thinking are identified and analysed, supported by evidence from the data and the theory discussed in the literature review. 5 Finally, Chapter Five concludes the research by addressing the research questions, presenting a summary of the key themes, and providing recommendations for educators. It addresses the limitations of the current study, and provides suggestions for areas of future research. 6 Chapter Two: Literature Review 2.1 Introduction The previous chapter introduced the background to the study and established the context of the research. This chapter summarises relevant literature that highlights the links between mathematics and culture. This is followed by a review of research studies investigating the development of functional thinking in early algebra. Section 2.2 introduces the sociocultural research framework and discusses mathematics as a cultural endeavour. It draws attention to the disconnect between mathematics education and the cultural worlds of Māori and Pāsifika learners. Section 2.3 examines relevant literature focused on the use of contextualised tasks aligned with non-dominant students’ cultural practices. It considers how culturally located tasks offer opportunities to both improve educational outcomes in mathematics, and build cultural knowledge and understanding. Section 2.4 defines the role of functional thinking in developing early algebraic understanding, looking particularly at the literature on generalising and representing functions. This section also outlines difficulties with functions, and the limitations on students’ awareness and ability to successfully generalise the functional relationship between two varying quantities, depending on the task. Finally, section 2.5 highlights research investigating growing patterns and the development of algebraic thinking in indigenous cultures. 2.2 Sociocultural framework The theoretical framework underpinning this research is a sociocultural view of learning mathematics grounded in the work of Vygotsky (1986), who saw the social environment as instrumental to children’s learning. A sociocultural framework emphasises relationships and culture as essential facets of learning and development (Nasir & de Royston, 2013; Planas & Valero, 2016). From this perspective, understanding mathematical learning requires a focus on how students participate in mathematical activities with other students, and how they draw on cultural artefacts and practices, to solve problems in a local context (Nasir & Hand, 2006). A central concept of sociocultural theory is that children construct and modify their mathematical capabilities and competencies through the cultural practices they engage in their everyday lives, because they are culturally and socially situated learners (Planas & Valero, 2016). Therefore a sociocultural framework draws our attention to the consideration of relationships between cultural and content knowledge in mathematics and provides a lens to understand and support students using knowledge from their cultural backgrounds to 7 construct new mathematical knowledge and understanding in the classroom (Nasir et al., 2008). Sociocultural theory highlights that learning is inherently cultural, and mathematics classrooms are cultural and social spaces. Different forms of knowing and being are validated, based on historical, economic, political, and social power structures that serve to perpetuate inequities in both schools and society (Davis, 2018; Nasir et al., 2008; Nieto, 2010). Rather than focusing on gaps in achievement researchers (e.g. Louie, 2017; Wager, 2012) question the gaps in opportunities for all students to learn mathematics in ways they see as relevant to their cultural identities and communities. 2.2.1 Concept of culture Nieto (2010) defines culture as: The ever-changing values, traditions, social and political relationships, and worldview created, shared, and transformed by a group of people bound together by a combination of factors that can include a common history, geographic location, language, social class, and religion (p. 136). Researchers (e.g. Nasir et al., 2008; Nieto, 2010) regard culture as a social construct because culture cannot exist outside of social interaction and culture is produced, as well as reproduced, between people in local contexts. Closely related to the concept that culture is socially constructed is the concept that culture is learned (Fecho, 2016; Nasir et al., 2014). This differentiates culture from ethnicity, as culture is not a passive legacy passed down through our genes or inherited, but is dynamic and learned through interactions with families, communities and other social groups (Nieto, 2010). 2.2.2 Māori and Pāsifika cultures Māori are indigenous to New Zealand, and are a dynamic, heterogenous population made up of divergent groups and cultural identities (Greaves et al., 2015). Pāsifika is an umbrella term used to describe a diverse, heterogeneous group of people who originate, or identify in terms of ancestry or heritage, from the Pacific Islands of Samoa, Tonga, Niue, Cook Islands, 8 Tokelau, Tuvalu and Fiji (Coxon et al., 2002; Samu, 2015). Each group is unique in terms of how they identify with particular ways of knowing, being, and viewing the world (Mahuika et al., 2011). There are, however, a set of common values which Māori and Pāsifika peoples share. For example, family and collective responsibility are an integral part of life for both Māori and Pāsifika peoples (Hunter et al., 2019). Other core cultural values include reciprocity, respect, service, inclusion, relationships, spirituality, leadership, love, and belonging (Ministry of Education, 2013b). 2.2.3 Culture and mathematics Mathematics is traditionally thought of as a discipline that is objective and independent of values and culture (Davis, 2018; d’Entremont, 2015). The perception that mathematics is culture-free, and consists of neutral, universal truths, is perpetuated by assumptions such as mathematical properties remain the same regardless of who is doing the mathematics, for example, five plus five is always 10 (Nasir, 2016; Parker et al., 2017). Research studies (e.g. Hunter & Hunter, 2018; Louie, 2017; Nasir, 2016; Parker et al., 2017) show both students and teachers frequently consider mathematics learning and their cultures as distinct entities. For example, Parker et al. (2017) describe the challenges involved in developing mathematics teachers’ cultural responsiveness when the belief that mathematics is culture-free has been normalised within mathematics education. Louie (2017), and Nasir (2016), show that mathematics educators often frame mathematics as a culturally neutral, fixed body of knowledge to be received and focus on developing universal approaches to instruction, rather than culturally specific ones. In the New Zealand context, Hunter and Hunter (2018) report statements from middle years Pāsifika students that indicate a strong belief their cultural background is non-mathematical. Every culture has mathematics and the mathematics students know, understand, and come to school with is linked to the particular cultural practices and cultural identity of the learner (Nasir et al. 2008; Wager, 2012). Students from non-dominant cultures enter school with rich mathematical experiences, backgrounds, and knowledge (Nasir et al. 2008; Wager, 2012). For example, Barton and Fairhall (1995) identify mathematical aspects of Māori culture implicit in geometric patterns in raranga (weaving), kowhaiwhai (painting) and whakairo (carving). Hunter et al. (2019) provide examples of the mathematics embedded within 9 Pāsifika homes and communities such as food preparation, making traditional clothing, and patterns in drumming. However, educators frequently under-utilise or dismiss the cultural assets of diverse students (Aguirre et al., 2017; Wager, 2012). When all children are expected to conform to the dominant groups’ practices emphasised in school mathematics, there is a disconnect between diverse students and what they perceive to be another group’s mathematics which is alien to their reality (Aguirre et al., 2017; Wager, 2012). Many non-dominant students believe that mathematics has been developed and is owned by a community they are not part of (Barta et al., 2012). Louie (2017) argues that a key product of the dominant culture characterising mathematics education, (i.e. the model of everyday instruction and classroom practices), is exclusion. Students who do not conform to the dominant stereotypes about what people who are good at math are like, typically White or Asian males from economically privileged backgrounds, are excluded from developing positive mathematical identities through narrow definitions of what counts as mathematical activity and mathematical ability. Indigenous students in Matthews et al.’s (2005) study on indigenous perspectives of mathematics education in Australia, perceived mathematics as an assimilation process and a subject where they had to become white in order to succeed. Similarly, students interviewed by Hunter and Hunter (2018), examining practices which have marginalised Pāsifika students in mathematics classrooms, expressed “a belief that to be successful in mathematics you must enter a “white-space” (p. 5). A “white-space” is a pervasive space in the education system representing dominant white values, knowledge, culture and language as the culturally neutral norm, thereby marginalising or alienating non-dominant learners from their own cultural and mathematical identities (Battey & Leyva, 2016; Davis & Martin, 2018; Hunter & Hunter, 2018). 2.3 Contextualisation of tasks A mathematical task is defined as a set of problems or a single complex problem, that focuses students' attention on a particular mathematical idea (Stein et al., 1996). Mathematical tasks directly determine what learning opportunities are made available to students and are significant in establishing how students come to “view, develop, use, and make sense of mathematics” (Anthony & Walshaw, 2009, p. 13). 10 Tasks embedded within cultural contexts provide opportunities for non-dominant students to see their experiences reflected in school mathematics, and to recognise that the activities they engage in at home and in their communities involve mathematics which is meaningful and valued (Lipka et al., 2005; Nasir et al., 2008; Wager, 2012). Meaney and Lange (2013) report on students transitioning between home and school contexts. They highlight the potential limitations on learning caused by discontinuities when the two contexts have significant differences in what counts as valued mathematical content and knowledge. These researchers contend that having to mediate between in and out of school knowledge systems results in students more likely to be unable to perform well at school in mathematics topics they master easily at home. On the other hand, drawing on connections between the mathematics culturally diverse students learn in the classroom and the mathematics embedded in their everyday lives, repositions students as having valid cultural capital in their own mathematics classrooms (Hunter & Hunter, 2018; Wager, 2012). Matthews et al. (2005) emphasise that contextualising mathematics has the potential to change the educational environment so that indigenous students believe they can perform well in the education system. Both Leonard et al. (2010), and Martin (2009) undertook research on opportunity and access in mathematics for underrepresented minority students in the USA. They report that when non-dominant students mathematics knowledge is valued alongside school mathematics knowledge students develop their mathematical identity and disposition, and feel empowered as learners and doers of mathematics. Children create meanings for mathematical concepts when they work within contexts that already have meaning for them. There are rich opportunities for students to develop new mathematical knowledge, and construct progressively more abstract understandings, as they generate formal mathematics from cultural ideas (Lipka et al., 2005; Nasir et al., 2008; Wager, 2012). As part of a long term project on mathematics in cultural contexts, Lipka et al. (2005) undertook a case study in rural, predominantly Yup'ik villages where students simulated building a fish rack structure used to dry salmon. The cultural context enabled rich explorations into how area can change while perimeter stays the same, and students explored properties of a rectangle, developed geometric proof that their bases were in fact rectangles, and investigated how quadrilaterals are related. The students participated in challenging mathematics based on a practical problem that community members of their Yup’ik villages solve every time they build a fish rack. 11 Researchers (e.g. Beatty & Blair, 2015; Davis & Martin, 2018) report that aligning tasks with non-dominant students’ cultural practices not only improves educational outcomes in mathematics, but strengthens diverse students’ cultural identities. Davis and Martin’s (2018) research focused on teaching practices and assessment that stigmatises African American students. They argue that for mathematics education to be liberatory for African American students, they must receive an education that produces growth not only in their mathematical skills, but builds on their cultural knowledge base. Beatty & Blair (2015) explored mathematical content knowledge based on the Ontario curriculum expectations and the mathematics inherent in indigenous cultural practices through Algonquin loom beading. The research focus was both the students’ mathematical thinking and cultural connections. Lessons in the study emphasised the cultural importance of looming, providing an opportunity for students to connect to their own cultural heritage. Community members, educators and students expressed pride in the mathematical thinking the children were demonstrating, such as patterning, multiplicative thinking and spatial reasoning, alongside the development of deeper cultural knowledge and understanding. 2.4 Functional thinking Functional thinking entails generalising relationships between two or more varying quantities, representing and justifying these relationships in multiple ways (such as words, symbols, tables, or graphs), and reasoning with these generalised representations to understand and predict function behaviour (Blanton et al., 2015; Stephens et al., 2017; Wilkie & Clarke, 2015). Functional thinking provides an important entry point for developing algebraic understanding, and is regarded by many mathematicians as a powerful, unifying strand, because functional thinking is threaded through all of mathematics, is a crucial part of mathematical development, and leads to a deeper understanding of the structural form and generality of mathematics (Blanton et al., 2018; Kaput, 2017; Kieran et al., 2016). Researchers have found that elementary school children are capable of deeper functional analysis than previously thought, and that functional thinking begins at grades earlier than typically expected (e.g. Blanton & Kaput, 2011; Brizuela et al., 2015; Kieran et al., 2016). Blanton & Kaput’s (2011) examination of children’s capacity for functional thinking found, for example, the types of representations students use, the ways students organise and track data, and how they express functional relationships, can be scaffolded in instruction beginning with the very youngest students, at the start of formal schooling. 12 2.4.1 Pathways of student thinking related to functions Learning trajectories are used in mathematics education to provide a research based representation of the ways student thinking in a particular domain develops over time (Fonger et al., 2018; Sarama & Clements, 2019). The levels of thinking in the progressions of a learning trajectory are not intended to be interpreted as stages students progress through in a linear sequence, but as a depiction of the growing sophistication shown in their reasoning (Stephens et al., 2017). Several researchers (e.g. Blanton et al., 2015; Markworth, 2010; Stephens et al., 2017; Wilkie, 2014) have developed learning trajectories specifically for functional thinking, describing the typical development of progressions of student thinking in generalising functional relationships. Generalisation is at the core of functions, and involves noticing a commonality in terms of a sequence, and deliberately extending the range of reasoning from specific situations, to more general ideas and conclusions identifying patterns, structures and relationships (Blanton et al., 2017; Kaput, 2017; Wilkie, 2016). Three types of functional thinking are evident from the learning trajectories, and serve as a framework to plan for, interpret, analyse and assess the kinds of functional reasoning found in mathematics classrooms: recursive, covariational, and correspondence thinking. Recursive thinking involves looking for a relationship in a single sequence of values, and indicating how to obtain a number in a sequence given the previous number or numbers (Stephens et al., 2017). When students display recursive thinking, they consider the relationship between successive terms in a pattern by referring to a sequence of distinct, particular instances only, and add the constant from term to term to extend the pattern (Blanton et al., 2015; Miller, 2016; Stephens et al., 2017). For example, Stephens et al. (2017, p.154) asked students to describe patterns they noticed in a task involving a growing pattern of seats and tables being joined for a birthday party. “The number of people is going up by 2s” was evidence of recursive thinking, because the response of adding two each time suggests the student was attending to only one variable, the people, and therefore only the recursive structure of the pattern. Students engaging in covariational thinking analyse how two quantities are coordinated and vary in relation to each other (Blanton & Kaput, 2011; Blanton et al., 2015; Stephens et al., 2017). For instance, Blanton and Kaput (2011) report on a cutting string task, where children 13 could describe the covarying relationship between the number of cuts on a piece of string and the resulting number of pieces of string when the string is folded in a single loop. Students were not looking for a recursive pattern such as adding two every time, but a relationship between the two variables: “Every time you make one more snip it’s two more” (p. 51). A critical marker of this level is that, similar to the recursive level, children could describe a relationship within specific cases but not as a generalised functional relationship over a series of instances (Blanton et al., 2015; Carraher & Scheilman, 2016). With a correspondence approach the focus is on identifying the correlation between the variables in one set and the related variables in another set (Wilkie & Clarke, 2015). Students must pay attention to the correspondence between the two variables and identify an explicit rule so that they can calculate a variable no matter which term they are looking at (Wilkie, 2014). For example, Wilkie (2014) reports on students giving a rule for an upside-down T plant growing pattern in words: “if you multiply the day number by 3 and add 1 more, you will be able to find the total number of leaves for the plant on any day” and symbols: “‘t’ is the total number of leaves on the nth day. The rule for the ‘upside-down T’ plant is t = 3n + 1.” (p. 25). A distinction of children’s thinking at this level is their focus on the structure of the pattern, their awareness of what constitutes a functional relationship, that is, a correspondence between two variables, and explicitly stating a function rule which describes a generalised relationship (Rivera & Becker, 2011; Stephens et al., 2017; Wilkie, 2014). Figure 1 provides an example of a covariation and correspondence approach to understanding functional relationships in a sequence. 14 Figure 1 A Covariation and Correspondence Approach for a Growing Pattern (as cited in Wilkie, 2012, p. 4) 2.4.2 Layers of generalisation As students’ functional reasoning develops, so does their capability to articulate the underlying structure of growing patterns algebraically (Cooper & Warren, 2008; Radford, 2010). Radford (2010) undertook a six-year long program of research with students as they moved from Grade 2 to Grade 6, to determine their transition from non-symbolic to symbolic expressions of algebraic thinking in pattern generalising activities. From his research, Radford (2010) theorised that the ways students express relationships and convey generalisations develops through three levels of algebraic generality—factual, contextual, and symbolic generalisation. As students notice pattern structures as recursive, covariational or correspondence relationships, they express these structures in factual, contextual, or symbolic forms (Warren et al., 2016). Radford (2010) views the process of generalisation as ‘objectification’, where students progress from arithmetic thinking to algebraic thinking, and recognise functional relationships in more abstract and analytic ways (see Figure 2). At an early level, factual generalisation is in the form of a concrete rule which allows children to calculate a numerical value for particular instances. For example, “there will be 10 triangles, and since there are three matches for each triangle, there will be 30 matches altogether” (Kanbir et al., 2018, p.100). Contextual generalisation focuses on more descriptive terms such as “you add the next figure from the top row”, and is language driven to explain the generalisation (Radford, 2018). The generalisation still refers to material objects in the sequence, but has moved to a new layer of generality where students’ attention has shifted from specific numbers to the variables and their relationship (Radford, 2010, 2018; Twohill, 2018). Symbolic generalisation requires a different perspective on the mathematical objects involved, and 15 students use algebraic notation, including letters, symbols or signs, to express the generalisation (Cooper & Warren, 2008; Kieran et al., 2016; Radford, 2010). The focus is now relational, and children represent their thinking symbolically, without reference to specific or situated instances (Twohill, 2018; Wilkie, 2019). Figure 2 Pattern Generalisation as Objectification, Showing Layers of Generality and Different Types of Thinking (as cited in Wilkie, 2016, p. 338) 2.4.3 Representations to support reasoning and generalising Researchers (e.g. Blanton, 2008; Blanton et al., 2018; Cañadas et al., 2016), provide evidence that elementary school children can develop and use a variety of representational tools to help them reason with functions, describe recursive, covarying, and correspondence relationships, symbolise relationships, and express generalisations. Individually, students use representations as a tool to make sense of ideas and explore the problem in their own way. Socially, representations are a way for children to communicate their reasoning with others as they represent, explain and justify generalised relationships in diverse ways (Blanton, 2008; Stephens et al., 2017; Tripathi, 2008). Investigating student- created representations also gives educators and researchers insights into students’ functional thinking and emerging conceptual understandings (Blanton, 2008; Stephens et al., 2017; Tripathi, 2008). Representation is a dynamic process, and students move through different phases in their choice of representation, showing various levels of complexity in the ways they represent the relationship between two terms in a growing pattern (Blanton et al., 2015; Bobis & Way, 16 2018; Stephens et al., 2017). Blanton (2008) discusses representation as referring to both the process and the product of expressing an idea. For instance, children may represent their thinking with pictures or tables showing the process of their developing awareness of the functional relationship. As students develop their understanding of the relationship between the variables they represent a generalisation of the relationship - the product of their thinking - using representations such as words or symbols. Blanton (2008) describes children’s representations transitioning from direct modelling, to mathematising, to mathematical understanding, as children’s representations develop from the more concrete to more abstract. 2.4.4 Forms of representation Different forms of representation can be helpful to support the development of algebraic reasoning. These representations include concrete, verbal, numerical, graphical, contextual, pictorial, or symbolic components (Tripathi, 2008). Through drawing, students may identify the underlying functional relationship between two variables, and express a generalisation. Cañadas et al. (2016) report on children using drawing to make the specifics of a problem explicit, for example, students drew desks and people to represent the context of a growing pattern. They used both drawings and natural language to express and justify the relationships between the two varying quantities: “I draw desks first and then I draw the people and then I counted by 2’s.” (p. 95). Conversely, Moss and McNab (2011) designed a teaching intervention to support second grade students understanding of linear functions, through geometric and numeric representations of growing patterns. All of the students had worked with repeating patterns, but none with growing patterns. Results showed that when students used visual representations of geometric and numeric patterns they were able to notice the constant in visual arrays, represent generalisations in their natural language, and identify and express two part rules for growing patterns (e.g. y=ax+b). Moss and McNab (2011) concluded that when visual representations were prioritised students were better able to find, express, and justify functional rules. Using concrete materials to manipulate both variables in a growing pattern gives students the opportunity to model the growing pattern, examine the pattern structure and construct generalisations (Cooper & Warren, 2008). Twohill (2018) investigated the strategies that nine and ten year old students used when constructing general terms for shape patterns. Prior to constructing the pattern terms of tiles, two students tended strongly towards recursive 17 thinking with comments such as “each time you’re adding two”. After using materials, one student identified the pattern as a top and bottom row containing n + 1, and n, tiles respectively. The other student identified the terms as containing “t diagonal pairs, with an additional tile on the top right corner” (p. 224). A common representation for solving functions problems is creating a t-chart, or function table, where students make two columns of data and record corresponding entries for the independent and dependent variables (Blanton & Kaput, 2011). Blanton et al. (2015) report on children in the early grades initially creating function tables as a means to organise covarying data, but by Grade 2, they were beginning to use tables as a tool for thinking about the data and reflecting on the relationships and patterns they could see in the table. Similarly Blanton and Kaput (2011), discuss the shifts from students in the early grades using t-charts as a place to record numbers, to students in the middle grades using t-charts as a tool that can be used to determine relationships in data and an important structure in mathematical reasoning. For instance, a student in a task where students were to find the number of body parts a growing snake would have on day ten and day n stated: “I know that on day 10 the snake will have 101 body parts and I know that on day n the snake will have n x n +1. I know this because I used my t-chart and I looked for the relationship between n and body parts” (p. 11). This suggests that the t-chart’s structure helped the student use it as a tool to compare data and find relationships. As students are plotting a graph they can directly see how relations change, rather than, for example, reading a table showing a limited number of discrete input and output values. Caddle and Brizuela’s (2011) exploration of fifth grade students discussing linear graphs found that using a graph prompted different reasoning than other representations, and gave a more complete view of the function. Similarly, Brizuela and Earnest (2017) examined how young, multi-ethnic, students worked with multiple representations to choose the best deal between a grandmother either doubling a child’s money, or tripling his money and then taking away $7. Brizuela and Earnest (2017) found students questioned their understanding about the nature of the relationship between the variables as they constructed a graph and interpreted the best deal, and the functions became more explicit in a graph than with other representations. 18 Students’ natural language descriptions are another form of generalising and representing mathematical concepts. Cañadas et al. (2016) investigated second grade students articulating their ideas about functional relationships. They found the use of natural language is helpful for students in the early grades because of their familiarity with it, and students used natural language to express relationships as a recursive pattern (counting by twos) or a functional relationship (doubling). Interestingly, Stephens et al. (2017) and Blanton et al. (2015) both unexpectedly found that students were generally more successful representing generalisations using variable notation than using words. For example, in Stephens et al. (2017) 66% of students provided a correct function rule for a growing stars pattern in variables: x · x = y, and only 32% in words: “The picture number times itself equals the number of stars” (p. 157). This clearly challenges the notion that variables as a varying quantity should not be introduced until secondary school. Researchers (e.g. Blanton et al., 2015; Brizuela et al., 2015; Stephens et al., 2017; Wilkie, 2014) underline the importance of students using letters or symbols to represent variables and generalise functional relationships through explicit rules. Brizuela et al. (2015) provides evidence that children as young as six can use variable notation in meaningful ways to express relationships between quantities and represent generalisations. They argue that allowing variable notation to become part of mathematical language as children represent their algebraic ideas gives them opportunities to develop a deep and powerful means of representing generalised relationships. Wilkie (2014) reports on upper primary students being keen to explore and experiment with letters to represent variables in algebraic equations. This tended to flow naturally from students’ own attempts to create a number sentence for their rule which often contained a mixture of numbers and words, for example: “The answer is the day × 3 + 1” (p. 26). Both Wilkie (2014) and Brizuela et al. (2015), consider students should be given opportunities to use variables and variable notation early in their formal schooling, in order to give children multiple opportunities to fully explore and develop their understandings over time. Researchers (e.g. Cañadas et al., 2016; Neilsen & Bostic, 2018; Stephens et al., 2017; Tripathi, 2008) provide evidence of the benefits of using multiple representations in order to develop a deeper, richer and more flexible understanding of functional relationships. When children have multiple ways to represent an idea they can choose representations that are intrinsically meaningful to them (Blanton, 2008; Daryaee et al., 2018). Using multiple 19 representations enables students to develop a deeper understanding of underlying mathematical structures, because different mathematical representations highlight different aspects of mathematical relationships (Neilsen & Bostic, 2018; Stephens et al., 2017; Tripathi, 2008). Each representation provides a different way for students to examine and compare the relationships, and students learn about and deepen their understanding of functions when they are explored while making connections across diverse representations (Carraher & Scheilman, 2016; Daryaee et al., 2018). 2.4.5 Difficulties with functions: from patterns to generalisation Although young students are capable of sophisticated functional thinking, there are a number of common difficulties related to the development of functional thinking reported in the research literature. Students often have difficulty transitioning from looking at the relationship between successive terms in a pattern as recursive, to focusing on the relationship between variables and viewing patterns as functions (e.g. Blanton et al., 2015; Isler et al., 2015; Stephens et al., 2017; Wilkie, 2014). Research identifies numerical and visual geometric growing patterns as a common vehicle for supporting the development of functional thinking (Blanton & Kaput, 2011; Cooper & Warren, 2011). However, the growing patterns typically used, and the ways in which the pattern structures are presented, can limit students’ awareness and ability to successfully generalise the functional relationship between variables and across instances (Beatty, 2014; Miller, 2016). Beatty (2014) shows when students work with growing patterns (such as Figure 3), they are able to describe the pattern and extend it to the next position based on additive reasoning, but have difficulty predicting values for terms further down the sequence (i.e., the 10th position, 25th position, 100th position, nth position). 20 Figure 3 A Typical Growing Pattern and Recursive Approach to Finding the Next Position (as cited in Beatty, 2014, p. 1) Many students over-generalise the applicability of proportional methods to growing pattern contexts, believing that proportional reasoning can be applied to all linear relations because they increase (or decrease) at a constant rate (e.g. Ayalon & Wilkie, 2019; Lannin, 2003). Lannin (2003) provides an example of joined cubes in a row and “smiley” stickers on rods of ten cubes. A student claimed that because a rod of ten cubes has 42 stickers, you could multiply 42 by two for a rod length of 20. The student was unaware that when she used this method she was overcounting the extra stickers where the two rods joined. In this case the student was focused on the numeric relationships (examining the increase in the number of stickers in a table), without connecting her reasoning to the context and considering the functional relationship between the variables. Moving beyond particular cases and expressing generality can create a challenge for students. Lannin’s (2005) research shows students relying on empirical evidence to support general statements. This difficulty appears to be due to the traditional focus in mathematics on calculating only a particular instance of a situation, rather than determining a general relation. Lannin (2005) describes students using guess-and-check strategies, and experimenting with various operations and numbers provided in the problem situation, to construct a generalisation. This trial and error approach led to students guessing a rule without considering why the rule might work, and attempting to find a formula to fit a particular instance of the pattern. Lannin (2005) emphasises the importance of students constructing and justifying a general relation in the problem context, and developing arguments which are independent of particular instances. 21 The role of variable as a varying quantity is an essential tool to develop and express functional relationships, but is one with which students may struggle (e.g. Blanton et al., 2017; Isler et al., 2015). Students’ typical difficulties with variables include believing that variables stand for a fixed unknown quantity, label, or attribute (e.g., l stands for leaves, d stands for day, or x is always three), rather than as a symbol that can stand for any real number in a functional relationship (Blanton et al., 2015; Blanton et al., 2017; Isler et al., 2015; Wilkie, 2014). Other conceptions regarding variables include thinking that two different variables (e.g. x, y) in the same equation cannot represent the same value, and believing that the value of a variable has something to do with its position in the alphabet (Brizuela et al., 2015; Wilkie, 2014). 2.5 Growing patterns and indigenous cultures The majority of studies into the development of students’ functional thinking are with students from majority backgrounds, or if they feature indigenous students then tasks are generally decontextualized or shared context rather than drawing on culturally located tasks (Hunter & Miller, 2018). It appears that there is limited research that focuses on indigenous students and the development of algebraic thinking through engaging with their culture within the mathematics curriculum. Miller’s (2014) study with Year 2 and 3 students in an urban indigenous school in North Queensland investigated the role of culture in young indigenous students mathematical generalisation of growing patterns. Results indicated that the type and context of the pattern impacts on indigenous students’ abilities to access the structure and relationship between the variables. Young Australian indigenous students were more successful extending and generalising growing patterns that came from the natural environment (e.g. identifying the relationship between possum tails and eyes) than they were extending and generalising growing patterns represented by decontextualized geometric shapes. Miller selected growing patterns for tasks deliberately to ensure that the functional relationship was transparent and explicitly represented. This was achieved by using materials where the variables were explicit (i.e. pattern term cards and coloured tiles) or could not be physically separated (i.e. plastic toy kangaroos and crocodiles). Miller (2016) conjectured that representing growing patterns in this manner assisted students to attend to both variables in the pattern, and potentially pushed them towards functional thinking rather than recursive thinking. 22 Further evidence demonstrates the importance of educators and researchers understanding students’ cultural representations, such as gesture in indigenous contexts, and providing opportunities for students to engage in and express their own mathematical understandings in culturally appropriate ways. Miller’s (2014) research explored how young indigenous students use gesture to generalise growing patterns, and brings attention to the importance of cultural interpretations of gesture and actions within the classroom. Miller provides two case studies of students using gesture to support their explanations, articulate their understanding of the structure of growing patterns, and generalise the rule of growing patterns. Miller argues that these students are demonstrating contextual and symbolic generalisation within Radford’s (2010) three layers of generality, and these young indigenous students are using gesture to represent generalisations of geometric growing patterns, rather than alphanumeric symbolism. If these cultural signs are missed or misinterpreted the true understanding of students’ knowledge will potentially be unseen. In the New Zealand context there is some evidence that drawing upon the mathematics implicit in Pāsifika and Māori patterning provides a powerful means of developing culturally diverse students’ early algebraic reasoning and understanding of functional patterns (Hunter & Miller, 2018). Hunter and Miller’s (2018) research concentrates on the use of culturally located patterns from Pāsifika and Māori cultures to develop young culturally diverse students’ understanding of functional patterns and support generalisation. Hunter and Miller focused on a pattern from a Cook Island tivaevae (a communally sewn traditional quilt) which grew in multiple directions. Their evidence showed that when mathematics is embedded in a cultural context, in this case the structure of a tivaevae pattern, young culturally diverse students are able to make a meaningful connection to the mathematics, begin to see covariation, develop their understanding of growing patterns, and articulate generalisations as they see the structure of the pattern growing in multiple ways. 2.6 Summary Despite mathematics being positioned as a value and culture free subject area, researchers (e.g. Hunter & Hunter, 2018; Nasir et al. 2008; Wager, 2012) have shown that mathematics is a cultural product, and is closely tied to the cultural identity of the learner. As this literature review depicts, setting mathematical tasks in contexts centred on non-dominant students 23 traditions, experiences, and cultures, gives more equitable opportunities to participate in, and develop, higher level mathematical thinking (Anthony & Walshaw, 2009; Hunter & Miller, 2018). All children can be successful in mathematics when their understanding is linked to meaningful cultural referents (Ladson-Billings, 1997). Evidence from research studies highlights that young children are capable of sophisticated sense making of functional relationships, and generalising and representing these relationships in diverse ways (Blanton et al., 2015). However, in contrast to the research, Māori and Pāsifika students are well behind their peers in national and international measures of mathematics achievement (e.g. Educational Assessment Research Unit and New Zealand Council for Educational Research, 2018; Ministry of Education, n.d.). There appear to have been limited studies which have investigated young non-dominant students’ understanding of growing patterns, and little is known about Māori and Pāsifika students’ representations of growing patterns, and how they move through the stages of mathematical generalisation. The objective of this study is to provide insight into the representations Māori and Pāsifika students use when engaging with contextual functional tasks, and how Māori and Pāsifika students generalise culturally located tasks involving functions. 24 Chapter Three: Methodology 3.1 Introduction The previous chapter discussed the literature related to the current study. This chapter outlines the research design and methods used in the study. Section 3.2 provides a justification for the selection of design based research and qualitative methods used. This section also details the Ula model, and how the current study aligns with this Pāsifika theoretical framework. Section 3.3 describes the role of the researcher, and data collection methods are explained in section 3.4. The participants and research setting are introduced in section 3.5. Section 3.6 outlines the research project and the instructional sequence that forms the basis of the study. Section 3.7 describes the data analysis, and discusses the validity and reliability of the findings of the research. Finally, section 3.8 elaborates ethical considerations. 3.2 Justification for methodology The choice of methodology was influenced by the aim of this study. This was to provide insight into how Māori and Pāsifika students generalise culturally located tasks involving functions, and the representations Māori and Pāsifika students use when engaging with contextual functional tasks. Design based research is a prominent methodology in mathematics education research, and is appropriate for developing research based solutions to complex problems in educational practice (Plomp, 2013; Prediger et al., 2015). Several key features define design based research, namely “it is interventionist, theory driven, context-specific, collaborative and contains a … focus on local impact and theory generation” (Crippen & Brown, 2018, p. 490). Design based research was selected as the most appropriate methodology for the current study for the reasons outlined in the following paragraph. Firstly, design based research aligns with the sociocultural perspective that underpins the current study. Both design based research and sociocultural theories emphasise the social aspects of learning, such as collaboration, the active construction of knowledge, and the integration of cultural experiences into the learning process (Prediger et al., 2015; Steffe & Thompson, 2000). Secondly, design based research is well suited to studying real-life issues in naturalistic environments (Anderson & Shattuck, 2012). The current study is classroom 25 based because there are complexities and social interactions in natural settings that would not be present if the study took place out of context (Anderson & Shattuck, 2012). Thirdly, a goal of design based research is to investigate the possibilities for educational improvement by supporting and studying the development of new or different forms of practice (Cobb et al., 2015). The goal of the current study is to understand how students develop the generalisation and representation of functional thinking through culturally located tasks. An interventionist methodology such as design research that aims to bring about the intended developments in order to study them is therefore appropriate. Design based research is descriptive and explanatory by nature, so design based researchers often make use of qualitative methods to study learning in the design intervention (Reimann, 2016). The essence of nearly all qualitative studies is a focus on people, and narrative information about their perceptions, actions, beliefs or behaviours (Merriam & Tisdall, 2016; Yin, 2016). A qualitative approach to data collection and analysis was appropriate for this study, in order to provide an insight into Māori and Pāsifika students lived experiences in an upper primary school mathematics classroom. A qualitative approach allowed the researcher to draw on student voices which not only provided insights into participants points of view, but also privileged the voices and experiences of Māori and Pāsifika students. 3.2.1 Pāsifika framework In order to reflect a culturally appropriate theoretical framework for the research a Pāsifika research methodology, the Ula model, was drawn upon (Sauni, 2011). The Ula model (see Figure 4) is a metaphor for collaborative engagement based on fa’asamoa: Samoan cultural principles, values and beliefs (Sauni, 2011). 26 Figure 4 The Ula Model for Collaborative Engagement (as cited in Sauni, 2011, p. 57) The circle of flowers represents the use of cultural values throughout the research study (Sauni, 2011). The space inside the ula represents the va (the space between), which is not an empty space but holds the ula together through relationships (Anae, 2010; Sauni, 2011). The Ula model also supports the current study to be responsive to Māori participants through integrating core cultural values such as whanaungatanga (whanau or family type relationships), manākitanga (caring) and kotahitanga (unity). To teu le va - cherish, nurture and care for the relational spaces - is a way of affirming the importance of relationships, and engaging with both Māori and Pāsifika participants within the same research project (Naepi, 2015). In the current study building trusting relationships, and creating a comfortable atmosphere for the participants and the researcher to engage in meaningful dialogue, were prioritised. Reciprocity was encouraged by the researcher sharing their personal identity with participants, listening carefully and respectfully, and utilising a strength based approach – participants were viewed as experts when sharing their experiences, views and understandings. Being open and approachable, and sharing a sense of humour when 27 appropriate, also helped students feel at ease during interviews and classroom observations (Sauni, 2011; Vaioleti, 2006). 3.3 Researcher Role In qualitative research studies the researcher is “the primary instrument for data collection and analysis” (Merriam & Tisdell, 2015, p. 15). Accordingly, the researcher in the current study was the sole collector of data. This enabled the researcher to maximise the efficiency of data collection and the quality of the data collected (Merriam & Tisdell, 2015). For example, immediately responding to and adapting questions in an interview, or checking with participants to validate the researcher’s interpretation of events. During classroom observations the researcher’s role was observer as participant, primarily observing and gathering information, but having some level of interaction with participants (Merriam & Tisdell, 2015; Mertler, 2019). This meant the researcher sat in with the class and participated in small ways, as a means for generating a more complete understanding of the groups’ activities, but was not integrally involved in the lesson. The relationship between the researcher and the teacher was professional and collaborative. Unlike traditional studies in education that pose the researcher as expert, design based research and the Ula model allow for a reciprocal relationship as the teacher and researcher share knowledge and resources (Jung & Brady, 2015; Sauni, 2011). During the research lessons the teacher had primary responsibility for teaching the lesson, but the researcher was available to assume a tuakana (older sibling) role if required, and provide support, advice, or further information in the moment. For example, helping the teacher decide how she might select and sequence students solution strategies in order to highlight important mathematical ideas. Participants understood the researcher’s role as an observer and collector of data, and data was openly collected (Merriam & Tisdell, 2015). The students knew the researcher as a mentor in mathematics, and were used to the researcher working in classrooms supporting teachers for pedagogical change in mathematics. This meant that participants felt comfortable when observations took place. However, participants who know they are being observed can behave in ways differently than they normally would (Merriam & Tisdell, 2015; Yin, 2016). 28 For example, the participants in the current study may have tried to present themselves in a favourable manner or regulate their behaviour in reaction to the presence of the researcher (Merriam & Tisdell, 2015). The researcher’s experiences in using an inquiry approach to teach mathematics with Māori and Pāsifika students meant familiarity with expected classroom practices and potential outcomes of the lessons. However, this strength can also be seen as a weakness, due to bias or over-familiarity with the research context leading to critical data being missed (Merriam & Tisdell, 2015). When the researcher is the primary instrument for data collection and analysis, it is important to be aware of biases and assumptions that might impact the study, and monitor how they may be shaping the collection and interpretation of data (Merriam & Tisdell, 2015). 3.4 Data Collection Collecting data from different sources in the learning environment is consistent with both design based and qualitative research methodologies (Merriam & Tisdell, 2015; Prediger et al., 2015). Using multiple methods for data collection produces rich and detailed information, and enhances the validity of findings (Merriam & Tisdell, 2015). Data collection tools utilised in the current study were interviews, field notes, video recorded classroom observations, and photographs of student work. 3.4.1 Interviews Interviewing is a valuable way of gaining in-depth information related to participants’ experiences, perspectives, and constructions of reality (Flick, 2018; Seldman, 2019). Interviews are used to find out from people things which cannot be directly observed, such as thoughts, feelings and intentions (Merriam & Tisdell, 2015). Conducting interviews in culturally appropriate ways is essential within Māori and Pāsifika contexts. Prior to interviews in the current project participants were given a brief explanation about the purpose of the study, the conduct of interviews, confidentiality, and consent. During interviews a talanoa (conversation) format was employed (Vaioleti, 2006). The interview questions were asked in a conversational rather than inquisitorial manner, allowing 29 respectful, reciprocal interactions. Each interview lasted approximately 15 minutes and took place in a quiet breakout space within the participants’ classroom. All interviews were video recorded and wholly transcribed for coding and analysis. Two types of interviews were used in the current study: semi-structured individual interviews and task-based interviews. 3.4.1.1 Semi-structured interviews. Semi-structured interviews were undertaken before and after the intervention (see Appendix A1). Semi-structured interviews sit in the middle of a continuum between structured and unstructured interviews (Merriam & Tisdell, 2015). The questions in semi-structured interviews are open-ended and flexibly worded (Merriam & Tisdell, 2015). In the current study, interview questions exploring participant’s perceptions of connections between their cultural identities and mathematics were developed from the literature review. The interview protocol provided a framework of ideas to investigate with participants, while also leaving room for participants to share their experiences beyond the expectations of the researcher. Semi-structured interviews provided space for reciprocity, as the researcher could respond to participant’s ideas in the moment, and participants could elaborate on the points of each question that were meaningful to them. 3.4.1.2 Task based interviews. Task-based interviews have been used by researchers in qualitative research in mathematics education to gain insights into participants developing knowledge, and ways of explaining, reasoning, justifying and representing mathematical ideas (Assad, 2015; Pepper et al., 2018). Task based interviews were utilised for the assessment tasks, and additionally, subsequent to lessons in the teaching sequence (see Appendix A2). During task-based interviews participants were asked to think aloud while reflecting on a previously completed task, in order to gain an in-depth perspective into shifts in student thinking which might not have been obtained from observations in the classroom setting. A semi-structured interview protocol allowed for open-ended prompting if required, depending on the judgement of the researcher in response to participants descriptions. 3.4.2 Observations Observations are important in qualitative research because they provide researchers with a first-hand account of the activity being studied while it is occurring (Cohen et al., 2018; 30 Merriam & Tisdell, 2015). Researchers have the opportunity to notice behaviours that might otherwise be taken for granted or go unnoticed by participants, which leads to richer, and more valid and authentic data (Cohen et al., 2018; Merriam & Tisdell, 2015). As with other forms of data collection, observational data must enable the research questions to be answered (Cohen et al., 2018). In the current study eight semi-structured observations were made. Semi-structured observations are conducted when the researcher has considered what will be observed and recorded beforehand, and developed loose categories of data to gather (Cohen et al., 2018). Key themes that emerged from the literature review served as prompts to guide the observations (see Appendix B). This meant the researcher had a lens for the observations, but could also adapt the focus in the moment based on what emerged during the course of the lesson (Punch & Omacea, 2014). Observations were recorded in the form of field notes and video recordings. Field notes were written up as soon after the observation as possible in order to add details and provide a summary of what had been observed. Alongside factual descriptions of what happened, the field notes had a reflective component. Reflective comments were based on, for example, initial interpretations, and tentative themes and ideas that emerged during the lesson. In this respect the researcher was engaging in preliminary data analysis alongside data collection, and could note things to ask, observe, or look for in the next round of data collection (Merriam & Tisdell, 2015). 3.4.3 Video recording of lessons Video recording is a qualitative research method that captures complexities of social activity, which is not possible through observation alone (Cohen et al., 2018; Wang & Lien, 2013). Three features of video-recording underpin it’s strengths in qualitative research: video recording provides a real-time sequential record; captures authentic behaviours in their situational context; and is a permanent record that can be viewed multiple times (Cohen et al., 2018; Wang & Lien, 2013). In order to gather a rich picture of classroom learning, and to provide a broad scope for data collection, all lessons in the current study were video recorded. Two video cameras were used during observations. While participants worked on the problem in small groups, a video camera was focused on two groups. The students included in these groups varied in different lessons. The purpose was to investigate the ways participants were engaging with the task, 31 building conceptual understanding, and reasoning about the structure of the patterns and relationships. When the teacher was facilitating a larger group discussion there was one camera recording the teacher and participants, in order to document the collaborative discourse and shifts in thinking that were occurring. Video recording the lessons allowed many details to be captured which would otherwise have been missed (Basil, 2011; Wang & Lien, 2013). For example, video recording captured non- verbal cues, reactions and gestures when participants were working in their small groups. The video records allowed the data to be analysed more than once, which helped to overcome the “fleeting nature” of observation (Basil, 2011, p.251). This allowed for reflection on what had been observed, and the possibility of reinterpretation, because the researcher wasn’t restricted by memory of the lesson (Wang & Lien, 2013). All video recordings were downloaded at the conclusion of the lesson, and wholly transcribed. 3.4.4 Student work Digital photographs of student’s responses to tasks were taken during lessons, as a means of capturing participants varied forms of representation and development of generalisations. This provided concrete evidence of how students were engaging with the pattern structures and the types of representations students were using. Names and faces were not photographed to assure anonymity. 3.5 Participants and research setting The research was conducted with a group of twelve students from one Year Six to Eight class (10-12 year olds), in a low socio-economic, high poverty, urban school in New Zealand. Ethnicities of the twelve students were Māori (40%), and Pāsifika (60%). The study participants were selected from a larger group of possible participants by purposive sampling. Qualitative research methods are likely to produce large amounts of data, and therefore the number of participants was limited to keep the study manageable (Merriam & Tisdell, 2015; Punch & Omacea, 2014). The teacher and the researcher selected the group of twelve based on who was likely to be available and willing to participate, and able to clearly communicate their experiences and opinions (Merriam & Tisdell, 2015; Punch & Omacea, 2014). 32 The school was selected because they are part of a longitudinal mathematical research project called Developing Mathematical Inquiry Communities (DMIC), designed to address the persistent underachievement of Māori and Pāsifika students in mathematics (Hunter & Hunter, 2018). The teacher was an experienced educator who has long-term involvement with DMIC professional development, and core Māori and Pāsifika values underpin her pedagogy and classroom practices. She regarded her active participation in this research as professional development and a means to reflectively inform her instructional practice. 3.6 The research study schedule Phase one The first phase of the study began with individual semi-structured interviews with participants and the teacher. Additionally, a pre-intervention assessment was administered to ascertain what students already knew about representing and generalising functional relationships. Participants showed their algebraic thinking through parallel assessment tasks, one contextualised and one decontextualised, representing the same function. Questions were intended to ascertain students’ understanding of growing patterns, including their ability to predict further positions in the pattern and describe, in general terms, the relationship between the pattern and its position. The assessment tasks mirrored the types of tasks and discussions that occurred during the teaching phase, and aligned with Level 4 learning outcomes of the New Zealand Curriculum (see Appendix C). The teacher and the researcher analysed participant responses to the assessment tasks, and collaboratively built an understanding of the progressions of thinking that could be seen, and the shifts in thinking to be developed through the intervention. Data from the pre-intervention assessment and interviews, evidence from the academic research discussed in the literature review, and the teacher and researcher’s personal and professional experience, were used to construct a series of contextual tasks (see Appendix D1-8). The tasks were designed to build on current student understandings, and increase in complexity over a sequence of lessons. The mathematics in all the tasks was embedded in a cultural context relevant to the students’ backgrounds. Patterns were drawn from Māori and Pāsifika cultures and chosen both for their context, and the ways they lent themselves to multiple ways of seeing and representing the pattern’s growth. For example, it was likely that some students would see the pattern growing recursively, while others would view the pattern in a way that lent itself to finding an explicit 33 rule (Markworth, 2010). Table 1 provides an overview of the cultural context, pattern and function tasks used in the teaching intervention. Table 1 Cultural Context, Pattern and Function Type of Tasks Used in the Lesson Sequence After the tasks were drafted, a subset of the tasks were piloted in a class with similar aged students at the same school. The teacher and students participating in the pilot work were not participants in the study. The pilot work provided the researcher with an understanding of how students made sense of the task components and helped with testing the appropriateness of the tasks, for example, the language used, the mathematical content, and the cultural context. The results of the pilot work were discussed with the teacher. Context Pattern (visual) Possible function type Task 1 Sāsā 𝛾 = 3 𝑥 + 1 Task 2 Ngatu 𝛾 = 8 𝑥 + 4 Task 3 Vaka 𝛾 = 6 𝑥 − 1 Task 4 Titi 𝛾 = 6 𝑥 + 3 Task 5 Tukutuku panel 𝛾 = 4 𝑥 − 6 Task 6 Tivaevae 𝛾 = 24𝑥 + 4 Task 7 Fala 𝑦 = 𝑥2 + 𝑥 + 4 Task 8 Kapa haka y = 𝑥(𝑥 + 1) 2 34 Phase two The second phase of the project involved teaching the series of problems. The instructional sequence that formed the basis of the research was comprised of eight tasks taught over eight one-hour lessons. The teaching intervention took place over a five-week period in Term 4 (October / November). The number and length of lessons was intended to give participants enough opportunities to engage with the concepts, and to gain traction toward algebraic understanding. Each lesson followed the same structure. First, the teacher launched the problem in a whole group setting, then participants worked in small groups to discuss the problem, look for relationships, and represent the relationships in multiple ways including words, symbols, or variable notation. After participants had enough time to explore the problem in small groups, the teacher would facilitate a larger group discussion of students’ approaches, connecting students thinking to the big mathematical idea which underpinned the lesson. The collaborative and iterative nature of design-based research allowed the teacher and the researcher repeated opportunities to discuss and reflect on the students’ responses to each task in the sequence. After each lesson the teacher and the researcher reviewed the lesson, analysed classroom events and student responses, and made any modifications required for the next lesson based on, for example, what was noteworthy about how children were talking about or representing the functional relationship. Additional data collected at this stage included audio recordings of the collaborative planning meetings and field notes regarding reflective discussions between the teacher and researcher. Phase three The final phase of the data collection consisted of participants completing a post-intervention assessment, and a concluding semi-structured interview. Table 2 summarises the research activities and data gathering strategies implemented during each phase of the study. 35 Table 2 Summary of Research Activities and Data Gathering Strategies Implemented During Each Phase of the Current Study Research activity Data gathering strategy Phase One Individual semi-structured interviews Two pre-intervention assessment tasks Collaborative planning of teaching unit and series of tasks with teacher Pilot tasks Semi-structured interviews video recorded and transcribed Think-aloud video recorded and transcribed Photographs of student work Phase Two Algebra intervention (8 lessons) Task based interviews with participants Reflective discussion and collaborative planning after each lesson Semi-structured observations Obervational field notes: descriptive notes and reflective comments Group work video recorded and transcribed Large group discussion video recorded and transcribed Photographs of student work Task based interviews video recorded and transcribed Audio recordings Researcher notes Phase Three Post-assessment tasks Final student semi-structured interviews Think-aloud video recorded and transcribed Photographs of student work Semi-structured interviews video recorded and transcribed 3.7 Data Analysis Data analysis is the process of making meaning out of the data (Merriam & Tisdell, 2015). Analysing the data in this study meant making sense of the nature of students’ generalisations and representations, when patterns and relationships they know from their cultural experiences were the basis for functional thinking. 36 In both qualitative and design based research, two phases of data analysis are carried out: ongoing and retrospective analysis (Merrian & Tisdell, 2015). In the current study ongoing analysis was an interactive and iterative process throughout the data collection phase. For instance, analysis of the pre-intervention assessment informed design of the tasks. After lessons, the researcher and teacher’s collaborative analysis informed the teaching of the following task, and the next stage of data collection. This analysis continued in a cycle throughout the series of tasks. Additionally, data was collected and analysed concurrently. Transcribing interviews and video-footage throughout the lesson sequence allowed the researcher to highlight items of potential interest and begin to informally identify themes from the data. Once all the data from assessments, transcripts from interviews and video observations, field notes, and student work were collected, retrospective and formal analysis began. 3.7.1 Coding and developing themes Thematic analysis was the method used for “systematically identifying, organizing, and offering insight into patterns of meaning” across the data set (Braun & Clarke, 2012, p. 57). Thematic analysis is a flexible method of qualitative data analysis and allows the researcher to derive meaning and understanding from the data, in order to answer the research questions being addressed (Braun & Clarke, 2012; Merriam & Tisdell, 2015). The systematic analysis of the data began with coding. Initially the data were coded with respect to functional thinking, Māori and Pāsifika cultures, and culturally located tasks. Codes were a combination of descriptive and interpretative. In the next phase analysis shifted from codes to themes in a manner consistent with the thematic analysis proposed by Braun and Clarke (2012). The basic process of generating themes involved gathering the codes into possible themes, then “collapsing or clustering codes that shared some unifying feature together, so that they reflected and described a coherent and meaningful pattern in the data” (Braun & Clarke, 2012, p. 63). The subsequent phase involved reviewing the developing themes in relation to the coded data set to determine that they did in fact address the research questions. Eventually the following themes were identified: generalisations to support functional thinking; representations to support functional thinking; contextual tasks; and culturally located Māori and Pāsifika learners. See Appendix E1 for the thematic 37 analysis table used to group codes into themes, and Appendix E2 for illustrative excerpts of coded transcripts removed and grouped on to a table. 3.7.2 Validity and reliability All research involves analysing data in a reliable way, to produce conclusions that are valid (Merriam & Tisdell, 2015). Reliability, in terms of replication, is challenging in qualitative research because human behaviour is dynamic and the social environment being studied relies on context (Merriam & Tisdell, 2015). Replication of a study will not produce the same results. However, qualitative studies can convey consistency, transferability, and trustworthiness (Flick, 2018; Merriam & Tisdell, 2015). Transferability can be achieved through thick descriptions, detailing the setting, participants, and themes. This technique allows readers to experience the events being described, and judge for themselves how the findings could be transferred to other contexts. Additionally, thick descriptions provide enough evidence to prove that the results are consistent with the data collected, and the conclusions are trustworthy and make sense (Merriam & Tisdell, 2015). Using multiple sources of data helps to eliminate bias that can result from relying on one research method. In the current study many sources of data collection were used and the data collection method and analysis was systematic. An audit trail detailing how the study was carried out, and how the findings were drawn from the data, strengthens reliability (Merriam & Tisdell, 2015). Triangulation of multiple data sources also supports confidence in the data (Merriam & Tisdell, 2015). In the current study, data has been triangulated wherever possible. For example, data from interviews and classroom observations were cross checked and compared, in order to verify that both sets of data were telling the same story (Atkins & Wallace, 2012). Additionally, the iterative nature of design based research strengthened the credibility of the research findings. In the current study, the ongoing discussion, reflection, and modification of lessons, made it possible to test developing theories from earlier lessons in later lessons. Retrospective analyses were contrasted with the informal analyses conducted while the study was in progress to strengthen the credibility of the research claims. 38 Diversity amongst participants allows for a greater range of application of the findings (Merriam & Tisdell, 2015). In the current study participants were of Samoan, Tokelauan, Cook Islands and Māori descent. Some were New Zealand born, and others Pacific born. They varied in terms of home language, gender, age, and time at school in New Zealand. Identifying patterns that exist across a range of people increases the chance that patterns can be applied to another example, compared to collecting a narrower range of data (Holley & Harris, 2019). Member checks are the procedure where a study’s findings are shared with participants, who can give feedback to correct or otherwise improve the accuracy of the study (Merriam & Tisdell, 2015; Yin, 2016). In the current study tentative interpretations of interviews and observations were checked with participants, who were asked if they accurately represented their realities. Collaboration between the researcher and the teacher enabled facts to be checked and interpretations to be corroborated or revised. Similarly, the researcher discussed themes and emerging findings with colleagues and supervisors for peer review. Asking them to comment on whether the findings were plausible based on the data adds another layer to the study’s validity. 3.8 Ethical considerations The research was designed and conducted in a responsible manner, in accordance with the Massey University Code of Ethical Conduct for Research, Teaching and Evaluations Involving Human Participants (Massey University, 2015). The project was reviewed and approved by the Massey University Human Ethics Committee prior to data collection. Ethical considerations taken into account included respect for participants with informed consent, respect for privacy and confidentiality, and social and cultural sensitivity. Written consent was obtained from all participants, including the school Principal, Board of Trustees, students, parents or guardians, and the teacher (see Appendices F1-F5). This research involved children under the age of fifteen years old, therefore consent from their parents or guardians was sought and obtained. A detailed information sheet was provided alongside the consent form to ensure that participants had clear details of why the research was being conducted, what was involved throughout the research, and what they were 39 consenting to (see Appendices G1-3). All participants in this study were allocated pseudonyms, and care was taken to exclude any identifying information about the teacher, students or school within any written reports. Participants also had the right to withdraw from the research at any time. Sensitivity to social and cultural issues was observed at all times by the researcher, for example, maintaining the daily opening and closing karakia, the classroom routines, the social groupings selected by the teacher, and respecting any silences during interviews. Risks that required consideration included the time commitment of the teacher and her students to the project. Disruptions were avoided as lessons under study took place as part of the normal classroom programme. Meetings were held at times and locations that suited the teacher so as not to burden her with an increased workload. 3.9 Summary This chapter has outlined the research design and methods used in the study, including the rationale for selecting design based research and qualitative methods of data collection and analysis. The Ula model for Pāsifika engagement provided a culturally appropriate framework to inform the research. A variety of methods to collect data were used, including interviews and classroom observations. Data was analysed using thematic analysis, identifying codes and generating themes. Triangulating data supported the credibility of the interpretations, and conducting the research in a thoroughly documented and ethical manner ensured the reliability and validity of results. The findings and discussion of the study are presented in Chapter Four. 40 Chapter Four: Findings and Discussion 4.1 Introduction The previous chapter provided an overview of the research design and methods used in the study. This chapter presents the findings and discussion in relation to how Māori and Pāsifika students generalise culturally located tasks involving functions, and the representations Māori and Pāsifika students use when engaging with contextual functional tasks. Section 4.2 outlines the initial functional understandings of the twelve Māori and Pāsifika students involved in the study. Section 4.3 draws on the data analysis from the series of eight contextual tasks to describe student learning and provide insight into the development of students’ capabilities in representing and generalising functional relationships. Section 4.4 highlights connections between the cultural contexts of the tasks, and the cultural worlds of Māori and Pāsifika learners. Finally, section 4.5 presents student understandings of functional relationships at the conclusion of the teaching intervention. 4.2 Students’ initial understanding of functional relationships Prior to commencing the teaching intervention participants compl