Indigenous education policy, practice and research: unravelling the tangled web

An abundance of research in Indigenous education has not resulted in significant systemic change in relation to Indigenous education in Australia. In this paper we examine convergence and divergence across the policy, practice and research realms with the aim of identifying key sites of opportunity for innovation and change. Through analysing how research and evidence is produced and included/excluded in Indigenous education policy settings, the complexities of how different types of evidence are considered rigorous and relevant were found to be clearly implicated with broader social and political discourses with relation to Indigenous peoples and interests. Whilst we argue for an Indigenous based evidence approach that centres Indigenous agency and solutions, we propose that deeper conversations about Indigenous voice and diversity is needed in implementing such an approach. We re-visit some key policy cycles that resembles the new co-design approach announced by the Australian Government in 2019 and consider the implications based on published literature to date for Indigenous education.


Introduction
An issue plaguing experts in policy studies since its inception as a field of study in the 1950s is how best to approach policy development so it will result in effective implementation, especially with regard to Indigenous communities (Kinchin et al., 2017). It 1 3 is well documented that in colonised countries, such as Australia, policies across various domains continue to fail First Nations peoples (Aitken, 2009;Porter, 2017). In light of this special issue on what counts as 'evidence' in Indigenous research, we attempt in this paper to contribute to the much-needed discussion of the complex debates that have occurred in Indigenous education in Australia, including convergences and divergences in how evidence is understood or contested in policy, practice and research.
Critique of what has been an overwhelming failure to deliver effective policy reform in Indigenous education has been undertaken by scholars such as Hogarth (2018), Maxwell et al., (2018) and Fogarty et. al., (2018). This paper will examine critical examples of policy failure, but the real emphasis is more transformative as we take a macro view of several dimensions in Indigenous education that need considerable focus and action with Indigenous people at their core. What counts as evidence should privilege Indigenous voices, intellectualism, sovereignties, strengths and aspirations. To start working more constructively together, as Indigenous and non-Indigenous Australians, in the quest to improve educational provision for Indigenous young people is, of course, a primary aim. Still, a further purpose is to ensure all students in Australian classrooms understand the richness of the 65,000 + years of Indigenous knowledges, histories and stories of this place, and our shared histories.
In providing critical and applied analysis of how evidence in policy, practice and research within Indigenous education interact, contend with, and sometimes contest or even ignore each other, we will weave past and current policy trends into the analysis. We attend to questions of who produces Indigenous evidence and how governments then use this evidence, particularly in Indigenous education and policy settings. This is a policy futures paper intended to unpack these questions using scholarly sources from both Indigenous and Western paradigms, introducing the concept of codesign to guide productive and informed understandings of this term and how it might impact Indigenous education policy, practice and research in the future.
Author one, Shay, is from Wagiman Country (Daly River, Northern Territory) through her mother and grandmother. She was born off Country but is fortunate to know who she is and where she is from. Shay was born in Brisbane and raised around Southeast Queensland where she has many community ties. She is an experienced youth worker and secondary classroom teacher. Author two, Sarra, is an Indigenous Professor at the Queensland University of Technology (QUT). She is of Aboriginal heritage from the Birrigubba nation and Torres Strait Islander heritage, and her research work utilises Indigenous knowledges and frameworks with theoretical frameworks to contest prevailing assumptions and stereotypes that contribute to the lack of success of Aboriginal and Torres Strait Islander young people in schools. Author three, Lampert, is a White non-Indigenous Australian originally from Canada. She has been involved in work around embedding Indigenous perspectives in the curriculum since 1996 during which she has seen many policies come and go.
Indigenous education policy, practice and research:… Research, research, and more research-but produced by whom and for whom does it serve?
One of Australia's most influential educational scholars, Lester-Irabinna Rigney published a seminal article in 2001 that provided deep analysis of the ways notions of science and objectivity continue to reject and undermine Indigenous people's knowledges and intellectual sovereignty. Rigney (2001) discerned that the concepts of truth constructed by Western discourses, underpinned by scientific research paradigms, only seek to reinforce colonial hegemonies that centre Western knowledges and reinforce these cultural norms within knowledge production. Indeed, Smith et al. (2019) proposed that not only has Western research bolstered scientific falsehoods under the guise of truth and objectivity, but they also note how much research involving Indigenous people is focussed on 'Indigenous damage' caused by the "supposed aftermath of colonization (supposed because settler colonialism continues to violently shape Indigenous life)" (p. 13). Dominant epistemologies in Indigenous education research can be, and are persistently shaped by, the broader narrative about where problems are located and how to resolve them.
There is an abundance of scholarship in the social sciences and increasingly in the natural or physical sciences that debates objectivity and neutrality in knowledge production (Levitt et al., 2020). Many scholars, particularly those who are positioned within constructivist paradigms, acknowledge the subjectivities for Indigenous people within every aspect of knowledge production from identifying research problems, designing a study, analysing the data and re-presenting that data. Factors such as race, gender, age, ability, sexuality, socio-cultural and socio-economic histories and status of the researcher are well acknowledged to affect the ways that knowledge is produced. Many scholars are now contributing to a historical understanding of just how biased most taken-for-granted "regimes of truth" are (Checketts, 2014) and how pervasive "the cultural lens which denies the true reflection of Aboriginal and Black consciousness" (Woolombi Waters, 2018, p. 1). Indigenous Australian academic scholars such as Graham (2014); Martin (2003); Moreton-Robinson (2013); Nakata (2007) and Rigney (1999), and many more, have theorised and developed their scholarship to contest ideas that have refused to acknowledge Indigenous knowledge paradigms and indeed Indigenous ontologies and axiologies. The ascent of Indigenous scholarship has been paramount in progressing debates beyond the ideological and observational to theoretical, paradigmatic, and methodological.
It is through the work of Indigenous scholars that the existence of Indigenous theories, methodologies, and scholarship within the academy today is possible. Furthermore, this work enables rigorous intellectual debate and contestation about concepts of evidence and who are legitimate 'knowers' within research contexts that focus on Indigenous people. As such, in understanding that Indigenous scholars have only recently been included in the Australian academy, it is important to explore current Indigenous researcher representation in the field of education in the context of exploring the current evidence base in Indigenous education.
In Australia, the only country colonised by the British without a Treaty (Burney, 2018), Indigenous Australians currently make up approximately 3.3% of the population (Australian Bureau of Statistics, 2018; Markham & Biddle, 2018). Indigenous inclusion in higher education is only relatively recent. The emergence of an Indigenous Australian academic workforce resulted in some representation but currently is nowhere near parity in relation to the total Indigenous population. The data from the University Australia Indigenous Strategy Report (2017) reported that only 0.79% of the academic workforce identify as Indigenous. This was a marginal increase from 0.73% representation in 2005 (Universities Australia, 2017). With growth at this rate, we could expect only to see parity in the academic workforce (assuming the Indigenous population doesn't grow at all in this period) in 46 years' time. The report does not disaggregate the Indigenous academic statistics into discipline areas, but it does provide some evidence that, like the Indigenous teacher workforce, Indigenous academics appear to be underrepresented in education faculties nationally as well.
With limited Indigenous professional representation in the teacher workforce (More Aboriginal and Torres Strait Islander Teachers Initiative, 2014) and underrepresentation of Indigenous academics trained in the discipline of education employed in Australian universities, there is clearly an issue about diversity in knowledge production in education studies. As Buckskin (2013) stated "teachers are good role models, and we need our children to be exposed to positive role models in their schools and communities. You can't be what you can't see".

Evidence-based policy or Indigenous-based evidence?
The advent of evidence based or evidence informed policy ostensibly seeks to reduce the level of and/or improve the quality of political influence in policy development (Althaus et al., 2018). Notions of what counts as evidence in the public policy literature are well traversed. Marston and Watts (2003) outline that while policy development should of course be based on the best evidence available, they discern that not all evidence is created equal nor are notions of evidence neutral. As outlined earlier in the paper, the contestation of Indigenous knowledge paradigms and even the inclusion, much less the prioritising, of growing Indigenous scholars in this country is not yet evident. Thus the politicisation of what is accepted and refuted as evidence will continue to be plagued by ideologies that are steeped in Australia's colonial roots, intertwined with racial hierarchies that continue to accept, normalise, and promote western values (Ma Rhea, 2015).
There are many different perspectives of how the idea of evidence is constructed, often dependent upon the discipline or policy context. For example, Burns and Schuller (2007) conclude that evidence and what it constitutes is a philosophical debate about rigour, often leaning towards research that provides measures of causation. They recognise that there is no concordant definition, which means that policy makers and practitioners are likely to draw from their own discipline training and personal dispositions when deciding what counts as evidence. Thomas and Pring (2004) also discuss the values attributed to certain kinds of evidence. These values can extend beyond epistemic debates to judgements about methodology, methods, and contestability of findings. Harding (1993) was one of the first feminist theorists to highlight problems associated with so-called objective knowledge, recognising the role of standpoint in what counts as evidence. Critical race scholars have also added to conversations about the racialisation of evidence, including Milner (2020, p. 252), who explores how, "these historical and contemporary biases may force Black researchers to follow colonized, White-centric ways of knowing that dissuade them from pushing against a grain that has not necessarily advanced what we know…".
In the Australian context, Nakata (2007) extended the idea of Indigenous Standpoint as more than just a perspective. Nakata explains how standpoint produces evidence as well as how standpoints compete against each other. In the context of Indigenous education, how evidence is defined, and by whom, is vital, particularly because of the persistence of colonial and racial settler ideologies that continue to pervade much of the discourses in policy, practice and research. Moreover, what debates about what counts as evidence often lack is adequate recognition that the lived experiences, voices, knowledges, solutions and aspirations of Indigenous peoples who are the subjects of these policies and research are frequently excluded, disregarded and deliberately silenced (Brown, 2019;Shay, 2016). Indeed, Masta (2019) and Brown (2019) go much further, writing that the connection between Indigenous educational disadvantage is a product of colonial dispossession. In doing so, they suggest that "Indigenous educational disadvantage, and the policy failure to adequately address this inequality, are forms of structural violence" (Brown, 2019, p. 65). Globally, many Indigenous scholars refer to the Euro-western approaches that have controlled the provision and quality of education to, and for, Indigenous peoples (Whitinui et al., 2015).
In exploring what conditions make evidence informed policy happen, Head (2016) states that "some kinds of evidence are inevitably seen as more relevant than others for underpinning policy positions" (p. 472). The values attributed to certain kinds of evidence are also connected to how policy problems are defined. A recurring problem in policy settings is that sometimes the evidence doesn't match the actual policy problem (Burns & Schuller, 2007). In Indigenous education policy settings, Ma Rhea (2015) outlines that "Indigenous education was, and still is, subject to the whims of the majority via a complex system of regulations and funding arrangements negotiated between the federal and state levels of government" (p. 68), essentially reminding us of the lack of Indigenous input into the entire Indigenous education policy cycle.
Politicisation is inherently intertwined with how democracies are operationalised and therefore decision making on what constitutes evidence can be subject to public opinion and scrutiny (Althaus et al., 2018). The impact of politics and the media on Indigenous policy broadly has been well recognised. The media in Australia has been critiqued for some time now for its influence on social and political discourses about Indigenous peoples, cultures and issues facing our communities (Mesikammen, 2016). In unpacking the politicisation of what counts as evidence in Indigenous policy settings, Maddison (2012) uses the example of the Northern Territory Emergency Response (or commonly known as the Northern Territory Intervention) policy led by the then Prime Minister John Howard and Indigenous Affairs Minister Rob Brough. Both of these senior politicians cite eminent Indigenous health researcher and advocate Pat Anderson and her co-authored report the Little Children are Sacred (2007) as the 'evidence' of alleged extensive child abuse in remote communities in the Northern Territory to enact the intervention. This is in spite of the fact that both authors of the report "vigorously contested this claim" (Maddison, 2012, p. 269). The Australian Government wilfully ignored the expertise of the authors and went so far as to suspend the Racial Discrimination Act 1975 to enable the intervention to occur (Hunyor, 2009). Without getting into the specifics of this particular policy and the impact it has had on remote Aboriginal communities in the Northern Territory, what this example outlined by Maddison (2012) exemplifies is a recent, practical example of the politicisation, misinterpretation, dismissal and, it could be argued here, racialisation of Indigenous scholarship and agency of what is constructed as evidence in Indigenous policy settings to justify decisions made by governments and policy makers.
The essence of the example highlighted above is to illustrate how what counts as evidence, and by whom it is produced, demonstrates an operationalisation of underlying racism, broader ideologies in Indigenous affairs that cement paternalism and colonialism and resistance to Indigenous self-determination and sovereignty (Gray & Beresford, 2008). In understanding the role of evidence in Indigenous policy, we need to face the reality that the vast majority of research in Indigenous education in this country has been done by non-Indigenous researchers. Overwhelmingly, the majority of the existing body of evidence, if it has been used to inform Indigenous education policy to date, has not been produced by Indigenous people and therefore not what could be considered Indigenous-based evidence (Maxwell et al., 2018).
In the example provided above, those who are tasked with making decisions demonstrated limited understanding, or depth in understanding, of the problem. In addition, the quality of evidence, who it has been produced by and what has influenced the research design may not recognise the urgency of the problems. Policy recommendations often have not codesigned methodologies, research, strategies or solutions and seem not to have been proactive in seeking out Indigenous scholarship in policy development and enactment. As Rigney (2001) articulates, producing robust Indigenous knowledges is not a case of just adding Indigenous peoples and "giving it a stir" (p. 1); the power of Indigenous scholars and scholarship designed to bring Indigenous "philosophies, ideas and imagery" (p. 9) is fundamental to producing counter-narratives, contesting neo-colonial discourses and moving towards Indigenous intellectual sovereignty, forging new relations and understandings of what Indigenous futures in this Country might be.

Indigenous-based evidence: research to policy pipeline
The impactful National Aboriginal Education Committee (NAEC) is often cited as the first significant Indigenous leadership body to influence education policy (Holt, 2021). Organised in 1973 as a response to inequities in educational outcomes between Indigenous and non-Indigenous students, NAEC was the main consultative body for over 15 years, leading to the development of the National Aboriginal Education Policy, which continues to this day (Holt, 2021). The impact of the NAEC has Indigenous education policy, practice and research:… been considerable, despite the challenges of maintaining momentum and continuing to advocate for change when governments are constantly changing and funding arrangements are regularly withdrawn or reinvented (Wilson & Wilks, 2015). Despite the plethora of task forces, committees and policy advisors on Indigenous education, there is little evidence that these have produced any promised improvements in educational outcomes for Indigenous students (Arabena, 2017;Lowe et al., 2019b;Schwab, 2018). It would be impossible to over-emphasise just how many issues papers and consultation reports have been funded and produced that should have indeed made a difference. Donovan (2015) refers to the "consistent stream of research" (p. 613), but which of this research has informed policy? How does research find its way into policy, and what happens from there? This is remarkably hard to determine.
The Australian Institute of Aboriginal and Torres Strait Islander Studies (AIAT-SIS) Code of Ethics (2020) is direct in stating, "at every stage, Aboriginal and Torres Strait Islander research must be founded on a process of meaningful engagement". It is unlikely any education policy would now be produced without an Indigenous advisory body or reference group and at least some minimal awareness of the need for consultation or a senior Indigenous researcher. Hunt's (2013) Issues Paper on policy engagement with Aboriginal and Torres Strait Islander communities outlines some of the history and debates around engagement, including issues around trust and the long history of unequal power relationships represented in policy research. Although we now take for granted seemingly consultative processes such as inviting people onto reference groups and spending long periods on consultation, equal participation in policy design is likely an overstatement. There are repeated concerns that Indigenous policy is still deficit-based, tokenistic (Maxwell et al., 2018) and sometimes misrepresentative of the people it seeks to represent. The challenge of attaining collaborative leadership is persistent, and there continues to be concerns about consultation as a panacea in policy settings. Furthermore, rigorous, Indigenous-led evaluations of the impact of policies and recommendations are few and far between (Guenther et al., 2019).
In 2019 The Australian Government Productivity Commission sought to develop a more cohesive approach to undertaking evaluation of policy and programs affecting Indigenous people (Australian Government Productivity Commission, 2020). The Productivity Commission in Australia estimated that in 2012-2013, 30.3 billion dollars was spent on programs to address Indigenous disadvantage (Steering Committee for the Review of Government Service Provision, 2014). Although this appears to be a significant amount of funding, the report published in 2014 further outlined that only $5.6 billion (or 18.6% of the total expenditure) was invested in Indigenous-specific programs, meaning that the remaining funds were used for mainstream services provided to all Australians. Moreover, in examining Government expenditure on Indigenous programs, the Productivity Commission does not provide information about whether these programs (Indigenous or mainstream) are led by Indigenous people or have Indigenous input into where the funds are being directed and what programs are being prioritised. Not only is research overwhelmingly being undertaken by non-Indigenous researchers, but the Australian Government has also identified only a very small proportion of funding allocated to Indigenous programs, indicating a deeply seeded inability to put Indigenous people in positions that enable Indigenous leadership, governance and decision-making.

Codesign -the future of Indigenous education?
In 2019, the Australian Government announced that they are "committed to considering models for Indigenous voices at local, regional and national levels" (Australian Government, 2019) and that 'codesign' is the process that will be utilised to achieve local, regional and national Indigenous based decision making. It is anticipated that this process of codesign will extend to policy domains that affect the lives of Indigenous people. In Indigenous education, codesign has already started to feature in policy vernacular. For example, the Queensland Department of Education published a document, "Our commitment to Aboriginal Peoples and Torres Strait Islander peoples", intended as a statement to signal the Department of Education's objectives in their commitment to delivering educational outcomes for Indigenous peoples. In the document, the Department of Education explicitly states, "Our approach will: develop relationships with, and connections to, community to foster local decisionmaking and co-design" (Queensland Government, 2019).
Whilst the concept of codesign in Indigenous policy broadly is new, the approach of working collaboratively with Indigenous peoples in education has been implemented under government programs using different terminology for some time. The Aboriginal Student Support and Parent Awareness (ASSPA) program is a crucial example. The program aimed to improve educational outcomes for Indigenous students by increasing opportunities for parent participation in their children's schooling, with the parliamentary committee stating that they had heard "substantial evidence of the successful role played by ASSPA over its years of operation" (Parliament of Australia, n.d.). Funding was provided to deliver a wide range of local programs, resources and equipment to share decisions about how resources are best allocated. Back in the 1990s, it was documented that Indigenous parents were wary of programs that were primarily consultative and that there was a general aspiration for there to be a shift to a principled self-determination approach that would enable Indigenous people to have a meaningful role in education that would include decision-making (Stewart, 1999). From the little literature published about ASSPA, there are reported case studies of where successful change was implemented (for example Dwyer, 2002;Stewart, 1999), and it was often cited that having a resourced, localised model that brought Indigenous parents and schools together achieved perhaps some of the aspirations of the Government's newly labelled codesign approach. Despite conflicting reports about success stories and criticisms through an internal review report published in 2004, the Government decided to end the ASSPA program rather abruptly (Parliament of Australia, n.d.). The Australian Government replaced ASSPA with the 'Parent School Partnerships Initiative (PSPI)' despite their acknowledgement that there was still evidence of significant educational inequalities for Indigenous students (Australian National Audit Office, 2008). Packer (2005) reported that the change from ASSPA to PSPI had a substantial negative impact, with a survey by the Australian Education Union revealing that the new approach had "effectively left Indigenous parents and communities out of school activities" (p. 48) and it had resulted in thousands of Indigenous students being without tutors. The Government reported on the policydecision and its implications, which revealed that the shift to PSPI was intended to move to 'school-based initiatives', thus demonstrating the shift away from parent-driven initiatives. Is codesign a return to a 30-year old ambition to give Indigenous people agency in educational settings, or is it something different?
Codesign is a term that draws from the literature on participatory design, design thinking and public sector innovation (Blomkamp, 2018). It is well established in the codesign literature across disciplines that there is no one shared definition of what it is and how it can be operationalised in policy or research contexts. The vast majority of published scholarship on codesign in Indigenous policy purports that it is a promising shift for Indigenous policy development and enactment (Dillon, 2021;Driese & Mazurski, 2018;Verbiest et al., 2019). There is also consensus that a codesign framework that enables Indigenous leadership and governance in policy settings will assist in shifting power dynamics that continue to position governments and non-Indigenous bureaucrats as having policy and program solutions to complex issues that are often the legacy of racist policies and colonialism. However, there are also significant gaps identified in the literature that illustrate the level of investment needed in research and policy/ program training and development to understand what good codesign with Indigenous peoples looks like in practice and if a codesign approach is successful in addressing complex policy issues.
Dreise and Mazurski (2018) acknowledge that in Indigenous policy settings, the aspiration of codesign is influenced by historical and contemporary relationship dynamics between Indigenous people and Government and non-Indigenous people. Additionally, there is always a risk that the benefit and process of codesign may serve governments well but risks perpetuating practices that dismiss Indigenous knowledge and processes and are focussed more on policy needs and government defined Indigenous needs. Dillon (2021) proposes that there is "limited rigorous commentary from governments and Indigenous interests, respectively, as to why codesign is seen as important and worth pursuing in Indigenous policy contexts" (p. 7). Nonetheless, the Government has signalled a shift to what is considered a participatory empowerment model. If it encompasses features of codesign noted in the literature, such as positioning groups as experts in their own lives (Blomkamp, 2018), shared decision making (Verbeist et al., 2019), coproduction of outcomes (Dillon, 2021) and diverse Indigenous voices (Dreise & Mazurski, 2018), codesign can potentially be a welcome shift across Indigenous education policy, practice and knowledge production. With so many risks identified, primarily resource (mainly time) demands of the process on both sides, power differentials stemming from racialised positioning of those involved in codesign and a lack of evidence base about what effective codesign looks like in practice (Blomkamp, 2018;Dillon, 2021;Driese & Mazurski, 2018;Verbiest et al., 2019), it is with caution that codesign can be conceived as being a panacea to unravelling the complex web.

Indigenous voices-the heart of transformative discourses in Indigenous education across policy, practice and research
Indigenous agency and diverse voices are critical in unpacking this complex web between policy/practice/research. Not including diverse Indigenous voices in these areas reinforces the historical and contemporary exclusion and homogenising of Indigenous voices in these domains. Furthermore, Indigenous voices should not just be included but should be visibly leading this work. It is only through distinctly and unwaveringly centring Indigenous voices, sovereignty and scholarship consistently across all policy/practice/research that we may see outcomes change, as our Elders and activists have been saying for many, many years. It is beyond the scope of this paper to unpack broader political agendas on the inclusion of Indigenous voices more widely in the political arena. Still, as Indigenous education does not exist in a vacuum and is so heavily influenced by political and social discourses, it is essential to recognise the championing and fight for political sovereignty and voice that has been culminating for decades. Fredericks and Bradfield (2021) detail the Government appointed and funded consultation on constitutional change about the inclusion of Indigenous Australians that resulted in the 'Uluru Statement from the Heart' (From the Heart, 2020). With over 250 Indigenous people representing diverse communities culminating in an agreement and proposal for an Indigenous voice to parliament, the then Prime Minister (Turnbull) as well as Indigenous Affairs Minister (Scullion) rejected the proposal, despite their Government seeking Indigenous perspectives on the reform (Fredericks & Bradfield, 2021). This is an extraordinary example of the broader discourses that exist in how Indigenous people's voices are so precariously and selectively read.
With limited political representation, coupled with a lack of a Treaty, limited governance representation on matters that affect Indigenous lives and livelihoods has resulted in a unilateral policy approach in Australia. Policy approaches favour the same institutions, such as schools, that have created and re-produced disadvantages for some groups and advantages for others. One could suggest that policymakers produce education policies to inform systems that are a reflection of themselves, these being features of a culture that represent dominant western ideologies in schools (Tait, 2016). The lack of inclusion of Indigenous voices, agency and aspirations in policy broadly as an issue has been raised in political and scholarly spaces for some time now (Gillan et al., 2017;Hogarth, 2017). Community engagement and the importance of working collaboratively and holistically with Indigenous families have featured heavily in Indigenous education policies nationally (Shay & Lampert, 2020), with the National Australian Standards for Teachers even requiring teachers to "understand and respect Aboriginal and Torres Strait Islander people" in standard 2.4 (Australian Institute for Teaching and School Leadership Limited, 2013). These policy reforms may result from tireless advocacy of Indigenous educators and non-Indigenous allies that recognise the lack of Indigenous voices informing all aspects of policy (development and delivery) concerning Indigenous education. However, we argue that it is time for deeper conversations about voice-like evidence, not all agents are created equal (Lowe et al., 2019a). Indigenous diversity needs to be visibly represented, including people in metropolitan, regional and remote communities and inclusive of diversity in, for example, gender and sexuality (Ryan, 2020). Indigenous multiculturalism is seldomly attended to in great depth when analysing wicked policy problems such as Indigenous education. For instance, geographical diversity, cultural diversity, gender, linguistic diversity and socioeconomic diversity, are all fundamental factors that impact how policy is developed and implemented to address persistent issues in diverse communities. Moreover, a lack of diverse Indigenous voices being fairly represented across the scope and breadth of research, policy and practice may contribute to the continuing poor delivery of education to Indigenous students, their families and communities. Privileging Indigenous voices, even in codesigned policy, means that what counts as evidence will be informed by, and not merely about, Indigenous people.

Conclusion
In this paper, we have attempted to unpack the convergence and divergence of evidence-based policy, practice and research in Aboriginal and Torres Strait Islander education, to identify key opportunities for these mechanisms to interact constructively with the purview of improving the current situation. We recognise, as have Lowe et al. (2019b) and Schwab (2018) that the abundance of research in the field of Indigenous education to date has not produced evidence and data that have resulted in significant improvement of outcomes in Indigenous education. Lowe et al. (2019b) have synthesised some studies to illustrate the evidence to date. However, it is well established that Indigenous voices are still relatively absent in all aspects of Indigenous education (Blair, 2015;Gillan et al., 2017), including knowledge production (Universities Australia, 2017). As Kukutai and Taylor (2016, p. 2) point out, Indigenous peoples, have inherent and inalienable rights relating to the collection, ownership and application of data about them, and about their lifeways and territories.
In this paper, we have discussed knowledge production, how Indigenous people have been excluded from knowledge production and the opportunities that have been missed as a result of this exclusion. Walter & Anderson (2013, p. 132) note the "fractured relationship between first world Indigenous people and researchers" due, amongst other things, to the ways Indigenous peoples have been over-researched with limited tangible outcomes for Indigenous people resulting from this research. We propose that there are several reasons why non-Indigenous research (that addresses policy problems and otherwise) may or may not be useful in Indigenous education policy settings and suggest codesign would produce much more ethical, moral, legitimate and robust forms of evidence. We propose that in Indigenous education policy settings, we should shift from normative notions of evidence-based policy to Indigenous-based evidence to inform policy and practice towards change.
While we recognise that there are power dynamics even within Indigenous communities over who gets a say, much effort has been focussed on getting some form of meaningful representation in spaces where decisions are made and Indigenous voices have been absent. We propose that it is time for more in depth and critical conversations about diversity within Indigenous communities or, indeed, in this Covid era of isolation and online platforms, how communities can be more meaningfully involved in influencing policy and in how responses can occur.
Systemic changes are needed to ensure broader scale change in Indigenous education; the failure to deliver the Close the Gap imperatives each year is evidence of this. The new codesign approach promises more collaborative, participatory processes in Indigenous policy, some of which have been done better in the past. The question is whether an approach that centres on Indigenous intellectualism, evidence, voices, and aspirations will be more effective-what we have argued in this paper as a critical ingredient to unravelling the policy, practice and research web in Indigenous education.
Funding Open Access funding enabled and organized by CAUL and its Member Institutions. The authors received funding for a project 'co-designing Indigenous education policy in Queensland' from the Australian Research Council Discovery Indigenous funding -2024