Research-practice Partnerships for School Improvement

The Learning Schools Model

Mei Kuin Lai, Stuart McNaughton, Rebecca Jesson, Aaron Wilson

• Publisher: Emerald Publishing Limited
• Year: 2020
• ISBN:
  9781789735734
• Topic: Education
• Subtopic: Multicultural Education

Thereis an increasing focus on research-practice partnerships that adopt research designsaimed at improving educationalpractice while advancing research knowledge. There is now a need for books thatprovide a theoretical and practical account of successful research designs thathave been tested and replicated over time and contexts. This book addressesthis need by providing thefirst comprehensive account of the Learning Schools Model (LSM), a design-basedresearch-practice partnership that has been tested over 15 years and across contexts and countries(n=5). This model has successfully built teacher and school capacity andimproved valued student outcomes for primarily indigenous and ethnic minority students from lowersocio-economic communities.

Thequality of research into the model has been recognised locally andinternationally. The International Literacy Association reprinted a paper onthe original model in their volume "Theoretical models and processes of Reading(6th Ed)". The authors won the University of Auckland's Research ExcellenceAward (2015), awarded for research of demonstrable quality and impact, for their research into the Model.

This book addressesseveral gaps in the existing literature on research-practice partnerships. Firstly, understanding applications incontexts beyond the USA where much of the seminal work is located adds to ourcollective understanding of contexts in terms of constraints and enablers. Secondly, we provide a theoretical accountof partnership development and demonstrate how these are practically developedin situ to address the known need for stronger theoretical understandings of partnershipdevelopment and better training in developing partnerships. Finally, our bookdemonstrates how research can be both responsive to context and yet have robustand replicable research designs that improve valued student outcomes over timeand contexts. This in turn provides an alternate research approach forcountries where randomisedcontrol trials are often the "gold standard" for interventions.

Chapter One

Ambitious Aims: Research for Solutions and Knowledge

Improving educational practice while advancing research knowledge is a lofty aim. Meeting these twin objectives is fundamentally important yet difficult to achieve in practice, particularly when the focus is on meaningful and lasting changes in educational outcomes in schools. The literature is awash with cautionary tales of research having little impact on practice. Yet there are pressing problems in school effectiveness and in particular, inequities between groups of students, that feel intractable. Both between and within countries we can see ongoing patterns of disparities at the same time as we can see shining examples of effective practices (OECD, 2015).

Educational research should be contributing better to solutions. The reasons why it has not are, in part, due to the questionable relevance of educational research to practice (Snow, 2015, 2016). They also reflect just how difficult educational challenges are; five of which are variability, scalability, capability, acceleration and sustainability. These pose substantial challenges for researchers in school reform; and much has been written about them.

It is in this landscape that our work germinated. Our motivation as researchers is twofold: to improve the valued outcomes for students, primarily from culturally and linguistically diverse communities, who have been historically under-served in education; and to advance research knowledge both locally and internationally. We have not been able to do this without a real partnership with the local communities, schools and students, respecting and drawing on their expertise in the design and implementation of the joint work. These motivations have culminated in a whole-school intervention model called the ‘Learning Schools Model’ (LSM), a design-based research-practice partnership that has been tested and replicated over 15 years and across diverse contexts and countries. This book provides a research- and theory-informed, yet practical account of the model and its application across contexts.

In this introductory chapter, we begin by positioning our work within the global challenges facing whole school interventions and within the wider call for different approaches to partnering with schools to improve practice and outcomes. We provide compelling reasons for adopting these new approaches, before describing how our book contributes to addressing the current gaps and avenues for future research in the current literature on these approaches. We then introduce the LSM, our contribution to addressing these gaps.

Solving the Big Five – Variability, Scalability, Capability, Acceleration and Sustainability

Five well-known challenges – variability, scalability, capability, acceleration and sustainability – pose challenges for researchers in school interventions focussed on improving valued student outcomes.

Variability

The first, variability, is inherent in education, at every level; from granular to macro-units of analysis (McNaughton, 2011). Variability in effects of teaching and learning outcomes can be seen within classrooms and between classrooms, within schools and between schools, within and between clusters of school and districts and within and between countries.

There are three major explanations for these phenomena. One is the nature of teaching. Teachers are professionals whose very humanness means that they use their tools and deploy instruction in ways that reflect aspects of their knowledge, skills, values and goals. There is plenty of room in what constitutes this personal professionalism for there to be idiosyncratic actions. The second explanation focuses on the degree of specification or prescription in the worlds of teaching: in curricula, syllabi, lesson plans and programmes or in the degree of professional accountabilities either professional or external. The third is the resourcing of teaching such as physical or social resources, the leadership available and how these impact on individual or collective preparation and delivery of instruction. We will have more to say later in this book about these sources of variability, where we address the paradox that variability is both a boon to, and a constraint on effective school change.

A form of variability occurs in programmes or interventions that go beyond a first experimental demonstration. It occurs as the designers, or those tasked with implementing, take what has been demonstrated to be effective to a larger scale. Scalability, that is repeating known-to-be effective instruction across multiple sites, activities and programmes, is a very real problem. The recent evaluation of Reading Recovery in the USA shows this starkly (May et al., 2014). Reading Recovery sets a ‘gold standard’ for early intervention in literacy. It has been shown repeatedly to be effective for a large percentage of the target children and provides a ‘Response to Intervention’ means for identifying those for whom a more clinical intensive intervention is needed. Despite an overall strong effect size for Reading Recovery on a variety of outcome measures, the variation in the effect size from school site to school site is substantial. It ranges from a negative effect size, meaning the Reading Recovery intervention children in that site actually got worse than their control group peers, to effect sizes greater than 2.0, almost unheard of in social science research interventions.

Reading Recovery is an exception to the general picture of educational interventions, in the sense that it consistently produces high average or median effect sizes, as well as having this substantial variation. Many experimentally tested programmes in education only ever show weak or small effects when implemented across groups of schools. For example, in the USA since the Education Science Reform act in 2002, the Institute of Educational Sciences has funded about 90 methodologically strong randomised control trials (RCTs), 9 out of 10 of which (88%) found no or weak effects (Coalition for Evidence-Based Policy, 2013).

There are methodological reasons why few show important educational effects (see Schochet, Puma, & Deke, 2014). One of the most obvious of the problems is integrity in implementation. Programmes are hard to put in place reliably if the criterion is being consistent with the original programme design. Durlak and DuPre (2008) recently reviewed 500 health and education programmes for children and adolescents in studies from 1976 to 2006. There were generally low levels of compliance to the programme design.

Each person who is directly responsible for the intervention on the ground is a source of variability, and weakness in implementation. This is true in Reading Recovery where being able to use the complex assessment and instructional procedures flexibly and adaptively for each child can be difficult. In addition, there are features of context across the different sites of an intervention, which may be influential. Durlak and DuPre (2008) identified 23 contextual factors that influence implementation. These include differences from the initial experimental site that change the implementation needs; these could be in the characteristics of teachers and students, such as language or knowledge. There can be differences between the original demonstrations that occur under carefully controlled and well-resourced conditions with the more open complex conditions that typically operate which may be less well funded. In the case of Reading Recovery, advocacy by a principal and the status accorded the intervention in a school, as well as resourcing in terms of time and funding at a district level are influential.

Scalability

Scalability is a form of replication; in Murray Sidman’s (1960) terms it is ‘systematic replication’. In educational settings it is impossible to have as it were a true replication (‘direct replication’) – that is, repeating the same experiment in exactly the same way under exactly the same conditions. Schools are not laboratories and except under very special circumstances cannot be controlled to make each look like the other for replication purposes. So each step in scaling an intervention should be considered a systematic replication, does it work under these new conditions, with these new groups of teachers and students across these sites that are known to vary in systematic ways (McNaughton, Lai, & Hsiao, 2012).

The finding, that integrity in the ‘treatment’ design is often compromised, due to methodological and contextual issues on the one hand, and variability in how professionals carry out the intervention on the other, is well established (Schochet et al., 2014). The old adage applies: ‘there is many a slip twixt the cup and the lip’; the distance between what an intervention was designed to do and what it might actually do on the ground and across several grounds is fraught with obstacles.

Capability

The role of leaders and teachers in the aforementioned ‘slippage’ introduces the third big challenge: it is building capability on the ground to engage in the core activities specified by the evidence-informed design of an intervention. Here is the nub of this challenge, what should be specified? One approach to variability and scalability is to increase treatment integrity, focussing on the specificity of the design or programme itself. The clearer the specification, the more the integrity. Teachers are both human actors and professionals, with all that entails in terms of knowledge, skills, values and beliefs (McNaughton, 2018). This means that they interpret and reconstruct new ideas, new programmes, new methods and new designs through their existing human and professional roles. The more prescribed the actions, the more likely the actions are to be carried out in keeping with the original design features.

Two aspects of this assumption pose problems for research that is able to change practices on the ground. One is in how teachers are positioned and enabled to act. Increased specification to increase adherence to a programme results in teachers being used as technical experts, technicians who follow the prescribed steps. But this is not how teachers of the twenty-first century should be positioned according to Linda Darling-Hammond and John Bransford (2005). For these and other writers teachers need to be more like adaptive experts. More like a Reading Recovery teacher who is flexibly and adaptively able to tailor instruction to the needs of individual learners. Their bespoke instructional designs draw on rich, evidence-informed judgments and an articulated knowledge base of content, learning and instruction. Prescription also does not work well in a context of the increasing diversity in classrooms in many countries, and is less useful in developing the complex repertoire of skills and knowledge needed for life and work leaving school in the twenty-first century, let alone the design skills needed to effectively use digital tools across curricula.

An alternative view of teachers might lead to an approach that favours loose specification. When an intervention comes to town the focus should be on developing teachers’ and leaders’ understanding of the underlying principles of the design. This might position teachers as more like adaptive experts, but it carries problems too. The looser the specification the more room there is for interpretation and idiosyncratic enacting; much of which may not add value to the original design leading to greater variability and the challenges of taking to scale we have already identified. A different approach, which we explore in this book, is to reconsider the nature of capability. It is not only a capability to carry out an instructional design. In essence, it is the capability to be a co-design expert, in partnership with others including the research partners.

Acceleration

The focus of the developing co-design expertise is most often the improvement of valued student outcomes. That is, learning that is valued by all partners in an educational site. This is not only valued student achievement outcomes, but those sets of skills and knowledge valued across cognitive and social and emotional domains. In the case of achievement the challenge is often to not just improve overall levels, say in mathematics or in literacy for the ‘average’ student. With students from communities traditionally not well served by schools the challenge is to produce accelerated learning such that all students and not just some subgroups make more than the expected rates of gain. The accelerated learning needs to shift distributions of achievement for students resulting in the entire distribution of interest approximating the national expected distribution.

This fourth challenge is in part because of the problem that school interventions are often tasked to solve: to address long-standing inequalities in achievement at scale. Existing research shows how difficult this task is. Gains through interventions are typically small and need to be accumulated over long periods of time (e.g. Borman, 2005). If a group is on average two years behind where they would be expected to be nationally, a not uncommon finding for some groups of students, making even expected gains is not good enough. Students need to make more than just an expected rate of gain. This is a known problem recognised by the designer of a successful literacy intervention, Reading Recovery. Clay’s (2013) developmental argument was that in order for an early intervention programme to be functional for an individual, it needed to change the rate of acquisition to a rate of progress faster than the cohort to whom the individual belonged. This was needed so that over the brief but intensive period of the individualised intervention a learner would come to function within the average bands required for their classroom.

The issue for students from social and cultural groups who have not been well served by school is not the same as in Reading Recovery in that the target is not for a group of students to come to function as a group within average bands. Rather, in the ideal case the distribution of students needs to approximate an expected distribution, in our case of New Zealand (NZ) students, the NZ national distribution. The probability of being in the lower (or indeed the upper) ‘tail’ of the distribution should be no more or no less than what would be expected for the population as a whole.

There is an added problem facing school interventions that are designed over several years to produce accelerated gains in achievement. It is the presence of summer effects where there is differential growth in learning over the months when schools are closed (Cooper, Charlton, Valentine, & Muhlenbruck, 2000). Students from poorer communities and minority students make less growth than other students over this period contributing to a widening gap in achievement. The challenge is to design powerful school effects that over time are greater than these summer effects. This is an added challenge to meeting the criteria of effective acceleration.

Sustainability

The final big challenge is the issue of sustainability. The field of educational interventions is littered with examples of interventions that were not sustainable (Coburn, 2003). The usual meaning of sustainability is that the programme or design continues to be effective, after the point at which the original intervention conditions have stopped. Does the intervention continue in the school or schools over time, with new teachers and leaders and with new cohorts of students whose demographic characteristics might also be changing? The answer to that question is not unlike the answer to the question of scalability: not necessarily or consistently.

Identifying when an intervention is ‘finished’ can be problematic in itself (see Chapter Six), but the idea is that evaluation periods, and the specified time over which professional learning and development (PLD) has put the intervention ‘in place’, define the beginning and end. Although, in the case of interventions such as Reading Recovery or another well-known intervention, ‘Success for All’, the intervention never really ceases because the intervention is systemic. It includes an infrastructure of more expert personnel who have continuous contact with the teachers and leaders on the ground providing ongoing refreshing of the original programme and integrity to its specifications.

There is form of sustainability that is even more challenging. It is the idea that interventions ultimately are designed to change what, and possibly even how, students learn over time. Early intervention in the life course of a student is often seen as the most cost effective form of an educational change, early means the changes are put in place and will sustain. Yet much about what we know in education tells us that early interventions in and of themselves often have weakening effects over time and even then any lasting effects are very dependent on conditions that learners subsequently encounter (McNaughton, 2011).

This is a problem of proximal and distal causation or chains of causation. That is, whether the learner from an early intervention continues to progress appropriately depends on the teaching they then subsequently receive. It also depends on what they have learned and how that enables (or does not ena...

Table of contents

• Cover
• Title
• Chapter One. Ambitious Aims: Research for Solutions and Knowledge
• Chapter Two. The Learning Schools Model (LSM)
• Chapter Three. Collaborative Data Analysis
• Chapter Four. Partnerships for Design and Sustainability
• Chapter Five. Resourcing and Professional Learning and Development (PLD)
• Chapter Six. Sustainability of the LSM
• Chapter Seven. Learning to Learn
• References
• Index