NAPLAN data on writing: a picture of accelerating negative change.

Wyatt-Smith, Claire ; Jackson, Christine

Introduction

Australia's NAPLAN has become what White (2013) has characterised as 'a political football' (p. 1) and a multimillion dollar investment in education. More than a decade after the country's first move to large scale standardised testing, NAPLAN investment has generated national longitudinal data across Years 3-9 which offers a wide range of opportunities to examine educational outcomes across successive phases of education. The measurement goal of NAPLAN has come to the fore in schooling policy, research, and classroom practice, gaining widespread public attention through popular media. Currently however, it remains a largely untapped educational resource for informing improvement at classroom level. Indeed, little is known about how teachers stitch together instruction for writing for NAPLAN assessment purposes and their wider program of writing instruction within and across year levels. Similarly, there are no large scale research studies of the consistencies (and inconsistencies) between individual students' NAPLAN reported literacy results in writing and school reports of achievement. The challenge is to explore and optimise the potential of NAPLAN assessment instruments and the data, both within a year and longitudinally, for informing improvement strategies. To this end this paper takes a concentrated focus on reported NAPLAN writing results and reveals that concerning percentages of students are falling below the benchmark.

The emerging interest in data

A hallmark of education in the past three decades has been the strengthening commitment to census or whole cohort standardised testing, as distinct from sample testing. In this period successive governments at state and federal levels have demonstrated insatiable appetites for data as part of moves towards evidence-based education policy. This development had been evident earlier in several countries including the United States, the United Kingdom, Japan, Singapore and China. At the time of writing, NAPLAN results across Australia are widely understood in policy, schools and the wider community as central to public accountability. NAPLAN has been normalised to the point where it is accepted as a fixed feature of schooling policy, even though critical discussions about its impact on, and more specifically, benefits to students and teachers has been limited. For example, it is not uncommon to hear teachers talking about 'teaching The NAPLAN' as though the assessment had its own prescribed curriculum content, knowledge and skills.

While debates continue as to whether NAPLAN assessments in Year 3, 5, 7 and 9 are properly characterised as low or high stakes, it is clear that they have generated high levels of public engagement and discussion about intra- and inter-state and-school performance. In part, this engagement can be traced back to the ambitious goal for literacy and numeracy that all Australian Education Ministers agreed to at a meeting of the Ministerial Council on Education, Employment, Training and Youth Affairs (MCEETYA): 'That every child leaving primary school should be numerate and be able to read, write and spell at an appropriate level' and the related sub-goal:

'That every child commencing school from 1998 will achieve a minimum acceptable literacy and numeracy standard within four years' (Department of Employment, Education, Training and Youth Affairs, 1998a).

This initiative was consistent with the belief that children in Australian schools should be entitled to learn to write and read and further, that the teaching of writing and reading is core business in schooling. This is not a new claim. While debates about effective strategies for teaching literacy have spanned several decades, there is widespread recognition in research, policy and practice that, 'The teaching of writing, like that of reading, is a fundamental responsibility of schooling, and what teachers do is very important in ensuring that children learn to write, as well as read' (Christie & Derewianka, 2008, p. 214).

Looking beyond the field of literacy education to the field of assessment and evaluation, there is an already strong and growing body of assessment research on standards-driven education reform including large-scale standardised testing (Lingard, Thompson & Sellar, 2016). Internationally, the research shows how this combination--standards and standardised testing--applied to various knowledge domains, and in the case of Australia, literacy and numeracy, have provided powerful policy levers for reform (Broadfoot, 2014; Klenowski & Wyatt-Smith, 2012). Further, the inevitable competitive dimension of testing and reporting has fueled an increasingly strong move in Australia and elsewhere to the marketisation of education. This is evident, for example, in how schools and education sectors showcase test scores in various ways, including on websites, billboards and other marketing avenues, promoting literacy and numeracy results as proxies for school quality.

Interestingly, the strengthened testing movement has occurred at the same time as attention has turned to issues of teacher quality and quality teaching. In the fields of educational assessment and literacy, the teacher has long been recognised as playing the central role in quality learning and assessment practice (Black, 2014). However, classroom assessment under the control of the teacher has typically been regarded as falling in the domain of formative assessment or assessment for improving learning (Earl, 2013). Learning and achievement data that teachers routinely collect has been long distinguished from the achievement data generated from standardised testing. Broadly speaking, the latter is typically used by education systems and sectors for measurement and accountability purposes. This longstanding distinction reflects the system needs for high reliability in publicly reported grades.

In what follows we propose that it is timely to go beyond the traditional distinction between the goals of measurement and improvement to see the two, in the context of NAPLAN, as potentially coming together through a sharpened focus on standards inclusive of benchmarks. Consideration of these matters is timely: in 2015, a second cohort of Year 9 students had experienced four episodes of testing across Years 3, 5, 7 and 9. We can now look to reported data to examine it for what it reveals about the patterns of literacy capabilities of young Australians.

To this end, in the following section we look to the genesis of standardised literacy testing in education policy in Australia as a backdrop for considering later in the paper the data on writing in particular. Underpinning the discussion is the view that educational outcomes in literacy are of high priority, connected to the broader issues of students' equity of opportunity in schooling and long term educational trajectories into employment and civic contribution.

NAPLAN in historical perspective

In the early 1990s, the historical precursor of NAPLAN was in the form of state-based tests. The original, stated purposes of the state-based testing program were twofold: early diagnosis of students at educational risk of not making satisfactory progress, and timely intervention for improvement. National goals and standards were developed in Australia through two significant agreements between the federal Minister and all state and territory Ministers for Education: the Hobart Declaration on Schooling (MYCEETYA, 1989) and the Adelaide Declaration on National Goals for the Twenty-First Century (MCEETYA, 1999). To support the achievement of these goals, the Ministers endorsed the country's first National Literacy and Numeracy Plan which included:

1. Assessment of all students by their teachers as early as possible in the first years of schooling;

2. Early intervention strategies for those students identified as having difficulty;

3. The development of agreed benchmarks for Years 3, 5, 7, and 9 against which all children's achievement in these years can be measured;

4. The measurement of students' progress against these benchmarks using rigorous state-based assessment procedures, with all Year 3 students being assessed against the benchmarks from 1998 onwards, and against the Year 5 benchmark as soon as possible;

5. Progress towards national reporting on student achievement against the benchmarks, with reporting commencing in 1999 within the framework of the annual National Report on Schooling in Australia; and

6. Professional development for teachers to support the key elements of the Plan. (Department of Employment, Education, Training and Youth Affairs, 1998b, p. 10).

The emphases of the Declarations and the National Plan were firmly on learning and equity of opportunity through early diagnosis, ongoing monitoring, improvement of student learning and teachers' professional development. This reflected the policy recognition of literacy and numeracy as central to student learning in the curriculum, with the early years focusing on oracy and learning to read and write, and in the latter years, reading and writing to learn. Literacy and numeracy were understood as the means by which young people accessed the curriculum; the corollary being that literacy and numeracy could present powerful barriers to young people's academic success. Further, the promise of the National Plan, expressed in Literacy for All: The Challenge for Australian Schools (Department of Employment, Education, Training and Youth Affairs DEETYA, 1998a) was one of harnessing the power of assessment for diagnosing learning needs and informing learning improvement approaches. Teachers' assessment capabilities and in particular, their abilities or 'know-how' in using classroom evidence and assessment data were identified as central to this and as such, areas for professional development.

Another focus of the National Plan was the endorsement of Australian literacy benchmarks by the Ministerial Council on Education, Employment, Training and Youth Affairs (MCEETYA) in two stages, April 1998 and April 2000, for Year 3, 5 and 7. This was the result of an agreement in July 1996 by the then Ministers for Education in the States, the Territories and the Commonwealth that the development of national benchmarks were for 'use in reporting minimum acceptable standards of literacy and numeracy achievement in support of the national goal'( Wilson, 2000, p. iii). The intended utility of the benchmarks was to inform teachers about 'the portable outcomes that should be available from their pedagogies' (Freebody, in Wyatt-Smith, 1998, p. 26).

DEETYA (1998a) claimed that 'the benchmark at each year level is intended to set a minimum acceptable standard: a critical level of literacy and numeracy without which a student will have difficulty in making sufficient progress at school' (p. 23). Officially, the literacy benchmarks for Years 3, 5, and 7, published in 1998 and 2000 respectively, have informed the NAPLAN testing benchmarks or 'National Minimum Standards' (NMS). Undoubtedly, these are influential with parents, schools and students today. At the time of writing, however, a Year 9 benchmark has not been published. As discussed later in this paper, this missing benchmark means that the teaching workforce and parents and students do not have an external reference point for considering quality at this key transition in schooling.

The move away from state-based cohort literacy and numeracy standardised testing to national cohort testing in 2008 occurred without a concurrent transfer of the focus on professional development. The preparedness of the teaching profession to use reported test results data for informing teaching, diagnosing learning needs and developing improvement strategies continues to be problematic. Some Australian researchers have pointed to the negative impact of NAPLAN. This includes concentrated periods of instructional time being given to teaching to the test and over-practice of tests and items (Cumming, Wyatt-Smith & Colbert, 2016; Comber & Cormack, 2013). According to Comber (2012), the impact can be intensified in 'low-socioeconomic and culturally diverse primary school communities' (p. 133). However, the prospect was there from the earliest days of state-based testing for the tests to model quality assessment for the teaching workforce and further, be a basis for generating exemplars to inform how teachers assess and score writing. This has been the approach taken in New Zealand, for example, in sample testing in the National Education Monitoring Program (NEMP) where schools and teachers are strongly engaged with test administration as professional learning through to marking and using the tests for informing learning and teaching improvement.

In the last decade Australian Territories and States have developed data analytic systems to support teachers understanding of the NAPLAN data (for example, SMART--New South Wales; SunLANDA--Queensland; EARS--Western Australia). At the time of writing, however, there are no available exemplars showing features of writing at or above the NMS to illustrate writing benchmarks. This is the case even though assessment research points to the utility of exemplars to illustrate standards (Sadler, 1989, Wyatt-Smith & Klenowski, 2014). The provision of exemplars, and especially those drawn directly from student samples, could be useful as part of a strategy to drive improvement by showing teachers, students and parents about the range of writing standards across year levels.

The following discussion considers the promised potential of the country's first National Plan, National Minimum Standards and NAPLAN more generally and examines the reported data on writing to determine if the potential has been realised.

The NAPLAN Writing domain

It is important to acknowledge the value and uniqueness of the Australian writing assessment that is NAPLAN. Currently there is no recognised international standardised writing assessment against which Australia can benchmark students' levels of writing competence. As acknowledged by Hamp-Lyons (2004) and Christie and Derewianka (2008), the Program for International Student Assessment (PISA), Trends in International Mathematics and Science (TIMMS), and Progress in International Reading Literacy Study (PIRLS) assess the domains of reading literacy, mathematics and science literacy from a global perspective, but 'writing', understood from a basic skills and genre-based perspective, is not assessed by PISA, PIRLS or TIMMS. In terms of an important diagnostic tool, the NAPLAN writing assessment provides one of the few large-scale system opportunities to explore Australian students' capabilities in this domain.

Various countries including the United States grapple with a streamlined approach to the testing of writing. Jeffery's (2009) comment on a call to action, based on persistent discrepancies between state and national assessment results, highlights the difficulty of streamlining a national writing assessment that benchmarks standards of writing proficiency. Jeffery's study was based on the need for greater uniformity in US academic standards. The concern was that the variability and complexity between state writing assessments resulted in opaque explanations of how descriptions of quality (for example, in the form of standards or rubrics) relate to writing proficiency definitions. This variability and complexity made it difficult to 'determine exactly what a student has to do to produce a 'proficient' response' (Jeffery, 2009, p. 16). As suggested earlier, Australia faces the same challenges in formulating and promulgating clear expectations of quality.

In addressing challenges in communicating expectations of quality, Australia has sought to streamline the national writing assessment through adopting what can be described as an analytic approach to formulating assessment criteria (Diederich, 1974). Broadly speaking, a main focus in this approach is the identification of specific criteria, with appropriate weighting or scores attached, so that ultimately in the scoring, the total equals the sum of the parts. However, the current lack of transparency when it comes to benchmarks hampers teachers' efforts to link benchmark standards, curriculum and pedagogy, and achievement standards. This has led some states and territories to mapping NAPLAN results and A-E (or other five-point scales) for school reporting, to examine patterns between the two. Anecdotal evidence suggests a need for systematic investigation of these patterns between the two sets of reported results and at a conceptual level, the two reporting frameworks (A-E and benchmark standards). This is vital given that achievement standards in the Australian Curriculum are not defined beyond the year level expectation generally accepted as constituting 'C'.

NAPLAN writing assessment has two important features. First, it is informed by an internationally respected genre-based language theory (Halliday and Hasan 1985). Central to this theory was an interest in language as integrally connected to culture and specifically language in use. From this theory followed a functional approach to literacy including the work lead by Australian scholars of applied linguists, including Martin, Rothery and Christie (Cope & Kalantzis, 1993; Cope et al., 1993) and more recently applied in classrooms (Love & Humphrey, 2012); second, its design is criterion-referenced where every mark a student scores has a particular meaning on a marking scale, as mentioned above. This approach to NAPLAN design through to scoring means that students, teachers and parents can have access to information about individual scores with considerable diagnostic potential.

It is well documented that historically, one of the significant barriers to realising the diagnostic potential of NAPLAN results in general, and writing results in particular, is the return of results late in the school year. The limited opportunity for schools to use the results has been acknowledged by the Australian Government in response to the Senate Education and Employment References Committee Report (Department of Education and Training 2014). The Committee agreed in principle that 'the quick turnaround of test results should receive the highest priority in the design of NAPLAN Online' (p. 3) and accordingly, it is to be implemented from 2017 over a two-to-three year period. The government focus on the changed reporting schedule is arguably critical if there are to be opportunities for schools to utilise returned data to inform student learning at individual, group and cohort levels. The change to earlier reporting brings potential benefit to schools, teachers, parents and students though its impact currently remains unknown.

The following discussion draws directly on the 2015 NAPLAN National Report published by ACARA of the results for the writing domain across all Australian states and territories. The focus starts with the 2015 results across Years 3, 5, 7 and 9, and then attention turns to the 2015 writing results against the reading results. Finally, the analysis extends to the 2009 cohort starting with their Year 3 writing results through to their 2015 Year 9 writing results, to consider the performance pattern in writing over time for this cohort.

What the data reveal about writing

The report of the national analysis of NAPLAN results is released via the National Assessment Program (NAP) arm of the ACARA website, http://www.nap. edu.au/results-and-reports/national-reports.html. The first stage of the NAPLAN summary information is released in August each year prior to the distribution of reports to parents. The full National Report is released later in the year. It supersedes the preliminary report as it contains granular detail such as: gender, Indigenous status, language background other than English, parental occupation, parental education, and location (metropolitan, provincial, remote and very remote), and shows results at each year level and for each domain of the test. The National Report is derived from a significant sample of the Australian population.

Table 1 below presents a breakdown of each state and territory as a percentage of students who participated in NAPLAN and their related levels in the NAPLAN testing in writing from the 2015 National Report. Of those students who participated in the NAPLAN writing test, the percentages are presented in relation to three levels or national standards: 'Below National Minimum Standard', 'At National Minimum Standard' or 'Above National Minimum Standard'. The official stance of ACARA (2015) in their reporting of results is that, 'exempt students are deemed not to have met the national minimum standard' (ACARA 2015) and therefore contribute to the percentage for below NMS.

The information in Table 1 and Table 2 is for 2015 only. It does not represent a longitudinal data set tracking a cohort. Readers are also advised that the percentages are represented against Band Levels related to the targeted testing years. For example, the NMS for Year 3 is Band 1, Year 5 is Band 3, Year 7 is Band 4 and Year 9 is Band 5. The approach taken in the following discussion is to focus exclusively on those students who sat the NAPLAN writing test in 2015 in these year levels. That is, we removed the percentage of students who were reported to be officially exempt in our representation of the writing data to avoid the possible distorting effect on the writing results.

Close examination of the writing data in Table 1 reveals a growing trend of students who are performing below NMS. This growth of accelerating negative change starts from Year 3 with 2.6% of Australian students below the NMS for writing and increases over 15% to 17.7% by Year 9. This means that nearly one in every five Australian students is performing below NMS by the time they are in Year 9.

Examining the Northern Territory results as an illustrative case, the growth of accelerated negative change for writing starts with 25.7 % of Year 3 students below NMS and peaks at 52.6% of Year 9 students below NMS. This means that over half of fourteen to fifteen year old students in the Northern Territory do not have the requisite skills to meet the NMS of writing.

While the test itself could be subject to scrutiny from various perspectives, the reported national data point strongly to the need for inquiring into the teaching of writing in all Australian states and territories. Further, it raises the serious issue of young people's employment readiness and employability, inevitably impacted by low levels of basic skills in writing.

Writing and reading in perspective

As indicated earlier, 'literacy' is a continuing focus for policymakers and has remained a high priority in education policy for some decades. The definition of 'literacy' has traditionally targeted reading and reading comprehension, and for Australian students this is undeniably an area that needs improvement. The Australian PISA results in 2000 and 2009 in the major domain of reading literacy acknowledged a decline equating to 4Vi months of schooling for Australian fifteen year olds (Thomson, Hillman, De Bortoli, 2013, p. 15). It is indisputable that improving 'reading' is a significant issue, especially as this extends to young people's engagement with and enjoyment of reading. However, current NAPLAN data in writing show a more significant concerning national pattern.

To illustrate, both reading and writing results are considered. Table 2 shows the Australian Reading results from the 2015 NAPLAN testing in Years 3, 5, 7, and 9. In regard to the below NMS percentages, overall the Australian reading results demonstrate relative stability: the greatest gap in results is between Year 7 and Year 9 with 2.9% of students in Year 7 below NMS increasing to 5.9% by Year 9. This contrasts with 15.1% increase in students falling below NMS between Year 3 to Year 9 for writing.

Longitudinal Tracking of students

The data for the longitudinal tracking in Table 3 was taken from the National Report of NAPLAN for 2015 results released via the National Assessment Program (NAP) arm of the ACARA website, http://www.nap. edu.au/results-and-reports/national-reports.html Once again, the 2015 results for all years demonstrate the trend of accelerating negative change when we track the cohort of students who started their first year of testing in Year 3 in 2009, Year 5 in 2011, Year 7 in 2013 and Year 9 in 2015. When reviewing the longitudinal snapshot for this cohort an important caveat is that the genre for the 2009 test was the narrative genre and in the subsequent years, a persuasive genre was chosen. Based on the 10 criteria for both tests, there are four criteria (Audience/Text Structure/Ideas/Character and setting) that differ between the genres. The other six criteria that relate to language and grammar remain the same and therefore, it is arguable that they create a sufficiently robust basis for tracking whether students have improved in these areas.

If we once again turn our focus to the below NMS, there is a decline in the 2009 cohort's results across the years of testing. The most significant statistic is that across all states and the ACT the percentage of students in the below NMS category nearly doubles from Year 7 to Year 9. In 2014 the commentary on the decline in writing results included reported 'prompt issues'1 referring to the suitability of the choice of writing task in the test. Alternatively, the below NMS percentage could reflect the oft reported middle years slump (Lingard et al., 2001) associated with disengagement from school. These possible explanations are worthy of investigation. However, the key point is that if the data are accepted as accurate, then they should serve to trigger inquiry in policy, research and practice: Why is this happening and what catalysts or changes are necessary to prevent further decline for these students fast headed to un/employment (Goss et al., 2015)? Additionally, what action is needed for improving writing outcomes and turning the trend of decline for future cohorts? The following section begins this exploration.

The Australian Curriculum and Writing Standards

The call for a greater focus on the teaching of writing has been acknowledged internationally with the National Commission on Writing (NCoW) in 2003 calling for reform in the United States, arguing that improvement was needed, 'if students are to succeed in college and life' (National Commission on Writing, 2003, p. 7). In 2010, based on the recommendations of the NCoW, the Chief State School Officers released the Common Core State Standards (CCSS). The grade-level standards in CCSS were formulated to 'provide a road map for the writing skills students need to acquire, emphasise the use of writing to support content learning and the comprehension of text, and assess students' mastery of CCSS objectives through the use of written responses'(Graham & Harris, 2015, p. 459).

Prioritising writing for all subjects acknowledges from a policy perspective the critical attention given to writing and 'the importance of teaching it effectively at every grade level' (Graham & Harris, 2015, p. 461). The CCSS 'also provides benchmarks for what students need to master along the way' (p. 459), allowing a more tangible opportunity for teachers and students to see what the standards are at various development stages that will support understanding and engagement with teaching the writing process. A related policy position has been adopted in Australia, with the official stance being that all teachers are responsible for teaching literacy, though the absence of a literacy curriculum, and comprehensive literacy standards and literacy benchmarks tend to undermine the impact of this position. Also at play in the story of Australia's efforts to improve writing is that currently, the connections between standards for subject English and literacy standards are not made clear, including as the latter relate to NAPLAN.

Currently, there is no large scale data on the teaching of writing in Australian classrooms. Similarly, there is no longitudinal data about teachers' pedagogical practices to support all students in their development of writing skills. This gap in knowledge is noteworthy given the oft reported concerns over poor academic writing skills of undergraduate students in academic teacher education programs, recognised in research and more widely (Teacher Education Ministerial Advisory Group, 2014; Louden, 2005). Teacher understanding of language features and their confidence in teaching students writing skills, particularly the language strand of the Australian Curriculum, has been raised by several authors (Harper & Rennie, 2009; Jones & Chen, 2012).

In this curriculum, writing is predominantly addressed through the domain of English. In subject English writing is presented to teachers in three strands: Language, Literacy and Literature. Writing across the other curriculum subjects is woven through domains such as Languages (see Systems of Language, Communicating and Socialising); Mathematics (see Number and Place Value); Digital Technologies (see Processes and Production Skills); Music, Drama, Visual Arts, Humanities and Social Sciences (see Historical Knowledge and Understanding). In the main, the domain of writing is included in the curriculum as part of the 'Content Description' or as a bullet point in the 'Elaborations'. The effect of this is that it leaves writing instruction--the 'how' writing is to be taught--to the teacher.

Writing is a challenging and difficult skill to master. There has been much debate around what is the best course of action to support teachers in the teaching of writing. Applied linguistic scholars including Christie and Martin (2007), Hasan (2005) and Humphrey, Droga & Feez (2012) have pointed to the critical contribution of teachers' 'Knowledge About Language' from the level of the word to the whole text, and the implications this has for teachers' confidence and their ability to assist students in the classroom to achieve acceptable standards of writing. In Unsworth's (2014a, 2014b, 2014c) recent work, this argument has extended beyond print to multi modal texts. Myhill (in Christie & Derewianka, 2008, p. 1) reiterated this concern particularly at the secondary level where there is 'a dearth of systematic exploration of the linguistic characteristics of children's writing'.

Recent research has also challenged how prepared pre-service teachers are to teach language and their confident understanding of how language works in the context of writing (Harper & Rennie, 2009; Jones & Chen, 2012; and Christie & Derewianka, 2008). Love, Sandiford, Macken-Horarik, and Unsworth (2014) have argued that teachers need a new kind of professional knowledge base about language. They discussed the need for a metalanguage or 'grammatics' to support the analysis of the traditional and emerging forms of text. This work builds on the earlier work addressing the literacy demands of curriculum and the related conceptualisation of curriculum literacies (Cumming & Wyatt-Smith, 2001), with implications for assessing achievement in all subject areas (Wyatt-Smith & Cumming, 2003).

As discussed in Love, Macken-Horarik and Horarik's study (2015), teachers have diverse starting points for the teaching of language, making 'a common knowledge base much harder to establish' (p. 173). As suggested earlier, leading language theorists and scholars including Halliday (1979a, 1979b, 1985a, 1985b; Halliday & Hasan, 1985) and Christie (1995) have provided a critical foundation for the widely recognised language-in-use framework to the study of language and specifically, the teaching of writing. Broadly speaking, drawing on the work of Halliday, genre theory focuses on two main approaches: the text type and process model. Depending on the genre or language process required for a writing task, these approaches allow identifiable and recognisable generic, structural and language features to be employed. Genre theory has been influential in the framing of subject English and literacy as one of the seven General Capabilities.

A challenge facing teachers in writing instruction can be traced back to this influence. The acceptance of genres in the Australian Curriculum involves a tacit acceptance of the related underpinning theorising of language and related instructional approaches to teaching writing. While the authors recognise that an official curriculum cannot be expected to comprehensively address how to teach writing, the preceding, discussion of concerning numbers of students falling below benchmark opens the opportunity to reconsider the extent of teacher knowledge about writing instruction in curriculum areas. At issue is the potential benefit of deep professional engagement with pedagogical frameworks and related assessment approaches consistent with the orientation to writing taken in the Australian Curriculum. Vygotsky's (1986) work around the Zone of Proximal Development described the gap between a child's actual development and his or her potential development that can be achieved when assisted. Vygotsky subscribed to the notion of deliberate supported or scaffolded learning under the guidance of a more 'expert' other. This consideration of expertise or connoisseurship, certainly from an assessment perspective, supports the notion of a greater need for a sharpened focus on standards and illustrative exemplars that provide a fine-grained analysis of genre and a common knowledge base of grammar to support teacher understanding of how to teach writing. A developmental focus of what writing standards look like across the years is needed to avoid an appearance of 'robotic posturing of genre'(Exley, Woods & Dooley, 2013, p. 60) in rigid applications of content descriptions and elaborations in the Australian Curriculum: English (AC:E).

In an initial move towards progressing standards the AC: E includes various samples, categorised as 'Above Satisfactory, Satisfactory and Below Satisfactory' with accompanying annotations. These samples go some way towards supporting teachers' knowledge of standards and expected features of quality. The extent of the challenge of developing such knowledge should not be underestimated, however. The Grattan Institute's report (Goss et al., 2015) identified that many Australian teachers 'struggle to accurately interpret curriculum standards and use them to evaluate their students' learning' (p. 15). The report also discussed that teachers' ability to grade against the national A-E Standard is very weak with some provision of examples of teacher's norming student assessment to their own classes rather than external standards (Goss et al., 2015). In light of these observations we identify a critical disconnect between the annotated samples of quality in the Australian Curriculum, the NAPLAN assessment criteria, the A-E reporting standards, and the benchmarks more generally. Teachers are expected to make connections across these three frameworks that have not been formally connected at system levels. These disconnects hamper national efforts to develop with the teaching profession clear understandings of quality and in turn, impact efforts to realise the potential of NAPLAN to secure real improvement.

Conclusion

This paper has presented and discussed NAPLAN writing data within 2015 year cohorts, and explored the pattern of data across Years 3, 5, 7 and 9 for the 2015 Year 9 cohort. The discussion has brought to light, negative change in writing across the years of schooling for students in years 3, 5, 7 and 9 in the 2015 cohorts, and further, that the percentages of students assessed as below benchmark is increasing across these years. It has also brought to light the critical need to focus attention on the connections teachers need to make in assessing writing. These include connections among Australian Curriculum achievement standards, NAPLAN assessment criteria, and mandated five-point scale reporting (for example, A-E).

At the time of writing there are no published Year 9 benchmarks for writing we propose an empirical approach to address this omission. This includes selecting actual student writing samples at the current band level cut scores to develop Year 9 descriptors for the benchmarks. This work would entail undertaking pairwise comparison. Additionally it is now timely to revisit the Years 3, 5, 7 benchmarks once again in reference to marked student scripts. Taken together this work would provide an opportunity for empirical data to clarify benchmark standards and offer a test of reasonableness for both the benchmarks and current cut scores.

We call for a national conversation about what NAPLAN is really delivering for students, parents, the wider community, and for education systems and whether we are utilising the national scale and longitudinal data sets across Years 3-9 efficiently. A key question facing the education community is whether accountability has become so powerful as a political priority that the equity purpose, and indeed, the learning and teaching purposes, have become subsumed by and made subservient to testing for accountability reporting as the primary goal in and of itself. Instead there is a need to go beyond the traditional distinction between measurement and improvement, mentioned in the introduction, to see both in the context of NAPLAN as potentially coming together through a sharpened focus on the meaning of the reports and an alignment with progressing student achievement. This effort would be supported by the development of a distinctive, Australian survey of the teaching of writing. Currently we are developing such a survey to investigate writing instruction in Australian classrooms and also how teachers assess writing using benchmark criteria.

Finally we revisit the visionary MCEETYA priorities for Australian students. Had the vision been realised, the discussion of the data in this paper would have been necessarily different. It is not the case that every child leaving primary school is able to write at an appropriate level and further, many children who commenced school in the period 1998 to the present have not achieved a minimal acceptable level of writing. Indeed, the paper has revealed a concerning picture of accelerating negative change across the years of schooling in writing. However it would be too easy to attribute this change to issues of teacher quality and quality teaching. The actions we have called for reflect our understanding that realising improvement in writing will require a multi-pronged approach as we have outlined. This includes a systematic review of messages about quality and standards available to the profession and more fundamentally, clarity around what benchmarks look like in student work samples.

References

Australian Curriculum, Assessment and Reporting Authority. (2015). NAPLAN Achievement in Reading, Persuasive Writing, Language Conventions and Numeracy: National Report for 2015. Retrieved from http://www.nap.edu.au/ results-and-reports/national-reports.html

Australian Education Council. (1994). A statement on English for Australian schools. Carlton: Curriculum Corporation. Australian Education Council. (1994). Statement & Profile on Mathematics for Australian Schools. Curriculum Corporation, Victoria.

Black, P. (2014). Foreword. In V. Klenowski & C.M. Wyatt-Smith (Eds.), Assessment for Education: Standards, Judgement & Moderation (pp. vii-ix). London: Sage.

Broadfoot, P. (2014). Educational assessment evaluation and research: the selected works of Mary E. James, The Curriculum Journal, 25(3), 461-463.

Christie, F. (1995). The teaching of story writing in the junior secondary school. Report 2 of a Research Study into the pedagogic discourse of secondary Subject English. A study funded by the ARC. Melbourne, Vic.: University of Melbourne.

Christie, F. & Derewianka., B. (2008). School Discourse: Learning to Write Across the Years of Schooling. Continuum Discourse. London: Continuum International Publishing.

Christie, F. & Martin, J.R. (2007). Language, Knowledge and Pedagogy: Functional Linguistic and Sociological Perspectives. London: Continuum.

Comber, B. & Cormack, P. (2013). High stakes literacy tests and local effects in a rural school. Australian Journal of Language and Literacy, 36(2), 78-89.

Comber, B. (2012). Mandated literacy assessment and the reorganisation of teachers' work: federal policy, local effects. Critical Studies in Education, 53 (2), 119-136.

Common Core State Standards. (2016). Retrieved from http:// www.corestandards.org/.

Cope, B. & Kalantzis, M. (1993). The Powers of Literacy: A Genre Approach to Leaching Writing. London: The Falmer Press.

Cope, B., Kalantzis, M., Kress, G., Martin, J. & Murphy, L. (1993). Bibliographical Essay: Developing the Theory and Practice of Genre-based. In B. Cope & M. Kalantzis (Eds.), The Powers of Literacy: A Genre Approach to Leaching Writing Literacy (pp. 231-247). London: The Falmer Press.

Cumming, J.J. & Wyatt-Smith, C.M. (2001). Literacy and the Curriculum: Success in Senior Secondary Schooling. Melbourne: ACER.

Cumming, J.J., Wyatt-Smith, C. & Colbert, P. (2016). Students at risk and NAPLAN: the collateral damage. In Lingard, B., Thompson, G. & Sellar, S. (Eds.), National Lesting in Schools. An Australian assessment. Abingdon, Oxon: Routledge.

Department of Education and Training. (2014). Australian Government response to the Senate Education and Employment References Committee Report Effectiveness of the National Assessment Program--Literacy and Numeracy. Retrieved from https://docs.education.gov. au/system/file s/doc/other/response_to_naplan_senate_ report_0.pdf

Department of Employment, Education, Training and Youth Affairs. (1998a). Literacy for all: The challenge for Australian schools. Commonwealth literacy policies for Australian schools. Canberra: Department of Employment, Education, Training and Youth Affairs.

Department of Employment, Education, Training and Youth Affairs. (1998b). National Report on Schooling in Australia. Retrieved from http://scseec.edu.au/archive/ Publications/Publications-archive/National-Report-onSchooling-in-Australia/ANR-1998.aspx

Diederich, P.B. (1974). Measuring Growth in English. Urbana, Illinois: National Council of Teachers of English.

Earl, L. (2013). Assessment as learning: using classroom assessment to maximize student learning. Thousand Oaks, California: Corwin Press.

Exley, Beryl E., Woods, Annette F. & Dooley, Karen T. (2013). Thinking Critically in the Land of Princesses and Giants: The Affordances and Challenges of Critical Approaches in the Early Years. In Pandya, Jessica & Avila, Julianna (Eds.) Moving Critical Literacies Forward: A New Look at Praxis Across Contexts. Routledge, London, pp. 59-70.

Goss, P., Hunter, J., Romanes, D. & Parsonage, H. (2015). Targeted teaching: How better use of data can improve student learning. Grattan Institute. Retrieved from http:// grattan.edu.au/report/targeted-teaching-how-better-useof-data-can-improve-student-learning/

Graham, S. & Harris, K. (2015). Common Core State Standards and Writing: Introduction to the Special Issue. The Elementary School Journal, IIS (4), 457-463.

Halliday, M.A.K. (1979a). A Grammar for Schools? In M.A.K. Halliday (Ed.), Working Conference on Language in Education: Report to Participants (pp. 184-186). Sydney: Sydney University Extension Programme.

Halliday, M.A.K. (1979b). Differences between Spoken and Written Language: Some Implications for Literacy Teaching. In G. Page, J. Elkins & B. O'Connor (Eds.), Communication through Reading: Proceedings of the Fourth Australian Reading Conference (pp. 37-52). Adelaide: Australian Reading Association.

Halliday, M.A.K. (1985a). An Introduction to Functional Grammar. London: Edward Arnold.

Halliday, M.A.K. (1985b). Spoken and Written Language. Geelong: Deakin University Press.

Halliday, M.A.K. & Hasan, R. (1985). Language, Context and Text: Aspects of Language in a Social Semiotic Perspective. Geelong: Deakin University Press.

Hamp-Lyons, L. (2004). Writing assessment in the world. Assessing Writing, 9(1), 1-3.

Harper, H. & Rennie, J. (2009). 'I had to go and get myself a book on grammar': A study of pre-service teachers' knowledge about language. Australian Journal of Language and Literacy, 32 (1), 22-37.

Hasan, R. (2005). Language, Society and Consciousness. London: Equinox.

Humphrey, S., Droga, R. & Feez, S. (2012). Grammar and Meaning. Newtown, NSW: Primary English Teachers' Association Australia.

Jeffery, J.V. (2009). Constructs of writing proficiency in US state and national writing assessments: Exploring variability. Assessing Writing, 14(1), 3-24.

Jones, P. & Chen, H. (2012). Teachers' knowledge about language: Issues of pedagogy and expertise. Australian Journal of Language and Literacy, 35 (1), 147-168.

Klenowski, V. & Wyatt-Smith, C.M. (2012). The impact of high stakes testing on learning: The Australian story. Assessment in Education: Principles, Policy & Practice, 19(1), 65-79.

Lingard, B., Ladwig, J., Mills, M., Bahr, M., Chant, D. & Warry, M. (2001). The Queensland school reform longitudinal study. Brisbane: The State of Queensland (Department of Education).

Lingard, B., Thompson, G & Sellar, S. (2016). National Testing in Schools. An Australian assessment. Abingdon, Oxon: Routledge.

Louden, W., Rohl, E., Gore, J., Mcintosh, A., Greaves, D., Wright, R., Siemon, D., & House, H. (2005). Prepared to Teach: An investigation into the preparation of teachers to teach literacy and numeracy. Mount Lawley, Western Australia: Edith Cowan University.

Love, K. & Humphrey, S. (2012). A multi-level language toolkit for the Australian curriculum: English. Australian Journal of Language and Literacy, 35(2), 173-191.

Love, K., Macken-Horarik, M. & Horarik, S. (2015). Language knowledge and its application: A snapshot of Australian teachers' views. Australian Journal of Language and Literacy, 38(3), 171-182.

Love, K., Sandiford, C., Macken-Horarik, M. & Unsworth, L. (2014). From 'Bored Witless' to 'Rhetorical Nous': Teacher Orientation to Knowledge about Language and Strengthening Student Persuasive Writing. English in Australia, 49(3), 43-56. Retrieved from http://search. eb scoho st. com/login, a spx?direct=true&db=eric&AN =EJ1051302&site=ehost-live\nhttp://www.aate.org.au/ journals/2 014-vol-49-no-l-to-3

Ministerial Council on Education, Employment, Training and Youth Affairs. (MCEETYA). (1989). The Hobart declaration on schooling. Retrieved from http://www. scseec.edu.au/EC-Publications/EC-Publications-archive/ EC-The-Hobart-Declaration-on-Schooling-1989.aspx Ministerial Council on Education, Employment and Youth Affairs. (MCEETYA). (1999). The Adelaide Declaration on National Goals for Schooling in the 21st Century. Retrieved from http://www.scseec.edu.au/archive/Publications/Publications-archive/The-Adelaide-Declaration.aspx National Commission on Writing. (2003). The neglected R: The need for a writing revolution. New York: College Board. Retrieved from http://www.collegeboard.com/ prod_downloads/writingcom/neglectedr.pdf Sadler, D.R. (1989). Formative assessment and the design of instructional systems. Instructional Science, 18(2), 119-144.

Teacher Education Ministerial Advisory Group (TEMAG). (2014). Action Now: Classroom Ready Teachers (pp. 1-118).Department of Education and Training, issuing body. Retrieved from http://nla.gov.au/nla.arc-150576 Thomson, S., Hillman, K. & De Bortoli, L. (2013). A Teacher's Guide to PISA Reading Literacy. Melbourne: ACER. Unsworth, L. (2014a). Point of view in picture books and animated film adaptations: Informing critical multimodal comprehension and composition pedagogy. In E. Djonov & S. Zhao (Eds.), Critical multimodal studies of popular culture (pp. 201-216). London: Routledge.

Unsworth, L. (2014b). Interfacing Visual and Verbal Narrative Art in Paper and Digital Media: Recontextualizing Literature and Literacies. In G. Barton (Ed.), Literacy in the Arts: Retheorising Learning and Teaching (pp. 55-76). New York. NY: Springer.

Unsworth, L. (2014c). Multimodal Reading Comprehension: Curriculum Expectations and Large-scale Literacy Testing Practices. Pedagogies: An International Journal, 9(1), 26-44.

Vygotsky, L.S. (1986). Thought and Language (A. Kozulin, Ed & Trans.). Cambridge, Massachusetts: MIT Press. White, L. (2013). NAPLAN under fire. Education Review. Retrieved from http://www.educationreview.com.au/2013/ 02/naplan-under-fire/Wilson, B. (2000). Foreword. In Curriculum Corporation, Literacy Benchmarks Years 3, S & 7 Writing, Spelling and Reading (p. iii). Victoria: Curriculum Corporation.

Wyatt-Smith, C.M. (1998). Interrogating the benchmarks. English in Australia, 123, 20-28.

Wyatt-Smith, C.M. & Cumming, J.J. (2003). Curriculum literacies: Expanding domains of assessment. Assessment in Education: Principles, policy and practice, 10(1), 47-59.

Wyatt-Smith, C.M. and Klenowski, V. (2014). Elements of Better Assessment for the Improvement of Learning. A Focus on Quality, Professional Judgement and Social Moderation, in C. Wyatt-Smith, V. Klenowski and P. Colbert (eds), The Enabling Power of Assessment: Quality, Ethics and Equity. Dordrecht, The Netherlands: Springer International.

Note

(1.) The 2015 writing assessment was criticised for the selection of the writing question. The media referred to the concern as 'prompt issues', http://www.theaustralian.com.au/ national-affairs/education/marked-down-how-one-toughquestion-skewed-the-naplan-results/news-story/48274cc8 75b227cf39b0ea3255454802?=

Claire Wyatt-Smith and Christine Jackson

Australian Catholic University

Claire Wyatt-Smith is a Professor of Educational Assessment and Literacy and Institute Director of the Learning Sciences Institute Australia. Her research focuses on professional judgment, standards and moderation, with an aligned focus on curriculum and literacy education. Her publications address teachers' assessment identities; large-scale standardised testing and its impact on learning; assessment adaptations for students with disabilities and assessment and new technologies. Claire's research has attracted funding from the Australian Research Council and she has undertaken numerous government-funded large-scale longitudinal projects.

Christine Jackson works for the Learning Sciences Institute Australia, ACU. She has worked in schools and the education sector for over 15 years. She has taught English in Australia and overseas and has worked in policy and project management roles for the Government and Independent Sectors. Christine has worked as an Assessment and Reporting Manager in all aspects of test development and the reporting of writing assessments. Christine is currently undertaking postgraduate research in literacy assessment and educational measurement.

Table 1. 2015 NAPLAN writing results--percentage of students by state,
territory and Australia in each 'standard'

State         Below National Minimum Standard (%)
            Year 3   Year 5      Year 7      Year 9
            Band 1   Band 3      Band 4      Band 5
                     and below   and below   and below

NSW         2.0      4.7         10.9        18.9
VIC         0.7      2.4         6,9         12.2
QLD         3.4      8.0         13.3        20.8
WA          4.2      7.7         12.7        15.8
SA          3.8      8.1         10.5        19.6
TAS         2.8      7.9         13.6        20.3
ACT         2.1      4.2         7.7         13.9
NT          25.7     38.4        46.0        52.6
AUSTRALIA   2.6      5.9         11.0        17.7

State       At National Minimum Standard {%)
            Year 3   Year 5   Year 7   Year 9
            Band 2   Band 4   Band 5   Band 6

NSW         4.1      11.0     18.5     21.9
VIC         2.6      8.2      16.0     20.2
QLD         6.7      14.5     19.5     23.1
WA          5.6      12.8     18.0     19.2
SA          6.8      15.2     18.8     22.5
TAS         6.4      13.7     20.3     23.7
ACT         4.2      10.1     17.0     19.3
NT          12.2     15.4     16.1     16.1
AUSTRALIA   4.8      11.7     18.1     21.5

State       Above National Minimum Standard (%)
            Year 3      Year 5      Year 7      Year 9
            Band 3-6+   Band 5-8+   Band 6-9+   Band 7-10

NSW         92.3        82.6        69.2        57.7
VIC         93.9        86.5        74.9        65.1
QLD         88.4        76.1        65.6        54.6
WA          89          78.3        68          63.8
SA          87.2        74.5        68.8        55.9
TAS         89          77          64.6        54.7
ACT         91.7        83.7        73.5        64.2
NT          60.3        44.2        35.5        29.1
AUSTRALIA   90.7        80.6        69.2        59

Table 2. Comparison of 2015 NAPLAN reading and writing: percentages of
students by state, territory and Australia in each 'standard'

State               Below National Minimum Standard (%)
                    Year 3   Year 5   Year 7   Year 9
                    Band 1   Band 3   Band 4   Band 5

NSW                 3.0      4.6      2.6      5.9
NSW-Writing         2.0      4.7      10.9     18.9
Vic                 1.7      2.8      1.9      4.0
Vic-Writing         0.7      2.4      6.9      12.2
Qld                 3.6      4.9      2.6      6.7
Qld-Writing         3.4      8.0      13.3     Z0.8
WA                  5.8      6.9      4.0      5.6
WA-Writing          4.2      7.7      12.7     15.8
SA                  4.3      6.1      3.1      6.3
SA-Writing          3.8      8.1      10.5     19.6
Tas                 5.4      6.5      3.9      7.3
Tas-Writing         2.8      7.9      13.6     20.3
ACT                 2.8      2.8      1.5      3.4
ACT-Writing         2.1      4.2      7.7      13.9
NT-READING          27.0     30.3     25.2     31.6
NT-WRITING          25.7     38.4     46.0     52.6
AUSTRALIA-READING   3.6      4.9      2.9      5.9
AUSTRALIA-WRITING   2.6      5.9      11.0     17.7

State               At National Minimum Standard {%)
                    Year 3   Year 5   Year 7   Year 9
                    Band 2   Band 4   Band 5   Band 6

NSW                 7.1      13.3     12.5     17.3
NSW-Writing         4.1      11.1     18.5     21.9
Vic                 5.3      10.8     10.7     15.4
Vic-Writing         2.6      8.2      16.0     20.2
Qld                 8.4      14.0     13.0     19.5
Qld-Writing         6.7      14.5     19.5     23.1
WA                  9.3      14.8     13.7     15.2
WA-Writing          5.6      12.8     18.0     19.2
SA                  8.4      15.1     12.9     18.2
SA-Writing          6.8      15.2     18.8     22.5
Tas                 9.4      16.0     14.9     19.3
Tas-Writing         6.4      13.7     20.3     23.7
ACT                 5.9      9.4      8.4      11.8
ACT-Writing         4.2      10.1     17       19.3
NT-READING          14.3     18.1     19.5     19.0
NT-WRITING          12.2     15.4     16.1     16.1
AUSTRALIA-READING   7.4      13.2     12.4     17.1
AUSTRALIA-WRITING   4.8      11.7     18.1     21.5

State               Above National Minimum Standard (%)
                    Year 3      Year 5      Year 7      Year 9
                    Band 3-6+   Band 5-8+   Band 6-9+   Band 7-10

NSW                 88.2        80.5        83.4        75.3
NSW-Writing         92.3        82.6        69.2        57.7
Vic                 90.2        83.5        85.2        78.1
Vic-Writing         93.9        86.5        74.9        65.1
Qld                 86.6        79.8        82.8        72.3
Qld-Writing         88.4        76.1        65.6        54.6
WA                  83.7        77.1        81          78
WA-Writing          89          78.3        68          63.8
SA                  85          76.6        82.1        73.4
SA-Writing          87.2        74.5        68.8        55.9
Tas                 83.5        76.1        79.8        72.1
Tas-Writing         89          77          64.6        54.7
ACT                 89.3        85.8        88.4        82.2
ACT-Writing         91.7        83.7        73.5        64.2
NT-READING          56.8        49.6        52.9        47.2
NT-WRITING          60.3        44.2        35.5        29.1
AUSTRALIA-READING   87.2        80.1        83          75.2
AUSTRALIA-WRITING   90.7        80.6        69.2        59

Table 3. Longitudinal Tracking of 2009-2015 cohort Year 3, 5, 7, 9:
percentages of students by state, territory and Australia in each
standard'

State       Below National Minimum Standard (%)
            Year 3   Year 5   Year 7                      Year 9
            Band 1   Band 3   Band 4                      Band 5
            2009     2011     and below                   and below
                              2013                        2015

NSW         1.5      3.5      9.4         [right arrow]   18.9
VIC         0.7      3.0      6.6         [right arrow]   12.2
QLD         4.2      8.2      9.5         [right arrow]   20.8
WA          3.6      8.2      8.8         [right arrow]   15.8
SA          2.3      7.7      8.8         [right arrow]   19.6
TAS         2.2      8.4      12.0        [right arrow]   20.3
ACT         1.2      3.4      6.9         [right arrow]   13.9
NT          24.3     36.2     41.9        [right arrow]   52.6
AUSTRALIA   2.4      5.6      9.1         [right arrow]   17.7

State       At National Minimum Standard (%)
            Year 3   Year 5   Year 7   Year 9
            Band 2   Band 4   Band 5   Band 6
            2009     2011     2013     2015

NSW         5.0      8.8      17.9     21.9
VIC         3.8      10.1     15.6     20.2
QLD         9.7      14.9     17.9     23.1
WA          7.8      14.2     16.8     19.2
SA          6.9      15.6     17.3     22.5
TAS         7.6      17.9     20.1     23.7
ACT         4.9      10.6     14.8     19.3
NT          13.1     15.2     16.8     16.1
AUSTRALIA   6.3      11.8     17.1     21.5

State       Above National Minimum Standard (%)
            Year 3      Year 5      Year 7      Year 9
            Band 3-6+   Band 5-8+   Band 6-9+   Band 7-10

NSW         92.2        86.2        71.3        57.7
VIC         92.5        84.3        75.7        65.1
QLD         84.2        75.3        70.9        54.6
WA          87.3        76.3        73.1        63.8
SA          89.1        74.7        72.3        55.9
TAS         88.9        72.4        66.6        54.7
ACT         91          83.1        75.9        64.2
NT          60.9        46.5        38.8        29.1
AUSTRALIA   89.4        80.7        72.2        59