When assessment defines the content— understanding goals in between teachers and policy

Education policy development internationally reflect a widespread expansion of learning outcome orientation in policy, curricula and assessment. In this paper, teachers’ perceptions about their work are explored, as goals and assessment play a more prominent role driven by the introduction of a learning outcomes-oriented system. This is investigated through interviews of Norwegian teachers and extensive policy analysis of Norwegian policy documents. The findings indicate that the teachers are finding ways to negotiate and adjust to the language in the policies investigated in this study. Furthermore, the findings show that the teachers have developed their professional language according to the policies. The teachers referred to their self-made criteria and goal sheets as central tools in explicating what is to be learned. In many ways, the tools for assessment, thus determine the content of education as well as what is valued in the educational system. in a featuring are teachers’ perceptions of the relation between nation-ally defined goals and their practices of assessment framed by a learning outcomes-oriented curriculum?


Introduction
In this paper, we investigate Norwegian educational policy regarding learning outcomes and assessment in schools through teachers' perspectives. Illuminating the perspectives of teachers is important since education policies conceptualise education in ways that influence what teachers do and how they perceive themselves (Ball, 2003). Education policy development internationally reflects a widespread expansion of learning outcome orientation in policy, curricula and assessment (Shepard, 2000(Shepard, , 2007Kellaghan & Greaney, 2001). Research has illuminated how these developments have changed the perceptions of quality (Kellaghan & Greaney, 2001;Adam, 2004) and shifted the focus from input indicators to outcome indicators (Fuller, 2009;Young, 2009). Researchers have argued that the scope of interpretation, action and evaluation in the teaching profession have been reformulated and constricted (Young, 2009;Forsberg & Pettersson, 2014).  argued that curriculum policy has defined teachers as 'agents of change', as also seen in Norway  where prescriptive curriculum has been replaced by an enhanced focus on teacher agency. Mausethagen and Smeby (2015) summed up trends in research showing that younger teachers are more accepting of a stronger emphasis on performance and seem to balance accountability and autonomy better than their older counterparts. In a study by Wilkins (2010), for example younger teachers reported that they are aware of the conflicts between the demand for accountability and the desire for autonomy but are generally comfortable with the balance between them.  argued that policy rhetoric in newer curricula place the teacher at the forefront of curriculum development, implying a return to teacher autonomy in curriculum making. They further asserted that curricula focussing on dialogic pedagogies, active learning, individualisation in learning and learner autonomy enhance the professional role of teachers. This advocates for that the teachers to produce modern citizens who are well equipped to thrive in a knowledge society as members of a globalised workforce .
In countries with a strong central state, such as Norway, the state actively constructs and redefines concepts and words to create political consensus (e.g. in the form of a national curriculum). The concepts and words used in the curriculum often express a consensus about education at the overall policy level, while questioning and challenging this consensus when the curriculum is to be operationalised in the local school and classroom context. In contemporary society, where texts are spread rapidly and digitally, the state governs through stable concepts as well as short-lived, high-profile buzz words such as learning outcomes, competency goals, learning culture, learning environment; and assessment for, of, and by learning (Stenersen & Prøitz, 2020).
The curriculum structures the educational course and pedagogical interests to ensure a high-quality learning environment for children based on the goals, content and evaluation of schooling (Lundgren, 2006). Conceptually, a national learning outcomes-oriented policy expressed in the form of curriculum documents and accompanying regulations and guidelines serves as a set of policy instruments for the governance of education and as a pedagogical platform for professional practice (Ewell, 2005;Aasen, 2007). Learning outcomes-oriented education systems have predominantly been associated with Anglophone countries with stronger curriculum and assessment traditions. Today, the focus on learning outcomes has spread to Nordic countries and continental Europe, challenging long-standing didactic traditions in education (Hopmann, 2015;Mølstad & Prøitz, 2019, Prøitz & Nordin, 2019. A growing demand for evidence in decision-making and the subsequent need for assessment and data have likewise driven the development of the knowledge society (Lundahl & Waldow, 2009;Ozga et al., 2011). The development can be seen as a response to a globalised world, an economy in which new technologies have changed production, and a society characterised by heterogeneity in cultures and beliefs. These shifts have led to changes in power structures, which in turn influence how education is governed (Lundgren, 2006).
Viewing curricula as representative of basic principles of cultural and social reproduction, there is a need to consider the current changes and critically scrutinise how curricula are formed and function today (Lundgren, 2006). Curriculum control is complicated, and curricula reform may involve mixes of different approaches for steering and controlling education (Sundberg & Wahlström, 2012). In this paper, we explore teachers' perceptions about their work in a learning outcomes-oriented system featuring a more prominent role of goals and assessment. Accordingly, we ask the following question: What are teachers' perceptions of the relation between nationally defined goals and their practices of assessment framed by a learning outcomesoriented curriculum?

International and national drivers towards learning outcomes orientation
Researchers have interpreted global education developments during the last 30 to 40 years as shifts in ideology, in perceptions of quality, and in focus, changing from inputs to outcomes. In Anglophone countries, a major shift in political ideas from issues of equality to excellence, accountability and choice in education has been occurring since the late 1970s (Kellaghan & Greaney, 2001;Fowler, 2012). In the United States, the authorities adjusted their focus from engineering optimal mixes of school inputs to regulating outputs in the 1980s, followed by a rising accountability script emphasising performance and the crafting of the No Child Left Behind (NCLB) Act in 2001 (Fuller, 2009). Researchers have described this development as a shift in perceptions of quality in education, moving from an input-oriented approach towards an outcomes-oriented approach (Kellaghan & Greaney, 2001;Adam, 2004). These developments can be seen in relation to the discovery of tools for measuring the attainment of defined learning outcomes within the institutional effectiveness movement of the 1920s and the assessment movement of the 1980s, which featured government calls to examine the effectiveness of funding of public institutions of education in the United States (Ewell, 2005;Shepard, 2007). By the 1990s, this approach developed further as systems for institutional and programmatic quality assurance and accreditation (Ewell, 2005).
Evaluation, assessment or control of education is not new, but the capacity of national systems to observe the whole field and make comparisons between data  on learning outcomes is new. A particular feature of the European development has been the increased involvement of the Organisation for Economic Co-operation and Development (OECD) in educational policy since the 1990s; in particular, PISA has evolved into an important tool for justifying change or providing support for chosen policy directions (Hopmann, 2008;Pettersson, 2008;Simola et al., 2011). The impact of PISA on national policies varies; for example, England has been less 'marked' due to its long-term investment in high-stakes testing and its sophisticated system of data production and use . Norway has been more strongly 'marked' by PISA, perhaps due to a tradition of focussing on inputs and processes rather than measuring results and outcomes (Prøitz, 2014).
The developments in Norway must be seen in relation to drivers of education policy during the last 20 years as well as certain particularities of the Norwegian system prioritizing aspects like process and inputs as well as results and outcomes within a hybrid and softer model of governing by results (Prøitz, 2014;Tveit, 2019). The government's educational reforms in the early 2000s increasingly focussed on learning outcomes. With the introduction of the Quality Reform in Higher Education in 2003 and the Knowledge Promotion reform in compulsory and upper secondary education and training in 2006, the Norwegian term laeringsutbytte (learning outcomes) became widely used to cover a variety of aspects in education (Prøitz, 2015). Since 2005 priority has been given to the development of a national qualification's framework for lifelong learning with reference to the European qualification's framework for recognition of qualifications. However, there has also been a long tradition of scepticism towards formal assessment in Norway (Lysne, 2006). The past decade of innovations and practices in educational assessment in Norway has been described as evoking ideological disputes over educational assessment (Tveit, 2014), indicating that issues of learning outcomes and assessment have not been taken lightly. The Norwegian school system has included few traditions for working with predefined learning outcomes as opposed to the Anglo-American tradition (Hatch, 2013). The practice of assessing student performance according to predefined goals and standards is relatively new (Telhaug, 2005;Engelsen & Smith, 2010;Skedsmo, 2011).
The more results-oriented approach within Norwegian education can be linked to a 1988 OECD report recommending a stronger focus on result measurement and the need for a quality assurance system in education (Hatch, 2013;Møller & Skedsmo, 2013;Prøitz, 2015;Pettersson et al., 2017). Despite several governmental efforts to create such a system, none emerged until the first introduction of a national test in 2004 and the Knowledge Promotion reform of 2006, among other things legitimised by evaluations of the education reforms of the 1990s (Haug, 2003) and 'below average PISA results' of 2001 (Hatch, 2013;Møller & Skedsmo, 2013;Prøitz, 2014). The 2006 reform introduced an outcomes-oriented education policy including elements such as a national outcomes-oriented curriculum, national tests, a national quality-assessment system, new regulations for assessment, and a stronger emphasis on decentralisation, governing by 'competency goals', and local accountability (Aasen et al., 2012). In the last 10-15 years, education reform has placed stronger emphasis on assessment, partly by introducing externally induced national tests and a new curriculum with defined learning outcomes, but primarily by revisions of the regulations for assessment emphasizing students' rights to 'formative assessment' and 'assessment for learning'. The current national curriculum does not directly regulate teachers' classroom assessment practices, provide any criteria, or explicate the relationships between goals in the national curriculum and assessment. Teachers' assessment work is rather guided by students' rights defined in the regulations and the Education act, the national examination system, and guidance and support material provided by the national authorities. As such the Norwegian model of assessment has been characterised as a softer approach to a learning outcomes-oriented education system (Hatch, 2013;Prøitz, 2015;Tveit, 2019).
In the following, we describe the curriculum context of Norway before presenting the theoretical framework of the study. The curriculum context provides a broader understanding of curricular development in Norway and the movement towards a more elaborate learning outcomes orientation in the national curriculum. Next, we explain the methods and analysis, followed by a presentation of the findings and a concluding discussion.

Curriculum context in Norway
Norwegian state governance of education through the national curriculum is a longstanding tradition (Telhaug, 2005), although several curricular reforms illustrate a shift in the focus and measures of government over time. The national curriculum from 1974, known as Mønsterplan 74, was written for the individual teacher and recognised the teacher's authority to make judgements concerning the teaching content. Traditionally, the teacher determined the instructional methods, but this reform introduced the idea of the teacher deciding which content to emphasise. The next curriculum of 1987, Mønsterplan 87, emphasised the responsibility of locallevel educators (Engelsen, 2009). From 1990 to 1995, the influence of educational experts was reduced in favour of a strong political centre, resulting in a shift of power from the professional to the political arena. This change established the idea of a stronger state and combined central management with social democratic confidence in the state (Telhaug, 2005). During the 1990s, most of the educational system was changed through several educational reforms. The 2006 Knowledge Promotion reform brought a range of new elements into Norwegian education and began a stronger learning outcomes-oriented education policy. The main intention of the reform was to transfer more power to the local level, with a significant share of action and responsibility allocated to local school authorities and schools (Aasen et al., 2012).
The current curriculum reform (LK06) is highly oriented towards learning outcomes through its competence goals, which define what the students are to learn however also indicate content to be taught (Mølstad & Karseth, 2016). In line with this emphasis on outcomes, national testing of basic skills was introduced in 2004. While the authorities do not publish league tables, the media announce their national rankings of schools on an annual basis (Tveit, 2014). In recent years, policies emphasising student outcomes and external control of teachers' work have challenged the trust in and autonomy of teachers (Mausethagen, 2013). More recent policy studies have portrayed a complex picture of policy expectations towards teachers as change agents of education and curriculum reform , where they are to be both controlled and autonomous, indicating that the policies shift between different parameters of expectations regarding the roles of teachers and teaching. These studies also identify yet another belief about what can be accomplished through aligned learning outcomes as an 'alignment between teachers' competences and practices and students' future prospects and life opportunities (Mølstad & Prøitz, 2019).
Like many nations, Nordic countries have introduced various public management approaches in education that emphasise the combined power of performance measurement, goal setting, and a growing number of regulations and guidelines to increase accountability in mobilising teachers' efforts and improving student achievement (Fuller, 2008;Mintrop, 2012;Prøitz & Nordin, 2019). Several Nordic studies have investigated the effects of these relatively new ways of governing on teachers' experiences and practices (Mausethagen, 2013;Prøitz, 2014;Mølstad, 2015). Although learning outcomes-oriented education and curriculum reform prioritise the student and student learning, few studies have investigated how teachers and students are framed and defined as the main object of change and the central target actor group for curriculum reform (Prøitz et al., 2016. In understanding state-based curriculum-making, it is essential to examine the teaching profession in the specific country (Mølstad, 2015). Historically, Norwegian teacher education has not been associated with research institutions such as universities (Skagen, 2006). Skagen (2006) noted that the content of teacher education has been controlled by the government and criticised for its lack of research-based teaching.
In a case study of Norwegian and Swedish teachers' autonomy, Helgøy and Homme (2007) pointed out that Norwegian teachers are characterised by old professionalism and that the strong input regulations in Norway limit the individual teacher's autonomy. However, the authors concluded that, at a collective level, teachers are still autonomous and that national standards and control in education are accepted as tools for securing professional knowledge and status (Helgøy & Homme, 2007). Additionally, Mausethagen (2013) found that Norwegian teachers' work is influenced by accountability policies (e.g. national testing) despite the low-stakes context of the country. More recent research on this theme has offered various explanations for this observation, including the ways in which relations to parents and peers influence teachers (Camphuijsen et al., 2019;Wermke & Prøitz, 2019). For example, teachers emphasise the parental perspective in discussions about student performance data, and teachers and school leaders refer more to their knowledge about students and parents than to research and information from data and statistics in their interpretation of national test results (Wermke & Prøitz, 2019). In summary, the Norwegian teacher profession has been portrayed as having old professionalism and autonomy at the collective level while also being affected by the accountability movements introduced by the 2006 curriculum reform. More recently, researchers have also identified parents and peers as important influencers.

Theoretical framework
The way in which a curriculum is conceptualised and defined by researchers depends on the theories they employ and the context of their analysis (Sivesind, 2008). Hence, the distinctiveness of a national educational system and the geographical identities of researchers are important to understand the meaning of a curriculum as well as what it includes and excludes. Although recent research has shown that curriculum reform worldwide seems to follow common general ideas emphasising learning outcomes and assessment (Meyer, 2007), institutional differences and distinctive national cultures result in varying frames of reference in the field of curriculum-making.
Research has identified a shift from content-oriented models of education to learning outcomes as indications of competence. The central and northern European curricula have traditionally been dominated by the core concept of Didaktik, which is defined as the art or study of teaching (Gundem & Hopmann 1998). Hopmann and Riquarts (2000) explained,

since the days of Comenius and Ratke, Didaktik has been the most important tool for planning, enacting, and thinking about teaching in most of northern and central Europe. Indeed, it is impossible to understand German, Nordic and central European schooling without appreciating the role and impact of Didaktik. (p. 3)
An important aspect of Didaktik is the primacy of the content, illustrated by its position on top of the well-known Didaktik triangle, with the teacher on the bottom left and the learner on the right (see e.g. Hopmann, 2007). Simply stated, the Didaktik model deals with the following three questions: (1) What is to be taught? (2) How is content to be taught and learned? (3) Why is the content to be taught and learned? While Didaktik addresses all of these questions, the first and third dominate the model (Künzli, 2000). Didaktik implies that in the prescribed curriculum content is essential for teaching and then why it is to be taught. As such, teachers are supposed to first and foremost focus on the content of the curriculum.
Furthermore, the position of the content is underpinned by the way 'a didactician looks for a prospective object for learning and asks himself what this object can and should signify to the student and how the student can experience this significance' (Künzli, 1998, pp. 39-40). This position implies that the significance of learning content is the most important aspect of education. Künzli has argued that all other questions and problems, such as class management or individual and social learning, are subordinate. Content is not defined as a fixed body of knowledge to be learned but rather as a range of possibilities to be explored through reasoning and interactions (Karseth & Sivesind, 2010). Moreover, the curricula based on Didaktik have been oriented towards overall purposes and subject content rather than objectives. The objective-driven model is designed to develop specific capabilities directly connected to the needs of society (Westbury, 1998;Ross, 2000).
Whereas content and purpose are the priority in the Didaktik model, objectives and expected learning outcomes are the cornerstones of an objective-driven curriculum model (Ross, 2000). Outcomes, as stated above, are learning results that students are expected to demonstrate at the end of significant learning experiences; moreover, desired learning outcomes are expressed in terms that clearly illustrate how students' achievement can be measured. When outcome descriptions or behavioural objectives are prioritised, content is primarily seen as a means to achieve the outcomes (Andrich, 2002). This approach assumes a direct and often linear relationship between objectives and learning activities and performance. However, it is not clear how learning outcomes are to be understood, as they often communicate the results of educational efforts while establishing the premises for the resulting learning activities-the processes. In short, distinct traditions underpin different understandings and uses of learning outcomes, leading to debate over whether they are mainly limited to expressions of performance or if they include broader educational process goals (Prøitz, 2010(Prøitz, , 2014. Some practitioners have identified the lack of conceptual clarity as a problem (cf. Kennedy et al. 2009), while others have warned against definitive and narrow learning outcomes, as they may limit the interpretive space needed in education practice (Lundgren, 2006). Studies on learning outcomes have shown that there is one established and dominant understanding of learning outcomes that is usually employed in policy and management (Hargreaves & Moore, 2000;Hopmann, 2008;Ozga, 2009; Understanding goals in between teachers and policy 297 2014). This involves understanding learning outcomes as the end product of education. Kennedy et al. (2007) defined a learning outcome as 'a written statement of what the successful learner is expected to be able to know, understand and or be able to demonstrate after a completion of a process of learning' (p. 5). Some scholars have disagreed, suggesting other definitions and perceptions, often with an emphasis on the more process-oriented aspects of working with learning outcomes (Allan, 1996;Hargreaves & Moore, 2000;Eisner, 2005;James & Brown, 2005;Prøitz, 2010Prøitz, , 2014. Researchers on outcomes in higher education have also highlighted how learning outcomes might be useful in individual teaching and build towards something that is assessable, while the learning outcomes themselves are not necessarily suited to assessment (Jackson, 2000;Entwistle, 2005;Hussey & Smith, 2008). Hussey and Smith (2008), in particular, have critiqued the application of learning outcomes at a systemic level by noting that 'the further away from students and the teacher in a classroom, the more remote, generalized and irrelevant statements of learning outcomes become ' (p. 114). Several studies have shown how different understandings of learning outcomes relate to different conceptualisations of learning and what is valued and validated as learning in education as well as the consequences for relevant assessment approaches.
In sum, based on the theoretical perspectives presented here, we apply the understanding of different types of curricula combined with perspectives on autonomy and agency to investigate how assessment and goals are constructed in between policy and practice.

Methods
The study draws on a combined set of data on comprehensive content analysis (Bowen, 2009;Cohen et al., 2011) of key policy documents and interviews with 12 teachers in three lower secondary schools in three municipalities in Norway, conducted as part of the larger research project Tracing Learning Outcomes Across Policy and Practice (LOaPP; Prøitz et al., 2016).
This paper presents a further corroboration of patterns identified in an overview content analysis of key policy documents framing education reform and development in Norway in the reform intensive period of 2010-2016 (Prøitz & Mølstad, 2017). The content analysis was based on simple word counts as a point of departure for further analysis. Prøitz and Mølstad focussed on 53 central concepts considered characteristic of the shift in Norway. These concepts included learning outcomes, competence, knowledge, learning, result, goal, curriculum, assessment, evaluation, decentralisation, locally and accountability. For methodological reasons, relevant synonyms, antonyms and competing concepts (e.g. process, didactics, freedom, autonomy, bildung, critical reflection, centralisation and government, guidelines, regulation and inspection) were also included. The analysis resulted in an overview of the frequencies of concepts used in the key policy documents, enabling the identification of clusters of high-frequency concepts (more than 600 incidences), medium-frequency concepts (200-600 incidences), low-frequency concepts (less than 200 incidences) and concepts with no incidences. Another observation based on the word count was that the high-frequency concepts were clearly separated from the lower-frequency groups by having substantially higher frequencies (most were above 1000 incidences); as such, they stood out as definitional concepts and were thus, interpreted as policy-defining concepts. Based on incidence count six key concepts were identified: goal (1785), competency (1668), knowledge (1594), learning (1250), development (1645), and support (989) (see Diagram 1). A central observation relevant for this paper is how the topics of teachers, teaching and teachers' work were addressed with considerable frequency in relation to the identified policy-defining concepts. This finding led to focussing this paper on expressed policy expectations directed towards teachers and teaching for further corroboration and analysis. It also formed the basis for the primary focus of this paper on how the policy-defining and highest frequency concepts goal and competency are understood by teachers and how the teachers work with these reform aspects in curriculum planning and teaching.
Interestingly concepts that we might consider highly relevant in the context of a learning outcomes-oriented curriculum reform (e.g. learning outcomes, assessment, test, grade, criteria, evaluation and summative assessment) were in the medium-, low-, and no-frequencies clusters. We could interpret this finding as policy avoiding difficult and potentially contradictory concepts within the Norwegian context traditionally characterised by scepticism towards assessment (Lysne, 2006). Alternatively, it could illustrate classic policy behaviour of defining visions and giving direction at an overall level by the use of broader concepts while leaving the complexities of assessment (and evaluation) to practitioners (Hargreaves et al., 2002).
This study is qualitative and as such does not aim for generalisation to a larger population of teachers, schools, or municipalities. The study does offer analytical generalisations by providing transparency and theoretical interpretations involving a reasoned judgement about the extent to which its results can be used as a guide for predicting what might happen in similar contexts (Kvale & Brinkmann, 2009 Understanding goals in between teachers and policy 299

Interview material
The selection of teachers and municipalities for this study can be characterised as purposeful and, to a certain extent, of maximum variation, with the purpose of 'documenting unique or diverse variations that have emerged in adapting to different conditions, and to identify important common patterns that cut across variations' (Palinkas et al., 2015, p. 534).
The 12 Norwegian language and science teachers who were interviewed work in three lower secondary schools geographically located in three different municipalities. The selected schools vary in terms of location (urban to rural), ethnicity (from multi to homogenous student group), student and teacher numbers (from high to low), different local governing systems and focus on assessment. However, the schools are similar in that they are all public schools that score average on national tests and examinations. Highly relevant for this study, they are also similarly governed by national regulations and a national curriculum.
The teachers were recruited by the research team during a 6-month period of ethnographic field work drawing on a set of field reports from each school. Norwegian teachers are responsible for teaching clusters of school subjects such as Norwegian language, English, and social science or science and math. Teachers in lower secondary school mostly follow one class from Grade 8 through Grade 10. As a result, they assess their own students over the 3-year period and set the final grades for the students' diplomas at the end of Grade 10. By the end of Grade 10, all Norwegian students undertake one written examination and one oral examination. The written examination can be in Norwegian language, math or English. The oral examinations can be in varied subjects such as science, social science, religion, Norwegian or English. Students and teachers do not know in advance which examinations the students will take. That determination is made annually based on principles of randomisation by education authorities (nationally for the written examinations and locally for the oral examination). This assessment system relies heavily on teachers' capacities to interpret the curriculum and assess and grade their students accordingly. Hence, how teachers understand the national curriculum and assessment becomes highly important. Teachers teaching the largest subjects in terms of hours (Norwegian and math) are often head teacher with overall responsibilities for the wellbeing of the individual students as well as the general learning environment of the class over the three years of lower secondary education.
The teachers interviewed were experienced teachers of varied age and genders. The majority of the Norwegian language teachers were head teachers. All teachers were working in teams with other teachers either defined by the school subject or defined by the grade level. At the time of the interview, the teachers in schools East and West were teaching in Grade 10 while the teachers in school North were teaching in Grade 9. However, all teachers interviewed had experiences from or were teaching at more than one grade level, giving them substantial knowledge and experience from teaching at all grade levels in several school subjects in lower secondary school (see Table 1 for overview of the data material).
From the beginning of the study, all the participants were informed about the aims of the research project and details of their involvement, and they all provided formal Understanding goals in between teachers and policy 301 voluntary consent. The interviews were conducted by two experienced researchers by the end of the foregoing ethnographic study to prepare for further studies of the teachers' classroom practices of video recordings. The interviews were conducted in a suitable room at the teachers' schools using a semi-structured interview guide developed for the LOaPP project that covered a range of questions and themes (see Appendix 1). In this study, we focus on the part of the interviews that cover the themes on how teachers work with curriculum planning, teaching and assessment, how they select themes for teaching, if and what kind of assessment they prefer to use, and why and how. The interviews were audio-recorded after receiving participant consent, and the audio recordings were transcribed verbatim. The respondents were anonymised in the transcripts and in the analysis of the material.

Document material
The document analysis builds directly on the findings of the comprehensive content analysis of key Norwegian policy documents (for a detailed method description, see Prøitz & Mølstad, 2017;Mølstad & Prøitz, 2019). The initial content analysis involved a simple word count of a priori and theory-based categories and concepts derived from research on the latest education reform period in Norway. The approach led to an overview of a substantial number of documents in a systematic fashion and the identification of trends and patterns within them (Weber, 1990). This made it possible to make inferences that could then be corroborated using other methods of data collection as well as further in-depth content analysis (Stemler & Bebell, 1998). In total, the material covered about 2600 pages published from 2010 to 2016. The purpose of choosing this period was to investigate the development of the comprehensive national education reform (KL 2006), which was extensively documented in the period 2005-2012. In accordance with the selected focus of this paper, we zoomed in on three central policy documents for a further in-depth analysis with especially high relevance for teachers' work in schools among the total document corpus of the LOaPP project 1 (1) the subject plan of Norwegian language, (2) the subject plan of science, and (3) the core curriculum.

Choice of focus and analytical approach
Interpreting text involves searching for underlying themes in the material being analysed (Bryman, 2012). The analysis in this paper was conducted to investigate which themes were covered most often and the way concepts were used to illuminate these topics. We systematically included extracts of texts and transcripts related to the research question connected to secondary schooling, focussing on teachers' understanding of goals and competency in curriculum and assessment work. The definitional and high-frequency policy concepts (goal and competence) of the total document corpus guided the identification of extracts in both documents and the interview transcripts. All extracts included in this article were translated by the authors. The next step of analysis was an in-depth exploration of the transcribed interview material, seeking to understand how the teachers perceive and reflect on their practice in relation to curriculum work. Following this, we conducted another round of document analysis and exploration based on the topic illuminated by the teachers in the interviews. This enabled a combined analysis of the interview and document material, where the analysis of the first informed the analysis of the other.

Findings
The teachers were concerned with the relation between assessments, goals and planning lessons to ensure learning for the students, which is also an important topic in the policies. We have chosen to present the findings based on which topics the teachers focussed upon the most. In the following, we highlight the topic of curriculum, the purpose of assessment, and the relation between pupils and assessment in the final planning process.

Understanding the national curriculum
In the interviews, we asked the teachers how they prepare and plan for their teaching. We also asked them how they relate to the national curriculum and how they work with assessment. As the national curriculum is a central guidance point for the teachers, an example from the Grade 10 science curriculum (p. 14) is provided here. The example also illustrates how the learning outcomes in the national curriculum are presented in so-called competency goals: The goal of the education is that the student shall be able to: • Explain the main elements of the evolution theory and explain observations that supports the theory • Describe the structure of animal and plant cells and explain the main features of photosynthesis and cell respiration. (p. 14) Another example of competence goals can be found in the Grade 10 Norwegian curriculum: The goal of the education is that the student shall be able to: • Orient themselves in large amounts of text on screen and paper to find, combine and evaluate relevant information in working with the subject

• Write different types of texts by pattern of sample texts and other sources. (p. 15)
These examples illustrate the type of learning outcomes that the teachers are to use, including both the content and a description of student understanding. The learning outcomes also indicate the methods teachers need to employ to ensure the development of the targeted skills.
In the interviews, the teachers reflected on these questions and explained how they use the competence goals in their planning. One teacher explained how they used the competency goals of the national curriculum to increase the students' awareness of the goals and the topic:

Understanding goals in between teachers and policy 303
The teachers also reflected on the process of getting accustomed to using competence goals in their planning of teaching, in contrast to previous practices and earlier reforms without defined learning outcomes: In these quotes, the teachers expressed that they had an existing process for using the national curriculum, and it has taken time to adjust to this reform. They also highlighted that they have developed a particular way of using the curriculum in their planning of teaching. They elaborated on how they had changed their planning of teaching in relation to the current curriculum reform by increasingly focussing on competence goals. This underscores the role of the curriculum in teaching planning and shows that the teachers' planning and teaching practices have changed with the curriculum reform.
Several teachers addressed the issue of making assessment criteria, and one stated, 'The challenge is to make specifically good enough assessment criteria so that they [the students] know what will be looked for, and that is okay for me to use' (Teacher B). The teachers said that the students become aware of what will be assessed due to their use of assessment criteria. Furthermore, some teachers also explained that they do not necessarily use or discuss the curriculum with the pupils; instead, they use the assessment criteria as a way to present the curriculum.
It is interesting that the teachers reflected on assessment and criteria when asked about the curriculum in the interviews, as the word 'assessment' is found, for example, only seven (three relevant for this study) times in the science curriculum and three (one relevant for this study) times in the Norwegian subject curriculum. This focus is illustrated in the following quote regarding the science curriculum (p. 3): In science teaching, science appears both as a product of current knowledge and as processes that concern how science knowledge is built and established. The processes include hypotheses, experiments, systematic observations, discussions, critical assessment, argumentation, arguments for conclusion, and dissemination. (Teacher C) The word 'assessment' in this quote is used in relation to a description of what the science teaching is to accomplish in terms of the content and teaching methods and how this should contribute to students' learning what comprises science knowledge. Here, assessment is not about assessing students; rather, it is a method used in science that the students should know about. The relevant quote from the Norwegian subject plan refers in general terms to how assessment is described in the Education Act and its regulations for assessment. This conceptualisation of assessment underscores the fact that, while assessment is not a central part of the national curriculum in Norway, it is thoroughly covered in the Education Act and regulations.

The purpose of assessment
In the interviews, we asked questions about what types of assessments the teachers used and how they worked with them. Overall, the teachers spoke about assessment as a dynamic process; however, they also expressed a practical perspective: 'I do not think there is one form [of assessment] that is best suited, but I think one has to choose written assessment since it takes the least time' (Teacher E). Another teacher said, 'I think that science is not a subject of writing, but it is good to have one written test per semester, but also other assessments such as reports or writing a subject article, which I think are good but exhausting assessment forms' (Teacher G). This latter perspective highlights that teachers use student assignments for conducting assessments despite how they view the subject characteristics. This indicates that teachers give assessment priority over the characteristics of the subject based on pragmatic choices related to learning outcomes and the need to assess student performance, as further elaborated later in the discussion. The same teacher further elaborated on this issue: 'Parents and pupils complain about grades. So, then it is important to have the documentation in order' (Teacher G). This quotation illuminates two different purposes for assessment: helping the students to learn and providing documentation for the parents. This highlights how the teachers work with assessment partly as an element of teaching and learning and partly to secure documentation for grading. While the teacher referred to the need for documentation for students and parents, assessing documents for grading is also a requirement based on national regulations for assessment and a central element in the national education inspection system (Fylkesmannens tilsyn).
Furthermore, the teachers also referred to student peer assessment as a way of allowing students to understand where they are in their learning trajectory: 'They often assess each other. I really believe in that' (Teacher F). The teachers frequently described peer assessment as a part of how they work with students' understandings of where they are and what they need to work on:

So, then they write down what they have to work on and how they plan to do it, and I go around and read it and discuss with them. So, then we have an agreement for how they should reach their goal. (Teacher A)
The teachers reported that they used peer assessment to help students understand how to reach their goals.
Similar purposes for assessment can be found in the policy documents, as in the following excerpt from the Core curriculum/Prinsipper for opplaeringen: 'Assessment and guidance shall contribute to strengthen their [the students'] motivation for learning' (p. 3). This is further elaborated a few pages later: 'The school shall contribute to the students' reflections concerning their own learning, understanding their learning processes and gain knowledge independently' (Core curriculum/Overordnet del, p. 12).

Students' relation to assessment and goals
The teachers reflected on the appropriateness of using goals with the students. Specifically, they indicated that the curriculum goals may not be appropriate to all grade levels: 'In Grade 10, we discuss the curriculum goals a lot and show them. In Grades 8 and 9, they have intermediate goals (delmål)'. They argued that intermediate goals are necessary since the goals of the national curriculum are too advanced for the students. As such, the teachers indicated that the curriculum goals are not appropriate directly in teaching in Grades 8 and 9, but they still described how they use them in their teaching in Grade 10.
This way of describing practices related to varied types of goals is interesting. It could mean that in the first 2 years of lower secondary school teachers interpret and adapt national goals to make them understandable for their students before applying them more directly in their planning, teaching, and assessment in the final year. Grade 10 is a high-stakes year for the students in terms of final grades for the diploma being set and for the teachers due to the national and final examinations. These descriptions of practices indicate a temporal element in how the teachers adapt their work across the three years of lower secondary schooling according to student age and maturity as well as the proximity of the goals of the national curriculum in the final year and the examination system. The analysis displays how the teachers to a growing extent focus on and make use of national goals directly in their teaching and in dialogues with students as they progress.
The schools in this study use varied types of tools for clarifying goals and criteria for students. One example is the 'goal sheet' (målark). This document provides goals for learning coupled with information about the planned lessons, the learning activities and criteria for assessment. Goal sheets were commonly used by the teachers we interviewed, and they elaborated on their use in planning teaching, learning activities, and assessment: 'We make goal sheets for the pupils where both learning goals…., it is first knowledge goals and then learning goals and then the central concepts-and how they are assessed and the assessment criteria' (Teacher B). The goal sheets are used regularly in teaching, and pupils will ask for them if they are not provided (see Figure 1 for an example of a goal sheet).
The top part of this goal sheet provides the students with the competence goals from the national curriculum. The next part lists the intermediate learning goals that the teachers have defined to cover the competence goals. This is followed by a description of how the students are to show that they have gained the described competences. The last section is a list of concepts relevant for the topic in question. The teachers indicated that students usually comment on being asked about things outside the frame of the goal sheet. As such, the goal sheets can be understood as a sort of agreement that delimits and defines what is to be learned and assessed between the students and the teacher.
The teachers focussed on how they help the pupils understand what is going to be assessed: 'We usually go through what they will be assessed on since it is written on the goal sheet. They use that to prepare for tests: what am I really going to be assessed on?' (Teacher I). However, the informants also problematised the use of goal The same teacher continued, 'It becomes an outcry if there is something on a test which is not on the goal sheet: Because if it is not on the goal sheet I do not need to know it'.
Other teachers were more critical about constructing learning goals based on the competence goals in the national curriculum. The following informant problematised this in the following way:  Understanding goals in between teachers and policy 307 The national curriculum also highlights the aspect of working with goals in relation to the students: 'Education shall encourage students by for example making the goals clear for them and to provide varied and goal-oriented activities' (Core curriculum/ Prinsipper for opplaeringen, s 6). At the policy level, as this quote illustrates, it is necessary to have a process involving the competence goals before they are used in teaching. Furthermore, the national curriculum also highlights the important role of assessment: The school and the teachers have to balance the need for good information about the pupils' learning and undesirable consequences of different assessment situations. Unfortunate use of assessment can weaken individual confidence and hinder the development of a good learning environment. (Core curriculum/Overordnet del,p. 17) Summing up this section, we find that both policy and teachers' reflections demonstrate an understanding of the importance of working with the national curriculum to make it understandable for the students, but to varying degrees. Even so, in the interviews and documents, the teachers reported making use of defined learning outcomes from the national curriculum directly in their teaching when final grading and national examinations are approaching in the final year. This finding is supported by analysis of classroom video observations from the same schools and with the same teachers in the LOaPP project (Semb & Prøitz, 2019;Silseth). In this process, the combined use of goals and assessment is highly important.

Planning process
The Core curriculum highlights the importance of the teachers' competence in education planning: 'The teachers' and instructors' competences contain several components where professional expertise, the ability to organize learning work and knowledge about assessment…' (Prinsipper for opplaeringen, p. 9).
The teachers highlighted that the way they work with the competence goals is connected and dependent on the specific subjects. The teachers also described a process of collective work between teachers within the same subjects in developing learning goals based on the competence goals in the national curriculum. They spend time working with the competence goals to make them more concrete before presenting them to the students: The competence goals are broken down to learning goals. We make the language easier… on the bottom of the goal sheet are the central concepts, which we pick up in the lessons, so that the learning goals are not meaningless. (Teacher G) The teachers also expressed a belief that the competence goals are too broad and provide room for interpretation, which makes it impossible to use them in lessons or for planning. One teacher discussed the development of goals that are not too complicated for the pupils: This quotation illuminates that teachers prioritise the process of developing goals for the students so that they can understand what to learn and what is to be assessed, while the contents of the school subjects are expected to be developed throughout the process. This generally highlights that the teachers placed greater importance on clarifying what is to be learned by clarifying goals and using assessment tools rather than reflecting on how they work with the curriculum in their own teaching.

Discussion
The analysis of this paper indicates that teachers are finding ways to negotiate and adjust to the language in the policies investigated in this study. Furthermore, the findings show that the teachers have developed their professional language according to and further beyond the policies. In the following, we discuss how assessment and goals are constructed in between teachers and policy.
Curricula in Norway have been based on the Didaktik model, in which the planning of teaching starts with the question of what is to be taught, implying an emphasis on the content. Conversely, the latest curriculum has been designed on an outcomes-based model emphasising the competency of students in terms of what students know and has learned. With this in mind, it is interesting to discuss our findings regarding what the teachers and policies focus on. The teachers reported that they spend time during both lesson planning and in the lessons to clarify, elaborate and explain the national and intermediate teacher-defined goals to the students. This focus on goals, which gives direction to teaching, is supported in and required by the policy documents, indicating that the teachers focus on the national goals in educating students. Hence, the focus on goals by teachers and policy in this study strongly resembles and is more in line with an outcomes-oriented model for education than with the Didaktik model. This is particularly underscored by one teacher, who said that subject content is established through the process of working with goals.
In many ways, it is hard to research change in teachers' practices based on curriculum reforms. In this study, we cannot claim that we are directly documenting changes in teachers' practices, but we have documented how teachers themselves report having changed their planning and their teaching practices. Accordingly, we argue that, based on the historical context provided combined with the investigation of policy and teachers' descriptions of their work, we can shed some light on a possible change in practice and hence the professional language of the teachers. The teachers described this change, highlighting the process of adjusting to a curriculum with learning outcomes defined as competency goals. The changing role of textbooks is especially interesting. One of the teachers noted that they used to start with the textbook and work through it, but now the competence goals define which parts of the books are to be used in teaching. This indicates a change in the teachers' focus from what is to be taught to what is to be learned.

Understanding goals in between teachers and policy 309
Another interesting observation from the material is how strongly the teachers prioritised assessment when they were asked about the curriculum and goals. It is well documented, although also heavily debated in the literature, that outcomes-oriented education requires clearly defined assessment structures (Shepard, 2000;Eisner, 2005). In the Norwegian reform, this was initially not provided by the national curriculum, and to a large degree the question of assessment was left to the teachers to solve (Prøitz, 2010). Since then, several revisions of the national regulations for assessment have been made, and the government has developed guidelines for assessment supplemented by a range of tools and criteria for assessment. Still, most policy messages on classroom assessment concern guidance for how to develop good assessment practices rather than requiring certain practices of regulating classroom conduct. Regulations typically define assessment for learning as a student right to be secured by teachers' practices rather than define what teachers must do or what goal sheets or criteria to use. Consequently, the teachers in this study referred to their self-made criteria and goal sheets as central tools in explicating what is to be learned. In many ways, the tools for assessment thus determine the content of and what is valued in education. As such, this study highlights a case where teachers have been encouraged to develop elements such as their own assessment criteria and goal sheets. One can thus argue that the Norwegian teachers have been ensured room for teacher agency within the learning outcomes frame. Further this study indicates that the teachers have been able to use their professional agency to develop their practices, which implies that the teachers are at the forefront as curriculum developers, as others also have found (Priestley & Minty, 2012).
There is a question regarding whether the direct employment of learning outcomes from the national curriculum and the usage of assessment tools can be interpreted as teacher agency and an example of teacher autonomy manifesting as new professionalism in curriculum work framed by the new learning outcome-oriented national curriculum in Norway, as proposed by Priestly and colleagues (2012). The findings of this study imply that teachers make use of national goals directly in classroom situations when the final examinations approach, while they interpret and provide students with intermediate goals earlier in the education trajectory. This could indicate that teachers act without autonomy and with restricted agency following their students' path towards final grading and examination defined by the curriculum and the examination system. However, another interpretation could be that teachers regard the goals as rather open for interpretation and develop varied teaching and learning activities with tools to support student learning through more process-oriented approaches to assessment until the end point of Grade 10. The choice of action seen in this study could be interpreted as teacher agency where the professionals manoeuvre between governmental requirements and controls (e.g. national examination) and their own professional beliefs in planning and teaching combined with the requirements of the subject content as well as student and classroom characteristics.
The way in which the teachers problematise their assessment tools and adapt and revise their planning and reported practices based on students' reactions is also interesting. According to the teachers, goals are helpful in guiding learning, but they can also be troublesome when students become obsessed with them, resulting in a narrowing of the curriculum and the process of content building. These concerns have been raised in the outcomes literature repeatedly for a long period of time (Hussey & Smith, 2003Eisner, 2005;James & Brown, 2005;Young, 2009), and as such our study empirically confirms previous research on learning outcomes. What might be even more interesting in this case would be to examine how the teachers have interpreted and adapted to a national curriculum reform as well as their concerns about students' understanding of defined goals and the parents' perspective when it comes to assessment. This could involve students and parents more closely in the curriculum work of teachers and thus raise other issues of autonomy and accountability in education besides curriculum that are more directly related to central state governance.

Funding
The article is funded by the Norwegian Research Council. The article is a part of the research project Tracing Learning Outcomes through Policy and Practice (LOaPP), grant number: 536560. NOTE