A relational perspective on students' experiences of participation in an ‘Interest-based physical education’ programme

While many students participate autonomously in physical education (PE), research shows that students who do not participate in leisure sport are less likely to perceive PE positively. Attempting to optimise the reciprocal student (↔) PE relationship and secure equal opportunities for learning in PE, schools in a Norwegian county developed an ‘Interest-based PE’ programme offering students a choice of two PE approaches: a sports approach (SA), focused on sports activities, and an explorative approach (EA), focused on alternative movement activities. Based on a process-relational understanding of adolescent development and learning, this study seeks a deeper understanding of changes in the PE experiences of students in the programme. Sixteen students (ages 17–18 years) who had participated in Interest-based PE for 18 months participated in qualitative semi-structured one-on-one interviews where they reflected on their relationship with PE prior to and during the programme. Data were subject to inductive interpretive thematic analysis showing that ‘the role of sports in PE’ framed student ↔ PE relations and that this sports discourse regulated the relations, also within the Interest-based PE programme. The separation of students into an EA and a SA accentuated the sports discourse and students’ sports competencies, contributing to segregation on the basis of students' confidence, competence and ability in sports. Based on this study, we question the assumption that differentiation programmes, such as Interest-based PE, will optimise student ↔ subject relations if these relations remain governed by the sports discourse, rather than the PE curriculum.


Introduction
School represents a major developmental asset in young people's lives (Eccles and Roeser, 2009) and equity in education is a pillar in the educational systems of many countries, including Norway (Utdanningsdirektoratet [Norwegian Directorate for Education and Training], 2015). However, not all teachers feel convinced that their students benefit equally from today's physical education (PE) and research suggests that sports-active students are more appreciative of PE than their peers who do not participate in leisure sport (Erdvik et al., 2019a;Kjønniksen et al., 2009;Koka and Hein, 2003;Moen et al., 2018;Prochaska et al., 2003;Säfvenbom et al., 2015;Viira and Koka, 2012), and that these students 'reap most of the benefits' from the subject (Säfvenbom et al., 2015: 629). This research indicates a sports discourse in PEa discourse that contradicts the subject's purpose and contributes to the importance of sporting skills in students' ability constructions (Aasland et al., 2019;Kirk, 2010;López-Pastor et al., 2013). Disparity between the sports discourse and current social and political expectations for PE in schools is a central source of critique in today's PE research not only in Norway (Aasland and Engelsrud, 2017;Aasland et al., 2016;Erdvik, 2020;Säfvenbom, 2010;Säfvenbom et al., 2015;Solesnes, 2010) but also internationally (Kirk, 2010;Tinning, 2010).
While teachers may not necessarily be familiar with the academic debate concerning the role of the sports discourse in shaping contemporary PE, many teachers appear to notice that their sportsactive students are more likely to benefit in their PE class. In an attempt to level the learning field for all students, PE teachers at nine schools in Norway implemented a didactic differentiation programme in PE, known as 'Interest-based PE' (Tangen and Husebye, 2019). Through this programme, students were provided with the opportunity to choose between two different approaches to learning in PE: a sports approach (SA) which offered students traditional sporting activities and ballgames, and an explorative approach (EA) which offered students less sportscentred and more alternative and playful activities. The programme was developed on the basis of the national curriculum in PE (Utdanningsdirektoratet, 2015) and sought to optimise the relations between students and the PE context, and thus to spark positive development and learning among more students (Tangen and Husebye, 2019).
While prior research has applied quantitative methods to assess the effectiveness of Interest-based PE in terms of achieving developmental aims (Erdvik et al., 2019b), we lack important knowledge about students' broader and deeper experiences of the programme (e.g. Tangen and Husebye, 2019). To narrow this gap, this study interviewed students about their experiences of the Interest-based PE programme. More specifically, the research question pursued in this paper was: 'how did the Interest-based PE programme intervene in the relations between the students and the PE subject? ' To be able to present and discuss our work we will first provide a brief overview of the relational developmental systems meta-theory (RDS; Lerner, 2018;Overton, 2015) that is the theoretical framework of this study. RDS offers an infrequently applied process-relational perspective to the understanding of students' development and learning in PE. In the next section, we then provide a short outline of PE in Norway, before we give a brief overview of research on how students express their relationship with the subject. Finally, we present the Interest-based PE programme and our analysis, which prompted a new question about what had triggered the teachers' need for didactic differentiation in the first place. In the final part of the paper, we discuss the unintended and problematic consequences of Interest-based PE.

Theoretical framework
The present study is anchored in RDS and thus a process-relational paradigm (Lerner, 2018;Overton, 2015), acknowledging that human development and learning cannot be understood without focusing on developmental processes, interaction and thus relational analysis (Lerner, 2018). From an RDS perspective, human development and learning are 'reciprocal, understood as resulting from mutually influential reciprocal person ↔ context relations within a certain culture and time of history' (Säfvenbom et al., 2018(Säfvenbom et al., : 1992. Because students' development and learning in PE are situated in context and time, a relational approach incorporating contextual and historical perspectives is considered advantageous. Such a perspective considers the plasticity in both the student and the context as significant for optimising learning and developmental processes. Yet, according to the RDS perspective even plasticity should be considered a relational phenomenon. Relative plasticity in the student ↔ PE relationship is not only considered in the context of student narratives but also in relation to the subject's history and origin. RDS offers this perspective to the study of students' PE experiences (Lerner, 2015(Lerner, , 2018Lerner et al., 2011) by analysing the relations between (a) current individual student characteristics; (b) current contextual specificities of PE; and (c) historical or developmental aspects relating to both the student and the PE subject. The historical perspective is important not only because students' PE narration shapes their current understanding of the subject, but also because the PE subject's past may influence current approaches to PE. As such, a PE class represents the relationship between students, who may have years of experience with the subject, and the PE context that has been part of the educational system for decades.
From an RDS perspective (Lerner, 1991(Lerner, , 2018, change in the bidirectional (↔) relation between the students and the subject (as offered by the teacher) is considered the basic process of both the students' development and learning in PE and the development of PE as a school subject. Change in the relationship between a student and PE can occur if the student adapts to expectations in the subject, based on the teacher's interpretation of the subject (e.g. master a volleyball serve), or if the teacher adapts PE to meet the student's needs or competencies. In adopting the RDS perspective in PE, researchers focus on the rules, or the developmental regulations (e.g. Brandtstädter, 2006), which govern the exchange between the students and the subject. In PE, this exchange should be governed by the epistemic objects as stated in the PE curriculum. That said, prior research suggests that teachers find it hard to anchor their teaching in the PE curriculum (Redelius et al., 2009(Redelius et al., , 2015, and that they may act according to alternative rules of regulation, such as the promotion of fitness or health (e.g. Walseth et al., 2017;Webb et al., 2008).
As long as the relations between the individual student and the PE context are characterised by adaptive developmental regulations that benefit both the students and the subject (e.g. the PE teacher), RDS maintains that all students may experience positive movement experiences (and thus positive development and learning) in PE (Lerner, 2018). Optimising student ↔ subject relations to ensure students' access to positive movement experiences in PE is important, as this could result in long-term physical, psychological and social benefits, which may encourage continued participation and learning, and thus more positive movement experiences for the student (Agans et al., 2013). However, achieving such optimisation requires not only students' effort to adapt to the context of PE but also a PE subject that accepts diversity among studentsa PE that allows students to flourish and adapts to support diverse students in their attempts to thrive (Lerner et al., 2008).

Developmental regulations in (Norwegian) PE
Historically, a number of different regulations have governed the relationship between students and PE in Norway. From 1848, when the subject was introduced for the purpose of preparing young men for military service (Augestad, 2003), it would take nearly 100 years for PE to become a mandatory subject for all. Since then, the subject's focus shifted to health promotion, underscoring the importance of students' hygiene (Augestad, 2003) and physical fitness (Aasland et al., 2016). In later years, as the sports movement gained increased importance in Norwegian society, sport activities became an increasingly central element of PE (Aasland et al., 2016;Augestad, 2003).
Today, PE is governed by the Norwegian Education Act (Opplaeringslova [the Norwegian Education Act], 1998) and the national curriculum (Utdanningsdirektoratet, 2015). Through 13 years of mandatory education, PE is expected to promote students' self-worth and inspire lifelong enjoyment of physical activity (Utdanningsdirektoratet, 2015). To achieve these purposes, students are supposed to work towards specific curricular competency aims. In broad terms, these aims emphasise fair play and collaboration, bodily learning, self-management and implementation, as well as competence and understanding. The national curriculum in PE does not emphasise specific standards of student achievement or students' relative development of sport competence (Utdanningsdirektoratet, 2015).
While the competency aims of the national curriculum are relatively clear, researchers have argued that students' general ability to achieve the purposes of PE may also depend on the quality of students' PE experiences (e.g. Kjønniksen et al., 2009;Säfvenbom et al., 2015). Even though several studies suggest that PE is appreciated by the majority of adolescents in Norway (Moen et al., 2018;Säfvenbom et al., 2015), a representative study by Säfvenbom et al. (2015) shows that as many as 44% of the PE students disapprove of its current approach or would like the subject to be taught differently. This aligns with international research showing that students who do not thrive in PE feel alienated (Carlson, 1995;Spencer-Cavaliere and Rintoul, 2012), experience a lack of choice (Flintoff and Scraton, 2001) and do not consider the subject to be personally meaningful (Carlson, 1995;Spencer-Cavaliere and Rintoul, 2012). Performativity culture and male dominance have been identified as threats to students' thriving in PE (Allender et al., 2006;Beltrán-Carrillo et al., 2012), and students who are not involved in leisure sports have been identified as less likely to experience the subject positively (Erdvik et al., 2019a;Kjønniksen et al., 2009;Koka and Hein, 2003;Moen et al., 2018;Prochaska et al., 2003;Säfvenbom et al., 2015;Viira and Koka, 2012). While most would agree that sports are an important part of PE (Annerstedt, 2008), Kirk (2010: 41) has argued that PE is informed by the physical culture of sports in ways that have led to the institutionalisation and reproduction of a PE practice referred to as 'physical education as sport techniques'. Researchers have argued that this approach to PE does not promote positive development and learning equally among adolescents, and that the subject should be more sensitive to adolescent diversity (Erdvik et al., 2019a(Erdvik et al., , 2019b.

Interest-based PE as a means to optimise student ↔ PE relations
The idea that optimising person ↔ context fit enables individuals to become more active producers of their own positive development is essential to RDS. One approach to optimise the relational fit between students and PE is to adjust the PE context via the provision of student choice. Several studies have shown that students' opportunity to choose in the context of PE is associated with more positive PE experiences, as indicative of positive student ↔ subject fit. For instance, Mitchell et al. (2015) found that female PE students' ability to choose activities, and thus choose whether they would (or would not) like to have a sports-centred PE, promoted positive PE experiences and increased their participation in the subject. Increased enjoyment and engagement in the subject as a result of activity choice has also been identified in the research of Smith et al. (2009), Condon and Collier (2002), and Lagestad (2017), and the belief in this association was also integral to Interest-based PE.
In our study, Interest-based PE was a local didactical differentiation programme centred on the students' choice between two interest-based approaches to PE, both of which were framed by, and practiced in accordance with, the current national PE curriculum in Norway (LK-06; Utdanningsdirektoratet, 2015). Interest-based PE represented the teachers' attempt to optimise the relationship between the individual student and PE to spark positive development and learning in more students (Tangen and Husebye, 2019). Processes of optimisation were believed to occur as students were allowed to choose between two different interest based approaches to PE, known as EA and SA. It is important to note that the competence aims in the curriculum remained unchanged and that only the approach to help students achieve these competence aims differed in these two PE approaches. On the one hand, SA offered participation in traditional sporting activities and ballgames, and allowed the students to play different sports according to traditional rules, techniques and logics of these sports. On the other hand, students who chose EA were offered a less sports-centred and more explorative and playful approach to movement activity. Teachers who taught EA would for instance use more games (e.g. 'tag', 'red light green light' and 'hunter hawks') or modified sports (e.g. use multiple balls, play while attached to a fellow student, change emphasis from competing, winning and making goals to cooperating, contributing and making each other better) in their classes to encourage meaningful participation for a group of students that was not necessarily inspired by the logic of sports (for more information, see Erdvik et al., 2019b andTangen andHusebye, 2019).

Sampling procedures
Interest-based PE was implemented in four lower and four upper secondary schools in eastern Norway. Participants in our qualitative study were final year students (ages 17 and 18 years) at one school that was randomly drawn from the four upper secondary schools. At this school, Interest-based PE was taught by four PE teachers: two of them taught EA together, and the other two taught SA together. Throughout the study, each class was taught by the same teachers. The number of students per teacher did not change with Interest-based PE, and the number of teachers per class reflected the teachers' decision to implement EA and SA in relatively large classes 1 . At the time of the interviews, students had been involved in Interest-based PE for one-and-a-half years. With an additional 11 years of experience with 'traditional PE' (i.e. PE before the introduction of Interest-based PE), students were assumed to be familiar with, and able to reflect upon, their experiences of both traditional and Interest-based PE.
At the one randomly drawn upper secondary school, a total of 83 students were informed about the purpose of our study at an oral information meeting where they were also provided with written information letters. At the meeting, students were invited to volunteer for one-on-one interviews by adding their names to a list, which would assist subsequent quota sampling (Robinson, 2014). During the information meeting, nine students from SA (three boys and six girls) and 11 students from EA (no boys and 11 girls) signed up for interviews. The fact that fewer boys volunteered from EA reflected the fact that most of the boys chose SA rather than EA: in the Interest-based PE study as a whole, 103 boys and 65 girls chose SA, whereas 52 boys and 128 girls chose EA (see Erdvik et al., 2019b). At the school where the qualitative study took place, 22 boys and 13 girls chose SA, whereas five boys and 43 girls chose EA. However, the representation of both genders from each PE approach was considered important, as boys and girls typically experience PE somewhat differently (Cairney et al., 2012;Säfvenbom et al., 2015). As no boys from EA had volunteered for interviews, a second announcement was organised by the teachers in class, increasing the total number of volunteers to eight boys and eight girls from SA, and two boys and 15 girls from EA. After having performed interviews with 16 students who had been randomly drawn from these four pools of participants, the data collection had reached the point of saturation. The final sample thus included participants of both genders from both EA and SA, who were students at the same school. More specifically, this study was based on interviews with six girls and two boys from EA, and four girls and four boys from SA.

Interview procedure
All interviews were conducted by the first author. The interviews took place in a suitable room at the school during school hours, and lasted between 67 and 112 min with the majority of interviews lasting 85 min or more. The semi-structured interview guide, which had been tested in two pilot interviews prior to this study, referred to three main topics that were guided by major questions as well as possible follow-up questions about: (a) PE in general (e.g. 'can you describe a typical PE lesson?', 'how do you experience such a lesson?'), (b) Interest-based PE (e.g. 'how did you feel that PE did or did not change after you started with the explorative/sports approach?'), and (c) the purpose of PE (e.g. 'what do you learn in PE?') 2 .

Thematic analysis
The six phases of thematic analysis proposed by Braun and Clarke (2006) were used to conduct an inductive interpretive thematic analysis. Tape-recorded interviews were transcribed verbatim by the first author, and re-read (Phase 1) before initial line-by-line coding procedures (Phase 2). The qualitative data analysis software NVivo was used to systemise the research material and to support the coding process. The inductive coding process was characterised by the authors' attempt to remain open to the data and not, at this stage, limit coding to preconceived concepts and theory, accompanied by the acknowledgement that 'data are not coded in an epistemological vacuum' (Braun and Clarke, 2006: 84). The line-by-line codes were structured into higher order codes, which reflected students' experiences of 'change' in PE, as well as their emphasis on the importance of being 'good', 'skilled', 'suited', 'positive' or 'engaged' in the subject. All codes appeared to be sufficiently elaborated after the interview with the 16th student, and the last four interviews did not result in additional codes, indicating that the data collection had reached the point of saturation (Phase 3; Fusch and Ness, 2015). The descriptive nature of the codes made the generation and interpretation of themes at this stage difficult; a second level of interpretation was necessary to provide a deeper understanding of students' experiences of Interest-based PE. As such, all interviews were re-read and subject to manual focused coding (Phase 4). The focused codes were then re-studied in relation to initial codes to ensure that the essence of the interviews was maintained, and to facilitate the identification of major themes in the data material. During this procedure, students' experiences of Interest-based PE were identified by two subthemes: 'it hasn't changed that much' and 'you get to be with people at your own level'. The content of both subthemes reflected a common, major theme: 'the role of sports in PE'.
As such, understanding 'the role of sports in PE' appeared essential for achieving a deeper understanding of students' experiences of Interest-based PE (Phase 5). After the completion of the analyses, this inductively derived theme was substantiated by student quotes that formed the basis for the formal write-up of research findings, which were then presented and discussed from an RDS perspective (Phase 6).

Trustworthiness
The Interest-based PE programme was based on a group of educators' experiences from their own teaching. It was developed in a county of Norway as a local didactic programme by ordinary PE teachers in collaboration with PE teacher education lecturers from a local university college. The programme represents one of many local, experience-based, didactic PE actions performed every year in many schools nationwide and worldwide. What these didactic programmes often have in common is a desire among staff to improve the PE subject, to reach specific groups of students or to make PE a better place for all. Like other local, experience-based didactical projects developed in the everyday life of teachers, Interest-based PE had limited access to scientific expertise and financial funding, and was thus not developed according to the standards of a randomized controlled trial or as a classical intervention study. Nonetheless, because these programmes are developed by PE teachers in the context of their everyday lives, they have high ecological validity (Bronfenbrenner, 1979;Schmuckler, 2001).
The authors of this paper were introduced to the 'Interest-based PE' programme just before it started up in the fall of 2014 and did not influence its content or implementation. The interviews for this study were performed when students had participated in Interest-based PE for 18 months. During interviews and analysis, we made efforts to self-disclose potential preconceptions, and looked for disconfirming evidence when working with the data material in particular (Brantlinger et al., 2005). Being part of a research group also allowed for the discussion of analysis and interpretations with research group members and co-authors in a way that strengthened the trustworthiness of the study. In addition to discussions within the research group, findings in this paper have been discussed in other settings pertaining to PE teachers and colleagues working within PE teacher education in Norway (see Brantlinger et al., 2005). To get into the contextual and relational depth of how students experienced Interest-based PE, the 16 interviewees were all recruited from the same school. Interviews with students from more than one school could possibly have contributed to more diversity in student responses and more nuanced findings, yet it could also have harmed the contextual understanding, and thus the validity of the study.

Ethics
The study was approved by the Norwegian Centre for Research Data. The school principal allowed the research, and the purpose of the study was explained to the students and their teachers at the information meeting. Both the students and the teachers were informed that participation was voluntary. This information was repeated to each student before the individual interviews. All participants provided informed consent prior to their participation, and they were informed that they could withdraw from the study at any time without providing a reason. Pseudonyms were used for anonymity.

Findings
As mentioned in the 'Method' section, we identified that 'the role of sports in PE' was a major theme in both EA and SA students' experiences of traditional PE, as well as Interest-based PE. Our analysis revealed two subthemes expressed by the students as: 'it hasn't changed that much' and 'you get to be with people at your own level'.

'It hasn't changed that much': Relational change in light of students' histories with PE
Both Interest-based PE approaches were intended to create a change in the student ↔ subject relationship. Yet, many students did not feel that there was any change. While Heidy, in the SA class, 'assumed' there had been a change 'because the people who are there [in SA] all like PE', Hannah, who also had chosen SA, claimed thateven though she was separated from some of her former PE classmatesthe subject had 'not changed that much'. Robert, who had switched between SA and EA, also reported that SA was no different from traditional PE. This was confirmed by Christine who also switched from SA to EA and claimed that SA allowed students to play sports according to a familiar and traditional PE logic and practice: However, Christine also felt that any change in EA was not persuasive and after all quite temporary: I thought it was very fun in the beginning [when introduced to EA], but I don't think that the teachers have that much of a plan anymore (…) I chose EA because it was supposed to be fun, not normal sports and stuff.
[But] we very often just get a choice between playing volleyball and playing football (…) our [EA] has kind of tapered off (…) it feels like it's the same again. (Christine, EA) According to Christine, EA teachers resumed traditional PE teaching methods during the programme, which could explain why several other EA students struggled to identify the real change in PE: 'I can't really remember what was the big difference' (Susan, EA). 'It's really the same' (Mats, EA). 'I would say it's quite similar, actually. Nothing has changed much' (Eve, EA). All in all, there is much to suggest that the students experienced their former PE to be very centred around traditional sports, and that the logic of traditional sports continued to govern the relationship between the students and the PE subject even in Interest-based PE. Because the focus on traditional sports in SA resembled 'traditional' PE, and because the nature of EA gradually changed from being 'something new' to the 'traditional' sports logic of PE, students in both groups did not consider Interest-based PE to be a true change in the PE subject. This means that the change from 'traditional' to 'interest-based' PE was insufficient to change the student ↔ PE relationship, as adaptive student ↔ PE relationships still required the students to accept a logic of sports in PE. According to the students, Interest-based PE did not include a reflective emphasis on current curricular objectives in PE, meaning that the traditional discourses, as rules and regulations, governing PE were never challenged, and that changes in the PE context did not alter students' already established relationships with PE. Consequently, students felt that their choice between EA and SA served only to separate the 'good students' who performed well in traditional PE from those who did not.
'You get to be with people at your own level': Two approaches to learning or two levels of sport competence?
Many of the interviewees experienced EA and SA as two levels of a sportified PE, rather than two different activity approaches to PE.
Those sports people are there [in SA], and those who are very eager and like to do sports and stuff, the ones who have always really liked PE, (…) they are in SA. And those who kind of really haven't been so good, they try their best [in EA]. (Alice, EA) According to the students, Interest-based PE split them into two groups; those who were eager and good at sports became SA students, while those who were not so good at sports became EA students. Thus, according to the students, Interest-based PE did not actually offer two different approaches to PE, but only divided sports-based PE into two levels of ability. Alice (EA) explained that 'It's good because you get to be with people on your own level, instead of being with the ones who are super good'. Alice's reference to different 'levels' among the students seems to be a description of their relative athletic ability, unrelated to the competence aims in the PE curriculum. As such, 'super good' PE students refers to the students who are skilled at sports and now participate in SA, suggesting that students consider their ability in PE to be based on their sport competency. The idea of EA and SA as two levels of a sportified PE was articulated by both EA and SA students. However, somewhat surprisingly, students in both approaches largely perceived this as a positive development in the PE subject: You don't have those differences between 'oh, you're good, I want you on my team', and the ones who are really bad at it [where it's like], 'No, I don't want you'. I think we're all more at the same level now.
That's what's so nice about EA, that there's no difference between the good and the less good, because everyone is good no matter what. But it wasn't like that before. (Theresa, EA) While students in EA felt that their new classmates were more respectful towards them because they were all at the same level of sports competence, SA students enjoyed their new PE group, as they now were able to play sports alongside other students who were enthusiastic about PE (that is, sports in PE): In my previous PE class [before Interest-based PE], my classmates were not really athletic, they were more like really good at school rather than sports. (…) It was more just playing around kind of, those PE classes. (…) But there's kind of more competition and more speed in PE now than there was back then. (Christopher, SA) The students' enthusiasm regarding the increase in shared interests and abilities within their groups does indicate a relational change. However, SA students' references to 'enthusiasm' in PE were essentially about enthusiasm for playing various sports, which they themselves enjoyed. Their enthusiasm was noticeably related to students' physical ability and sports skills and not to learning or being good at school. The fact that SA students perceived their classmates in SA to be more 'enthusiastic' and 'skilled' at various sports meant that PE no longer had to be just 'playing around'. 'The students in SA take it [PE] a bit more seriously. (…) those who actually are interested in learning techniques and real sports' (Hannah, SA). Like Christopher and Hannah, the students perceived SA to be more 'serious' about skill acquisition than traditional PE had been, and they appreciated this shift in focus: Now I feel that we learn for real (…) how to play football. Rather than playing football just to play football (…) It's a bit more intense now than before. And now I have slightly more focus on pushing myself. (Julie, SA) Students' prior experience of 'playing football just to play football', combined with their new sensation of 'learning for real', suggest that students' own understanding of the purposes of PE do not correspond with those of the national PE curriculum. An implicit understanding that PE is intended to develop students' sports skills was identifiable to varying degrees in all student interviews, and it became clear that students made no real distinction between the domain of sports and the domain of PE. Rather, our analysis suggests that the students perceived EA and SA as a division between two ability levels of the same type of PE.
I think it's very nice [to choose] because it affects your PE. (…) You can choose something that suits you better. (…) So, I think those who chose SA… it's very nice for them too because they can finally have a PE class where they work with students who are on their level. (Theresa, EA) Theresa illustrates how some students held ideas that they may not have been suited for participation in sports activities, and thus unsuited for SA. This lack of relational fit largely seemed to rely on students' evaluations of their individual sport skills. Eve (EA), who had a hip injury and was advised by her doctor not to run during PE, shared the idea that some may not be suited to participate in SA: 'I feel that [SA] has a more like… demanding PE, or sort of… more things that I couldn't have been part of. And then it was kind of just natural for me [to choose EA]'.
There is much to suggest that students' understanding of the role of sports in PE, as well as their own perceived athletic competence and ability, guide their choice of PE approach. EA students believe that their sports-eager peers benefit from their absence, and that the SA students enjoy not having to show consideration for their fellow students who show less eagerness, are less skilled at sports and who now participate in EA. This idea was confirmed by the SA students who enjoyed their new, 'serious', and 'intense' PE, which allowed them to learn sports in a serious way together with other students who, like them, were eager and genuinely interested in learning to play sports in PE. Because neither the Interest-based PE classes nor the teachers who taught them succeeded in challenging the students' ideas of 'PE-as-sports' and the students' ideas of EA and SA as two different levels of the same old PE, the idea of PE-as-sports seemed to remain unquestioned. In other words, the same dynamic that governed the relationship between students and traditional PE survived the transition to Interest-based PE essentially unchanged. There is much to suggest that the dominant conception of PE-as-sports reduced Interest-based PE to no more than a pure differentiation programme based on students' sporting skills and achievement potential in organised sport activities.

Discussion
Interest-based PE was developed and started by PE teachers who experienced that students who were and were not eager about sports did not benefit equally from the subject and represented an attempt to optimise student ↔ PE relationships. Yet, this study shows that the students essentially experienced 'Interest-based PE' as a 'two-level PE', making it clear that the students' chose EA or SA based on their perceived ability to achieve high standards of sports competence and physical performance. Our identification of sports achievement as the prevailing benchmark for student capability in Interest-based PE indicates that it did not succeed in optimising student ↔ PE relations among all students.
Our study shows that the exchange between the PE subject and the students wasand continued to begoverned by a sports discourse, and as such, that this discourse determined the different students' relationships with the subject. The students' description of SA as a 'professionalised' form of traditional PE also suggests that studentswith over 11 years of experience with traditional PEhave strong associations with PE as a subject dominated by sports. The students' lived experience of the subject seems aligned with Kirk's portrayal of 'the idea of PE as sport techniques' (Kirk, 2010: 1). However, the Norwegian curriculum is clear in that improvement of physical performance and sports skills are not purposes of the PE subject (Utdanningsdirektoratet, 2015). Yet, like several other studies in PE (e.g. Moen et al., 2018;Smith and Parr, 2007), our study shows that not only the students but also the teachers seem to neglect the official goals of this subject, and that teachers do not relate to them as rules of regulation. It seems rather clear from our analyses that the sports discourserather than the learning objectives in the subject curriculumgoverned the student ↔ subject relationship also in Interest-based PE. The vital discrepancy between the written and taught curriculum was not targeted during the local didactic programme. It is also reason to believe that the students' histories with 'PE-as-sports' may have made them more accepting of what they experienced as a two-level PE, yet that this also rendered them unable to draw potential benefits from the Interest-based PE programme. What the students describe as benefits of Interest-based PE can at best be understood as a relief from some of the negative symptoms inherent to the sports discourse in PE. It appears that some of these symptoms were alleviated as Interest-based PE spared the students from having to deal with the diversity in sports competence present in their traditional PE classes. Yet, what occurred during the implementation of Interest-based PE was essentially a separation between those students who felt confident and those who lacked confidence when participating in sports. Therefore, in its ultimate effect, this didactical differentiation programme may not only have made students' sports competencies even more explicit in the context of PE, but it may also have contributed to student segregation on the grounds of student confidence, competence and ability in sports. Therefore, it may very well be that this programmewhich was intended to level the learning field for diverse students in PEmay unfortunately have had quite the opposite effect, preserving the sports discourse in PE and increasing the acceptance of this discourse among the students as well as their teachers.
The reproduction of 'physical education as sports techniques' (Kirk, 2010: 10) and related discourses in the context of PE is considered a major challenge to student learning in PE (Redelius et al., 2009;Wilkinson et al., 2013), and has led researchers to describe PE as 'backward-looking' (Stolz, 2014: 27), built on archaic notions of sport and pedagogy which have proven resistant to reform (Stolz, 2014). Siedentop, O'Sullivan andTannehill (1994, in Kirk, 2010) have explained the subject's resistance to reform on the grounds of PE's marginalised position and deep institutionalisation in schools. Research on PE teacher education has also shown that PE teachers play an important role in the reproduction of traditional approaches to PE, such as 'PE-as-sport-techniques'. For example, Mordal- Moen and Green (2014) have found that prospective PE teachers' beliefs and practices may be difficult to change through PE teacher education, and thus, that PE teachers are likely to reproduce the PE they were presented with during their early socialisation into the PE subject, as students themselves. As such, the PE teachers who taught EA and SA may have found it hard to challenge dominant ideas inherent in the sports discourse because this, on some level, would have required them to confront their most basic conception of what PE is and should be. For teachers who have been part of a subject dominated by a sports discourse it may have been easier to develop SA, which was centred on traditional sports, as opposed to EA, which had to be established on didactical approaches relatively distinct from classical learning practices in PE, such as the 'Demonstration-Explanation-Practice' method (Tinning, 2010: 43). This may also explain why EAwhich initially was experienced as different and newgradually returned to a more 'traditional' form of PE. The retrogression of EA suggests that teachers, at some point during the programme, returned to the idea of PE-as-sports, and as such, that Interest-based PE was never able to fully change the dynamic of the PE class. As a whole, these findings suggest that the sports discourse is still an embodied and highly 'sedimented practice' (Kirk, 2010: 50) in PE that many teachers and students may have found monolithic and unchangeable, despite the introduction of new activity approaches in PE.
Although this differentiation programme was developed as a means of drawing PE away from a sports-based discourse, the analysis presented in this study indicates that the teachers may have been unaware of the breadth and depth of the underlying discourse, and therefore, that this discourse was 'neither shaken nor stirred' (Mordal- Moen and Green, 2014: 430) by the programme content. On the contrary, its continued dominance challenged the relational fit between diverse students and learning objectives in PE and acted as a barrier to the promotion of adaptive developmental regulations and thus, positive movement experiences and equal education in PE. That said, it should be noted that a prior study on Interest-based PE (Erdvik et al., 2019b) found that EA was the preferred approach to PE among the majority of girls and students who were not active in sports. According to these findings, girls were 3.9 times more likely to choose EA compared to boys, while students not active in sports were 4.4 times more likely to participate in EA compared to their peers who were not active in leisure sport (Erdvik et al., 2019b). The qualitative analysis presented in the present study provides a more nuanced understanding of these findings. Given students' experience that EA (at least in the beginning) was less like traditional PE than SA, and that EA was the preferred option among girls and students who did not participate in leisure sport (Erdvik et al., 2019b), this suggests that student groups who typically report less positive PE experiences would like the subject to be taught differently (Säfvenbom et al., 2015). Yet, despite their choice, there is much to suggest that students' and teachers' histories with 'PE-as-sports' meant that Interest-based PE was always understood from the perspective of sports and therefore, that the 'new' approach to teaching EA was not long lived. Because EA was not explicitly provided as curriculum driven and because it did not target students' and teachers' reflexivity the sports discourse continued to dominate PE, counteracting the relational fit between diverse students and PE, and most probably preventing this programme from contributing to a levelled educational field.

Concluding comments
Despite good intentions to optimise the student ↔ PE relationship by offering two different PE approaches, students' perceptions of a two-level PE show that Interest-based PE did not succeed in levelling the learning field for the students. The programme seems to have manifested as a classic didactic differentiation programme, based on established premises taken for granted and not questioned. This leads to an uncomfortable, but inevitable question: is the implementation of Interest-based PE, and other similar programmes designed to level the learning field in PE, based on a primary misconception? While Interest-based PE was established to level the educational field, findings indicate that it served to make the sports discourse even more explicit by: (a) splitting the students who were competent at sports from those who were not; (b) neglecting to address the possibility that the sports discourse would continue to govern the student ↔ subject relationship in the new groups; and (c) neglecting the need to emphasise that learning objectives in the Norwegian PE curriculum do not include or concern students' sports achievements.
Based on prior research and the analysis presented in this study, there is reason to believe that the consequences of a dominant sports discourse triggered the need for a didactic differentiation programme, yet that this discourse also prevented any real change in the relational fit between students and the PE subject. Likewise, there is also reason to believe that other didactic differentiation programmes have been implemented for similar reasons. One example may be gender-based PE (e.g. Klomsten, 2013), which is intended to level the learning field among boys and girls. We believe that in all these programmes, teachers attempt to solve challenges caused by the sports discourse by engaging in didactic or individual differentiation. Yet, there is reason to believe that these differentiation practices would be unnecessary if only the formally stated learning objectives in the national PE curriculum were sufficiently emphasised from the earliest stages of students' PE careers.
This study has shown that offering different approaches to PE activity in itself is not sufficient to optimise students' relations with PE, and suggests that a perspective on students' understanding and learning more in accordance with the PE curriculum might better achieve equal opportunities for learning among all students. Efforts to promote adaptive developmental regulations and equal education in PE will most probably continue to fail if the sports discourse remains the dominant regulation of PE. This has important implications for future intervention and action research in the context of PE.

Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.