School refusal and school-related differences among students with and without diagnoses

School refusal is an increasing concern in the Nordic countries. In Norway, public recommendations to combat school refusal sometimes include making life at home less desirable, which reflects the notion that children choose to stay at home out of convenience. However, the mechanisms behind absenteeism are not trivial. A largely unaddressed topic is the compatibility of Nordic mainstream schools for students with neurodevelopmental or psychological diagnoses. This paper supports King and Bernstein’s definition of school refusal as related mainly to anxiety and emotional discomfort. We ask: Are there school-related differences between different groups of students who struggle with school refusal? The paper discusses the results of a web-based survey that included 256 guardians of children with substantial undocumented school absences. We found school-related differences between students with ADHD, autism, and psychological challenges and students without diagnoses who also struggle with school refusal. Compared to students without a diagnosis, autistic students strived socially more and had higher risk of being without friends at school, and students with ADHD were more exposed to bullying, strived socially more, and fewer of them had good relations to the teachers.


Introduction
In 2018, more than 22,000 students in Norwegian primary and secondary upper schools were absent for more than 1 month during the school year (Holterman, 2020, p. 7). Prior to that, Havik et al. (2015) had conducted a study in seven municipalities in Norway among students aged 12 to 13 years, and they found that 3.6% of them had absences related to school refusal. Although Norwegian schools do not keep statistics on how many pupils stay home due to school refusal, these studies indicate that the numbers are high. The consequences for those concerned can be serious. When their challenges are not addressed, these children are denied their right to education and may have difficulty getting back into the education system, which reduces their opportunities for self-realization, good jobs, and future economic security.
Genetics and heredity are fixed realities, but school refusal affects children with diagnoses in particular, and possibilities for adequate school-based preventive measures should be further explored. More knowledge about school-related determinants of school refusal in vulnerable groups is therefore pertinent. The suggestion that school-related factors are important corresponds with Raknes et al. (2017), who in a study involving 1,719 adolescents concluded that students who are anxious have poorer quality of life at school but better quality of life at home with their families.
A report from the Danish Institute for Human Rights (2017) underscored that autistic students have a particularly high risk of school refusal. This is supported by Munkhaugen et al. (2017), who concluded that it is reasonable to consider autism to be a risk factor for school refusal. In their study involving 78 autistic students aged 9 to 16 years, Munkhaugen et al. found that 43% of autistic children struggle with school refusal. They noted that when so many autistic students refuse to go to school, underlying causes need to be identified-especially differences in how school-related factors are associated more with absences of autistic students and students with ADHD compared to students with other diagnoses or students without a known diagnosis.
Our study used King and Bernstein's (2001) definition of school refusal, focusing on difficulties attending school due to emotional discomfort. The child or adolescent would like to go to school but cannot manage because he or she is afraid of being there. The child's guardians know their child stays home from school, but they are unable to persuade the child to attend. This definition considers some criticism of the school-refusal concept, stressing that refusal should not be seen as a deliberate decision made by the child, but as a consequence of the child experiencing the school as an unsafe place. The child should be neither blamed nor held responsible; the school should. This study also differentiated among neurodivergent (ND) children with diagnoses such as autism (ASD) or attention-deficit/hyperactivity disorder (ADHD), children with no diagnosis, and children with anxiety and other psychological challenges.
In this article, we use identity-first language (IFL) for children with an autism diagnosis, and person-first language (PFL) for children diagnosed with ADHD. This seemingly inconsistent choice reflects what appears to correspond with the current direction of the ongoing IFL/PFL debate and the currently dominant preferences of the respective communities (Botha et al., 2021;see also American Psychological Association, 2021). Recent times have seen the rise of the neurodiversity movement; a growing global response from predominantly autistic people to historic objectification and pathologizing. The neurodiversity movement questions the clear divide between neurodevelopmental diagnoses because many diagnostic traits overlap. Following the logics of the social and human rights definitions of disability, the movement uses the concept neurodivergent to contrast the notion of the neurotypical. In the latter, the majority structure a society where those belonging to the ND minority become "disabled" and have their rights violated in institutions such as schools (Lawson & Beckett, 2021). The neurodiversity movement rejects PFL and embraces autism as identity and pride. Mirroring the LGBT+ movement, autism-like sexual orientation-is not viewed as an illness but rather as a defining neurotype (Kapp, 2020;Yergeau, 2018). Whereas many with ADHD are joining the neurodiversity movement, the same linguistic shift is still not yet seen in that community. Therefore, for now, we use PFL for this group. We acknowledge that the abbreviation "ASD" is contagious in the current debate because "disorder" is undoubtably a pathologizing term. Until a better, more widely recognized term is in place, we will use it in this article-but not without declaring that we understand its bias.
Because ND children share some challenges with social interaction across diagnostic labels and communication, this study focused on factors such as social mastery, friendships, relationships with teachers, and bullying. We asked: Are there schoolrelated differences within the group of students who struggle with school refusal?

Methodology
In winter 2019, the research team contacted the leader of a Facebook group for guardians of children who refuse to go to school, called "School refusal/involuntary school absence." They gave us permission to share a web-based survey through their network. After the study received ethical guidance from the Norwegian Centre for Research Data (the Norwegian advisory body for data-protection issues), a link to the questionnaire was posted on the group's website in February 2020. Via the link, potential research participants were informed that the study was voluntary and required informed consent. To ensure full anonymity, no identifiable personal data or IP addresses were registered.
The link to the web-based study remained open for 3 days. At that time, the Facebook group had approximately 600 members, but not all members were active daily. In addition, members of this group may have shared links to the survey within their networks and other social media groups. The response rate relative to the population who saw the survey is therefore uncertain. A total of 256 guardians submitted completed questionnaires.
The sample in web-based studies may have biases related to gender, age, and education compared to the population (Stokes et al., 2019). Because the target group for this study was guardians with school-age children, the age range limited itself, and age bias was a lesser concern than was gender bias. Our sample included more boys (60%) than girls; comparatively, in Norway, more than twice as many boys than girls are diagnosed with ADHD and four times as many with autism (Surén et al., 2019). Few studies have documented gender differences in the incidence of school refusal, and we therefore have a feeble basis for assessing this gender bias.
A closed social media group can have a norm-forming effect on its members because the discourse in the group can influence their perceptions and attitudes towards the topic. In this study, we considered such bias to be a subordinate concern because the survey questions largely concerned facts and, only to a lesser extent, opened up for the respondents' subjective perceptions. Yet we remained aware that group norms on other issues, for example, attitudes towards school politics and child-protective services or a strong focus on particular diagnoses, may affect who chooses to join and remain in groups such as the one primarily targeted here.

Presentation of findings
Heyne's (2019) call for research on the links between school refusal and neurodivergence underscored the need for further exploration of possible school-based responses that would be adequate for this group. Neurodivergent children and adolescents are more vulnerable in terms of social adaptation and coping. Schools' capability to facilitate academic and social mastery presupposes those teachers and other adults around the student know the challenges that ND students face in everyday school life.

Diagnosis
The two main diagnoses represented in our data were ADHD and ASD. Under "other diagnoses," anxiety and other psychological challenges were most frequently listed. Although students who struggle with school attendance differed individually within these groups, they shared worries about being in school. Table 1 shows that only a quarter of the students in the sample had no known diagnosis, which means that 75% did. For the ND students, 23% had an ASD diagnosis, and 17% had an ADHD diagnosis. Adding the two, boys (43%) who avoided school were more often diagnosed with either ASD or ADHD compared to girls (34%). In Norway, the diagnostic prevalence of ASD is 0.9% at age 12 years (Surén et al., 2019), and of ADHD is 3.6% (Surén et al., 2018). Boys are more likely than girls to have both diagnoses. However, girls with these conditions more often may be misdiagnosed with psychological problems, or practitioners may give more diagnostic importance to comorbid psychological problems when assessing school-aged girls (Evans et al., 2010). The sample, as shown in Table 1, wherein more girls than boys were listed with "other" diagnoses, vaguely reflects these distributions. Compared to the national prevalence rates, it is evident that children with neurodevelopmental or psychiatric diagnoses were strongly overrepresented in this sample of children who avoid school (Surén et al., 2018(Surén et al., , 2019. Friends at school Personal friendships satisfy emotional and social needs. Friendships in adolescence are characterized by intimacy and confidentiality, which are important for further identity formation. Withdrawal, hurt, anger, or acting out can result if children and adolescents are rejected by their peers (Frønes, 2006). Adolescents mirror themselves in each other, and having friends strengthens self-esteem and increases social status. Many children and adolescents invest a lot emotionally, in terms of not only making friends, but also in keeping the friends they already have. If, despite such an emotional commitment, they do not feel part of the social community, they can perceive this as threatening their own integrity. Friendship is thus of great importance for well-being and the experience of belonging (Tetzchner, 2012).
As Table 2 shows, guardians reported that around three of five school-avoidant children in the sample had friends at school. Among the autistic students, however, the share was much lower, at only 44%. That almost six of 10 (56%) guardians of autistic students stated that their children do not have friends at school seems worthy of reflection. In comparison, one in 10 students in secondary schools in Norway had no friends in school (Amundsen & Garmannslund, 2018).  Beckman et al. (2016) pointed out that a high proportion of children with ADHD struggle socially. Table 2 also shows that more than one-third of the students with ADHD (37%) had no friends at school. Interestingly, this percentage was similar to school-avoidant children with no known diagnosis, indicating a commonality in school-avoidance patterns across diagnostic boundaries. A little less than one-third (32%) of guardians of children with "other" diagnoses stated that their children did not have friends at school. Although this percentage is somewhat lower than those in our sample with ASD or ADHD, it is still three times the national average (Amundsen & Garmannslund, 2018).

Relationships with teachers
Mutual respect and empathy are fundamental not only for the teacher and student to have a good relationship, but also for the teacher to appear to be a safe and authentic person. The literature has well documented the connections among learning, motivation, and relationships with the teacher (DuFour & Marzano, 2011;Evertson & Weinstein, 2011). Teachers should therefore meet the students with recognition, care, and empathy, and, at the same time, be fair and clear leaders. Nordenbo et al. (2008) referred to three key dimensions for teachers' competence: professional and didactic competence, relational and emotional competence, and management competence. Evertson and Weinstein (2011) concluded that these qualities are essential for academic and social learning in schools. Federici and Skaalvik (2013) pointed out how good relationships between teachers and students affect the students' motivation, effort, and self-perception, and the teachers who are perceived as cognitively supportive are also perceived as emotionally supportive. As shown in Table 3, around half (52%) of the guardians of autistic students in this study agreed that their children had good relations with several teachers. Among guardians of children without diagnoses, approximately the same proportion (53%) stated this. However, among guardians of children with ADHD, only 40% answered that their children had good relationships with the teachers. The finding that a very high proportion of students who struggled with school refusal also did not have good relationships with teachers is consistent with both Egger et al. (2003) and Havik et al. (2015). That is, those researchers also found a negative association between the student-teacher relationship and school refusal.
Well-being at school Kjaernsli and Olsen (2013) concluded that Norwegian students generally feel safe at school, and only a small group of school-aged children felt outside the social community there. Egger et al. (2003) claimed the most important measures that can be taken concerning school refusal are measures that make students feel safe at school. When Newman et al. (1992) associated the concept of engagement with school and schoolwork, they pointed out that students must feel a sense of belonging and identify with what is happening in school. A comprehensive range of assistance strategies are required for students who are anxious about being at school to experience belonging in a school environment they perceive as fundamentally unsafe.
The fact that such a large share of children in our survey, regardless of diagnosis, experienced school as an unsafe place deserves serious attention. Table 4 shows that an overwhelming majority (85%) of guardians of autistic children agreed that their children experienced school as unsafe. They shared this perception with guardians of students both without diagnoses (88%) and with other diagnoses (87%). Slightly fewer (73%) guardians of children with ADHD stated that their children experience school as an unsafe place. The proportion who experienced school as an unsafe place is very high relative to the general school-aged population. In comparison, the latest national youth survey showed that among the upper-secondary school student in general, about one in four answered that they often dread going to school (Bakken, 2021). This can also indicate that school refusal is linked to characteristics of the school and the school environment.
Coping socially Part of feeling safe at school often relies on social mastery of the school community. More than two-thirds of the guardians of the students in the sample did not believe their children experienced social mastery at school. Here, the two categories (ASD and ADHD) of ND children stood out as particularly vulnerable. Table 5 shows that the guardians of students with ASD overwhelmingly (84%) disagreed that their children experienced mastering the social environment at school, which is significantly more than for students without a known diagnosis (57%). This high share underscored the fact that known social challenges in ASD are not met with adequate mitigating measures to aid social coping in the school environment. The share of students with ADHD who did not experience mastering the social environment at school was also high, although a bit lower than for autistic students. Comparatively, three of four (74%) students with ADHD did not experience social mastery at school; this applied to around two-thirds (64%) of students with other diagnoses.
Exposure to bullying Although Wang et al. (2011) found that being bullied has a greater negative correlation with the school performance of boys than of girls, they also concluded that both boys and girls cope with bullying better when they have friends to support them. In light of this, the vulnerability represented by bullying is multiplied for the high number of school-avoidant students with no close friends. Autistic students seem particularly vulnerable to be affected by bullying because a higher share of them lack psychosocial support from friends. Recent theory on bullying emphasized the need to understand bullying as a result of structural conditions in the classroom, school, or local community. According to Faris and Felmlee (2014), bullying must be understood as a consequence of a competition or struggle for position within the adolescent group and especially a struggle to avoid social degradation. Schott and Søndergaard (2014) also pointed out that bullying occurs as a result of battling for position. Table 6 shows that more than half of the students in the sample had been exposed to bullying; this share was as high as seven of 10 (72%) among the students with ADHD. Restrained to a mainstream school setting, many students with ADHD may struggle with social interaction (Landau et al., 1998). When as many as six of 10 students with ADHD do not have good relations with the teachers (Table 3), the risk of their exposure to bullying is possibly heightened. Autistic students were slightly less exposed to bullying than was the sample as a whole, but those differences were small. Compared to the students with ADHD, some autistic students are possibly less preoccupied with positioning themselves in the student group and may be content to have some friends they hang out with in the school environment. The differences in this context are nevertheless so small that no definite conclusions can be drawn in terms of possible explanations. However, bullying came across as an important factor in school refusal also among children with no known diagnosis. The Norwegian Student Survey 2018 showed that 6% of students in Norwegian schools were bullied in some form by their fellow students (Wendelborg, 2019). Notably, Perren and Alsaker (2006) concluded that children who are victims of bullying are often less social and more submissive, withdrawn, and isolated than are the others, and they often have no friends. Further, Fosse (2006) drew a direct connection between bullying at school and mental illness in adulthood. She found that children with poor self-esteem are particularly vulnerable to being bullied. It should set off alarm bells that our data indicated that six of 10 students who refused to go to school were exposed to bullying.
Comparison of school refusers with and without diagnosis Our findings showed that children who struggle with school refusal are vulnerable to several factors related to the psychosocial school environment. Moreover, ND students appeared to be particularly vulnerable, indicating a major challenge for the inclusiveness of the Norwegian education system. Figure 1 shows that compared to children without a diagnosis, autistic students strive more socially and have a higher risk of being without friends at school. Similarly, Figure 2 shows that compared to children without a diagnosis, students with ADHD do not feel equally secure at school but strive socially more and are more exposed to bullying. The effects of this are probably strengthened by the fact that fewer of these students have good relations with their teachers. Aggregated burden of risk factors We have described how negative synergies can be produced in the interactions among the five risk factors presented in this article. For example, among the children who were bullied, only half had some friends in which to find support, and only 40% had good relationships with their teachers that they could resorted to for help. Aggregating the five risk factors by diagnostic group, Table 7 shows that more than one in four (27.5%) students with ADHD carried the burden of all the five risk factors, as did 16.7% of autistic students. In summing up these risk factors, the ND students stood out compared to the students with other diagnoses: nearly half of the autistic students (48.2%) and the students with ADHD (45%) lived with four or five risk factors, compared to around one in four (27.7%) with other diagnoses and one-third (36.7%) of the students with no known diagnosis.

Discussion
Building the future's competencies is a shared goal across the world, as in the EU and the OECD, and it is equally important in Norway. The OECD's "Education 2030 Program" outlines which knowledge, skills, and values the school should emphasize. Among them is the need to educate creative people who are open to new opportunities and solutions and who do not just passively adapt to the everyday life of which they are a part. The findings in this study indicate that schools are not facilitating variations in energy, creativity, and flexibility, thus preventing ND students from finding their place in the social community. Initially, we referred to Heyne's (2019) call for research on school refusal and neurodevelopmental diagnoses and how the two may be related. We asked whether it is possible to find differences between ND students and students with other diagnoses regarding school-related factors and school refusal problems. The results of the survey presented suggest moderate but significant differences between subgroups of school-avoiding students relative to certain school-based risk factors. They support the suggestion that autistic students more often lack friends in school and struggle with social mastery. Not all children have the same social needs, but the fact that almost six of 10 autistic students did not have friends at school suggests two things: first, they may struggle with communication and interaction difficulties; second, the psychosocial school environment does not cultivate an inclusive social setting. In both cases, the schools fail to facilitate for the children's equal right to social participation by providing neither a safe environment nor the individual support needed for the students to obtain social well-being in everyday school life.
Because sensory impressions in autistic children can be heightened, and the children therefore perceive them as overwhelming, it is important that the school environment be as predictable and safe as possible. A bombardment of sensory inputs in the classroom and schoolyard can lead to unnecessary stress and unrest, negatively affecting social well-being and learning equally. Spare cognitive capacity and inner security is a prerequisite for being able to dedicate mental resources to complex social interactions in a school environment (Øen, 2017).
Students with ADHD have poorer relationships with their teachers and are bullied more often. Aggregating the risk factors, ND students stand out as particularly vulnerable to both the number of, and presumably the negative synergies among, compiled risk factors. Overall, it is alarming to see the degree to which all the student groups experience school as an unsafe place, do not experience social mastery, are exposed to bullying, lack friends at school, and do not have good relationships with their teachers.
The fact that so many school-avoidant children have poor relations with their teachers appears to be an important issue regardless of diagnosis, but particularly important in the case of students with ADHD. In this study, as many as six of 10 guardians of children with ADHD stated that their children did not have good relationships with their teachers. The stories behind these number may differ, but regardless of that, schools appear to have failed to mobilize needed additional resources to support the students. Honneth (1995) pointed out that schools do not meet students' right to equal opportunities to succeed. He underscored that solidarity is based on an experience that others demand and recognize qualitative variations in contributions. Such rights are also reflected in the UN Convention on the Rights of the Child and the Sustainable Development Goals. The findings presented in this paper show that Norwegian schools have not sufficiently succeeded in facilitating student diversity. Welcoming differences and supporting social mastery in the school environment will lead to students not only gaining greater self-confidence and greater faith in their own abilities and resources, but also, to a greater extent, being met with recognition and respect from their fellow students and teachers.
Although school-related factors have an impact on school-refusal problems, the exact causal mechanisms at play remain largely undocumented. When relevant factors are allowed to interact and influence each other in negative ways, vulnerable students can easily end up in a vicious cycle. This suggestion supports Anvik and Gustavsen (2012), who in a study of adolescents who dropped out of upper-secondary schools, concluded that loneliness and bullying were the two main reasons.
The suggestion also aligns with Lie and Roe's (2003) study, which found that a good relationship between student and teacher is important for the student's performance, and with Rosier et al. (1994), who concluded that rejection from fellow students can lead to future problems such as drop-out and mental difficulties. Fosse (2006) found that poor self-esteem and lack of friendship can mean that students do not dare to tell about what happens when they are subjected to bullying. Considering this, our guardian-reported survey results are possibly even an underestimate of the burdens faced by children who, for good reasons, refuse to go to school. Some current advice from child-protection authorities suggests fighting school refusal by making it less desirable for school-avoidant students to stay home. If indeed an overwhelming majority of school-avoidant students experience school as an unsafe place where they face bullying, have no friends, or find little support in their teachers-or all the above-then such advice needs to be revisited. When these students are not in a position to fight back or to get sufficient help to stop what is happening to them in school, their experience of powerlessness can become so substantial that it can produce serious consequences also for their long-term mental health. Making life less pleasant at home will not help solve school-based issues. To the contrary, guilt-ridden children who already feel rejected by their schools and peers are likely to experience it as further punishment.

Conclusion
In this study, we investigated the differences within the group of students who struggle with school refusal. Our intention was not to pit vulnerable groups against each other, but to better understand the variations in some potential underlying mechanisms. In light of this, we conclude that ND students are particularly vulnerable in terms of social adjustment and inclusion. Students with ADHD have poorer relationships with teachers, whereas autistic students to a greater extent miss friends at school. The findings of this study suggest that students with diagnoses may be particularly vulnerable. This, and the fact that so many students experience school as an unsafe place to be, should be subject for more research that can help identify more adequate and student-friendly solutions.

Author biographies
Marie-Lisbet Amundsen has many years of experience as a pedagogical-psychological counselor. Her research is aimed at socially marginalized groups in society. Since 2010, she has been employed as a professor of special education at the University of South-Eastern Norway, where she is leading a research group and has teaching assignments related to the master's degree program.