Is it possible to get away from disability-based classiﬁcations in education? An empirical investigation of the Swedish system

Disability classifications is given a minor role in the Swedish national policy on special needs in schools. In the present study questionnaires are used to investigate the actual importance of medical diagnosis in obtaining special support as well as attitudes towards such practices. The study involves different levels of the educational system as well as different occupational groups within compulsory schooling and preschool. The results show that: 1) disability-based categories are seen as less needed in practice by chief education officers than by principals; 2) disability-based classifications have stronger support among school staff than in the guiding documents and among principals and chief education officers; 3) a disability-based approach has stronger support at the compulsory school level than at the preschool level; 4) the group most in favour of disability- based classifications is teachers and those most against are chief education officers, principals and SENCOs.

There are competing models regarding how to identify and define special needs. The characteristics of such models differ between countries (Hollenweger 2008) as well as over time in particular countries (e.g. Burke and Ruedel 2008). However, there is a lack of empirical analyses of educational classifications at a system level. Since Sweden is a country where there has been an effort to get away from disability-based classifications, it emerges as an especially interesting case to study. The purposes of the present study are to investigate a) the actual importance of disability-based classifications in school practice, and b) the attitudes towards disability-based classifications of different groups of personnel working on different levels and in different parts of the school system. The latter issue is of interest since e.g. Special Educational Needs Coordinators (SENCOs) have been expected to play a pivotal role in the creation of more inclusive practices and thus could be expected to be more negative to disability-based classifications.
It is common to differentiate between support systems that rest on the notion of a) specific disabilities and b) more general categorizations, e.g. special needs. In the former, categories that often have their origin in psychological and/or medical classifications are utilized in order to identify pupils eligible for support. Learning disability and specific reading disability are examples of such categories. Categories are either defined as being of a constitutional origin, e.g. specific reading disability or understood as if they are caused by constitutional factors even when the definition precludes a causal explanation, e.g. AD/HD as defined in DSM-IV (American Psychiatric Association 1994). Support systems characterized by disability classifications imply an understanding of educational problems as something caused by characteristics of the child, i.e. the disability rather than characteristics of the educational environment.
General categorical systems, on the other hand, do not utilize specific categories. In addition, it is often believed that such general categorizations promotes an understanding of educational problems as caused also by dysfunctional environments. In Sweden, for example, the notion of 'pupil in need of special support' is used. 'Pupil in need of special support' is defined as a category which a pupil can, so to speak, 'walk in and out of'; that is, it is not defined in relation to a specific trait or characteristic of the pupil. In a recently published book by Florian and McLaughlin, Disability Classification in Education (2008;also see Florian et al. 2006 andMcLaughlin et al. 2006), several characteristics of the classification systems in the U.S. and Great Britain are discussed. One distinction which is pointed out in several of the chapters is the one between classification systems which rests firmly on the notion of disability (for example in the US) and others in which disability classifications are challenged (for example in Scotland).
As stated above, classifications in terms of disabilities have also been challenged in the Swedish system. In addition, Sweden is an interesting case due to its efforts to situate educational problems in the context of schooling rather than solely in the minds and bodies of individual pupils, at the level of national school policy (Gö ransson, Nilholm, and Karlsson 2011). These efforts are illustrated in the avoidance of disability labels and the fact that in the late 1990s, three committees proposed a change of wording in the Education Act regarding 'students with special educational needs'. Hence, the official wording is now 'students in need of special support' -which implies a focus on the relational character of the support need and supposedly turns the focus partly away from an individual pupil's alleged shortcomings. Thus, we regard the Swedish system as an interesting object of study since 1) it involves, at least at the national policy level, quite a radical challenge to the use of specific classifications, and 2) it also challenges the view that the pupil is the locus of school problems How the classification of pupils within a school system is accomplished is of fundamental concern (Skrtic 1991). Classifications are part of professional languages, i.e. classifications have to do with which professional groups that should have the right to define problems in schools. Ridell (2008, 109) points out that 'understanding classification systems is extremely important because they reveal a great deal about dominant discourses and underlying relationships of knowledge and power.' In a similar vein, categorizations also have consequences for the identities and rights of the pupils involved. Classifications in special needs education involve a fundamental dilemma since the need to identify pupils/children in order to provide support also involves classifying them as 'not normal' or lacking in certain respects (Nilholm 2006;Wedell 2008).
Disability classifications tend to point out the individual as the locus of the problem and often involve stigmatization and non-inclusive educational solutions (Burke and Reudel 2008). Several scholars, while recognizing that disability categories carry some useful information, question whether such categories have substantial educational implications (Hattie 2009;Mitchell 2007;Norwich 2008). On the other hand, disability classifications make the need for support apparent and signify real differences between children (Clark, Dyson, and Millward 1998). In addition, some researchers argue that disability classifications do not necessarily localize the educational problem solely with the child/pupil, and instead can be used to address the whole learning situation (Simeonsson, Simeonsson, and Hollenweger 2008). However, an important point is how the system is actually used in practice (c.f. Ridell 2008). For example, in Sweden both individual educational plans and remedial plans for pupils in need of special support are supposed to take the whole learning environment into consideration. However, research suggests that in the actual educational plans, problems tend to become the problems of the individuals (Isaksson, Lindqvist, and Bergström 2007;Swedish National Agency for Education 2003).

The Swedish school system
Our focus in this study is on pre-school (age 1-6; optional but most children attend) and the comprehensive school within the compulsory school system (age 7-16). Preschool is primarily governed by the Education Act (Public Law 800 2010) and the Curriculum for Preschool (Government office 1998(Government office , revised 2010. The focus at the time of the study is on play and social goals.The Education Act (Public Law 800 2010), the School Ordinance (Public Law 85 2011) and national curricula regulate comprehensive schools. Schools and municipalities are goal-steered by central authorities, but have up to now been fairly autonomous in how the goals should be accomplished. Several of the steering documents have been revised since the time of this study. However, we have chosen to present the steering documents of today since the classification system is basically the same.
There are many goals regarding social values and knowledge in the documents that guide the Swedish school system (Gö ransson et al. 2011). Around one in ten pupils within compulsory schooling attend independent schools. Independent schools have to abide to basically the same juridical framework as municipal schools, the main contrast being that independent schools, while being financed by tax money in a similar way as municipal schools, are allowed to make a profit. At present, the school system in Sweden is undergoing changes involving, for example, a new grading system, new knowledge goals, more testing, and so on. However, no basic changes have been suggested as regards the classification system of support needs.
As a general rule, a disability classification is not needed in order to receive special support in the Swedish school system according to the legislation. There are three exceptions to this. Eligibility for special programs for pupils diagnosed with an intellectual disability (1.4% of pupils) and special schools (involving few pupils) is dependent upon a disability classification. Moreover, a National Agency for Special Needs Education and Schools coordinates state support for special needs education for pupils with disabilities. A disability classification is needed in order for schools to receive support from this agency. However, apart from this, no distinctions are made between difficulties because of disabilities, learning difficulties, or lack of opportunities or disadvantages (c.f. Ebersold and Evans 2008).
Instead, the concept of 'pupil in need of special support' is vital to the legislation surrounding eligibility for special support. In the Education Act Chapter 3 Section 8, it is stated that pupils at risk of not reaching the knowledge goals of the school or who experience other difficulties in school have a right to special support (Public Law 800 2010). More than 40% of the pupils receive support at least once during their nine years in comprehensive school, and around 17 percent of the pupils receive special support at any particular point in time (Gö ransson, Nilholm, and Karlsson 2011). No exact figures are available, but Nilholm et al. (2007) estimate that between 2.3% and 3.1% of pupils (aged 7-17) are educated in segregated settings most of the school day (including pupils in special programs for pupils diagnosed with intellectual disabilities). Children in need of special support in preschool are integrated in regular groups with a very few exceptions. The Swedish classification system of support is to a large extent professionally based. Professionals have a vital role since special needs are to be identified and handled at the school level. Principals have an overarching responsibility that special needs are met in an appropriate way.
In recent years increased accountability has become apparent in the Swedish system as seen in a) the emergence of independent schools; b) increased choice between different municipal schools; c) more explicit standards; d) increased testing, and e) more governmental inspections of the work of municipalities and local schools. The impact of these changes on the work with children/pupils in need of special support is not yet fully known.
To sum up, research into classification systems points out that such systems can have quite diverse characteristics. The Swedish system is interesting since the national policy level places less focus on disability labels and instead involves the environment as an important factor in the emergence as well as the handling of school difficulties. However, both internationally and in Sweden, there is a lack of studies of how classification systems function in practice. Thus, the present investigation is to our knowledge the first attempt to make an encompassing study of how a classification system is acted out in practice.

Method
Data from four different surveys were used to obtain information about the importance of disability-classification and the attitudes towards such classification practices among different groups of personnel.
More specifically, data from four different questionnaires were used. The questions analyzed in this paper were included in the questionnaires in order to be able to make a comparative analysis of different groups. As can be seen in Table 1, one of these was sent to all chief education officers in Sweden and one to all principals of independent schools in Sweden while the two others were sent to principals and school staff in one municipality. Thus, whole populations were investigated so no inferential statistics was used. The two questions analyzed in the present study used Likert-type response scales. In the following, more information about the participants and the questionnaires are provided.

Participants
Ten different groups of personnel participated in the study: 1) chief education officers of all municipalities in Sweden (questionnaire #1); 2) principals of all independent comprehensive schools in Sweden (questionnaire #2); 3) all principals of municipal comprehensive schools in one municipality (questionnaire #3); 4) all principals of municipal preschools in one municipality (questionnaire #3); 5) all subject teachers in municipal comprehensive schools in one municipality (questionnaire #4); 6) all class teachers in municipal comprehensive schools in one municipality (questionnaire #4); 7) all preschool teachers in one municipality (questionnaire #4); 8) all special teachers in municipal comprehensive schools in one municipality (questionnaire #4); 9) all SENCOs in municipal comprehensive schools and preschools in one municipality (questionnaire #4); 10) all assistants in municipal comprehensive schools and preschools in one municipality (questionnaire #4).
Chief education officers are responsible for the educational system of a municipality. Principals are responsible for pupils and children in need of special support at the school and/or pre-school level. At times, this responsibility can involve more than one school/preschool. Preschool teachers have pedagogical responsibility for children between 1 and 6 years of age, and class teachers are responsible for the education of pupils aged 7 to 13 in primary schools teaching most subjects within their class. Subject teachers teach one or more subjects to pupils aged 13 to 16 and are also responsible for grading. SENCOs work mostly in schools but sometimes in preschools and are educated to teach pupils in need of special educational support but are also supposed to develop the whole learning environment. Special teachers do not work in preschools but only in schools and are supposed to work directly with pupils and/or with small groups of pupils. Finally, assistant resource staff work closely with children and pupils in need of special support between 1 and 16 years of age.

The questionnaires
In the above mentioned four questionnaires several questions were asked about work with children and pupils in need of special support. For the purpose of this study two questions concerning actual classification practices and attitudes towards disabilityclassification practices were selected for closer analysis. The question about the importance of disability-based classifications for receiving special educational support was asked to the chief education officers and the three groups of principals whose responsibility it is to know about such matters (i.e. in questionnaires #1-3). The question about attitudes towards disability-based classifications was asked to different occupational groups working in schools (i.e. in questionnaire #4 and the three groups of principals (i.e. in questionnaires #2-3). Unfortunately, this question was not included in the questionnaire to the chief education officers.
Separate questionnaires (for more detailed descriptions of each questionnaire and the outcome of the analyses of other parts of the questionnaires, see references provided) were thus sent to: 1) the chief education officer in each Swedish municipality (N = 290, response rate 90.3%, year 2006) (Nilholm et al. 2007); 2) all principals of independent comprehensive schools in Sweden (N = 686, response rate 79.6%, year 2009) (Gö ransson, Magnusson, and Nilholm accepted); 3) all principals of all municipal preschools (N = 29) and comprehensive schools (N = 16) in one Swedish municipality with 55,000 inhabitants (N = 45, response rate 100%, year 2009) (Lindqvist and Nilholm in press); and 4) all municipal personnel in preschools and comprehensive schools in the same municipality (N = 1297, response rate 72.5%, year 2009) (Lindqvist et al. 2011). Respondents of the latter questionnaire include: 123 subject teachers, 147 class teachers, 199 preschool teachers, 35 special educational needs coordinators, 22 special teachers and 56 assistants. The response rate within each occupational category corresponded acceptably to the proportion of staff within that category as displayed in the statistics of the municipality.
All questionnaires were preceded by pilot studies and no problems were reported by participants in these pilot studies in interpreting either of the two questions analyzed in the present paper. The administration of the different questionnaires varied slightly due to demands particular to the questionnaire. Only the questionnaire to principals in municipal schools and preschools were distributed by e-mail.
Rather elaborate steps were taken in order to get an acceptable response rate. At least two reminders were sent for each of the questionnaires. In addition, personnel working in the municipality were allowed to respond to the questionnaire during their working hours. The chief education officers were personally contacted by phone or email in the final phase. In all, the response rates were exceptionally good which indicates that the respondents consider schools' work with children/pupils in need of special support to be important. All data from the participants was treated with confidentiality.

Questions analyzed
The two questions analyzed in this study were identically formulated in the questionnaires in which they appeared. The first one concerned the usage of disability-based classifications and the second one the desirability of diagnoses.
-What importance do you believe that a medical diagnosis has in order to obtain special support? -What importance do you believe that a medical diagnosis should have in order to obtain special support?
It was also pointed out to the respondents that the questions did not encompass pupils in the special program for pupils diagnosed with an intellectual disability. Further, we chose to use the concept of 'diagnosis' instead of the alternative option of 'disability' in the framing of the questions. The reason for this is that there has been quite a heated discussion in Sweden about the use of diagnoses (e.g. dyslexia, AD/HD, Asperger's syndrome, neuropsychiatric diagnoses) in schools. Thus, 'diagnosis' would be a more familiar word in the Swedish context than 'disability' when it comes to investigating the influence of classifications with their origin in psychiatry and medicine.
Since the notion of educational diagnoses occurs at times in the Swedish context, medical was added to diagnosis. We use the expression 'disability-based' classifications as an overarching concept covering both classifications explicitly using disability-classification as well as classifications in terms of diagnoses like AD/HD, dyslexia, and so on. As stated above, principals answered both questions, chief education officers only the first one and staff at the school level answered the second one only.

Results
The results are presented in Tables 2-4 where the means and standard deviations of groups are displayed. These outcomes will be further elaborated on below. The Statistical Package for the Social Sciences (SPSS) was used to analyze the data.
The importance of a disability-based classification in order to obtain support As stated above, this question about the importance of a medical diagnosis in order to obtain extra support was asked to chief education officers, principals working in independent schools, and principals working in municipal schools in a) comprehensive schools and b) preschools. As can be seen in Table 2, chief education officers judge it to be less important (m = 2,38) to have a medical diagnosis in order to obtain extra support than principals in municipal (m = 2,59) and independent schools (m = 3,08). These answers refer to the same age group of pupils. One interpretation which is substantiated by data reported by Giota and Emanuelsson (2011) is that chief education officers might underestimate the importance of medical diagnoses, alternatively that principals might overestimate it. However, principals are closer to actual decisions concerning support needs and the lower figure provided by chief education officers might reflect their concern with how policy ought to be. Compared to our data, Giota and Emanuelsson (2011) report a higher figure for principals of municipal schools but a similar figure to ours for principals of independent schools. Since we only have data from one municipality, it seems reasonable to assume that their investigation encompassing a representative sample of Swedish principals of municipals schools provides a better estimate. They detect a small difference between municipal and independent schools as regards the importance of diagnoses in order to obtain special support. It should be noted that this line of reasoning implies that there is quite a large difference between chief education officers and principals in schools in their estimates of the importance of medical diagnoses in order to obtain special support. It should also be noted that the estimate of principals of preschools is not higher than for the chief education officers. One interpretation of this is that diagnoses is of less importance in preschool than in school. Again, this estimate is based upon one municipality.
To sum up, despite the fact that medical diagnoses should be of minor importance in order to obtain special support in Swedish schools, the results suggest that in practice such diagnoses are of importance. There also seems to be a tendency for chief education officers to underestimate this state-of-affairs.

The importance a medical diagnosis should have
The answers by the nine groups who were asked the question concerning the importance a diagnosis should have are shown in Table 3. Some very interesting patterns emerge from the data. First of all, despite the fact that the Swedish educational system in general does not demand a diagnosis, a lot of the personnel believe that it should carry importance.
Secondly, all groups of teachers outside the support system are more positive towards the use of medical diagnoses than all the other groups with the exception of assistants. However, teachers in preschool are less positive than class teachers and subject teachers. Thirdly, the occupational groups least in favor of diagd noses are principals, SENCOs and special teachers. However, principals in independent schools are comparatively positive. SENCOs and principals in preschools are the two groups least in favour of medical diagnoses as a prerequisite to obtaining special support. Note. Range 1-4 (1 = Very low significance or no significance at all, 2 = Rather low significance, 3 = Rather high significance, 4 = Very high significance).
Thus, fourthly, there is huge variation between the different occupational groups regarding their opinions about the necessity of having a medical diagnosis in order to obtain extra support. Fifthly, it seems that on a general level, personnel working in schools are more positive towards the use of medical diagnoses than personnel in preschools. Finally, a comparison between Tables 1 and 2 reveals that all groups for which there are comparative data, i.e. principals in municipal schools, municipal preschools and independent schools, believe that diagnoses should carry less importance than they state that it actually does.
In Table 4, principals, teachers and resource staff (SENCOs, special teachers and assistants) are compared. For each group, the means of each group included in the category are weighed equally (i.e. disregarding the number of subjects within the group). As above, teachers appear as the group that attaches the most importance to the use of diagnoses falling slightly closer to 'rather high significance' than to 'rather low significance' while both the other combined groups score close to the latter alternative.

Discussion
To summarize, the main findings of this study show that disability-based classifications are more widespread and asked for in the Swedish school system than is implied in the steering documents. Thus, the ways to identify special needs as put forward in the steering documents are partly undermined. Moreover, there seems to be a lack of consensus at different levels and in different parts of the educational system regarding the need for medical diagnoses in school practice. Principals and support staff, with the exception of assistants, are less in favor of medical categories than teachers. In addition, personnel in preschools are less positive towards medical diagnoses than personel in schools. Taken together, a disability-based approach has fairly strong support at the school level with the exception of SENCOs and special teachers. These findings are very relevant to the field of special needs research since they plainly illustrate that classification systems have to be empirically investigated in order to be understood. It is obvious that the view of legislators and politicians do not translate neatly into school practice.
The differences found between different groups raise a need for further research concerning the micro-politics (Ball 1987) of educational practice regarding the use of medical categories and its consequences for inclusive education. One hypothesis is that principals, SENCOs and special teachers being least in favour of medical diagnoses might form alliances at the local level.

Why do different groups have different attitudes?
Different explanations of group differences can be suggested. Principals and special educators might be more knowledgeable about the legislation and/or be more prone than teachers to see other reasons (e.g. teacher shortcomings/unruly classrooms) than individual shortcomings for children´s difficulties. Further, teachers might feel a lack of knowledge regarding how to educate some of the children in their classrooms and believe that knowledge about disabilities will improve their teaching. However, more research using e.g. interviews or ethnography is needed in order to further explore reasons for group differences. The generally high quest for disabilitybased categorization is also interesting given research evidence suggesting that disability-based categorization has few educational implications (Hattie 2009;Norwich 2008). An important issue concerns how old models and attitudes can be changed in the more social direction implied in the steering documents. It is obvious that the findings point to the importance of discussions at the school level as regards what the legislation amounts to and how schools should reason when it comes to identifying special support needs. Moreover, there are differences between the studied groups in how they are educated and trained as regards issues in identifying school problems. It does seem that there is a need among these groups for further training in how to work according to the legislation.
As pointed out by Riddel (2008), discourse and power are involved in the use of classification systems. The discourse of medicine seems to be influential in the classification of pupils in Sweden despite the fact that this discourse has less influence at the level of policy. This discourse is voiced more by certain occupational groups than others. Teachers are the ones who are most positive to medical diagnoses in the present investigation. In the 'micro-politics' (Ball 1987) of schooling we can expect different groups to adopt diverse standpoints vis-à-vis the use of disabilitybased classifications. As is well documented, there is a tradition that educational problems are often regarded as individual problems in educational systems both in an international context (Ainscow 1998) as well as in the Swedish context (Gö ransson et al. 2011;Isaksson, Lindqvist, and Bergströ m 2007). We believe that this makes the educational system vulnerable to influences from medical discourses built on the logic that problems are to be diagnosed at the level of the individual.
Further, the use of medical diagnoses implies that schools hold less responsibility as regards educational problems. The neuropsychiatric movement which is reflected in the increase of diagnoses in each edition of DSM-4 as well as in the increased usage of diagnoses in a global perspective is obviously important as regards the medicalization of problems. While several of these diagnoses have been challenged (e.g. Baughman and Hovey 2006), no one can deny that they have become increasingly influential as regards analyzing and managing school problems. Unfortunately, we lack longitudinal data with which to compare the outcome of the present study. Given the tendencies outlined above, the tradition of individualizing school problems in the Swedish school system and the 'openness' to medical diagnoses revealed in the empirical data, we may in years to come witness the increased importance of disability classifications.

Should we get away from disability-based classifications?
Some low-incidence disability classifications are quite uncontroversial. However, how can we distinguish between disabilities, learning difficulties, or lack of opportunity or disadvantages (c.f. Ebersold and Evans 2008) for the large majority of children and pupils in need of special support? Obviously, lack of opportunity and/or disadvantages might manifest themselves as learning problems and/or disabilities We suggest that the notion of 'response-to-instruction' (Berkeley et al. 2009;Fletcher and Vaughn 2009;Fuchs and Deshler 2007;Fuchs and Fuchs 2006) could be a possible way forward in handling the dilemma involved in differentiating between disabilities and lack of opportunities/disadvantages in addition to being a way to contextualize educational problems. Furthermore, such an approach will reduce the amount of pupils with disability-diagnoses. However, the 'response-to-instruction' concept has only recently emerged in Swedish discussions.
Will abandonment of disability-based classifications make schools more inclusive? In the Nordic context, Haug (1998) has formulated an influential notion as regards inclusive education. One of his main points is that pupils should not be either excluded or pointed out. Given such an interpretation of inclusive education, one comes to the conclusion that disability categories should be avoided if possible. It is also the case that labels probably increase the likelihood that school problems will be viewed as individual shortcomings. This is hard to reconcile with the notion that differences should be celebrated.
Thus, while some disability classifications indicate real problems and thus are useful, others seem to run the risk of masking problems through not providing sufficient educational opportunities and thereby hiding problems with lack of opportunity. Additional critical issues regarding disability classifications concern whether such classifications 1) carry valuable educational information, 2) tend to localize educational problems within the child and lower expectations, and 3) become stigmas (cf. Norwich, 2008) and might become obstacles to more inclusive practices.

Some notes on the methodology used
As regards chief education officers and principals of independent schools whole populations were studied. As for the other groups studied, they are all from the same municipality. When it comes to estimates of how important medical diagnoses are in reality, data from Giotta and Emanuelsson (2011) provide a fairly representative sample of principals of municipal and independent schools which has been used in order to strengthen the analysis. However, the comparison with regard to the importance different groups believe that a medical diagnosis should have rests on whole populations within one municipality with the exception of principals of independent schools. It could be argued given the lesser importance attached to disability-based categories in this municipality (as estimated by the principals in the municipality) that occupational groups in other municipalities might attach even higher importance to disability-based categories (which is the case for principals in other municipalites).

Implications for policy and practice
One obvious implication is that policy needs to be coordinated with practice. Moreover, a system that strives to become more inclusive should be very hesitant to use disability-categories with regard to children lacking educational opportunities.