Do Spanish parents prefer special schools for their children with autism?

Do Spanish Parents Prefer Special Schools for Their Children with Autism? Javier Moreno, Antonio Aguilera and David Saldan˜a University of Seville Abstract: The social and communication difficulties of children with Autism Spectrum Disorders (ASD) pose a special challenge to educational inclusion. Previous research has suggested that, because of this, parents of children with ASD might be less favorable to educating their children in inclusive settings. In this study, 60 parents of children with ASD in the city of Seville (Spain) were interviewed about their perception of educational provision. Parents were from three different groups, according to the children’s educational placement: mainstream non-segregated settings (regular schools and sharing time with other children without disabilities), mainstream segregated settings (special classes in regular schools) and special schools. These contexts differ in teacher training, resources and contact with other children in ways that allow a comparison of the relative influence of these variables on parental perception. Overall parental satisfaction was high. However, parents in mainstream segregated settings were less satisfied than those in special schools. There were no differences between the satisfaction of parents with children in mainstream segregated and non-segregated settings. Results seem to indicate that it is resources and teacher training, rather than severity of the disorder, the classroom structure or fear of contact with other children without ASD, that determine positive parental perception. Over the past decades there has been an increasing tendency to educate children with disabilities in mainstream settings, with a growing emphasis in many educational systems on the need for full inclusion (Mitchell, 2004). As major stakeholders and clients of these systems, it is important that parents’ opinions and views are taken into account in this process. Firstly, if it is to be successful, parents must support inclusion as a part of major educational reform (Grove & Fisher, 1999; Turnbull & Turnbull, 1997). In addition, it has been found that there is a relationship between parental implication with inclusion and the academic performance of their children with disabilities (Palmer, Fuller, Arora, & Nelson, 2003). However, it is not clear that there actually is a favorable attitude of parents of children with disabilities towards inclusion. Although a great majority of studies do show positive attitudes (Bailey, Hebbeler, Scarborough, Spiker, & Mallik, 2004; Bennett, DeLuca, & Bruns, 1997; Leyser & Kirk, 2004; Lowenbraun, Madge & Affleck, 1990; Miller, Strain, Boyd, Hunsicker, Kinley & Wu, 1992; Seery, Davis & Johnson, 2000; Turnbull, Turnbull, Shank, Smith, & Leal, 2002), this is not always the case (Fox & Ysseldyke, 1997; Green & Shinn, 1994; Kohler, 1999; Palmer et al., 2003). As well as differences in educational systems and actual educational provision, there are various factors that could be underlying these different results. It has been found that demographic variables play a certain role in parents’ attitudes towards inclusion (Leyser & Kirk, 2004; Stoiber, Gettinger, & Goetz, 1998), with more favorable views held by college graduates, small families and married parents, when compared to high school graduates, larger families and single parents, respec- This study was possible thanks to a joint grant of the European Commission for the “European Year for Persons with Disability”, the Social Affairs Department of the Spanish Ministry of Labor and Social Affairs, and the support group Autismo Sevilla. The authors of this paper are very grateful to the participating parents and to Laura Flores, Silvia Lobato´n and Elisa Marcos, for their work during the collection of the data. Correspondence concerning this article should be addressed to: David Saldan˜a, Developmental and Educational Psychology, University of Seville, Avda. Camilo Jose´ Cela s/n, 41018 Sevilla (SPAIN). E-mail: [email protected] Education and Training in Developmental Disabilities, 2008, 43(2), 162-173 © Division on Developmental Disabilities 162 / Education and Training in Developmental Disabilities-June 2008 tively. The child’s age also influences the perception of inclusion, with responses from parents of older children usually more negative (Fox & Ysseldyke; Jenkinson, 1998; Leyser & Kirk; Spann, Kohler, & Soenksen, 2003). The present status of the children’s educational placement seems to also contribute to parental attitude, although findings here are not consistent. Whereas Jenkinson found that parents tended to favor their child’s present placement, i.e., parents of children in special educational settings were less favorable to inclusion, Leyser and Kirk found the opposite. Another relevant variable, also with unclear influence, is the severity of the child’s disability. Some researchers have pointed to parents of children with more severe disabilities as the most favorable to inclusion (Turnbull et al.), but others have found that parents of children with greater difficulties may feel that the benefits of inclusion will not be applicable to their children (Jenkinson; Leyser & Kirk; Palmer et al.). Children with Autism Spectrum Disorders (ASD) are often considered to be among those children with more severe or specific disabilities. Their social, communication and behavioral difficulties pose particular challenges to the process of inclusion (Simpson, Boer-Ott, & Smith-Myles, 2003). Few studies have explored attitudes towards inclusion specifically in this population. Most of the research has centered on parental satisfaction with educational provision (Kohler, 1999; Ruef, Turnbull, Turnbull, & Poston, 1999; Spann et al., 2003). Actual satisfaction with the educational provision in different educational settings can be related to attitudes towards inclusion, since it can provide insight into parents’ reactions to different types of educational placement as they are actually implemented. Unfortunately, not many of the studies with parents of children with ASD have considered differences in satisfaction in different types of educational placement, and therefore, the implications of their results for the process of inclusion are somewhat limited. One exception is a study by Kasari, Freeman, Bauminger, and Alkin (1999), with 149 parents of children with Down Syndrome and 113 with ASD. They found that, whereas parents of children with Down syndrome preferred full inclusion, those of children with ASD generally favored part-time mainstreaming. Half of the respondents in this group felt that their children’s specific needs could not be appropriately provided for in an inclusive program. At the same time, it was found that parents of children in special education programs were less satisfied than those in general education or early intervention. There were no differences between satisfaction of parents of children with Down Syndrome and ASD. The authors hypothesize that parents of children with ASD may be particularly preoccupied with the structure of the classroom, as opposed to parents of children with Down Syndrome, that are more worried about social role models for their children. Parents of children with ASD might also grant greater importance to specialized teacher training, which is to be found in special education settings, and be more worried about their difficulties in peer relations. However, it is difficult to determine which of these reasons is more likely to be influencing differences in attitude. One relevant limitation of the study is that it does not provide information about the severity of the children’s disabilities. In this study, it is difficult to determine if the parents view the disorder itself as a condition that renders children as specially unsuited for inclusion, or if it is the severity of the children’s needs that is modulating their perception. The present study attempts to provide some additional data on the perception of parents of children with ASD of different educational settings in order to understand factors influencing parental perception and potential limitations to the process of inclusion. In the context of a broader study in the city of Seville (Spain), parents of children with ASD were asked to rate the educational provision they were receiving. Seville is the largest city in Southern Spain, and as such is the capital of the Autonomous Region of Andalusia. It holds a population of approximately 700,000 inhabitants, with 95,861 children in Kindergarten, compulsory Primary and Secondary Education and Special Education. The structure of the educational provision for children with ASD in Seville is useful for determining the role of educational placement in parental perception and satisfaction. The system includes special and mainstream schools. The first include only children with Inclusion and Parents of Children with Autism / 163 special needs. In our study, the children with ASD were in special educational placements in which they only shared their time with other children with ASD. Teaching is provided by special education teachers, specifically recruited for these posts on the basis of a college degree in special education (although not specifically related to autism), who receive supplementary training in autism. In mainstream schools, children with disabilities can be found in two different situations. Some of them spend all their time in a support classroom (segregated mainstream placement). The contact of children in these units with the rest of the school is minimal, and in practice their educational placement is similar to a special school for children with autism within a regular school. Their teachers are also special education teachers, but they tend to lack the support of experts and training that those in special autism schools enjoy. The remaining children in mainstream schools (non-segregated mainstream placement) spend a varying amount of time in the regular classroom with other children, and with teachers with and without degrees in special education. This distribution allows testing for the hypothesis that parents with ASD are preoccupied with the structure of the classroom itself and potential difficulties with non-ASD peers, both of which are problems absent in segregated mainstream placement and special schools. These latter two contexts, however, are different in other respects. In special schools all teachers are special education teachers and resources are usually more specific and greater than in segregated mainstream placement. On the other hand, segregated and non-segregated mainstream placements have differing structure, while sharing the school they are in, its lack of specialization and specific resources for autism (see Table 1 for a summary of these educational placements). Data on the children’s severity of autistic symptoms and adaptive behavior, diagnosis and age were also obtained, in order to determine their possible influence on parental perception. In sum, the aims of this study were to: ●Determine parental perception of educational provision in different educational placements. ●Analyze the influence of diagnosis, severity and age on this perception. Method Participants Sixty parents of an equal number of children with ASD studying in schools in the city of Seville were interviewed. All of them were members of a local parents support group and were contacted through it. Forty seven percent of the children were in mainstream schools. Of these, seven children were in segregated mainstream placement, and 21 were in non-segregated mainstream placement. It was mostly the mothers of the children that were interviewed (70 %). In a lower proportion, both parents (22 %) or only the fathers (8 %) were present. In the cases where TABLE 1 Characteristics of types of educational placement for children with ASD in the study (yes ⴝⴙ,noⴝⴚ). Special Schools Segregated Mainstream Non-Segregated Mainstream In Regular School ⫺⫹⫹ Time shared with children without disabilities ⫺⫺⫹ Teachers with Special Education degree ⫹⫹⫺ Teachers with special training in autism ⫹⫺⫺ Additional specific autismrelated resources ⫹⫺⫺ 164 / Education and Training in Developmental Disabilities-June 2008 both parents were present, a single consensus score was established. Analyses of responses showed no differences related to the parent (mother, father or both) participating in the interview. Table 2 shows data for child and parent characteristics. No differences were found between the ages of children and parents of mainstream and special schools, or among the three subgroups of children (segregated mainstream, non-segregated mainstream and special schools) or in the educational level of fathers. However, mothers of children attending mainstream schools showed a higher educational level, ␹ 2 (2, N⫽60) ⫽7.50, exact p⫽ .016. This difference was not significant at the level of subgroup. Differences were also found among diagnostic labels of children in each type of school and placement subgroup, ␹ 2 (3, N⫽60) ⫽ 15.51, exact p⫽.001, and ␹ 2 (6, N⫽60) ⫽ 23.30, exact p⫽.001, respectively (see Table 2). The proportion of children with Autism was higher in the special schools, whereas all cases of Asperger Syndrome were in mainstream schools, in non-segregated special classrooms. Also, and as expected, the mean standardized scores of adaptive behavior and the mean raw scores on severity of autistic symptoms (see instruments below) of children in mainstream and special schools were significantly different, t(57) ⫽2.14, p⫽.036, and t(57) ⫽⫺3.26, p⫽.002, respectively. A multivariate analysis of variance of the sum score TABLE 2 Participants characteristics (mean with SD in brackets, range below) Total Mainstream Schools Special Schools Segregated Classes Non-Segregated Classes Child Age 11:05 (4:10) 12:06 (4:09) 11:0 (4:09) 10:10 (4:05) 3:09–21:02 7:06–20:0 3:09–21:02 4:05–21:02 Standardized score on adaptive behavior 53.1 (28.9) 28.2 (21.8) 60.2 (27.0) 38.2 (24.6) 5.0–104.0 5.0–58.0 5.0–104.0 1.0–81.0 Score on severity of autistic symptoms (max. 96) 40.1 (15.5) 56.3 (13.0) 35.4 (12.9) 52.9 (14.8) 15.0–74.0 40.0–74.0 15.0–69.0 14.0–94.0 Diagnosis Autism 30% 43% 24% 75% Asperger Syndrome 26% 0 33% 0 PDD-NOS 44% 57% 43% 25% Mother Age 41:01 (5:07) 43:06 (5:02) 40:02 (5:08) 40:08 (5:06) 31–52 38–51 31–52 29–53 Education No schooling 7% 0 10% 0 Elementary or Secondary 39% 43% 38% 72% University 54% 57% 52% 28% Father Age 42:04 (6:01) 45:05 (6:04) 41:02 (5:10) 43:0 (5:04) 32–56 39–54 32–56 27–54 Education No schooling 4% 0% 5% 3% Elementary or Secondary 46% 43% 48% 68% University 50% 57% 47% 29% Inclusion and Parents of Children with Autism / 165 of autistic symptoms and the standardized score of adaptive behavior was performed in order to compare differences among subgroups. Significant differences were found among the three groups, F(4, 110) ⫽5.33, p ⫽.001, ⌳⫽0.702. Post-hoc analyses showed that children in non-segregated mainstream classrooms had significantly lower scores on autistic symptoms than those in segregated mainstream classrooms (p⫽.006) and in special schools (p⬍.001). They also showed higher scores on adaptive behavior than those in segregated mainstream classrooms (p⫽ .025) and in special schools (p⫽.009). However, the differences were non-significant between children in segregated mainstream classrooms and in special schools either on adaptive behavior scores or severity of autistic symptoms (p⫽1.00). Procedure and Instruments All families in the main support group of the city were invited to participate. Conditions required for participation were a confirmed clinical diagnosis of ASD in the child and for him/her to be attending a school in the city of Seville. Initial telephone contact with the families led to a personal interview. The assessment of the level of satisfaction was obtained using a six-point likert-type scale of eight items (see Table 3). Each item was read out by a member of the research team to the parent, who had to provide a score from 1 to 6, with 1 as the lowest level of satisfaction (“very unsatisfied”) and 6 the highest (“very satisfied”). Internal reliability for all the items was computed and reached .94 (Cronbach’s ␣ ). The scale was complemented with a specific TABLE 3 Description of items on the parental satisfaction scale 1. Teachers’ attitude towards my child. Indicates the level of satisfaction of parents with teachers and whether a positive or negative attitude towards the child’s educational achievement possibilities are perceived. 2. Teachers’ attitude towards me. It evaluates satisfaction with the type and quality of personal and affective communication that, as parents, they have with the teachers of their children. 3. Teachers’ training in relation to my child’s difficulties. It estimates the degree of satisfaction with the educators’ perceived knowledge in relation to ASD and how to deal with them. 4. Stability of teachers in their post. Explores perception of the time teachers have been in their current post and school. Constant change of teachers had been a traditional complaint of parents of pupils with ASD in previous years in the Spanish educational system. 5. Availability of resources with which to help my child. Reflects perception of resources, both material (such as curricular materials or physical space related), and staff related. 6. Counseling in the way of dealing with a child with ASD. It assesses parents’ satisfaction with the amount and quality of communication of educational content. 7. The degree in which the school takes their opinion into consideration. Reflects perception of willingness to accept parental collaboration with the school 8. Activities and programs carried out with the child. Relates to perception of the content of educational programs and curricula. 166 / Education and Training in Developmental Disabilities-June 2008 question about the parents’ perception of the need for the child to change school. This item was intended to reflect possible situations of strong disagreement with the child’s education and dissatisfaction with educational provision. In order to assess the children’s adaptive behavior the Spanish version of the Inventory for Client and Agency Planning (ICAP) (Bruininks, Hill, Woodcok, & Weatherman, 1993) was administered. This instrument provides information on four areas of adaptive behavior: motor abilities, social, life, and community skills. The version used in this study has been adequately validated in Spain (Montero, 1993), showing sufficient discrimination among groups of persons with and without intellectual disability, and among groups of persons with intellectual disability in ordinary classrooms, part-time special education and full-time special educational settings, as well as different needs for supported residential settings. The Inventory for Autism Spectrum Disorders (IDEA) (Rivie`re, 1997) was used to assess the severity of children’s autistic symptoms. This scale is used widely by Spanish clinicians and covers twelve dimensions of the Autism Spectrum, grouped into four major categories: social development, language, cognitive and behavioral flexibility and representational abilities. Internal reliability for the IDEA in this study was .93. Results Parental Satisfaction The degree of overall parental satisfaction was high, with a mean score for all parents of 4.6 (SD ⫽1.4) (see Figure 1). Most parents’ mean scores centered round the highest values: fiftyseven percent of the parents had a mean score higher than five. Figure 2 shows a detailed analysis of parental satisfaction for each of the items in the scale. In all eight items means were in the higher scores, ranging from 4.0 to 5.2. Overall satisfaction, therefore, responded to consistently high satisfaction in all items. In agreement with these scores, the parents’ wishes to remove their children from the Figure 1. Overall parental satisfaction with educational provision. Inclusion and Parents of Children with Autism / 167 present school were not high. Seventy percent seemed to be overall satisfied with the current school, with only 26% claiming that the child would be best provided for in another one. AK-means cluster analysis with parental satisfaction responses was carried out. Following an exploratory analysis with 2, 3 and 4 clusters, three groups were selected as most illustrative. With a minimum distance between initial cluster centers of 8.54, convergence was reached following four iterations. The clusters were labeled as (I) parents moderately satisfied, (II) parents globally dissatisfied and (III) parents globally satisfied. The greatest group was composed of globally satisfied parents (66 %), followed by those moderately satisfied (20 %). The group presenting least satisfaction was also the smallest (14 %). Mean parental satisfaction scores are presented in Figure 3 for each cluster and item. Overall mean scores were 3.5 (SD ⫽0.5) for parents with moderate satisfaction, 1.9 (SD ⫽ 0.7) for the globally dissatisfied and 5.5 (SD ⫽ 0.5) for the globally satisfied group. It is interesting to note that, although groups are mainly reflecting degrees of satisfaction with educational provision, the group with moderate satisfaction actually scored lower on satisfaction with resources (item 7) than the less satisfied group. As expected, there was a relationship between parents’ wish to change their child from the current school and their cluster membership, ␹ 2 (2, N⫽58) ⫽13.51, exact p⫽.001. Parents globally satisfied seemed to be in disagreement with school change (88 %), whereas in the globally dissatisfied group the opposite was found (71 % favor change). Fortyfive percent of the moderately satisfied group wished to change the child’s school. Parental Satisfaction and Educational Placement There were significant differences between the satisfaction of parents of children in special (M⫽5.1, SD ⫽1.1) and in mainstream schools (M⫽4.0, SD ⫽1.6), t (46.7) ⫽⫺3.01, p⫽.001. Average differences in the different items ranged from 0.8 to 1.8. The differences were statistically significant in all items (p⬍.05), except for item 7 (“degree in which the school takes their opinion into consideration”) (see Figure 4). Figure 2. Parental satisfaction for each item of the scale. 168 / Education and Training in Developmental Disabilities-June 2008 A Kruskall-Wallis analysis showed significant differences among the three subgroups (segregated mainstream placement, non-segre- gated mainstream placement and special classes) in parental satisfaction, ␹ 2 (2, N⫽ 60) ⫽8.941, p⫽.011. In order to determine Figure 3. Scores for each type of parent group. Figure 4. Parental satisfaction for each item and type of school. Inclusion and Parents of Children with Autism / 169 the specific role of the actual school environment (special versus regular) as opposed to differences stemming from inclusion in regular classrooms and contact with other children, the satisfaction of the parents of pupils in segregated mainstream placement was then compared to that of parents of children in special schools. Overall significant differences were found between the mean scores of these groups of parents (M⫽3.3 and 5.1, and SD ⫽ 1.5 and 1.1, for segregated mainstream and special school placements, respectively, U⫽ 30.00, z⫽⫺3.01, p⫽.003). Mean scores were higher in special schools for all items (p⬍ .009), except for item 1 (“teachers’ attitude towards my child”). The same group of parents with children in segregated mainstream placement was then compared to the parents of the rest of the children attending regular schools (non-segregated mainstream placement). No differences were evident in parental satisfaction of these two groups of parents (for parents of children in non-segregated special placement, M⫽4.3, SD ⫽1.5, U⫽ 46.50, z⫽⫺1.44, p⫽.150). The only individual item in which there were significant differences was that related to “the degree in which the school takes their opinion into consideration” (item 7) (M⫽2.4, SD ⫽1.8 for segregated mainstream placement and M⫽4.6, SD ⫽1.5, for non-segregated, U⫽ 198.00, z⫽⫺2.64, p⫽.008). Differences in overall mean scores between special schools and non-segregated mainstream regular classrooms were also non-sig- nificant (U⫽252.00, z⫽⫺1.53, p⫽.125). There were, however, more positive perceptions of teachers attitude towards the children (item 1) (U⫽210.50, z⫽⫺2.63, p⫽.008), of teacher training (item 3) (U⫽198.00, z⫽ ⫺2.64, p⫽.008) and marginally of teachers’ attitude towards the parent (item 2) (U⫽ 226.00, z⫽⫺2.43, p⫽.019) in special schools. Parental group membership was related to type of school, ␹ 2 (2, N⫽60) ⫽9.68, exact p ⫽.007. Seventy-five percent of parents globally dissatisfied and of those moderately dissatisfied were in mainstream schools. Sixtyeight percent of those satisfied were in special schools. There was also a significant association between parental group and subgroups for educational placement, ␹ 2 (4, N⫽60) ⫽ 11.01, exact p⫽.023. Twenty-eight percent of parents globally satisfied were in non-segre- gated mainstream placement and 4% in segregated mainstream placement. Fifty percent of both those globally and moderately dissatisfied were in non-segregated mainstream placement, with 25 % in each segregated and non-segregated mainstream placement in both groups. Parental Satisfaction and Other Variables Age of the child did not seem to be related to parental satisfaction. There were no differences in satisfaction means among parents with children at different educational levels (Kindergarten, Elementary, Secondary and Post-secondary), ␹ 2 (3, N⫽60) ⫽2.59, p⫽ .459, nor were there differences in the children’s ages in the different clusters, F(2, 58) ⫽0.92, p⫽.404. There were no differences in the adaptive behavior of children, ␹ 2 (2, N⫽59) ⫽2.07, p ⫽.356, or in the severity of their autistic symptoms, ␹ 2 (4, N⫽59) ⫽5.06, p⫽.080, among parental clusters. Item and global scores for parental satisfaction in fathers with high school education was compared to fathers with college level education (three fathers with no formal education were excluded from this analysis). Differences were only found on item 2 (“attitude towards me”) (M⫽4.7, SD ⫽1.6 for college graduates and M⫽5.6, SD ⫽0.9, for high school fathers, U⫽266.50, z⫽ ⫺2.30, p⫽.021). No differences were found on an analogous analysis with mothers’ educational level. No relationship was found between parental educational level and cluster membership. Discussion Parents in this study have a positive perception of the education their children are receiving. Overall mean satisfaction is 4.6 on a scale of six, and only a quarter of the parents would like to change their child’s school. In a cluster analysis using all scores on the parental satisfaction scale, a majority of parents fell within fully or moderately satisfied groups. These findings are in line with those of studies car- 170 / Education and Training in Developmental Disabilities-June 2008