Contextual antecedents of co-teaching efficacy: Their influence on students with disabilities' learning progress, social participation and behaviour improvement

Teaching and Teacher Education 47 (2015) 218e229

Contents lists available at ScienceDirect

Teaching and Teacher Education journal homepage: www.elsevier.com/locate/tate

Contextual antecedents of co-teaching efficacy: Their influence on students with disabilities' learning progress, social participation and behaviour improvement Vasilis Strogilos a, *, Abraham Stefanidis b, 1 a b

Department of Early Childhood Education, University of Thessaly, Argonafton & Filellinon St., Volos 38221, Greece Department of Management, The Peter J. Tobin College of Business, St. John's University, 8000 Utopia Parkway, Bent Hall 326, Queens, NY 11439, USA

h i g h l i g h t s
We study teachers' attitudes towards curriculum modification, mixed ability groups, mainstream teachers' involvement.
We examine the role of the above factors in perceived co-teaching efficacy for students with disabilities.
Teachers consider that all three factors improve social participation.
Teachers consider that all three factors improve behaviour.
Teachers consider that mixed ability groups and mainstream teachers' involvement improve learning progress.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 28 February 2014 Received in revised form 5 January 2015 Accepted 12 January 2015 Available online 30 January 2015

This study examines whether the attitudes of co-teachers towards the efficacy of co-teaching for students with disabilities are related to their preferences towards the modifications of the curriculum, the participation of students with disabilities in mixed ability groups and the active involvement of mainstream education teachers in co-teaching students with disabilities. A survey questionnaire was used to explore 400 co-teachers' attitudes towards the efficacy of co-teaching. In general, the findings demonstrated that the attitudes of co-teachers towards specific activities and responsibilities are positively related to their attitudes towards the social participation, behaviour improvement and learning progress of students with disabilities. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Inclusive classroom Co-teaching Efficacy Curriculum modification Mixed ability groups Mainstream teacher

For the things we have to learn before we can do them, we learn by doing them. Aristotle (384e322 BC), Nicomachean Ethics.

1. Introduction International and national laws and policies have supported the practice of inclusive education for students with disabilities. The * Corresponding author. Tel.: þ30 242 100 6360. E-mail addresses: [email protected] (V. Strogilos), (A. Stefanidis). 1 Tel.: þ1 718 990 6713. http://dx.doi.org/10.1016/j.tate.2015.01.008 0742-051X/© 2015 Elsevier Ltd. All rights reserved.

[email protected]

development of inclusive schools where learning and participation is provided for all students seems to be an important goal for most education systems around the world. As a result, the number of students with disabilities in mainstream schools has increased and mainstream classrooms nowadays include diverse student populations. As Nind and Wearmouth (2006) indicate, inclusive schools are those which make major changes to their organization and processes to meet the needs of diverse populations. In response to these recent trends, many schools have promoted the development of different service delivery options for students with disabilities. Among those options is the growth of co-teaching between mainstream education teachers (METs) and special education teachers (SETs). According to Mastropieri and Scruggs (2006), co-teaching usually consists of a MET and a SET in a

V. Strogilos, A. Stefanidis / Teaching and Teacher Education 47 (2015) 218e229

mainstream classroom, where students with and without disabilities learn together. Roth and Tobin (2004) indicate that ‘coteaching is not just a way of going about the everyday work of accomplishing a teacher's task but, equally important, it is a way of changing the way one teaches… about developing as teachers while teaching (p. 165)’. As Thousand, Villa, and Nevin (2006, p. 241) support, ‘coteaching is a vehicle for bringing together people with diverse backgrounds and interests to share knowledge and skills to generate novel methods to individualize learning’. Similarly, Murphy and Beggs (2010) argue that co-teaching promotes a more democratic approach to teacher education and classroom teaching, and identify a positive impact for student teachers, classroom teachers, children and teacher educators. Although co-teachers have indicated several positive outcomes for themselves and their students with disabilities (Saloviita & Takala, 2010; Scruggs, Mastropieri, & McDuffie, 2007; Strogilos, Tragoulia, & Kaila, 2015; Thousand et al., 2006), research on the efficacy of co-teaching remains limited (Cook, McDuffie-Landrum, Oshita, & Cotheren-Cook, 2011; Murawski & Swanson, 2001). Research results on the efficacy of co-teaching is still considered sparse and inconclusive (Cook et al., 2011), since some studies have indicated a medium to strong positive relationship between student outcomes and co-teaching (McDuffie, Mastropieri, & Scruggs, 2009; Murawski & Swanson, 2001; Rea, McLaughlin, & WaltherThomas, 2002), while others have failed to identify benefits for students with disabilities in co-taught classrooms (Boudah, Schumacher, & Deshler, 1997; Murawski, 2006). 2. Theoretical framework The empirical research on co-teaching has focused on the impact of instructional activities and procedures (Magiera & Zigmond, 2005; McDuffie et al., 2009; Moin, Magiera, & Zigmond, 2009) and the responsibilities of co-teachers (Dieker, 2001; Harbort et al., 2007; Weiss & LIoyd, 2002) in order to evaluate its efficacy on the education of students with disabilities. Our theoretical framework consists of certain dimensions which have been identified in the research literature as important for the effective development of co-teaching. More specifically, this study conceptualises the efficacy of co-teaching in relation to inclusive practices within the mainstream class, elements that are considered important in the development of inclusive education. As Bandura (1997) argues, within organizations there are contextual influences, specific sociocultural factors, which have an impact on individual and organizational performance. The impact of these factors on efficacy is mediated through motivational and learning mechanisms operating at the individual and organizational level. As previous studies highlight (Nind & Wearmouth, 2006; Soukup, Wehmeyer, Bashinski, & Bovaird, 2007), meaningful participation of students with disabilities in the mainstream class is integrally linked with certain contextual variables that promote social and academic learning. Thus, the identification of contextual variables that are associated with students with disabilities' social participation, behaviour improvement and learning progress can be instrumental in enhancing our knowledge towards the understanding of a complex service delivery option such as that of coteaching (McDonnell, 2011). Based on the findings of past literature, we reconcile a theoretical framework that could potentially explain the scope and intensity of building efficient co-teaching environments. We empirically research whether the modifications of the curriculum, the participation of students with disabilities in mixed ability groups and the active involvement of METs in the education of students with disabilities could robustly explain the perceived efficacy of co-teaching in inclusive contexts. Thus, examining issues

219

related to the above activities and responsibilities are deemed important and were included in the development of this study, which aims at investigating the attitudes of co-teachers towards the efficacy of co-teaching and certain contextual antecedents of these attitudes. In designing this study we considered that the attitudes of coteachers towards the efficacy of co-teaching for students with disabilities could be related to their preferences towards the modifications of the curriculum during planning time, the participation of students with disabilities in mixed ability groups and the active involvement of METs in co-teaching students with disabilities. These instructional strategies and responsibilities, or contextual factors, promote learning progress and social participation according to the main values of inclusive education. Thus, in this study, the attitudes of co-teachers towards the efficacy of coteaching are being examined on the premise that the above activities and responsibilities promote the main values of inclusive education and thus justify the education of students with disabilities along with their typically developing peers. Since no other studies have investigated the efficacy of co-teaching in relation to certain antecedents, we consider the potential contribution of this study significant. 3. Co-teaching in Greece In this study, we investigate certain antecedents of the efficacy of co-teaching in Greek mainstream classes. Our motivation to look into the unique environment of Greece stems from the implementation of the country's recent policies that intend to integrate more students with disabilities in mainstream classes following the international inclusion movement. Greece was among the countries which agreed to follow the principles of the Salamanca Statement (UNESCO, 1994) for the education of students with disabilities within the mainstream education system. As a result, new legislation promoted the development of new pedagogical practices that promote inclusion (Law 2817/2000; Law 3699/2008). Among these practices, coteaching was introduced for the enhancement of the education of students with disabilities in mainstream classes. Since 2000, Greek legislation (Law 2817/2000) has provided a co-taught model in a move to improve the quality of education for students with disabilities within mainstream classes and to promote inclusion. However, research on the development of co-teaching as a means to inclusive education in Greece is very limited (Strogilos & Avramidis, 2013; Strogilos & Tragoulia, 2013; Strogilos et al., 2015). Since co-teaching is not necessarily understood or implemented in the same way across countries (Roth & Tobin, 2004), many differences concerning its development and praxis are evident in the Greek mainstream schools. Being responsible for all educational initiatives in Greek schools, the Ministry of Education promoted the development of co-teaching by employing SETs to directly work with a student with disabilities on a full time basis in all curriculum subjects in the mainstream class. Thus, these SETs are allocated by the Ministry not to a school or to a class, but to a student with a diagnosed disability, after the parents' request. It is obvious that the Greek Ministry of Education promoted the development of the ‘one teach one assist’ model of co-teaching, as a recent study by Strogilos & Tragoulia (2013) has indicated. Students who learn in co-taught classrooms include those with autism, sensory and physical disabilities, intellectual disabilities, health disorders and emotional and behavioural difficulties. Concurrently, although students with specific learning difficulties are accommodated in mainstream classes, the parents of these students cannot have a SET assigned to their children, since these students are partially educated in resource rooms. Similarly to the

220

V. Strogilos, A. Stefanidis / Teaching and Teacher Education 47 (2015) 218e229

US context (Scruggs et al., 2007), all students with disabilities in the Greek co-taught classes have a minimum academic and behaviour level, since Greek legislation (Law 3699/2008) still excludes students with severe disabilities from mainstream education. Due to the increasing implementation of co-teaching across the country, there is currently a lack of high qualified SETs such as those with an undergraduate or master's degree in special educational needs (SENs) for these posts. Yet, all SETs and METs hold at least a Bachelor's degree in Education. In this context, the Ministry has been employing qualified METs who have 400 hours of training in SENs or qualified METs with no qualifications in SENs to cover coteaching posts. It is worth mentioning that even though the Ministry of Education provides special education co-teachers with some training during their employment in co-taught classes, this training is neither exclusively focused on co-teaching nor available to METs. Finally, it should be noted that neither learning assistants nor health professionals are currently employed in Greek mainstream schools.

4. Review of the literature

Contrariwise, other studies have not identified significant benefits for students with disabilities in co-taught classrooms. In a meta-analysis of co-teaching research, Murawski and Swanson (2001) reported that co-teaching is a moderately effective procedure for influencing student outcomes; they identified an average effect size of 0.40 for all studies. The results of Murawski's (2006) study failed to identify significant differences for students with learning disabilities across resource classes, co-taught classes and mainstream classes without co-teaching; a finding which, according to the researcher, could be attributed to the uneven implementation of co-teaching due to lack of training. Vaughn et al. (1998) investigating the social functioning of students with learning disabilities, low to average achievers and high achievers in co-teaching and consultation/collaboration teaching settings, concluded that students in the consultation/collaboration setting demonstrated more positive outcomes concerning friendship quality and peer acceptance than students in the co-teaching settings. With regard to studenteteacher interactions, McDuffie et al. (2009, p. 507) observed that ‘students in non-co-taught settings interacted more with their teachers than did students in co-taught settings’, extending similar research findings identified by Magiera and Zigmond (2005).

4.1. Co-teaching efficacy in inclusive classrooms Research on the efficacy of co-teaching has taken into consideration three important dimensions, namely students' learning progress (e.g. Murawski, 2006; Rea et al., 2002), social participation (e.g. Boudah et al., 1997; Vaughn, Elbaum, Schumm, & Hughes, 1998) and behaviour improvement (e.g. Hang & Rabren, 2009; Rea et al., 2002). However, as mentioned above, research on the efficacy of co-teaching has provided different findings with regard to co-teaching outcomes. It appears that existing qualitative research provides some positive outcomes of co-teaching, whereas the results of experimental research on students' outcomes are mixed. As Scruggs et al. (2007) have indicated in their metasynthesis of qualitative research on co-teaching, most co-teachers believe that co-teaching has a positive effect on the academic and social performance of students with disabilities, due to the additional teacher attention and the benefit of imitating peers for appropriate behaviour. Walther-Thomas (1997) noted benefits for students with disabilities that grew stronger over time in twenty-three general and special educator teams in schools which were recognized for their innovative inclusive educational programs. In particular, she observed ‘positive feelings about themselves as capable learners, enhanced academic performance, improved social skills, and stronger peer relationships (p. 399)’. Similarly, conceptualizing coteaching as an inclusion program, Rea et al. (2002) identified better academic and behavioural outcomes for students with learning disabilities in co-taught classes than for those in pull-out programmes in two schools that followed these two different service delivery models. In Hang and Rabren's (2009) study, in which SETs co-taught with several METs, teachers and students also agreed that co-teaching improves the academic performance of students with disabilities and that it contributes positively to their behaviour. In addition, Welch (2000) highlighted, in a descriptive analysis of team teaching in which both teachers provided direct service to all students using various forms of team teaching, that students both with and without learning disabilities benefited in reading and spelling. Roth and Tobin (2004), using culturalehistorical activity theory research, argued that through co-teaching teachers are involved in an increasing range of actions and, according to their research (Roth & Tobin, 2002), three months of co-teaching can lead to important changes in individual classrooms and across a school.

4.2. Contextual variables for learning and participation in inclusive classrooms A number of contextual variables which are associated with the quality of education provided to students with disabilities in mainstream classrooms have been described in the literature. According to McDonnell (2011, p. 532) these include: ‘where instruction is provided, who provides instruction to students, how students are grouped for instruction, and what accommodations and modifications students use to perform successfully in various settings’. 4.2.1. Modifications of the curriculum Curriculum modifications are regarded as essential to accommodate the diverse needs of all learners under a universally designed curriculum (UDL) in which students' learning progress and participation increase in different learning areas (Rose, Meyer, & Hitchcock, 2005). Curriculum modifications include curriculum adaptations and augmentations. According to Soukup et al. (2007), curriculum adaptations are modifications that change the way the content of teaching is represented or presented through the use of supplementary aids (e.g. visual aids, technological means, advanced organizers), whereas curriculum augmentations refer to the additional instruction of specific strategies in order for students to acquire skills (e.g. learning how to learn strategies, cognitive strategies). The modification of the curriculum is considered an essential inclusive strategy for the education of students with disabilities in the mainstream class. As Janney and Snell (2006) indicate, the use of a range of instructional adaptations and modifications is required in inclusive classrooms, especially when it is combined with collaboration between co-teachers and the facilitation of positive interactions among students with disabilities and their developing peers. The authors consider that these elements are necessary in order to provide the development of an inclusive culture to diverse learners. Cardona (2003) holds that the mainstream classroom is the best place for all students, providing the appropriate teaching experiences within a peer group. However, she indicates that the success of inclusion is basically dependent on the willingness of schools and teachers to carry out the necessary instructional adjustments.

V. Strogilos, A. Stefanidis / Teaching and Teacher Education 47 (2015) 218e229

Despite the perceived importance of curriculum modifications, research has indicated absence or inconsistent use of these strategies (McDonnell, 2011). For example, Cardona (2003) found a moderate teacher acceptance of instructional adaptations by METs in Spain. According to the results of her study, METs do not perceive adaptations as highly feasible, effective or desirable. Kurth and Keegan (2012) evaluated the adaptations made by experienced and novice SETs, METs and teaching assistants in rural and urban schools in three US districts. Among their findings was that experienced educators created more simplified adaptations in comparison to novice teachers who created more functional alternative adaptations. In addition, METs' adaptations were of lower quality and clarity than those of SETs and teaching assistants. In addition, Lee, Wehmeyer, Soukup, and Palmer (2010) found that students with disabilities increase their engagement in academic-related responses, while they decrease their competing behaviours, and that teachers perform fewer classroom management activities if curriculum modifications are used. Yet, many research studies in this area report lack of curriculum modifications in co-taught classes for students with disabilities (Antia, 1999; Buckley, 2005; Strogilos et al., 2015). It appears that the education of students with disabilities in the co-taught class lacks appropriate teaching materials, differentiation in activities and opportunities for individualising the teaching process (Scruggs et al., 2007). 4.2.2. Participation in mixed ability groups The participation of students with disabilities in both whole class and small group work is regarded as essential for their social participation and academic progress. Teachers encourage the education of students with disabilities in whole class and small group arrangements as an alternative to one-to-one instruction. Although research has demonstrated that both approaches have benefits for the students with disabilities (McDonnell, 2011), many researchers hold that peers may be valuable sources of support for students with disabilities (Carter, Cushing, & Kennedy, 2009), since students can learn from one another through observational learning (Mechling, Gast, & Krupa, 2007). Other researchers have raised concerns about the low engagement of students with disabilities who are individually taught in inclusive classrooms (Magiera & Zigmond, 2005; Scruggs et al. 2007), while Nind and Wearmouth's (2006) review on the pedagogical practices that promote inclusion suggests that ‘the social and academic dimensions are integrally linked (p. 121)’. The authors also indicate that even in core subjects (e.g. maths and science) most research studies take a holistic perspective recognising the importance of the social context, whereas the individualised instruction is usually associated with ‘remediation’ approaches to the learning needs of pupils with disabilities. In their research study of an inclusive classroom where students with disabilities were encouraged to improve their abilities in literature selections through the use of literature circles, Blum, Lipsett, and Yocom (2002) found that these students improved both academically and socially. Similarly, Soukup et al.'s (2007) observational study showed that participation in whole class and small group work provides greater access to the general education curriculum, in comparison to individual physical arrangements (i.e. seated away from all other students). However, Logan, Bakeman, and Keefe (1997) found that students with disabilities display higher levels of engagement in the curriculum when they are taught in whole class rather than when they are taught in small groups. In their research review on the pedagogical approaches that can effectively include students with disabilities in the mainstream class, Rix, Hall, Nind, Sheehy, and Wearmouth (2009) highlighted that group work is included among the pedagogical practices that

221

enhance inclusion, since it is beneficial for the academic and social participation of students with disabilities. 4.2.3. The active involvement of METs in co-teaching students with disabilities One of the benefits of co-teaching is that METs and SETs bring their different training, perspectives and experiences in a class with diverse students. The input of the MET in a co-taught class is essential. As Murawski (2010) refers, these teachers have the general curriculum or subject area knowledge and they know what a typical student should be able to do in their class. Although SETs are the ones with the specialized training for students with disabilities, the input of the METs is considered equally important for the effectiveness of co-teaching. As Forlin, Keen, and Barrett (2008, p. 251) indicate ‘it is not legislation but the mainstream teachers themselves who are pivotal to the success of an inclusive model of education’. Research on the inclusion of students with disabilities in mainstream classes has reported that METs feel unprepared or even reluctant to educate students with disabilities in their classes. For example, in a research study conducted in eight schools by Idol (2006), METs reported that they prefer to accept students with disabilities in their classrooms when those students are accompanied by a teacher. Similarly, in an attitudinal study of 571 Queensland primary school teachers, Forlin (2001) found that even though METs did not appear to be overly stressed by inclusion, respondents reported several stressors. In particular, issues that related to a teacher's perceived self-competence and the behaviour of children with intellectual disabilities were the most stressful for teachers. In a recent review on the attitudes of METs, De Boer, Pijl, and Minnaert (2011) found that the majority of teachers have neutral or negative attitudes and that they do not feel knowledgeable, confident and competent to educate students with disabilities. More significantly, the researchers stated that they could not identify any studies that investigate the effects of teachers' attitudes on the social participation of students with disabilities in the mainstream school. Previous research has identified many inconsistencies in the implementation of METs' and SETs' roles in the co-taught classroom, such as lack of mutual planning time and inconsistency in instruction and behaviour management (Dieker, 2001; Fennick & Liddy, 2001; Hang & Rabren, 2009; Magiera & Zigmond, 2005; Strogilos & Tragoulia, 2013). As Fennick and Liddy (2001) argue, METs consider themselves responsible to plan the class curriculum and instruct all students, whereas SETs assume responsibility to modify the curriculum and to monitor the progress of the students with disabilities. Similarly, Magiera and Zigmond (2005) observed that METs spend significantly less time with students with disabilities when a SET is in their class. In addition, students with disabilities receive more individual teaching by the SET (Magiera & Zigmond, 2005), which sometimes may have counterproductive results for their social and academic inclusion (Strogilos & Tragoulia, 2013; Strogilos et al., 2015). 5. Aim of the study Co-teaching has gained popularity among teachers and researchers, but it seems that common sense has directed much of the debate, since there is currently little empirical research to support its efficacy. According to Cook et al.'s (2011) literature review, explanatory and descriptive research in this area provides some initial guidelines for effective co-teaching, but further research is necessary to better determine the effects of such an approach. In line with this, the current research aims to investigate co-teachers’ attitudes towards the efficacy of co-teaching for

222

V. Strogilos, A. Stefanidis / Teaching and Teacher Education 47 (2015) 218e229

students with disabilities. In particular, we examine whether the attitudes of co-teachers towards specific activities and responsibilities (i.e. the modifications of the curriculum, the participation of students with disabilities in mixed ability groups and the active involvement of METs in co-teaching students with disabilities) are related to their attitudes towards the efficacy of coteaching (learning progress, social participation and behaviour improvement) for students with disabilities in the co-taught class. Hence, the research questions of the current study are: 1. Do co-teachers who endorse the modification of the curriculum tend to agree that the efficacy of co-teaching for students with disabilities improves in the co-taught class? 2. Do co-teachers who endorse the participation of students with disabilities in mixed ability groups tend to agree that the efficacy of co-teaching for students with disabilities improves in the cotaught class? 3. Do co-teachers who endorse the active involvement of METs in co-teaching students with disabilities tend to agree that the efficacy of co-teaching for students with disabilities improves in the co-taught class?

6. Method According to Eagly and Chaiken (1993) attitudes are defined as ‘a psychological tendency that is expressed by evaluating a particular entity with some degree of favour or disfavour (p. 1)’. To explore the attitudes of co-teachers, a survey research in the form of a selfadministered questionnaire was used. The instrument was part of a larger investigation about co-teaching. In this research, we measured co-teachers' attitudes towards (a) the modification of the curriculum for students with disabilities, (b) the participation of students with disabilities in mixed ability groups, (c) the active involvement of METs in the education of students with disabilities in the co-taught class, and (d) the efficacy of co-teaching for students with disabilities. 6.1. Participants Having received the required approval from the Ethics Committee of the Greek Ministry of Education, we sent a participation invitation to all 58 Local Educational Authorities (LEAs) in Greece, in order to identify schools that employ co-teaching. Twenty five of the 58 LEAs provided contact information of the target schools and thus, only schools from these LEAs were included in the study. As a result, all mainstream schools that used co-teaching in these 25 LEAs were invited to participate in the study. Each of the 25 LEAs included in the study employed a minimum of 20 to a maximum of 120 co-teaching classrooms. The participating schools were located within fifteen prefectures geographically dispersed all over the country; eight of them were urban, while three were suburban and four were rural. The target audience for the survey were METs and SETs providing instruction to students with disabilities in co-taught classrooms, from kindergarten to secondary schools. In particular, an invitation email was sent to the head teachers requesting the schools' agreement for the survey. Upon agreement, we contacted head teachers with the request to provide either co-teachers' email addresses or to indicate co-teachers' preference for a postal questionnaire to the school. We decided to use both an online and a postal version of the survey, given the varying preferences of the participants. We assured anonymity and confidentiality in both delivery options. In total, 317 online and 483 postal questionnaires were sent out to 300 METs and 500 SETs during a three-month

research period. The online questionnaire was uploaded on a web-based platform, and a unique link was sent out to the acquired co-teachers' email addresses. For the postal survey we sent packages containing the survey, a cover letter and a pre-paid return envelope with no personally identifiable information. Three weeks after the original release of the survey, we conducted follow-up phone calls or sent a reminder email in order to increase the response rate. Having approached the co-teachers through the LEAs and having provided both online and hard copy surveys, we acquired a very satisfying response rate, considering other research attempts (Burns & Ysseldyke, 2009; Flewitt & Nind, 2007). We received a total of 181 online and 219 postal (n ¼ 400) usable responses. In particular, respondents included a total of 155 METs and 245 SETs, which resulted in a 51.6% and 49% response rate, respectively. The total response rate was 50%. Respondents were diversified in terms of their gender, education, teaching experience, school grade level, and students' type of disability. Eighty-seven percent of the respondents were females (n ¼ 348) and 13% were males (n ¼ 52). Of the 155 participating METs, 7 (4.6%) held a master's in mainstream education. Of the 245 SETs, 67 (27.4%) had received a bachelor's in SEN, 27 (11%) held a master's and 1 (0.4%) held a PhD, while 54 SETs (22%) had attended a 400-hour training in SENs. Ninety-six (39.2%) of the SETs had a bachelor's in mainstream education. The majority of the respondents were primary teachers (73%), while 23% and 4% were working at kindergarten and secondary schools, respectively. The mean age of the METs was 42.14 (SD ¼ 7.23) years, while the mean age of the SETs was 27.29 (SD ¼ 5.67) years. The teaching experience of the METs ranged from 1 to 34 years (M ¼ 15.99, SD ¼ 8.32). The teaching experience of the SETs ranged from 2 months to 8 years (M ¼ 2.77, SD ¼ 3.43). The years of co-teaching experience of the METs ranged from 2 months to 8 years (M ¼ 1.39, SD ¼ 1.22), while the years of co-teaching experience of the SETs ranged from 1 month to 11 years (M ¼ 1.37, SD ¼ 1.64). Respondents indicated that they taught students diagnosed with autism (n ¼ 248, 62%), intellectual disability (n ¼ 85, 21%), physical disabilities (n ¼ 27, 7%), hearing (n ¼ 27, 7%) and visual (n ¼ 13, 3%) impairments. The design of the national co-teaching programme in Greek schools allows us to consider the demographic characteristics of the participants representative of the national demographics. As we explained in Section 3, the Ministry of Education employs SETs with various qualifications in co-taught classrooms (i.e. SETs with a bachelor's or master's in SENs, METs with 400 hours of training in SENs, and teachers with a bachelor's in mainstream education). In an effort to compare the demographic data of our sample to the national population of co-teachers (N ¼ 2380 at the time of the research), we requested these data from the Ministry of Education. However, we were not granted access to such information. 6.2. Instrument A survey questionnaire was developed. We measured whether co-teachers' attitudes towards the modification of the curriculum, the participation of students with disabilities in mixed ability groups, and the active involvement of METs in co-teaching students with disabilities are related to their attitudes towards the efficacy of co-teaching for students with disabilities. Four scales were developed based on information from existing literature (e.g. Fennick & Liddy, 2001; Hang & Rabren, 2009; Koster, Timmerman, Nakken, Pijl, & van Houten, 2009; Lee et al., 2010; Soukup et al., 2007) and from previous qualitative investigations in Greek co-taught classrooms (Strogilos & Tragoulia, 2013; Strogilos et al., 2015). In particular, in the first scale, we asked the participants to indicate their agreement towards the modifications of the curriculum for the students with disabilities during planning time. We

V. Strogilos, A. Stefanidis / Teaching and Teacher Education 47 (2015) 218e229

used a 5-point Likert scale, where 1 was ‘strongly disagree’ and 5 was ‘strongly agree’. This scale was designed in order to measure co-teachers' preferences towards general modifications of the curriculum and not towards more specialized adaptations (e.g. give more time to or provide diverse activities for some students with disabilities). The scale included the following items: ‘Select appropriate teaching methods for the student(s) with disabilities’, ‘Select appropriate educational resources for the student(s) with disabilities’, ‘Modify the content of teaching for the student(s) with disabilities’ and ‘Select goals and objectives for the student(s) with disabilities IEP’. In the second scale, we asked co-teachers to indicate their agreement with regard to the participation of students with disabilities in different mixed ability groups in the class. A 5-point Likert scale was adopted (1 was ‘strongly disagree’ and 5 was ‘strongly agree’). We considered the participation of students with disabilities in mixed ability groups as an important instructional strategy and hence as a contextual dimension of a co-taught class. The scale included the following items: ‘Students with disabilities are integrated in whole-class activities’, ‘Students with disabilities receive individual instruction out of the co-taught classroom’ (reversed item) and ‘Students with disabilities are integrated in mixed ability small groups’. Both the ‘modifications of the curriculum’ and the ‘participation of students with disabilities in mixed ability groups’ scales were developed by the researchers. However, in order to design these scales, we took into consideration the observational studies of Lee et al. (2010) and Soukup et al. (2007) and ‘The Teaching Adaptation’ scale designed by Cardona (2003), which refer to curriculum modifications and instructional grouping arrangements for students with disabilities in mainstream classes. In the third scale, the co-teachers were asked to state their agreement to whether METs should have an active role in 9 different responsibilities (items) that refer to co-teaching students with disabilities (see Table 1). A 5-point Likert scale was adopted (1 was ‘strongly disagree’ and 5 was ‘strongly agree’). To develop this scale six items were adapted from Fennick and Liddy's (2001) ‘Collaborative Teaching Responsibilities’ scale, while two items were retained as they appeared in their original scale. Also, one item was developed by the researchers. Sample items included: ‘Cope with behaviour problems of students with disabilities’, ‘Provide individual instruction to students with disabilities’. Finally, for the scale that measured the efficacy of co-teaching, 19 items were used. We asked co-teachers to state their agreement with regard to the social participation, behaviour Table 1 Factor analysis of the mainstream education teachers' co-teaching responsibilities. Co-teaching responsibilities Demonstrate hands-on techniques to students with disabilities Cope with behaviour problems of students with disabilities Arrange physical environment for students with disabilities Teach study strategies to students with disabilities Monitor progress of students with disabilities Work with the group of students with disabilities Monitor behaviour of students with disabilities Assign work to students with disabilities Provide individual instruction to students with disabilities Eigenvalue Percent of variance

.80 .79 .79 .78 .77 .75 .74 .73 .57 5.04 56.01

Note: Extraction: principal component analysis; Rotation: varimax with Kaiser normalization.

223

improvement and learning progress of students with disabilities in the co-taught class. A 5-point Likert scale was adopted (1 was ‘strongly disagree’ and 5 was ‘strongly agree’). In this scale, all items were designed to measure the efficacy of co-teaching for students with disabilities compared with the efficacy of teaching in singletaught mainstream classrooms. We intended to collect information specific to the learning progress, social participation and behaviour improvement of the students with disabilities in the cotaught class. We adapted seven items from Koster et al.'s (2009) ‘Social Participation’ scale, two items from Fennick and Liddy's (2001) ‘Opinions for Collaborative Teaching’ scale and four items from Hang and Rabren's (2009) ‘Teacher's Perspectives Survey’ scale, while three items were adapted from Larrivee and Cook's (1979) ‘Survey of teachers' opinions relative to mainstreaming special-needs children’ scale and three items were developed by the authors based on previous qualitative research findings (Strogilos & Tragoulia, 2013; Strogilos et al., 2015). The three items developed by the authors were: ‘Student(s) with disabilities control their behaviour more effectively in a co-taught classroom than in a single-teacher general education classroom’, ‘Students with disabilities are more disciplined in a co-taught classroom than in a single-teacher general education classroom’ and ‘Students with disabilities in a co-taught classroom are more confident learners’. The survey questionnaire was originally developed in English. Following its development, two experts in survey educational research were invited to review several drafts of the questionnaire. Then, we asked three bilingual reviewers to translate our instrument in Greek. Standard back translation procedures were followed. The reviewers provided recommendations for improvements. To ensure the validity of the research instrument, we conducted a pilot-test with 24 co-teachers. The results of the pilot stage were used to further confirm validity, clarity and relevance of the items, while a few minor comments were addressed. A number of reversed anchor items were included in the survey in order to control for the possible development of response patterns by the respondents. 6.3. Analysis The first part of the data analysis included the calculation of descriptive statistics. We further conducted factor analyses in order to test the structure of the four scales (Johnson & Wichern, 2007). Following the factor analyses, we evaluated the internal degree of reliability with the Cronbach's alpha test. We then calculated the mean values and standard deviations of the variables and we performed correlations and hierarchical regression analyses (Cohen, Cohen, West, & Aiken, 2013). We also employed variance inflation factor (VIF) diagnostics in order to account for multicollinearity among the independent variables. The VIFs of the deriving models were lower than 2.62. The factor analysis of the ‘modifications of the curriculum for the students with disabilities’ scale rendered a one-factor solution that included all four items and accounted for 73.83% of the total variance. The Cronbach's alpha coefficient of the construct was .88. The factor analysis of the ‘students with disabilities’ participation in mixed ability groups' scale yielded a one-factor solution that included all three items and explained 47.93% of the total variance. The Cronbach's alpha coefficient of the instructional grouping arrangements construct was .62. The factor analysis of the ‘active involvement of METs in co-teaching students with disabilities’ scale provided a one-factor solution that included all nine items and accounted for 56.01% of the total variance (see Table 1). The Cronbach's alpha coefficient of this construct was .89. We assessed the structure of the ‘co-teaching efficacy for students with disabilities’ scale performing a factor analysis on the 19

224

V. Strogilos, A. Stefanidis / Teaching and Teacher Education 47 (2015) 218e229

items included in the construct. The factor analysis led to the extraction of three factors, which we named ‘social participation’, ‘behaviour improvement’ and ‘learning progress’ (see Table 2). The three-factor solution explained 61.02% of the total variance. The ‘social participation’ factor included four items and explained 25.87% of the variance, the ‘behaviour improvement’ factor included three items and explained 19.00% of the variance, while the ‘learning progress’ factor included three items and explained 16.14% of the variance. The Cronbach's alpha coefficients were .78, .78 and .60 for the first, the second and the third factor, respectively. Taking into account the exploratory nature of this research, we considered the Cronbach's alpha coefficients acceptable (Nunnally, 1967). In the hierarchical regression analyses, we included the factors extracted from the first three scales as independent variables, while the three factors that derived from the ‘co-teaching efficacy’ scale served as our dependent variables. Although our research focused

Table 2 Factor analysis matrix of co-teaching efficacy for students with disabilities. Components Social Behavioural Learning participation improvement progress .77 Students with disabilities make more friends in a co-taught classroom than in a single teacher mainstream classroom .75 Students with disabilities interact more with their classmates in a co-taught classroom than in a single-teacher mainstream classroom .68 In a co-taught classroom students with disabilities are more likely to be members of a group of friends .64 In a co-taught classroom students with disabilities are more frequently invited by their classmates to team games .57 Students with disabilities are more likely to participate in teamwork activities in a co-taught classroom than in a single teacher mainstream classroom Students with disabilities are more disciplined in a co-taught classroom than in a single-teacher mainstream classroom Student(s) with disabilities control their behaviour more effectively in a cotaught classroom than in a singleteacher mainstream classroom The behaviours of students with disabilities are better in a co-taught classroom than in a single-teacher mainstream classroom For students with disabilities, teaching in a single-teacher mainstream classroom is more efficient use of educational time than teaching in a co-taught classroom The instructional support provided to students with disabilities in a cotaught classroom is worse than that provided in a single-teacher mainstream classroom Students with disabilities learn more in a co-taught classroom than in a singleteacher mainstream classroom Eigenvalue 4.38 Percent of variance 25.87

.82

.80

.66

.79

.69

.59

1.31 19.00

1.03 16.14

Note: Extraction: principal component analysis; Rotation: varimax with Kaiser normalization.

on the curriculum modifications, grouping arrangements and METs' active involvement in the co-taught class, we considered it necessary to include teachers' characteristics as sources of variation for the efficacy of co-teaching. Thus, demographic characteristics of teachers (i.e. gender, age, education, co-teaching experience and role in the co-taught class -MET or SET-) were included in the analysis as control variables. Age, education and co-teaching experience were measured in years. Finally, in our analysis, we undertook the post-hoc diagnostic Harman single factor test to ensure that any common method variance issues would be limited (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). The test results did not reveal any unusual variations in our analysed data. 7. Results The descriptive statistics are presented in Table 3. With regard to their specific activities and responsibilities, the co-teachers indicated a moderate endorsement towards the curriculum modifications for students with disabilities (M ¼ 3.44, SD ¼ 1.03), the participation of students with disabilities in mixed ability groups (M ¼ 3.83, SD ¼ .75) and the active involvement of METs in the education of students with disabilities in the co-taught class (M ¼ 3.76, SD ¼ .63). In addition, based on the mean values, the coteachers agreed that co-teaching has a positive effect on students with disabilities' progress. In particular, the learning progress of students with disabilities in the co-taught class displayed the highest mean score (M ¼ 4.09, SD ¼ .68), followed by social participation (M ¼ 4.03, SD ¼ .59) and behaviour improvement (M ¼ 3.97, SD ¼ .69). Table 3 presents the correlation matrix of the studied independent and dependent variables. Several correlations were identified among the variables. We observed that the improved ‘Social Participation’ is positively correlated with the three independent variables, i.e. the curriculum modifications (r ¼ .13, p ¼ .05), the students' participation in mixed ability groups (r ¼ .18, p ¼ .01) and the active involvement of METs (r ¼ .11, p ¼ .05). Similarly, the ‘Behaviour Improvement’ variable was positively correlated with the curriculum modifications (r ¼ .15, p ¼ .01), the students' participation in mixed ability groups (r ¼ .15, p ¼ .01), and the active involvement of METs (r ¼ .12, p ¼ .05). Regarding the ‘Learning Progress’ variable, we observed a statistically significant correlation for the students' participation in mixed ability groups (r ¼ .17, p ¼ .01) and the active involvement of METs (r ¼ .21, p ¼ .05). We further assessed the observed correlations among the dependent and independent variables conducting three two-step hierarchical regression analyses. The regression analysis for the ‘Social Participation’, the ‘Behaviour Improvement’ and the ‘Learning Progress’ of the students with disabilities are presented in Table 4. The first step of the hierarchical regression analysis on the efficacy regarding ‘Social Participation’ provided a statistically significant model (F ¼ 2.05, p  .10). Age and the role of co-teacher explained 1% of the total variance of the ‘Social Participation’ variable. The second step regression model (F ¼ 3.99, p  .001) revealed that co-teachers’ age and role, the curriculum modifications, the students' participation in mixed ability groups and the active involvement of METs explained 6% of the total variance of the ‘Social Participation’ variable. The coefficients of the control variables age (b ¼ .16, p  .05) and METs' role (b ¼ .21, p  .01) were both positive. In addition, the coefficients for the curriculum modifications (b ¼ .11, p  .01), the students' participation in mixed ability groups (b ¼ .16, p  .001) and the active involvement of METs (b ¼ .09, p  .10) were positive. These findings suggest that the older teachers and the SETs tend to agree more that co-teaching improves the social participation of students with disabilities. In

V. Strogilos, A. Stefanidis / Teaching and Teacher Education 47 (2015) 218e229

225

Table 3 Descriptive statistics and correlations matrix between educators' attitudes and demographics. Variable 1. 2. 3. 4. 5. 6.

Social participation Behaviour improvement Learning progress Modifications of the curriculum Participation in mixed ability groups Active involvement of mainstream education teachers' in co-teaching 7. Gendera 8. Age 9. Education 10. Co-teaching experience 11. Roleb

Mean

Std. deviation

1

2

3

4

5

4.03 3.97 4.09 3.44 3.83 3.76

.59 .69 .68 1.03 .75 .63

1.00 .55** .41** .13* .18** .11*

6

1.00 .44** .15** .15** .12*

1.00 .06 .17** .21**

1.00 .08 .01

1.00 .11*

1.00

.87 33.05 16.48 1.38 .61

.34 9.61 1.43 1.49 .49

.04 .01 .02 .05 .09

.01 .03 .04 .03 .01

.04 .03 .00 .02 .07

.03 .04 .06 .14** .01

.08 .11* .02 .04 .09

.00 .15** .05 06 .16**

7

8

9

10

11

1.00 .00 .03 .02 .05

1.00 .03 .19** .75**

1.00 .15** .15**

1.00 .01

1.00

Note: n ¼ 400. * Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2-tailed). a 0 ¼ male, 1 ¼ female. b 0 ¼ mainstream education teacher, 1 ¼ special education teacher.

Table 4 Regression analyses on co-teaching efficacy for students with disabilities. Variable

Social participation First step

a

Gender Age Education Co-teaching experience Roleb Modifications of the curriculum Participation in mixed ability groups Active involvement of mainstream education teachers' in co-teaching R2 Adj. R2 D R2 F-statistic

Behaviour improvement Second step

First step

Learning progress

Second step

First step

Second step

b

t

b

t

b

t

b

t

b

t

b

t

.05 .17* .06 .03 .23**

1.04 2.17 1.13 .55 2.86

.04 .16* .06 .02 .21** .11* .16*** .09J

.53 2.94 1.29 .01 1.39 2.28 1.14 .25

.01 .14J .06 .05 .13

.14 1.75 1.16 .94 1.59

.01 .13 .07 .06 .11 .15** .12* .11*

.27 1.60 1.31 1.24 1.38 2.97 2.41 2.17

.05 .08 .02 .04 .13J

.96 .94 .29 .67 1.67

.04 .07 .02 .04 .11 .05 .14** .18***

.75 .86 .42 .68 1.41 1.01 2.88 3.63

.03J .01J e

2.05J

.08*** .06*** .05*** 3.99***

.01 .00 e .86

.06*** .05*** .05*** 3.36***

.01 .00 e .81

.07*** .05*** .06*** 3.83***

Notes: n ¼ 400; standardized regression coefficients are shown. J p  .10, *p  .05, **p  .01, ***p  .001. a 0 ¼ male, 1 ¼ female. b 0 ¼ mainstream education teacher, 1 ¼ special education teacher.

addition, those co-teachers who endorse more the curriculum modifications, the students' participation in mixed ability groups and the active involvement of METs tend to agree that the social participation of students with disabilities improves in the co-taught classrooms. Although the first step of the hierarchical regression analysis on the ‘Behaviour Improvement’ variable did not render a significant model (F ¼ .86, p ¼ .51), the second step of the regression led to a statistically significant one (F ¼ 3.36, p  .001). The curriculum modifications, the students' participation in mixed ability groups, and the active involvement of METs explained 5% of the total variance of the ‘Behaviour Improvement’ variable. The coefficients of the curriculum modifications (b ¼ .15, p  .01), the students' participation in mixed ability groups (b ¼ .12, p  .05) and the active involvement of METs (b ¼ .11, p  .05) were positive. These results suggest that those co-teachers who are more favourable towards the curriculum modifications, the mixed ability groups, and the active involvement of METs consider that the behaviour of students with disabilities improves in the co-taught class. While the hierarchical regression analysis on the dependent variable ‘Learning Progress’ provided no significant model (F ¼ .81, p ¼ .55), the second step model of the analysis was statistically significant (F ¼ 3.38, p  .001). The students' participation in mixed

ability groups and the active involvement of METs explained 5% of the total variance of the ‘Learning Progress’ variable. The coefficients for the students' participation in mixed ability groups (b ¼ .14, p  .01) and the active involvement of METs (b ¼ .18, p  .001) were positive. These findings indicate that those coteachers who favour the mixed ability groups and the active involvement of METs tend to agree that the learning progress of students with disabilities in the co-taught classrooms is higher than in the single-taught mainstream class. 8. Discussion In general, the findings of this study indicate that the endorsement of the above activities is positively associated with the perceived efficacy of co-teaching. The co-teachers who favour the curriculum modifications, the mixed ability groups and the active involvement of METs tend to agree that the social participation and the behaviour of students with disabilities improve in the co-taught class. Also, the co-teachers who favour the students' participation in mixed ability groups and the active involvement of METs tend to agree that the learning progress of students with disabilities improves in the co-taught class. In general, the co-teachers in this study appeared to display shared attitudes towards specific

226

V. Strogilos, A. Stefanidis / Teaching and Teacher Education 47 (2015) 218e229

contextual factors that promote perceived collective efficacy (Bandura, 1997). In theory, individuals' attitudes guide their behavioural decisions, and as Schwarz and Bohner (2001) indicate ‘numerous studies have shown that attitude-behaviour consistency is higher when the individual has direct behavioural experience with the attitude object (p. 447)’. Although we cannot predict a coteacher's behaviour across a broad range of situations from her/his attitudes, we can, as Schwarz and Bohner (2001) note, ‘predict that the individual will do “something” that is consistent with his or her attitude judgment (p. 451)’. 8.1. Social participation The social participation of students with disabilities has always been one of the main values of inclusive education. The results of the current study indicate that those co-teachers who endorse more the modification of the curriculum, the participation of students with disabilities in mixed ability groups and the active involvement of METs tend to agree that the social participation of students with disabilities improves. This is a very important finding due to the fact that research has extensively identified problems in the social position of students with disabilities in the mainstream class in several countries (e.g. Cambra & Silvestre, 2003; Koster, Pijl, Nakken, & Van Houten, 2010; Pijl, Frostad, & Flem, 2008). Also, in a recent literature review on teachers' attitudes towards the inclusion of students with disabilities (De Boer et al., 2011), no studies were identified and therefore no conclusions could be drawn with regard to the effects of teachers' attitudes on the social participation of pupils with disabilities. However, our results coincide with those from similar studies in co-taught classes in which teachers have extensively indicated improved social participation of students with disabilities (Strogilos & Tragoulia, 2013; Strogilos et al., 2015; Mastropieri et al, 2005; Schwab Learning, 2003; WaltherThomas, 1997). Yet, in the current study, the older, rather than the younger, teachers and the SETs, rather than the METs, tend to agree more that co-teaching improves the social participation of students with disabilities. It seems that older teachers value the presence of a SET in their classes and, for this reason, identify increased social benefits for students with disabilities. Similarly, it is important that SETs identify social benefits in the education of students with disabilities in the co-taught class, since this increases their willingness to develop strategies to educate these students alongside their peers. Although in Keefe and Moore's (2004) study some SETs reported that co-teaching is not an appropriate option for some students, Takala and Uusitalo-Malmivaara (2013) found that SETs frequently volunteered to co-teach in four Finish schools. The fact that specific inclusive activities and responsibilities of co-teachers are positively associated with their attitudes towards the improved social participation of students with disabilities indicates that these are important inclusive practices. As Rix et al. (2009) indicate, effective inclusion is influenced by МЕΤs' recognition of their responsibility for all learners. According to these researchers, the pedagogical approaches which can effectively include children with disabilities encourage participation through flexible grouping approaches and offer to the students opportunities to engage in meaningful concepts and activities.

taught class. The significance of this finding is further highlighted taking into consideration that in their recent research review, De Boer et al. (2011) identified negative attitudes of teachers towards the inclusion of students with behaviour difficulties in mainstream classes. Positive behaviour outcomes for students with disabilities have also been reported by co-teachers in Walther-Thomas' (1997) and Hang and Rabren's (2009) study. As Hang and Rabren (2009) argue, this positive finding could be explained by considering peers as suitable ‘behaviour models’ for students with disabilities. Also, in the Greek setting, in an observational study implemented by Strogilos & Avramidis (2013), it was found that students with disabilities were participating in classroom activities more actively, while misbehaving less, in co-taught than in non-co-taught classes. Our findings about behaviour imply that co-teachers consider that the improvement in the behaviour of students with disabilities is directly associated with certain contextual variables and it is not merely placed on SETs' behaviour management skills, as it has been referred to in other research studies (Buckley, 2005; Rice & Zigmond, 2000). The significance of this discovery is further enhanced, given the fact that the curriculum adaptations, the classroom arrangements and the METs' roles are all contextual variables that promote inclusive practices in the co-teaching setting. 8.3. Learning progress With regard to learning progress, the current study suggests that the co-teachers who favour the participation of students with disabilities in mixed ability groups and the active involvement of METs tend to agree that the learning progress of students with disabilities in the co-taught class is improved. Our results coincide with those in other research studies and research reviews which have identified positive perceived learning outcomes for students with disabilities in co-taught classes (Cook et al., 2011; Hang & Rabren, 2009; Scruggs et al., 2007; Strogilos et al., 2015). Dissimilar to the results on social participation and behaviour improvement, two of the three researched variables were related to the attitudes of co-teachers towards the learning progress of students with disabilities. Although the modifications of the curriculum are considered important for the learning progress of students with disabilities (Kurth & Keegan, 2012; Lee et al., 2010; Soukup et al., 2007), this study failed to identify a significant relationship between the modification of the curriculum and the learning progress of students with disabilities. A possible explanation for this could be the advanced academic level of the students with disabilities accommodated in the Greek co-taught classes. It is possible that due to the high functioning of students with disabilities, the participants did not hold that the modifications of the curriculum would be associated with students' learning progress. However, since the modification of the curriculum is considered an essential inclusive strategy for the education of students with disabilities in the mainstream class, further research would better clarify this finding. As Rose (2007) indicates, our understanding with regard to what works in the curriculum is currently limited and requires further investigation. 8.4. Implications for practice and research

8.2. Behaviour improvement Similar to the social participation results, those co-teachers who are more favourable towards the modifications of the curriculum, the participation of students with disabilities in mixed ability groups and the active involvement of METs consider that the behaviour of students with disabilities improves more in the co-

Friend, Cook, Hurley-Chamberlain, and Shamberger (2010) state that there are still many unanswered questions and many pieces need to be in place for co-teaching to be successful. This research, looking into co-teachers’ attitudes, examined a few of these pieces, namely three contextual factors, in order to identify their relationship with the perceived efficacy of co-teaching. Since co-

V. Strogilos, A. Stefanidis / Teaching and Teacher Education 47 (2015) 218e229

teachers’ attitudes towards certain activities and responsibilities are associated with their attitudes towards students' social participation, behaviour improvement and learning progress, we consider the theoretical framework of this study appropriate to further understand some of the antecedents of the co-teaching for students with disabilities. Given that there are no other attitudinal studies that examine the relationship between specific contextual elements and the coteaching efficacy, we consider this research significant. However, the total variances of co-teachers’ attitudes towards students' social participation, behaviour improvement and learning progress were explained by the independent variables to a relatively limited extent: 6%, 5% and 5%, respectively. Therefore, scholars and practitioners should consider that the three contextual variables included in our models can partially explain teachers' attitudes. We consider that further investigation that would examine other contextual variables in relation to the perceived effectiveness of coteaching is needed. Such variables could include the use of peer mediated learning (McDuffie et al., 2009), the effect of mutual planning time between co-teachers (Bryant Davis, Dieker, Pearl, & Kirkpatrick, 2012; Welch, 2000), the impact of training on coteaching (Boudah et al., 1997), the use of varied instructional models (Murawski, 2006), and even the roles and responsibilities of parents (Strogilos & Tragoulia, 2013). The results of this study provide important information that can have a significant impact on classroom practice and teacher preparation. Since, as Bandura (2000: 75) argues, ‘in many activities… people do not have direct control over social conditions and institutional practices that affect their lives’, educational policy support is necessary. The fact that teachers associate specific activities and responsibilities with co-teaching efficacy implies that training could be directed towards the development of these important contextual factors. Co-teachers need to be educated to design and implement efficient curriculum modifications to respond to individual student needs. The success of the efficient implementation of such modifications would be dependent on the adequate mutual planning time allocated to co-teachers and the administrative support provided by the schools. An emerging implication for research would, therefore, be the investigation of the relationships between mutual planning time and administrative support, and the efficacy of curriculum modifications. Likewise, co-teachers need to enhance the participation of students with disabilities in mixed ability groups, since this research provides evidence that teachers agree that such participation has positive implications for students' social participation, behaviour improvement and learning progress. Therefore, training should be provided to co-teachers on the development of practices that promote effective participation of students with disabilities in mixed ability groups. In particular, teachers need to be educated to provide meaningful means for students with disabilities to participate in whole-class teaching or to modify tasks and activities that could help them to participate in small mixed ability groups. This training should be based on the identification of the necessary educational and social prerequisites that enhance the education of these students in grouping arrangements. Since teachers agreed that mixed ability grouping is associated with co-teaching efficacy, further research needs to identify the educational and social factors that promote the education of students with disabilities in these arrangements. A third important implication of this study refers to the new expected responsibilities of METs. Since in this study co-teachers agreed that the active role of METs is positively related to the learning progress, social participation and behaviour improvement of students with disabilities, more consideration is needed on the involvement of METs in the education of students with disabilities

227

in the co-taught class. Training in the effective implementation of co-teaching, which traditionally focuses on the SETs, should equally significantly be directed towards the METs. In particular, this training should focus on the active involvement of METs so that they can provide the necessary support to the students with disabilities. METs should become familiar with the fact that, by being responsible to effectively plan and implement activities/tasks along with their special education colleagues, they could potentially enhance the efficacy of co-teaching for students with disabilities. Furthermore, as Jordan, Schwartz, and McGhie-Richmond (2009) refer, METs who believe that students with disabilities are their responsibility tend to be more effective with all of their students. This means that pre-service and in-service programmes for teachers could include appropriate knowledge and skills on the collaboration of METs with SETs to effectively educate students with disabilities. These programmes should focus on the responsibilities of both teachers in the education of students with disabilities and on specific teaching strategies that promote the academic and social participation of students with disabilities, such as the modification of the curriculum and their participation in flexible grouping arrangements. However, further research could clarify whether training on co-teaching has an effect on the attitudes of co-teachers. This training for the development of coherent inclusion practices should, according to Symeonidou and Phtiaka (2009), focus on local beliefs and assumptions about inclusion, while at the same time introduce the international trends of the inclusion movement. Finally, since co-teaching is an expanding practice that crosses cultural boundaries and countries, we consider our results important for the development of co-teaching internationally. Although, according to the Human Development Index, Greece is classified as an advanced country (United Nations, 2014), our results are important for developing countries, in which co-teaching tools are rarely employed, due to the fact that co-teaching could be used as an alternative to rather expensive and socially unjust forms of segregated education in which personnel and other resources are divided in different settings (e.g. special schools, special units). Similarly, our results could benefit other more advanced education systems in which the examined relationships have not been previously studied and the efficacy of co-teaching has been primarily qualitatively highlighted. 8.5. Limitations and future research The design and implementation of this research displays a number of limitations. First, due to the structure of the co-teaching programme in Greece, some of the SETs in the study did not have highly specialized training in SENs. Hence, our results could have been different if these teachers had received more specialized training. Future studies should investigate whether training on coteaching is a factor that contributes to the shaping of co-teachers' attitudes towards the efficacy of co-teaching. Second, many of the students in the Greek co-taught classrooms present mild learning difficulties and, thus, one must be cautious in generalizing the findings to populations of students who present more severe disabilities. Future research could investigate whether student characteristics have an effect on the attitudes of coteachers towards the efficacy of co-teaching. Furthermore, the development of a multi-level research design could possibly reveal more specific variations including both individual-level and schoollevel variables that would robustly contribute to the explanation of teachers' attitudes towards co-teaching efficacy. Third, our research investigated the relationships of teachers' attitudes towards co-teaching practices and efficacy, a design that does not necessarily provide conclusive evidence of causality

228

V. Strogilos, A. Stefanidis / Teaching and Teacher Education 47 (2015) 218e229

among the included variables. In addition, in the current research we examined the perceived efficacy of co-teaching with no attempt to measure its actual effectiveness on students' outcomes. The survey participants were asked to report their attitudes towards coteaching efficacy and three of its antecedents, a limitation that implies deviations from the real-life implementation of co-teaching practices. Furthermore, we acknowledge that the total variances of our dependent variables were explained by our independent variables to a relatively limited extent, a fact that may be attributed to the effect of a number of potential confounding variables, such as parents' support towards co-teachers. Hence, research could focus on the actual impact of co-teaching taking into consideration several other dimensions related to the students with disabilities as well as their parents. Furthermore, this line of research could be extended through the employment of observational techniques, while the use of purely qualitative investigations would likely address some of the complexities of co-teaching in more depth. Last, our results are based on data drawn from the Greek educational system and thus the inferred research conclusions are country-specific As a result, the generalisability of our conclusions about the efficacy of co-teaching in contexts where co-teaching methodologies are not systematically implemented may be limited (Schwarz & Bohner, 2001). We consider that future replications of our study in diverse educational contexts internationally would potentially refine the employed research tools while providing more robust research conclusions that would benefit from divergent economic, cultural, social, technological and legal specificities.

References Antia, S. D. (1999). The roles of special educators and classroom teachers in an inclusive school. Journal of Deaf Studies and Deaf Education, 4, 203e214. Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman. Bandura, A. (2000). Exercise of human agency through collective efficacy. Current Directions in Psychological Science, 9(3), 75e78. Blum, H. T., Lipsett, L. R., & Yocom, D. J. (2002). Literature circles a tool for selfdetermination in one middle school inclusive classroom. Remedial and Special Education, 23(2), 99e108. Boudah, D. J., Schumacher, J. B., & Deshler, D. D. (1997). Collaborative instruction: is it an effective option for inclusion in secondary classrooms? Learning Disability Quarterly, 20(4), 293e315. Bryant Davis, K. E., Dieker, L., Pearl, C., & Kirkpatrick, R. M. (2012). Planning in the middle: co-planning between general and special education. Journal of Educational and Psychological Consultation, 22(3), 208e226. Buckley, C. (2005). Establishing and maintaining collaborative relationships between regular and special education teachers in middle school social studies inclusive classrooms. In T. E. Scruggs, & M. A. Mastropieri (Eds.), Advances in learning and behavioral disabilities: Vol. 18. Cognition and learning in diverse settings (pp. 153e198). Oxford, UK: Elsevier. Burns, M. K., & Ysseldyke, J. E. (2009). Reported prevalence of evidence-based instructional practices in special education. The Journal of Special Education, 43(1), 3e11. Cambra, C., & Silvestre, N. (2003). Students with special educational needs in the inclusive classroom: social integration and self-concept. European Journal of Special Needs Education, 18(2), 197e208. Cardona, M. C. (2003). Mainstream teachers' acceptance of instructional adaptations in Spain. European Journal of Special Needs Education, 18(3), 311e332. Carter, E. W., Cushing, L. S., & Kennedy. (2009). Peer support strategies for improving all students' social lives and learning. Baltimore: Paul H. Brookes. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2013). Applied multiple regression/ correlation analysis for the behavioral sciences. Routledge. Cook, L., McDuffie-Landrum, K. A., Oshita, L., & Cotheren-Cook, S. (2011). Coteaching for students with disabilities: a critical analysis of the empirical literature. In J. M. Kauffman, & D. P. Hallahan (Eds.), Handbook of special education. London: Routledge. De Boer, A., Pijl, S. J., & Minnaert, A. (2011). Regular primary schoolteachers' attitudes towards inclusive education: a review of the literature. International Journal of Inclusive Education, 15(3), 331e353. Dieker, L. A. (2001). What are the characteristics of ‘effective’ middle and high school co-taught teams for students with disabilities? Preventing School Failure, 46(1), 14e24. Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Orlando: Harcourt Brace Jovanovich College Publishers.

Fennick, E., & Liddy, D. (2001). Responsibilities and preparation for collaborative teaching: co-teachers' perspectives. Teacher Education and Special Education, 24(3), 229e240. Flewitt, R., & Nind, M. (2007). Parents choosing to combine special and inclusive early years settings: the best of both worlds? European Journal of Special Needs Education, 22(4), 425e441. Forlin, C. (2001). Inclusion: identifying potential stressors for regular class teachers. Educational Research, 43(3), 235e245. Forlin, C., Keen, M., & Barrett, E. (2008). The concerns of mainstream teachers: coping with inclusivity in an Australian context. International Journal of Disability, Development and Education, 55(3), 251e264. Friend, M., Cook, L., Hurley-Chamberlain, D., & Shamberger, C. (2010). Co-teaching: an illustration of the complexity of collaboration in special education. Journal of Educational and Psychological Consultation, 20(1), 9e27. Hang, Q., & Rabren, K. (2009). An examination of co-teaching perspectives and efficacy indicators. Remedial and Special Education, 30(5), 259e268. Harbort, G., Gunter, P. L., Hull, K., Brown, Q., Venn, M. L., Wiley, L. P., et al. (2007). Behaviors of teachers in co-taught classes in a secondary school. Teacher Education and Special Education, 30(1), 13e23. Idol, L. (2006). Towards inclusion of special education students in general education. A program evaluation of eight schools. Remedial and Special Education, 27(2), 77e94. Janney, R. E., & Snell, M. E. (2006). Modifying schoolwork in inclusive classrooms. Theory into Practice, 45(3), 215e223. Johnson, R. A., & Wichern, D. W. (2007). Applied multivariate statistical analysis. New Jersey: Pearson Prentice Hall. Jordan, A., Schwartz, E., & McGhie-Richmond, D. (2009). Preparing teachers for inclusive classrooms. Teaching and Teacher Education, 25(4), 535e542. Keefe, E. B., & Moore, V. (2004). The challenge of co-teaching inclusive classrooms at the high school level: What teachers told us. American Secondary Education, 32,(3), 77e88. Koster, M., Pijl, S. J., Nakken, H., & Van Houten, E. (2010). Social participation of students with special needs in regular primary education in the Netherlands. International Journal of Disability, Development and Education, 57(1), 59e75. Koster, M., Timmerman, M. E., Nakken, H., Pijl, S. J., & van Houten, E. J. (2009). Evaluating social participation of pupils with special needs in regular primary schools. European Journal of Psychological Assessment, 25(4), 213e222. Kurth, J. A., & Keegan, L. (2012). Development and use of curricular adaptations for students receiving special education services. The Journal of Special Education. http://dx.doi.org/10.1177/0022466912464782. Larrivee, B., & Cook, L. (1979). Mainstreaming: a study of the variables affecting teacher attitude. The Journal of Special Education, 13(3), 315e324. Law 2817/2000. The education of persons with special educational needs and other regulations. FEK 14th May 2000, v. A0 , n. 78. Law 3699/2008. Special education and education of people with disability or special educational needs. FEK 2nd October 2008, v. A0 , n. 199. Lee, S., Wehmeyer, M. L., Soukup, J. H., & Palmer, S. B. (2010). Impact of curriculum modifications on access to the general education curriculum for students with disabilities. Exceptional Children, 76(2), 213e233. Logan, K., Bakeman, R., & Keefe, E. (1997). Effects of instructional variables on engaged behavior of students with disabilities in general education classrooms. Exceptional Children, 64, 481e497. Magiera, K., & Zigmond, N. (2005). Co-teaching in middle school classrooms under routine conditions: does the instructional experience differ for students with disabilities in co-taught and solo-taught classes? Learning Disabilities Research and Practice, 20(2), 79e85. Mastropieri, M. A., & Scruggs, T. E. (2006). The inclusive classroom: Strategies for effective instruction (3rd ed.). Upper Saddle River, NJ: Prentice-Hall. Mastropieri, M. A., Scruggs, T. E., Graetz, J., Norland, J., Gardizi., W., & McDuffie, K. (2005). Case studies in co-teaching in the content areas: Successes, Failures and Challenges. Intervention in Schools and Clinic, 40(5), 260e270. McDonnell, J. (2011). Instructional context. In J. M. Kauffman, & D. P. Hallahan (Eds.), Handbook of special education (pp. 532e544). London: Routledge. McDuffie, K. A., Mastropieri, M. A., & Scruggs, T. E. (2009). Differential effects of peer tutoring in co-taught and non-co-taught classes: results for content learning and student-teacher interactions. Exceptional Children, 75(4), 493e510. Mechling, L. C., Gast, D. L., & Krupa, K. (2007). Impact of SMART board technology: an investigation of sign word reading and observational learning. Journal of Autism and Developmental Disorders, 37(10), 1869e1882. http://dx.doi.org/ 10.1007/s10803-007-0361-9. Moin, L. J., Magiera, K., & Zigmond, N. (2009). Instructional activities and group work in the US inclusive high school co-taught science class. International Journal of Science and Mathematics Education, 7, 677e697. Murawski, W. W. (2006). Student outcomes in co-taught secondary English classes: how can we improve? Reading and Writing Quarterly, 22, 227e247. Murawski, W. W. (2010). Collaborative teaching in elementary schools. Making the coteaching marriage work! California: Corwin. Murawski, W. W., & Swanson, H. (2001). A meta-analysis of co-teaching research. Remedial and Special Education, 22(5), 258e267. Murphy, C., & Beggs, J. (2010). A five-year systematic study of coteaching science in 120 primary schools. In C. Murphy, & K. Scantlebury (Eds.), Coteaching in international contexts (pp. 11e34). New York: Springer. Nind, M., & Wearmouth. (2006). Including children with special educational needs in mainstream classrooms: implications for pedagogy from a systematic review. Journal of Research in Special Educational Needs, 6(3), 116e124.

V. Strogilos, A. Stefanidis / Teaching and Teacher Education 47 (2015) 218e229 Nunnally, J. (1967). Psychometric methods. New York: McGraw-Hill. Pijl, S. J., Frostad, P., & Flem, A. (2008). The social position of pupils with special needs in regular schools. Scandinavian Journal of Educational Research, 52(4), 387e405. Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: a critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879e903. Rea, P. J., McLaughlin, V. L., & Walther-Thomas, C. (2002). Outcomes for students with learning disabilities in inclusive and pullout programs. Exceptional Children, 68(2), 203e222. Rice, D., & Zigmond, N. (2000). Co-teaching in secondary schools: teacher reports of development in Australian and American classrooms. Learning Disabilities: Research & Practice, 15, 190e197. Rix, J., Hall, K., Nind, M., Sheehy, K., & Wearmouth, J. (2009). What pedagogical approaches can effectively include children with special educational needs in mainstream classrooms? A systematic literature review. Support for Learning, 24(2), 86e94. Rose, R. (2007). Curriculum considerations in meeting special educational needs. In L. Florian (Ed.), The sage handbook of special education (pp. 295e306). London: Sage. Rose, D., Meyer, A., & Hitchcock, C. (2005). The universally designed classroom: Accessible curriculum and digital technologies. Cambridge, MA: Harvard University Press. Roth, W.-M., & Tobin, K. G. (2002). At the elbow of another: Learning to teach by coteaching. New York: Peter Lang. Roth, W.-M., & Tobin, K. G. (2004). Coteaching: from praxis to theory. Teachers and Teaching: Theory and Practice, 10(2), 161e179. http://dx.doi.org/10.1080/ 1354060042000188017. Saloviita, T., & Takala, M. (2010). Frequency of co-teaching in different teacher categories. European Journal of Special Needs Education, 25(4), 389e396. http:// dx.doi.org/10.1080/08856257.2010.513546. Schwarz, N., & Bohner, G. (2001). The construction of attitudes. In A. Tesser, & N. Schwarz (Eds.), Blackwell handbook of social psychologyIntraindividual processes (pp. 436e457). Oxford, UK: Blackwell. Scruggs, T. E., Mastropieri, M., & McDuffie, K. A. (2007). Co-teaching in inclusive classrooms: a metasynthesis of qualitative research. Exceptional Children, 73(4), 392e416. Soukup, J. H., Wehmeyer, M. L., Bashinski, S. M., & Bovaird, J. A. (2007). Classroom variables and access to the general curriculum for students with disabilities. Exceptional Children, 74(1), 101e123.

229

Schwab Learning. (2003). Collaboratively speaking. A study on effective ways to teach children with learning differences in the regular education classroom. The Special Edge, 16(3), 9e21. Strogilos, V., & Avramidis, E. (2013). Teaching Experiences of students with Special Educational Needs in Co-taught and Non-co-taught classes. Journal of Research in Special Educational Needs. http://dx.doi.org/10.1111/1471-3802.12052. Strogilos, V., & Tragoulia, E. (2013). Inclusive and collaborative practices in cotaught classrooms. Roles for teachers and parents. Teaching and Teacher Education, 35, 81e91. Strogilos, V., Tragoulia, E., & Kaila, M. (2015). Curriculum issues and benefits in supportive co-taught classes for students with intellectual disabilities. International Journal of Developmental Disabilities, 61(1), 32e40. Symeonidou, S., & Phtiaka, H. (2009). Using teachers' prior knowledge, attitudes and beliefs to develop in-service teacher education courses for inclusion. Teaching and Teacher Education, 25(4), 543e550. Takala, M., & Uusitalo-Malmivaara, L. (2012). A one-year study of the development of co-teaching in four Finnish schools. European Journal of Special Needs Education, 27(3), 373e390. Thousand, J. S., Villa, R. A., & Nevin, A. I. (2006). The many faces of collaborative planning and teaching. Theory into Practice, 45(3), 239e248. UNESCO. (1994). The Salamanca statement and framework for action on special needs education. UNESCO. United Nations. (2014). Human development index. Retrieved from http://hdr.undp. org/. On Sept. 16, 2014. Vaughn, S., Elbaum, B. E., Schumm, J. S., & Hughes, M. T. (1998). Social outcomes for students with and without learning disabilities in inclusive classrooms. Journal of Learning Disabilities, 31(5), 428e436. Walther-Thomas, C. S. (1997). Co-teaching experiences: the benefits and problems that teachers and principals report over time. Journal of Learning Disabilities, 30(4), 395e407. Weiss, M. P., & LIoyd, J. L. (2002). Congruence between roles and actions of secondary educators in co-taught and special education settings. The Journal of Special Education, 36(2), 58e68. http://dx.doi.org/10.1177/ 00224669020360020101. Welch, M. (2000). Descriptive analysis of team teaching in two elementary classrooms: a formative experimental approach. Remedial and Special Education, 21(6), 366e376.