Perceptions of Primary School Children About the Roles and Responsibilities of Teachers in Co-Taught Classrooms

Abstract

Classrooms have become increasingly diverse worldwide. Respect for diversity, including the participation of students with disabilities/special educational needs in all aspects of school life, has made the need for adopting an inclusive education model pertinent. The present study has examined how Greek students with and without special educational needs perceive teachers’ roles and responsibilities, as well as their way of learning and participating during co-taught lessons. The sample consisted of 130 primary school children (aged 9–12 years), 16 of whom had special educational needs. Data were collected through the Co-Teaching Student Questionnaire and students’ drawings were analysed following the ‘Teacher Perception Coding List’. According to the results, students perceive that the general education teacher is mostly responsible for planning and implementing teaching, but both teachers, i.e., the mainstream and special education teacher, contribute to students’ learning and academic success. More specifically, students with special educational needs perceive the special education educator as a critical figure in the classroom, and they often seek support from him/her during the lessons. Also, the analyses of children’s drawings have revealed differences in the features (size, position, movements and age) of general and special educators. This research is original in that it offers a novel insight into how students perceive the co-teaching model and the role of teachers in Greek co-taught classrooms. It also gives valuable information on inclusive teaching practices that co-teachers could implement. Students’ perceptions of teachers’ roles and responsibilities can reveal challenges or successful practices of inclusion, thus acting as a valuable tool for both educators and researchers.

1. Introduction

Inclusive education is highlighted as an international and national policy goal (UN, 2006), and ensures that all students, including those with special educational needs (SEN), learn together in the same classrooms, thus promoting equal opportunities and social integration. Worldwide, there is an ongoing discussion about policies advocating for the simple integration of students with disabilities into regular classrooms and for authentic inclusion that changes educational practices, school culture and power relations (UNESCO, 2020). General education teachers (GETs) play a crucial role by creating accessible learning environments and modifying their teaching strategies to accommodate a varied student body. Special education teachers (SETs), on the other hand, create individualized learning plans, evaluate each student’s needs and offer SEN students immediate support. Both types of teachers collaborate to make sure that every student has access to the curriculum, can engage actively in school activities, and can reach their full potential (King-Sears et al., 2014; Rönn-Liljenfeldt et al., 2023).

Co-teaching has been promoted globally as a method for achieving inclusion, yet its impact is debated: while some research highlights its potential to reduce inequalities by promoting differentiated instruction and collaborative expertise, others warn that in real life it frequently perpetuates hierarchies, with GETs maintaining control and SETs being assigned to an “assistant “position (Zamkowska & Nogaj, 2024). Students with SEN are the main beneficiaries and consumers of inclusive education, which aims to eliminate barriers to learning and create an educational environment that meets the diverse needs of students in a tailor-made fashion. Students’ perspectives in a co-teaching context are under-documented in many countries. Examination of student perceptions and experiences may change implementation of co-teaching as a method of inclusion and improve access to the general curriculum for all students (King-Sears & Johnson, 2020).

1.1. Co-Teaching Models and Their Effectiveness

The ways in which GETs and SETs provide support to their students have shaped various models of co-teaching. Villa et al. (2013) described four different models of co-teaching: In ‘one teach, one assist’, one teacher gives instruction and has the main responsibility for the class while the other observes or assists. In parallel teaching, teachers divide students into two equal and heterogeneous ability groups (Friend & Bursuck, 1999), and each teacher teaches his/her group the same subject at the same time. Complementary teaching is implemented by one teacher that enhances instruction provided by the other teacher. Team teaching occurs when teachers share the teaching time equally and teach all students of the class, sharing responsibility for classroom planning, teaching and evaluation. The ‘one teach, one assist’ model which was first reported by Cook and Friend (1995) is currently implemented by the majority of teachers in many countries: one teacher teaches all students, and the other wanders among students and provides individual assistance to those in need. Within this context, it is customary for the ‘wandering teacher’ to be the SET (Friend et al., 2010).

Co-teaching has been gaining increasingly more ground in the field of education because it offers a wealth of benefits to both students and teachers. Carty and Farrell (2018) report that teachers managed to enhance their professional development by observing different teaching methods and adopting different teaching styles. Similarly, Brusca-Vega et al. (2011) and van Hover et al. (2012) found that teachers, when doing co-teaching, managed to enrich the learning environment by using a variety of teaching methods, materials and activities.

Co-teaching has been frequently reported to have a positive effect on students. Specifically, students with learning disabilities perform better in co-teaching environments, as their grades and course attendance increase compared to other classes (Rea et al., 2002). In a study by Bottge et al. (2015), co-teaching has been found to boost the academic progress of students with SEN. According to their findings, the concurrent presence of two teachers in the classroom enhanced the interaction among students and provided more opportunities for assistance and support, which had cascading positive effects for all groups of students. In another study by Rea et al. (2002), students with learning disabilities performed better in co-teaching classes, as their grades and attendance in class increased relative to students with learning disabilities in classes with a single teacher. Spörer et al. (2021) showed that students actively participated and collaborated more when there were two teachers as compared to single-teacher classes, which is particularly important for students with disabilities. In addition, Lochner et al. (2019) found that students in co-teaching classes developed socially and academically, engaging in activities that required critical thinking and collaboration. Embury and Kroeger (2012) point out that students, with and without SEN, prefer the support of two teachers, as they feel that learning is more personalized, tailor-made and effective compared to being taught by a single teacher.

Research, however, highlights some drawbacks in the implementation of co-teaching. Murawski (2006) found that in many co-teaching classes, teachers did not equally share their roles, they had different management styles, and their cooperation was limited. Moreover, in some cases, co-teaching becomes ineffective when the SET assumes a mainly supportive role without actively participating in teaching (Pancsofar & Petroff, 2016). Also, teacher cooperation has often been reported to be insufficient, with minimal mutual exchange of ideas and the SET not being treated as an equal partner in the educational process (McDuffie et al., 2009; van Hover et al., 2012). However, in order to reap maximum benefits from co-teaching models, mutual respect and joint programming between teachers are required (Qualls et al., 2025).

1.2. Co-Teachers’ Roles and Perspectives Within Inclusive Education

In a co-teaching situation, the roles and responsibilities of teachers are multidimensional and require flexibility. According to the literature, GETs are often viewed as the primary authority figures responsible for the whole class and the implementation of the curriculum, while SETs may be perceived as assistants or specialists focusing only on students with disabilities (King-Sears & Strogilos, 2018; Strogilos & Avramidis, 2016; Takala & Uusitalo-Malmivaara, 2012; Rönn-Liljenfeldt et al., 2023). The majority of students without SEN and/or disabilities tend to perceive that the responsibility for the evaluation of their academic performance lies primarily within the GET who monitors their progress and provides feedback on their school work. In contrast, most students with disabilities often consider that both the GET and SET share responsibility for assessing their learning progress. This dual perception likely reflects their experience of receiving instructional support and academic feedback from both professionals (King-Sears et al., 2014; Sundqvist et al., 2020).

Embury and Kroeger (2012) have explored students’ perceptions of their teachers in two integration classes in the United States. Interviews with middle-school students showed that the use of co-teaching strategies by both teachers influenced the way students perceived each teacher. More specifically, in one class where the co-teaching model exclusively implemented was ‘one teach, one assist’, students perceived the GET and SET as the main authority and his/her assistant, respectively. On the other hand, in a class where the complementary teaching model was implemented, students were familiar with teachers switching roles and responsibilities in a mutual way.

Sharma and Salend (2016) found that, when teachers mainly employed the ‘one teach, one assist’ co-teaching model, students with SEN and/or disabilities tended to interact more with the SET than students without SEN and/or disabilities. When children with disabilities interact more intensively and actively with the SET, they are less likely to interact with their peers (Feldman et al., 2015), thus creating the impression that students with SEN cannot be educated without the presence of the SET. More balanced roles shared between GETs and SETs would promote more equal interactions between teachers and students, which would, in turn contribute to the achievement of better learning outcomes.

King-Sears et al. (2014) conducted a survey to investigate how educators and students in a science class in the United States felt about co-teaching. The overwhelming majority of students (85.7%) said that the GET designed and prepared most of the lessons, whereas the remaining students (14.3%) said that both teachers were involved in designing and carrying out the lesson. All participants, however, presented converging views about both teachers collaborating to deliver education that satisfied the needs of the students.

In a study by Rönn-Liljenfeldt et al. (2023) that was conducted in primary schools in Finland, students reported that the GET is the one who teaches most of the time, plans the lessons and manages the classroom. When they need help, they feel that they receive it from both teachers. When comparing the responses of students with and without SEN, it was found that students without SEN reported better learning outcomes related to the GET, while students with SEN reported better learning outcomes stemming from either the SET alone, or both teachers.

1.3. Perceived Equity and Shared Responsibilities in Co-Taught Classrooms

Students’ social representations of the roles of teachers in inclusive classrooms reflect their perceptions, beliefs, and expectations, which are dynamically shaped by their everyday experiences and social influences. Their perceptions and expectations influence how they interact with teachers and their peers, and how they perceive inclusion (Moscovici, 2001). Moscovici (2001) stressed that students’ perception of the roles and the responsibilities of GETs and SETs should be expressed through symbolic expressions, such as images and narratives. Drawings, especially in children’s worlds, function as an effective means to express feelings, thoughts and perceptions, thus offering an important avenue of communication (Martikainen, 2019). In a study by Bessette (2008), elementary and middle-school students were asked to create illustrations depicting how their co-teachers collaborated. These drawings were next presented to the co-teachers, who responded to the images created by the students. Co-teachers’ responses differed in that the SEN co-teachers appeared to be especially troubled by being depicted more as helpers than as leaders. Indeed, 74% of middle-school students and 33% of elementary students in Bessette’s (2008) study opted to illustrate the ‘one teach, one assist’ co-teaching model. Several co-teachers mentioned that the lack of co-planning time contributed to the fact that considerably fewer students depicted collaborative teaching scenarios.

Understanding students’ perceptions of co-teaching is valuable, as this can affect and inform classroom dynamics, student engagement, and the effectiveness of inclusive practices. Furthermore, studying students’ views can help educators identify misconceptions, promote collaborative teaching, and foster a more inclusive mindset among all students. Specifically, Hang and Rabren (2009) found that studies that encompassed the viewpoints of both teachers and students regarding co-teaching showed that co-teachers often lacked awareness of their students’ perspectives, and that the experiences of students in a co-taught classroom differed significantly from teachers’ expectations.

1.4. The Role and Implementation of Co-Teaching in the Greek Educational System

Collaborative teaching was established in Greece more than 20 years ago. According to Greek legislation (Law 2817/2000; Law 3699/2008), all children with disabilities who attend co-taught classrooms should have an official diagnosis of special educational needs. An increasing number of students with mild learning disabilities (including medium-to-high-functioning autism, attention deficit hyperactivity disorder, mild intellectual disability, and physical disability, among others) are educated in mainstream schools, achieved by assigning one SET to one or more children with SEN for the whole or part of the school day in co-taught classrooms (Strogilos et al., 2023). The SET supports students by implementing an individualized education plan, as well as by enhancing their social participation in all school activities. He/she thus acts as a facilitator for the student he/she supports, while the GET provides whole class activities (Mavropalias, 2018).

A Greek study by Strogilos and Tragoulia (2013) highlights the difficulties and challenges in cooperation between SETs and GETs in co-teaching classes, with the most common teaching model being ‘one teach, one assist’. This model often ends up limiting the effectiveness of co-teaching, as it does not enhance the full participation and cooperation between the teachers. Another Greek study by Strogilos and Avramidis (2016) has found that co-teaching had some positive effect on students with intellectual disabilities and autism spectrum disorder in terms of their level of engagement and social relationships with teachers and peers. Koutsoklenis and Papadimitriou (2021) underline that the model of co-teaching has played a positive role in the education of students with disabilities and/or SEN, offering an open and inclusive environment of education. Nevertheless, a series of institutional and governmental state measurements are required to improve the benefits of co-teaching in Greek mainstream classrooms, such as the implementation of team teaching instead of the ‘one teach, one assist’ model, co-planning time added to co-teachers’ timetables, sufficient and continuous training in co-teaching for all educators regardless of years of teaching, and shared values and perspectives towards an inclusive education.

1.5. Aims of the Study

Within the Greek literature, there is a paucity of studies that have investigated co-teaching from the students’ point of view. Research on children’s own voices about how co-teaching works for them and how they interpret the co-taught classroom experience remains thin (Mavropalias, 2018; Strogilos et al., 2016; Strogilos & Tragoulia, 2013). This is the first research study in Greece that explores the perceptions of primary school children about co-teaching and the roles of co-educators in the mainstream classroom, using two sources of information, namely, questionnaires and drawing. More specifically, a set of three research questions were formulated to guide data collection and analysis in the current study:

• How do students (with and without SEN) perceive the roles and duties of SETs and GETs in the mainstream classroom?
• How do students (with and without SEN) perceive co-teaching in terms of equivalence, teaching atmosphere and way of learning?
• How do students (with and without SEN) conceptualize co-teaching in drawings?

2. Materials and Methods

2.1. Participants

The sample of the study consisted of 130 students from five public primary schools located in three different municipalities in Athens, the capital of Greece. Specifically, 21.5% of the participants attended the 4th grade of primary school (N = 28), 34.6% attended the 5th grade (N = 45), and 43.9% attended the 6th grade (N = 57). In terms of gender, the sample consisted of 68 boys (i.e., 52.3% of the total sample) and 62 girls (i.e., 47.7% of the total sample). The average age of the students was 10.5 years (age range 9–11.5 years). All students attended co-taught classrooms. Students without SEN made up 87.7% (N = 114) of the sample, while the remaining 12.3% (N = 16) had a formal diagnosis of SEN and received individual support from a SET in the form of the ‘one teach, one assist’ model. This type of support is usually offered to children with mild learning disabilities according to Greek legislation (Law 3699/2008). Participant selection was based on strategic selection, i.e., students at primary schools in three municipalities having the same socioeconomic status were asked to participate. The third author’s personal contacts with colleagues from neighbouring schools were utilized to contact the schools. The inclusion criteria were that the schools had co-taught classrooms in the 4th, 5th and 6th grade, had qualified SETs allocated to SEN students for the whole school year, and had been following the co-teaching model of ‘one teach, one assist’ for at least two school years.

2.2. Measures

For the student survey, the Co-Teaching Student Questionnaire (CTSQ) (King-Sears et al., 2014), which gives students the chance to indicate their response across a continuum, was employed. The CTSQ consists of four parts. Only the three parts of the CTSQ were used in the present study because they matched its purposes. The first part elicits demographic information about the participants (sex, grade, school). The second part entitled “Which co-teacher, or both?”, consists of ten statements that focus on the way students learn (1 item); teaching issues (4 items); and support that students receive from their teachers (5 items). Children were asked to choose one answer out of three options, namely, GET, SET, and both educators. Response distribution was calculated for each option. To make sure that students understood how to answer the questions in this section, two practice statements were supplied. The third part was entitled “How much do you agree?”, and consists of 15 statements in which students had to select an answer according to a 4-point Likert scale (1 = strongly disagree, 2 = disagree, 3 = agree, 4 = strongly agree). The items are related to students’ personal experiences in the co-taught classroom, and more specifically, the equivalence of teachers (3 items), the teaching atmosphere (4 items), and the way of learning (8 items). Internal consistency for the second and third part of the questionnaire as measured by Cronbach’s α was 0.70 and 0.90, respectively. While the value (0.90) for the third part of the questionnaire shows good reliability, the second part had a lower value (0.70); however this is considered to be sufficient for group-level research (Hussey et al., 2025; Šerbetar & Sedlar, 2016).

The questionnaire was translated from English into Greek by a researcher who had excellent knowledge of English. Next, two bilingual individuals in Greek and English checked the translation for possible contradictions and errors between the original English questionnaire and the translated text and made corrections. Back translation followed, where the Greek text was translated into English by an independent translator without knowledge of the original English version of the CTSQ. In the last step, the two bilinguals and the researchers leading the study compared the two English texts (the original and the result of back translation) and ensured that the content had not been altered.

Children’s drawings are an alternative, authentic way of evaluating a child’s perspectives (Bessette, 2008). They can be used as a complementary tool to assess children’s perceptions of teachers and their experiences of practices of inclusion in the classroom. For the analysis of the children’s drawings in the current study, the “Teacher Perception Coding List” created by Aykaç (2012) and adapted by Arslan Cansever (2017) was used. Arslan Cansever (2017) defines seven categories (themes) derived from the “Teacher Perception Coding List” as follows: gender (of teachers), external appearance, facial expression, physical characteristics, size, movements, and the objects held by teachers. The characteristics of the students’ drawings were evaluated within the framework of the above seven categories. Drawings were analyzed and interpreted by two authors (researchers) after studying carefully the coding scheme used in Arslan Cansever’s (2017) work. Specifically, the drawings were analysed according to the frequency and percentages obtained in the eight categories for each type of teacher (i.e., GET, SET) depicted in the drawing. To ensure reliability, the two authors initially rated 10 drawings independently, without exchanging any opinions during the initial phase. Agreement and disagreements associated with these ratings were determined until an agreement rate of 87% was reached. The rest of the drawings were assessed after taking into account the interaction between the two researchers who tried to avoid subjectiveness throughout the research. In addition, Cronbach’s a was calculated for the variables of the drawings, and it was quite satisfactory (a = 0.67).

2.3. Procedure

The research was conducted in March and April 2024 to ensure that students had sufficient experience in engaging in co-taught instruction since the beginning of the school year in September 2023. Initially, information letters were sent out to the principals of nine schools inviting them to participate in the research. Five schools agreed to participate. The school principals along with the teachers of the upper grades of each school informed parents about the aims and stages of the research, and took provisions for ensuring the confidentiality and anonymity of information related to the students. In addition, parents were informed that there were no expected risks for their child, whether he/she participated in the study or not. Subsequently, declarations of consent were obtained from the parents/guardians of 146 students, while 130 students finally took part in the study.

Data collection was performed by the principal researcher and one teacher (either the GET or the SET) during two teaching hours of a lesson. Before data collection, a meeting took place between the researchers and the teachers to familiarize them with the procedure. The CTSQ was first administered to the students with and without SEN in groups. Completion of the questionnaire lasted about 20 min. When all questionnaires were collected, the questionnaires of the children with SEN were marked with an asterisk on the back of the questionnaire by the researcher. The drawing task was conducted during the second teaching hour on the same day. Children were given a white A4 paper and were instructed to use their own writing implements (e.g., black pencil, crayons). They were given the following prompt by the principal researcher: “Draw the teachers of your classroom when they are working. You can paint anything else you would like from your classroom, such as your classmates and yourself”. On the back of every drawing, the student wrote down the initials of his/her name, his/her gender, grade level, and the first three letters of his/her classroom teachers. At the end of the allotted time, the drawings were collected by the principal researcher. The children were given more time if needed.

2.4. Data Analysis

Data analysis in this study was based on a methodological approach combining quantitative data obtained from questionnaires and student drawings. All analyses were conducted using SPSS (version 26). Descriptive statistics in the form of frequencies and percentages are presented for all the variables of the study. In order to explore the relation between SEN status, and the roles and responsibilities of GETs and SETs, a series of independent-samples chi-square (χ²) tests were conducted. Through cross tabulation, the observed frequencies were next compared to the expected frequencies, and the chi-square test determined whether the difference was statistically significant or not. The 95% Confidence Intervals were given as 95% CI [Lower limit of confidence interval, Upper limit of confidence interval]. Bonferroni corrections were adopted to control for Type I error. Drawings were analysed according to the frequency and percentages obtained in the seven categories (themes), and chi-square tests were next performed to identify differences in the perceptions of the students with and without SEN. Fisher’s exact test was used in place of the chi-square test when sample sizes were small, particularly when the expected frequency in any cell of a contingency table was less than 5 (Loukaidis, 2011).

3. Results

3.1. First Aim of the Study: Students’ Perceptions About the Role and Duties of Co-Teachers in the Classroom

Table 1. Perceptions of students about the role and duties of the teachers.

Statements	GET	SET	Both Educators
1. I learn better from…	60%	3.8%	36.2%
Teaching issues
2. The teacher who grades my work is…	85.4%	0.8%	13.8%
3. The teacher who organises the materials for the instruction is…	73.1%	4.6%	22.3%
4. The teacher who seems to plan the course the most is…	86.9%	5.4%	7.7%
5. The teacher who seems to be in charge of the lessons the most is…	70.8%	3.1%	26.2%
Support
6. When I need help, the teacher I ask…	26.2%	15.4%	58.5%
7. The teacher who walks around and helps the students the most is…	38.5%	23.1%	38.5%
8. The teacher who explains things most of the time is…	80.8%	3.8%	15.4%
9. The teacher who explains in different ways most of the time is …	60%	16.9%	23.1%
10. The teacher who explains things to me when I do something the wrong way is…	44.6%	11.5%	43.8%

Notes. GET = General education teacher; SET = Special education teacher.

presents the students’ perceptions about the role and responsibilities of the co-teachers in the mainstream classroom. The majority of the students reported that they learn better from the GET (60%) and both teachers (36.2%) (item 1). Regarding teaching issues (items 2 to 5), the majority of the students perceive the GET as the main figure who plans, teaches, and assesses in the classroom, while the next bigger percentage was that both educators were perceived as such. When comparing students with and without SEN, significant differences were observed in several items. Specifically, the students without SEN (63.2%) mentioned that they learn better with the GET, while students with SEN mentioned that they learn equally well with both teachers (43.8%) followed by the GET alone (37.5%) [χ²(2) = 9.07, p = 0.008; Cramer’s V = 0.53; 95% CI [2.28–3.33] for students without SEN; [1.33–2.16] for students with SEN]. Students without SEN considered that the GET is responsible for grading their assignments (88.6%), while students with SEN stated that they receive their grade from the GET or both teachers (GET: 62.5%, Both: 31.3%) [χ²(2) = 9.28, p = 0.006; Cramer’s V = 0.97; 95% CI [3.0–3.93] for students without SEN; [2.24–3.0] for students with SEN]. In the statement about who is responsible for the instruction materials, students without SEN indicated the GET (76.3%), while SEN students indicated the GET (50%) and both teachers (31.3%) [χ²(2) = 8.02, p = 0.013; Cramer’s V = 0.36; 95% CI [2.74–3.50] for students without SEN; [1.56–2.55] for students with SEN]. In the item about who seems to plan most of the lessons, almost all students without SEN indicated the GET (93%), while the opinions of students with SEN were divided; 43.8% indicated the GET, 37.5% indicated the SET, and 18.8% indicated both teachers [χ²(2) = 26.95, p < 0.001; Cramer’s V = 0.68; 95% CI [3.16–3.71] for students without SEN; [0.97–1.40] for students with SEN]. Figure 1 below displays proportions (%) of responses to item 4 targeting children’s perception of the person(s) in charge of lesson planning in class.

Regarding the items on teaching support, the majority of the students stated that they ask for help from both teachers (58.5%), and also indicated the GET as the person who explains things most of the time (80.8%) or explains things in different ways (60%). Students’ responses were divided among the three options in items 7 and 10 (these items targeted who walks around the classroom helping the students and who explains things when a student makes mistakes). Significant differences between students with and without SEN were observed in items 6 and 10. Specifically, students with SEN stated that they turn for help mainly to the SET (50%), while students without SEN ask for help from both teachers (61.4%) [χ²(2) = 12.73, p < 0.001; Cramer’s V = 0.45; 95% CI [2.52–3.10] for students without SEN; [1.61–2.38] for students with SEN]. The teacher who explains things when children make mistakes during the lesson is the GET for students without SEN (48.2%), while students with SEN indicated both teachers (50%) and the SET (31.3%) [χ²(2) = 8.19, p = 0.015; Cramer’s V = 0.20; 95% CI [2.0–2.87] for students without SEN; [1.56–2.43] for students with SEN].

3.2. Second Aim of the Study: Students’ Perceptions of Co-Teaching in Terms of Equivalence, Classroom Atmosphere and Ways of Learning

Regarding students’ perceptions about aspects of co-teaching, data from

Table 2. Perceptions of students about the equality of teachers, classroom atmosphere and the way they learn.

Statements	Strongly Disagree	Disagree	Agree	Strongly Agree
Equality
1. When the two teachers teach a lesson, I think they divide it in half so that one teacher does not do more work than the other.	21.5%	43.8%	26.2%	8.5%
2. One teacher is more responsible for our behaviour and the other teacher is more responsible for the lesson.	24.6%	34.6%	29.2%	11.5%
3. I believe that both teachers are equal in the classroom.	9.2%	15.4%	26.9%	48.5%
Classroom atmosphere
4. I believe that our teachers like to teach in this class.	4.6%	7.7%	46.2%	41.5%
5. I like that we have two teachers in the class.	3.8%	5.4%	36.2%	54.6%
6. Students seem to behave better when there are two teachers in the class.	23.1%	33.1%	28.5%	15.4%
7. The two teachers seem to feel comfortable sharing responsibilities when teaching together.	9.2%	22.3%	44.6%	23.8%
How I learn
8. I learn more when I’m in this class with two teachers.	13.8%	20.8%	33.1%	33.2%
9. I learn better with two teachers.	9.2%	25.4%	33.8%	39.8%
10. The two teachers use more ways to teach than when I am in other classes with only one teacher	10.8%	20.8%	40%	38.5%
11. One of the teachers explains things to me better than the other.	30.8%	15.4%	26.2%	27.7%
12. I would not prefer to learn with only one teacher in the classroom (reverse coding)	28.5%	31.5%	24.6%	15.4%
13. I wish all my classes had two teachers.	18.5%	19.2%	33.8%	28.5%
14. It is not difficult to have two teachers at the same time (reverse coding)	30%	35.4%	23.8%	10.8%
15. Having two teachers doesn’t confuse me sometimes (reverse coding)	33.1%	32.3%	17.7%	16.9%

indicate that most students strongly disagreed (21.5%) or disagreed (43.5%) that teachers divide the lesson in half to have equal amount of work. A similar pattern (i.e., disagreement) was exhibited about the statement that ‘One teacher is more responsible for our behaviour and the other teacher is more responsible for the lesson’. A great percentage of students (26.9% agreed; 48.5% strongly agreed) converged on the statement that both teachers are equal in the classroom (item 3). When comparing students with and without SEN, a significant difference was revealed in the statement about whether teachers in the co-taught classroom divide the lesson in half to do the same amount of work [χ²(3) = 7.37, p = 0.047; Cramer’s V = 0.90; 95% CI [1.28–2.33] for students without SEN; [1.69–2.55] for students with SEN]. More specifically, as one can see in Figure 2 below, half of the students with SEN seemed to strongly disagree, while students without SEN seemed to have rather milder objections to the statement.

Regarding classroom atmosphere, almost all students seemed to agree or strongly agree with statements 4 and 5, i.e., ‘Teachers like to teach in a co-taught classroom’, and ‘Students like to have two teachers’, respectively. Similarly, students agreed (44.6%) or strongly agreed (23.8%) that teachers feel comfortable sharing responsibilities when teaching together. Responses were more mixed and distributed across options for the statement that ‘students behave better when there are two teachers in the classroom’.

As far as the way students learn, most students agreed (33.8%) or strongly agreed (39.8%) that they learn better when they have two teachers in the classroom. When comparing students with and without SEN, a statistically significant difference was found [χ²(4) = 10.43, p = 0.023; Cramer’s V = 0.47; 95% CI [2.08–3.29] for students without SEN; [2.77–3.85] for students with SEN], which stemmed from the fact that a greater percentage of students with SEN expressed a preference for two teachers in comparison to the students without SEN (item 9). In addition, students also agreed (40%) or strongly agreed (38.5%) that teachers in the co-taught classroom use more teaching strategies in comparison to the non-co-taught classroom (item 10). Responses were more mixed for statement 11 that ‘One of the teachers explains things to me better than the other’ (students’ responses were distributed across all options). Although many students wished all their classes had two teachers (61.3%), some students disagreed (37.7%). In addition, some students felt hard or experienced confusion having two teachers (65.4%), while some students didn’t feel the same way (34.6%).

3.3. Third Aim of the Study: Students’ Conceptualization of Co-Teaching in Drawings

Regarding inter-rater agreement and reliability, of the 130 drawings, a sample of 30 randomly identified (23%) were evaluated by the first author of the study (first rater). To assess the inter-rater agreement, the same sample was re-scored a week after by a second rater (the third author), who was blind to the original scoring. Each rater coded each drawing in all the categories, separately for the GET and the SET. Inter-rater reliability was assessed by calculating Cohen’s kappa coefficient (κ 95% confidence limit). The results showed excellent agreement for both the GET (99.3% agreement, κ = 0.95) and the SET category (100% agreement, κ = 1.00).

Students’ drawings give insights into how children conceptualize co-teaching. They were joyful, creative, detailed, depicting scenes within the classroom. The majority of the students drew their teachers as ‘female’, (76.2% and 69.2% for the GET and the SET, respectively) (Figure 3), while male figures were less frequent (GET: 10.8%; SET: 23.1%). In addition, 13.1% of the students depicted the GET with an unspecified gender, using faces or heroes, while this tendency was encountered to a smaller extent (7.7%) respectively for the SET (

Table 3. Representation of gender of teachers in children’s drawings.

	GET		SET
	f	%	f	%
Male	14	10.8	30	23.1
Female	99	76.2	90	69.2
Unspecified	17	13.1	10	7.7
Total	130	100	130	100

Notes. GET = General education teacher; SET = Special education teacher.

).

Physical appearance of the teacher was one of the themes examined in children’s drawings (

Table 4. Representation of teachers’ physical appearance in the children’s drawings.

	GET		SET
	f	%	f	%
Skirt/dress	50	38.5	35	26.9
Suit	1	8	1	8
Trousers	16	12.3	22	16.9
T-shirt	20	15.4	16	12.3
Athletic suit	6	4.6	11	8.5
Jewelry	2	1.5	4	3.1
No clothes	35	26.9	41	31.5
Total	130	100	130	100

Notes. GET = General education teacher; SET = Special education teacher; f = frequency.

). Findings showed that 38.5% of the students drew the GET in ‘a skirt or dress’, while 26.9% preferred to draw him/her as a figure without clothes. In addition, a slightly lower percentage (26.9%) of students drew the SET in ‘a dress or skirt’, while 31.5% of the students drew him/her as a simple figure without clothing. The remaining characteristics had lower percentages. Figure 4a,b depict drawings which are representative of the physical appearance of the teachers.

Teachers’ facial expressions in children’s drawings are shown in

Table 5. Representation of the teachers’ facial expressions in the students’ drawings.

	GET		SET
	f	%	f	%
Angry	10	7.7	1	8
Happy	38	29.2	42	32.3
Excited	9	6.9	8	6.2
Confused	5	3.8	4	3.1
Sad	5	3.8	-	-
Worried	1	8	-	-
Smiling	37	28.5	45	34.6
Rationale	3	2.3	2	1.5
No expression	22	16.9	28	21.5
Total	130	100	130	100

Notes. GET = General education teacher; SET = Special education teacher; f = frequency.

. Students depicted the GET and SET mainly as being happy (GP: 29.2% versus 32.3%), smiling (28.5% v. 34.6%), or with no discernible facial expression (16.9% v. 21.5%), thus showing an overall positive social representation of the teachers (Table 5). Figure 5 shows the teachers being happy and cheerful.

Physical characteristics of the teachers are presented in

Table 6. Representation of teachers’ physical characteristics in the students’ drawings.

	GET		SET
	f	%	f	%
He/she wears glasses	19	14.6	6	4.6
Clean and groomed	9	6.9	-	-
Untidy hair	5	3.8	7	5.4
Well-groomed hair	39	30	22	16.9
Bald	-	-	1	8
With beard/mustache	7	5.4	3	2.3
Young person	12	9.2	49	37.7
Elderly	7	5.4	-	-
No physical characteristics	32	24.6	42	32.3
Total	130	100	130	100

Notes. GET = General education teacher; SET = Special education teacher; f = frequency.

. The GET is depicted mainly with well-groomed hair (30%), or with no physical characteristics (24.6%), while the SET is usually depicted at a young age (37.7%), with no physical characteristics (32.3%), followed by well-groomed hair (16.9%). In Figure 6, the GET (human figure on the left) has well-groomed hair, while the SET (human figure on the right) looks younger. It seems that teachers are positively perceived by students in terms of physical appearance.

The analysis of the size of the teachers in the drawings shows that most students captured GET (53.1%) and SET (57.7%) in their actual size in the classroom (

Table 7. Representation of teachers’ size in children’s drawings.

	GET		SET
	f	%	f	%
Bigger than the real one	50	38.5	44	33.8
Smaller than real	11	8.5	11	8.5
Actual	69	53.1	75	57.7
Total	130	100	130	100

Notes. GET = General education teacher; SET = Special education teacher; f = frequency.

). However, several students drew them larger than their actual size (GET: 38.5%, SET: 33.8%). Figure 7a shows a real-life size ratio, while Figure 7b shows teachers in a larger size than in real life. A possible reason for this discrepancy may be that bigger size reflects children’s respect towards the teacher as an important figure, while real size reflects a rather balanced, realistic approach.

The study also examined the physical actions of teachers, and relevant findings are reported in

Table 8. Representation of teachers’ actions in the classroom in the students’ drawings.

	GET		SET
	f	%	f	%
Writing on the board	2	1.5	-	-
Standing	70	53.8	48	36.9
Teaching	37	28.5	2	1.5
Walking around the classroom	-	-	34	26.2
Speaking to students	4	3.1	6	4.6
Sharing exercises	1	0.8	-	-
Caressing students	-	-	2	1.5
Playing with students	-	-	2	1.5
Sitting next to students	16	12.3	36	27.7
Total	130	100	130	100

Notes. GET = General education teacher; SET = Special education teacher; f = frequency.

. One can see that students mainly drew the GET standing (53.8%) or teaching (28.5%), while the SET often stands (36.9%), walks around the classroom (26.2%) or sits close to a student (27.7%) This pattern is in accordance with the ‘one teach, one assist’ model. In Figure 8, the GET teaches Mathematics, while the SET stands close to a student.

Table 9. Representation of objects held by teachers in the students’ drawings.

	GL		OP
	f	%	f	%
Table marker	15	11.5	1	0.8
Book	7	5.4	1	0.8
Photocopy	4	3.1	-	-
Ruler	4	3.1	-	-
Pen	1	0.8	-	-
Flowers	-	-	1	0.8
Doesn’t hold anything	99	76.2	127	97.7
Total	130	100	130	100

Notes. GET = General education teacher; SET = Special education teacher; f = frequency.

indicates the objects teachers held in the drawings. The majority of the students drew GET (76.2%) and SET (97.7%) without holding objects. Those who painted the GET holding something usually drew her/him as holding a marker (11.5%) or a book (5.4%). In Figure 9, both teachers are depicted to teach without holding an object. This pattern likely reflects the way children conceptualize teachers as “people first, workers second”.

Drawings of students with and without SEN were compared using the χ² criterion. There were significant differences between the two groups. Specifically, a considerably higher proportion of students with SEN (68.8%) tended to draw the SET at a young age compared to students without SEN (33.3%) [χ² (6) = 11.50, p = 0.046; Cramer’s V = 0.42; 95% CI [0.28–0.83] for students without SEN; [0.02–0.40] for students with SEN]. In addition, 81.3% of the students with SEN drew the SET in a larger size than the actual, while only 27.2% of the students without SEN did so, with the remaining 64% depicting SETs in their actual size [χ² (2) = 17.64, p < 0.001; Cramer’s V = 0.49; 95% CI [0.72–1.00] for students without SEN; [0.83–1.0] for students with SEN]. Figure 10 displays the distribution of the students’ responses across the three categories (Size: ‘larger than actual’, ‘smaller than actual’, actual), while Figure 11 displays a drawing by a student with SEN that depicts the SET in a considerably larger size than the GET. The student’s drawing also includes him/herself and two classmates.

Figure 12 below focuses on the disposition of the teachers and the instructional setting in the classroom. One can see that the GET stands in front of the students teaching them, while the SET is next to a child touching her shoulder. A heart painted over the teacher probably indicates the expression of love from the student towards the SET.

Finally, a student without SEN drew the two teachers having equal sizes, appearing in a ‘skirt or dress’, with well-groomed hair, and standing on each side of the board (Figure 13). This could be interpreted as the student perceiving both teachers as equal figures of authority, with the board being depicted as the shared centre of learning.

4. Discussion

The current study aimed to investigate the perceptions of primary school students with and without SEN regarding the roles, duties and cooperation of teachers in the co-taught classroom, as well as aspects of co-teaching, gathering data from students’ questionnaire responses and drawings. In relation to the first aim of the study, the findings reflect a distinction in the children’s perception of the learning gains stemming from GETs and SETs, depending on whether the children had SEN or not. Specifically, students overall perceived that they learn best from the GET and then from both teachers, although students with SEN perceived that they learn better from both teachers and then by the GET. In addition, distinctions were observed in children’s perception of the roles of the teachers, indicating a lack of a shared mental model of balanced duties across GETs and SETs; the GET is perceived as the main teacher who plans, assesses and is in charge of most lessons, while the SET has an assistive or supportive role, mainly being responsible to satisfy the individual needs of students diagnosed with disability or SEN. The above findings are in line with earlier research disclosing the SET’s role as an assistant in co-taught classrooms (e.g., King-Sears & Strogilos, 2018; King-Sears et al., 2014; Takala & Uusitalo-Malmivaara, 2012). When students with SEN were asked about teaching, planning and assessment issues, responsibilities were first assigned to the GET followed by both teachers, thus emphasizing the role of the GET as the primary authority figure in charge of the entire class and the implementation of the curriculum (Embury & Kroeger, 2012; King-Sears & Strogilos, 2018). This finding aligns with the co-teaching model of ‘one teach, one assist’ (Friend et al., 2010) which is implemented by the majority of teachers in many countries, including Greece. It is worth noting here that genuine inclusion promotes co-teaching as a way to foster equality, shared responsibility and meaningful participation for all learners (Zamkowska & Nogaj, 2024). Overall, the study’s findings highlight novel points that contribute to bridging the gap in knowledge and skills of stakeholders in inclusive education, with children’s questionnaire responses and drawings identifying specific ways in which inclusive teaching practices could be implemented more efficiently in the classroom. In terms of support, the study highlights the promising potential benefits of co-teaching when students with and without SEN are provided with equal access to both GETs and SETs. Students without SEN seem to ask for help from both teachers in the classroom, while students with SEN tend to seek help or advice primarily from the SET or from both teachers. When they make mistakes, students with SEN turn either to both teachers or the SET. This dual perception likely reflects a gradual shift from the ‘one teach, one assist’ model to a collaborative teaching practice, further suggesting that the SET could have an active role and work equally well with any student in need of assistance, and the same applies for the GET (Strogilos & Tragoulia, 2013). As Rönn-Liljenfeldt et al. (2023) suggest, ‘although SETs are not the lead teachers, the ‘one teach, one assist’ co-teaching model can be effective when active support is provided to all students’ (p. 11).

The second aim of the study has examined co-teaching in terms of equity, classroom atmosphere and ways of learning. Co-teaching and sharing responsibilities were highlighted as a valuable strategy for improving learning outcomes. Specifically, according to the findings of our study, the majority of the students liked having two teachers in the classroom. Their responses also converged on the fact that the classroom atmosphere in the co-taught classroom was positive, and that both teachers were equal, sharing responsibilities for behavior and lesson management. Although the majority of the students wished to have two teachers in all the classes, some students felt confused by this condition. The above findings indicate that students from classes where co-teaching was applied mainly had positive experiences and perceptions. It is worthwhile continuing to develop the co-teaching model moving on to more effective practices (e.g., team teaching). Similar findings have been reported by previous literature (e.g., Strogilos & Tragoulia, 2013; King-Sears et al., 2014; Rönn-Liljenfeldt et al., 2023). These studies show that teachers, when doing co-teaching, enrich the learning environment through exploiting their different personality styles and their variety of teaching methods and activities. Moreover, they try to keep a calm atmosphere in the classroom. By modelling collaboration and valuing diverse learning needs, co-teaching promotes a classroom climate of fairness, inclusivity, and respect for one another (Sharma & Salend, 2016). Furthermore, more balanced roles of teachers within the co-taught classroom would foster fairer interactions between educators and learners, leading to improved learning outcomes (Qualls et al., 2025).

The third aim of the study has attempted to conceptualize co-teaching through drawings. The latter have provided a useful independent means of identifying student–teacher relationships in classroom, and also offered complementary data about the role and duties of GETs and SETs, opening a window into the emotional and symbolic world of the children. Students portrayed their teachers mainly as being ‘female’, a finding which is in line with most previous research (e.g., Arslan Cansever, 2017; Aykaç, 2012). The presence of a significant number of female teachers in schools affect children’s perceptions because primary school children learn about gender roles by observing adults (Bessette, 2008). In addition, students portrayed their teachers most often as appearing in ‘skirts or dresses’, while some students preferred to draw them as figures without any clothes. This can be explained by gender acquisition roles while children go through socialization processes (Arslan Cansever, 2017). Furthermore, children represented both the GET and the SET as being ‘cheerful’, which implies that they had a positive image of them. As far as the physical characteristics of teachers are concerned, students drew the GET mainly with ‘well-groomed hair’ while the SET was consistently younger than the GET and with well-groomed hair. The latter can be explained because SETs in Greece are overall younger that the average age of permanently appointed teachers, especially since it is mainly recent university graduates that are hired each school year as ‘substitute teachers’ in the co-taught classrooms (Koutsoklenis & Papadimitriou, 2021). In addition, children preferred to draw their teachers without holding objects, giving emphasis on the “person” and not on the “object”. While children usually draw their teachers as bigger than actual (Aykaç, 2012), in our study the majority of the students drew their teachers in their actual size, giving a realistic representation of their figures. This can be explained in terms of the students’ chronological age, because the participants of the current study attended the upper three grades of primary school. However, when we focus on children with SEN, we see that they drew the SET in a larger size than the actual one, which further implies their admiration, respect, sense of authority or power, or their need to establish meaningful and positive relationships with him/her (Bessette, 2008). Another finding which deserves attention is students’ level of realism in the visualization of the teachers’ disposition in the classroom. More specifically, students were found to draw the GET standing in front of the whiteboard or near his/her desk, ‘teaching and relaying information’, while the SET was frequently portrayed to walk around the classroom or sit together at the desk with a child. This finding offers support to the ‘one teach, one assist’ model.

By combining the questionnaire data with the drawing evidence, a more holistic picture of students’ perspectives of co-teaching has emerged. More specifically, while the output of the questionnaire has quantified aspects of co-teaching across the sample, the drawings have added more depth to the findings by revealing emotions, symbols and fine-grained nuances that enable a better grasp of the students’ subjective experiences within the co-taught classroom. Children’s drawings included bright colours, smiling facial expressions and well-groomed figures, thus revealing a happy, collaborative classroom atmosphere, and a positive image of GETs and SETs. Students’ ratings of co-teaching in the questionnaire have been generally positive, and this finding seems to align with the drawing evidence of the study, further implying that drawings were a particularly salient correlate of children’s perception of co-teaching practices.

To sum up, although our study reflects the ‘one teach, one assist’ model, students in the co-taught classrooms generally felt that their learning needs were being met and that they benefited from having two teachers. However, the students were sometimes spilt between the GET and both educators, especially in terms of who does what, although in general they perceived that most tasks were carried out by both educators who have been collaborating in the classroom. Students with SEN seemed to feel closer to the SET, which is also reflected in the implementation of the legislative framework of parallel support in Greek schools. Our findings are consistent with earlier Greek studies (Mavropalias & Anastasiou, 2016; Strogilos & Tragoulia, 2013), but it seems that there has been progress in the implementation of co-teaching over the years. This is mainly reflected in the fact that our study has shown more balanced roles for the two teachers, as well as more complementary responsibilities as compared to previous relevant research. Although children seemed to be satisfied with the co-teaching practices, we have no quantifiable data about their benefits in terms of academic progress and social development. In the past years, the Greek educational system has witnessed a tendency to move away from the model of designating a SET to one or two students, implementing a different, more balanced co-teaching system where co-teachers undertake active instructional roles (King-Sears et al., 2014).

Finally, the present study on students’ perceptions of co-teaching can meaningfully contribute to high-quality, inclusive education that moves beyond tokenistic approaches. It can offer insights on the transition from the mere integration of students with SEN into mainstream classrooms to a “genuine inclusion”. Understanding how students in Greece perceive the effectiveness and benefits of co-teaching practices provides comparative evidence that may enrich international discussions. The current study is novel in that it reveals important avenues for exploring how students (with and without SEN) experience co-teaching practices in classroom and ways to capitalize on this pathway to promote inclusive education and the satisfaction of the students’ diverse learning needs. As reported by Qualls et al. (2025), in order to “understand how each partnership’s components function and what supports are needed, a view from inside the system instead of outside is needed” (p. 12).

5. Limitations

This study has several limitations which need to be taken into account. Given the small number of students with SEN in the sample, we were not able to do comparisons based on the gender and the grade of the participants, which would have probably revealed differences in the students’ perceptions of co-teaching. Furthermore, children’s responses may have been affected by their teachers being physically present in the classroom during data collection. Specifically, the teachers’ physical presence during data collection may have constrained students’ willingness to respond openly, instead encouraging responses shaped by social desirability. If SEN students’ identities or support needs were inadvertently revealed, or if their responses were evaluated differently, the reliability and validity of the data may have been compromised. Finally, time limitations did not allow us to collect data through similar research instruments from the teachers in the co-taught classrooms, which would have shed further light in their perspectives of co-teaching.

6. Conclusions and Recommendations for Future Research

Co-teaching has been promoted globally as a method for achieving inclusion (UN, 2006; UNESCO, 2020). It allows teachers to address diverse learning needs within the same classroom. Its implementation has positive characteristics but also presents difficulties and challenges. To the best of our knowledge, the current study is the first one conducted within the Greek educational context that has focused on the perceptions of primary school students with and without SEN about the roles and responsibilities of co-teachers in mainstream Greek classrooms. Although the students assessed positively the co-teaching model of ‘one teach, one assist’ in terms of equity, classroom atmosphere and support, there is much to be done to move beyond this model to more collaborative co-teaching practices (e.g., team teaching), and make the most of the special knowledge and skills of the teachers (Mavropalias & Anastasiou, 2016).

Future research could explore and evaluate students’ experiences of co-teaching in primary as well as in secondary school. Obtaining their feedback can serve as an effective approach to help co-teachers further improve their methods and make important modifications to co-taught instruction. Next, further research is needed to combine data from students and teachers, since co-teachers’ and students’ expectations and experiences of co-teaching practices in class may differ significantly (Bessette, 2008; Hang & Rabren, 2009). However, mutual respect and active collaboration between teachers are necessary to reap the greatest benefits, which is often a challenge in educational practice. To conclude, the upward trend in the proportions of students with disabilities and/or SEN internationally has contributed to the increase in attention paid to the presence of SEN in mainstream schools. As a result, teachers have the opportunity to teach diverse students and develop positive attitudes towards having a student with SEN in their classroom (Strogilos et al., 2023; Koutsoklenis & Papadimitriou, 2021). Of course, more research is warranted to investigate the effectiveness of different co-teaching models in classrooms and adapt them to the learning needs of students with and without SEN.