Inclusion of children with disabilities and special educational needs in physical education: an exploratory study of factors associated with Irish teachers’ attitudes, self-efficacy, and school context

ABSTRACT

Since the publication of the Education for People with Special Educational Needs Act in 2004, inclusion in Ireland has become a priority for the educational system. Despite the promotion of physical education (PE) as a key setting for facilitating inclusion in schools, research has shown that the children with special educational needs and disabilities are still excluded from PE. Research evidence shows that teachers’ attitudes and self-efficacy and the broader school context within which teachers work are among the most important factors facilitating inclusion in PE. Therefore, the aim of this study was to explore teachers’ attitudes towards inclusion, self-efficacy in delivering inclusive pedagogies, and perceptions about whether inclusion was supported within the schools. A cross-sectional design was used, and a questionnaire was sent to Irish PE teachers. Results of this study showed that teachers’ attitudes were positively correlated to the number of different types of disabilities in PE class. It was found that secondary school teachers had higher self-efficacy levels than their primary school colleagues. Findings from this study revealed that teachers who had frequent personal contact with SEND people showed more positive perceptions about the extent to which inclusion was supported in the school within which they worked.

KEYWORDS:

• Teachers’ attitudes
• teachers’ self-efficacy
• school context
• physical education
• Ireland

Introduction

Inclusive education in Ireland has become a priority over the last decades. Since the Education for People with Special Educational Needs Act (2004) that marked a fundamental shift in provision of inclusive education (Griffin and Shevlin 2007), the educational priority towards the education of children with Special Educational Needs and Disabilities (SEND) has advanced and shifted from segregated education to inclusive education (Drudy and Kinsella 2009). The need to provide a fully inclusive setting within the school has become, therefore, pivotal (Meegan and MacPhail 2006) for the Irish educational system, which has allocated resources and funding to support the education of SEND children in mainstream schools (Kenny, McCoy, and Mihut 2020). Over the past years, therefore, the number of students with SEND attending mainstream schools has increased (McConkey et al. 2016). In this context, it has become crucial the need to create an environment in which inclusion was fully promoted. In this sense, physical education (PE) has been constantly reported as a crucial setting for promoting the inclusion (Block and Obrusnikova 2007). PE not only provides a structured opportunity for SEND students to meet the daily physical activity guidelines (Carty et al. 2021), but it also fosters an ecosystem of complex interactions and relationships that enable SEND students to develop those skills needed to participate in the wider community setting (J.I. Butler 2006). However, despite the support that inclusion in mainstream schools has received in Ireland, children with SEND are not still fully included in school. In this regard, alarming signals come from recent research reporting that, over the past decade, the number of special classes in operation has increased from ∼700 to almost 1,800 (Shevlin and Banks 2021). Research has also highlighted that fact that students with intellectual or developmental disabilities in Ireland are put on short school days, rather than fully included in mainstream classes (Brennan and Browne 2019).

As teachers have always been at the forefront in the classroom (Leonard and Smyth 2022), they were soon deemed to be fundamental in establishing an inclusive learning environment within the classroom (Forlin 2010; Leung and Mak 2010). Researchers have argued that having more favourable attitudes toward individuals with SEND in general or PE classes creates a fruitful environment for teachers to provide more successful inclusive practices to students with SEND profiles (Meegan and MacPhail 2006). In this sense, teachers’ attitudes toward the inclusion of SEND children have become one of the most important key elements to be investigated by research (Avramidis and Norwich 2002; Folsom-Meek, Nearing, and Kalakian 2000; Kudláček, Sherrill, and Válková 2002; Morley et al. 2005; Stewart 1988; Tripp and Sherrill 1991; Vickerman 2002). And this research trend also became prominent within Ireland. An early cross-sectional study investigating factors affecting general teachers’ attitudes in Ireland reported, unsurprisingly, that positive teachers’ attitudes toward inclusion were crucial for building ‘solid foundations’ for an inclusive school setting (S. Butler and Shevlin 2001). A more recent study found that, despite having favourable attitudes towards inclusion, Irish teachers still perceived barriers (e.g. curriculum and assessment, lack of time, class size, and inadequate resources) affecting their attitudes and, overall, inclusion (Young, McNamara, and Coughlan 2017).

Attitudes, however, are not the only teachers’ characteristic that has been investigated to increase the inclusion of children with SEND profiles. Another key element that has been widely highlighted by researchers as crucial for fostering an inclusive environment for SEND students is teachers’ self-efficacy (Avramidis and Norwich 2002; Forlin 1998; M.L. Hsien 2007; Soodak, Podell, and Lehman 1998). Described as one of the strongest moderators of attitudes (Bandura 1977), teachers’ self-efficacy has been widely investigated by previous research. For example, it has been shown that primary school pre-service teachers may have higher levels of self-efficacy than secondary school pre-service teachers (Baker 2005). Furthermore, research has shown that pre-service teachers with higher levels of self-efficacy vis-à-vis inclusive education tended to have more favourable attitudes towards the inclusion of SEND children in their classrooms (Gao and Mager 2011). In relation to the Irish context, a recent qualitative study conducted at a general level reported that the majority of the teachers interviewed felt a lack of confidence regarding teaching students with SEND (Anglim, Prendeville, and Kinsella 2018), suggesting that low levels of self-efficacy may constrain the inclusion of SEND students.

However, despite research evidence is flourish regarding general teachers in the Irish context, little is known about PE teachers’ attitudes or self-efficacy levels towards the inclusion of SEND students. For example, the one and only study on Irish PE teachers’ attitudes was conducted almost two decades ago. In this quantitative analysis, Meegan and MacPhail (2006) investigated the associations between teachers’ demographic characteristics and their attitudes towards the inclusion of children with SEND profiles (Meegan and MacPhail 2006). Meegan and MacPhail concluded, unsurprisingly, that Irish PE teachers’ attitudes towards inclusion were mainly influenced by the severity of the disability and previous experience teaching SEND students. On the other hand, in relation to Irish PE teachers’ self-efficacy levels, the only study in Ireland was conducted by Tindall, Culhane, and Foley (2016) on pre-service teachers. This experimental study explored the impact of an inclusion-related programme on Irish pre-service PE teachers’ self-efficacy. The authors found that participants who attended the programme showed a significant increment in their self-efficacy levels toward the inclusion of children with SEND then their peers in the control group (Tindall, Culhane, and Foley 2016).

One limitation that has been predominant among research investigating teachers is the focus on one construct only – namely either attitude or self-efficacy. In their literature review Hutzler et al. (2019) highlighted the fact that of seventy-five papers included, only three studies investigate both teachers’ self-efficacy and attitudes (Hutzler et al. 2019). In this sense, considering that self-efficacy moderates attitude, the need to investigate both PE teachers’ attitudes and self-efficacy levels toward inclusion is necessary. Furthermore, researchers have also identified the school context as an assumption of inclusive education, suggesting that school ought to nurture and educate all students, regardless of differences in ability, ethnicity, culture, gender, language, and class (Kozleski et al. 2009). In this regard, it appears obvious that the school and its staff, including teachers, need to cooperate to transform the school and foster an inclusive educational setting (Savolainen et al. 2012). Notwithstanding the importance of a critical role that teachers play in the implementation of inclusive education (Forlin et al. 2010), research has also shown the importance of the school context in shaping teachers’ behaviours (Caprara et al. 2006), including those towards inclusion.

Therefore, the aim of this study is to investigate Irish PE teachers and explore how their demographic’s attributes and the characteristics of the school in which they work influence their (i) attitudes towards the inclusion of children with SEND profiles in mainstream PE classes; (ii) self-efficacy in delivering inclusive pedagogies; and (iii) perceptions of the context in which they work. The present study addresses the directions for future research within the Irish context suggested both by Meegan and MacPhail (2006) and Tindall, Culhane, and Foley (2016). The authors recommend that future studies needed to further investigate Irish PE teachers and their perceptions and attitudes regarding working with students with SEND profiles.

Materials and methods

Research design and participants

In June 2020, a request for participation in this cross-sectional study – including a description of the study aims and objectives and a link to an online questionnaire prepared in Survey Monkey – was sent to all schools across Ireland: 3,106 primary schools; 723 secondary schools; and 134 special schools. Schools were asked to forward the email containing the questionnaire link to the teachers teaching PE in the school during the current year. School email addresses were retrieved from the Irish Department of Education and Skills website. In total, 329 participants accessed the online questionnaire, of which 260 completed the survey partially or in full, resulting in a 79% completion rate. Ethical approval for this study was granted by the [University College Dublin, Human Research Ethics Committee – Sciences, Reference N: LS-E-19-177-Tarantino-Neville]. Participants provided informed consent prior to completing the online questionnaire, and no incentives were given for participation.

Measures

The questionnaire was divided into four main sections considering teachers’: (i) perceptions of the school context in which they worked; (ii) self-efficacy in providing inclusive pedagogies during PE classes; (iii) attitudes towards the inclusion of children with SEND in PE classes; and (iv) demographic and broader schools characteristics. Further details about the sub-sections of the questionnaire are described below.

Outcome variables

School context

Teachers’ perceptions about the school context within which they worked were ascertained using Makopoulou et al.’s (2021) ‘About Your School’ inventory. This 13-item inventory was developed for and used in a national evaluation of an inclusive PE teacher training programme that was delivered across England between 2013 and 2016 (Makopoulou et al. 2021). The inventory is based on Booth and Ainscow’s (2002) Index of Inclusion, as well as the notion of school as ‘learning organisations’ (Kools et al. 2020), and comprises three sub-scales concerning teachers’ perceptions of the: (i) general school context within which they work; (ii) the overall value attributed to the notion of inclusion at the school level; (iii) and quality of inclusion provided during the PE classes. The general school context sub-scale (4 items) refers to the teachers’ perceptions of the context in which they work, which comprises statements about the collegiality and the opportunities to share knowledge and practice (example question: ‘There are opportunities to cascade knowledge about inclusion to colleagues after CPD attendance’). The sub-scale about the sense of inclusion at the school level (4 items) refers to whether the school’s culture or ethos is inclusive, and the availability of resources provided to support inclusive practice (example question: ‘The PE curriculum is developed taking into consideration the different abilities and needs of all students’). Finally, the last sub-scale (3 items) refers to the quality of inclusive PE opportunities for pupils with SEND (example question: ‘All students in my class, including SEND students, learn together regardless of their (dis)abilities’). A 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree) accompanied each of the 13 items. Cronbach’s reliability score (α) for the school context inventory was .82, confirming the internal consistency of the scale (i.e. α ≥ .70).

Self-efficacy

Based on this definition, teachers’ perceptions of self-efficacy were ascertained using an inventory developed by Makopoulou et al. (2021), which is a refinement and extension of previous instruments conceptualised and validated by Humphries et al. in 2012 and Block et al. in 2013 (Block et al. 2013; Humphries et al. 2012). The inventory is based on Bandura’s (1977) definition of self-efficacy as an individual’s belief about how well he or she can execute a course of actions in a given situation (Bandura 1977). The resulting 10-item inventory comprised two sub-scales measuring: (i) teachers’ personal beliefs about teaching PE; and (ii) teachers’ confidence in supporting pupils with SEND profiles. The belief sub-scale (6 items) asks teachers to consider their efficacy to teach PE, as well as their readiness to adapt activities depending on individual students’ needs (example question: ‘How confident are you in your ability to change a task to make it easier for a student who is having trouble achieving in this task?’). The remaining 4-items asks teachers to reflect on their confidence to support students within PE classes, irrespective of ability level (example question: ‘How confident you are in your ability to support all students, including students with SEND, understand where they are in their learning, where they need to go next and how to get there?’). Each question was accompanied by a 5-point Likert scale ranging from 1 (not at all confident) to 5 (extremely confident). Cronbach’s α reliability score for the inventory was α = .94, confirming the inventory’s internal consistency.

Attitude scale

Teachers’ attitudes towards the inclusion of SEND children in PE were ascertained using the ‘Physical Educators’ Judgements about Inclusion’ inventory, which was developed and validated by Hodge et al. in 2002 (S.R. Hodge, Murata, and Kozub 2002). This inventory is based on Ajzen’s (1991) Theory of Planned Behaviour, which defines attitudes as the degree to which an individual has a favourable or unfavourable personal evaluation vis-à-vis an object, activity, or event (Ajzen 1991). The inventory is comprised of 15 items, which can be divided into three sub-scales focusing on teachers’: (i) judgements about inclusion in PE (example question: ‘Inclusion is an idealistic philosophy that will not work in PE’); (ii) acceptance of students with disabilities (example question: ‘I would readily accept teaching a student with a physical disability in PE’); and (iii) perceived training needs (example question: ‘It is important that I receive training on […] lesson planning for a variety of ability levels’). Each question was accompanied by a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). This inventory originally comprised questions regarding the following four categories of disabilities: hard of hearing; visual impairment; learning disability; and physical disability. However, given significant changes in the nature of and terminology used to define SEND profiles since Hodge, Murata, and Kozub (2002) original publication, in this study, we extended the categories of disability as follows: emotional and behavioural disorder; intellectual disability; specific learning disability; physical disability; autism spectrum disorders; and visual/hearing impairment. A reliability score of α = .81 confirmed the internal consistency of the inventory against the standard criterion of α ≥ .70.

Independent variables

Demographics

The final section of the online questionnaire comprised items regarding teachers’ demographics and broader schools’ characteristics. Teachers reported their age, gender, number of years teaching PE, number of years teaching SEND students, and number of SEND-related professional development courses they have undertaken. Teachers also reported the type of school in which they worked (i.e. primary, secondary, or special), total number of pupils in the school, average number of students in a PE lesson, and the different SEND profiles they taught in their classes.

Statistical analyses

We employed a linear mixed model procedure to estimate the effects of teachers’ demographic and schools’ characteristics on teachers’ (i) attitudes toward inclusion; (ii) self-efficacy on delivering inclusive pedagogy; and (iii) perceptions of the context in which they work. Imputed data were analysed using Statistical Analysis System (SAS® Studio, University Edition – SAS Institute Inc., Cary, NC, U.S.A.). The imputed dataset was analysed using the linear mixed model programme PROC MIXED. The precise effects of the independent variables were derived using least squares means estimates in the PROC MIXED procedure. For the categorical independent variables, the estimates represent differences in the adjusted mean changes between the comparison groups: for example, the difference in the adjusted mean changes in self-efficacy for teachers with an undergraduate or postgraduate degree. For the continuous independent variables, the estimates represent the changes in the outcome variable score when the standard deviation increases by one unit. However, to facilitate a robust interpretation and synthesis of the magnitude of the effect of the predictors on the outcome variables, the parameter estimates yielded by the PROC MIXED procedure were converted into effect sizes by dividing the estimate by the standard deviation of the outcome variable (Tabachnick and Fidell 2007). In particular, the estimates for the continuous variables were converted into correlation coefficients (r), and the estimates for the categorical variables were converted into Cohen’s d. The size of the effect was evaluated using Cohen’s guidelines. The correlation coefficients (r) were evaluated using the following scale: < .10 trivial; 0.10–0.30 small; 0.30–0.50 moderate; > 0.50 large (Cohen 1988). Whereas the effect sizes based on differences between means (Cohen’s d) were evaluated using the following scale: < .20 trivial; 0.20–0.50 small; 0.50–0.80 moderate; > 0.80 large (Cohen 1988). However, following Hopkins’ (1997) guidelines, the uncertainty in the estimate or effect in this study were interpreted in terms of their magnitude and respective upper and lower confidence limits (Hopkins 1997).

Results

Descriptive statistics

Table 1 summarises teacher and school characteristics. Teachers surveyed were primarily females (78%) and held an undergraduate teaching qualification (61%). Sixty-three percent of teachers were from primary schools, 32% were secondary school PE teachers, and the remaining 5% were special school teachers. Whilst twenty-eight percent of teachers reported having obtained a qualification in teaching students with SEND profiles, the majority of teachers (93%) reported having some form of contact with children with SEND profiles in the year prior to completing the questionnaire. In relation to broader school characteristics of the sample, the total number of pupils in teachers’ schools ranged from 12 to 1400, with an average of 444 pupils per school. There was a wide range of class sizes with teachers reporting numbers of students in PE classes ranging from 6 to 35, with an average value of 25. Teachers also reported the average number of children with SEND profiles in their PE classes was 7, with average exposure of teachers to three different types of disabilities in their lessons.

Table 1. Participants’ and schools’ characteristics.

Variable
Gender (n = 231)	n (%)
Female	180 (77.9)
Male	51 (22.1)
Academic Qualification (n = 230)
Undergraduate	140 (60.9)
Postgraduate	90 (39.1)
Qualification in teaching SEND (n = 231)
Yes	64 (27.7)
No	167 (72.3)
Type of School (n = 229)
Primary	145 (63.3)
Secondary	73 (31.9)
Special	11 (4.8)
Contact with SEND people in life (n = 229)
Frequent personal contact	64 (27.7)
Infrequent personal contact	87 (37.7)
Just see them around	63 (27.3)
No contact in the past year	17 (7.4)
	Mean (SD)	Median (Range)
Age (n = 160)	39 (9)	38 (23–62)
Number of years teaching (n = 160)	15 (9)	13 (1–38)
Number of years teaching PE (n = 160)	15 (10)	13 (1–38)
Number of years teaching SEND (n = 160)	13 (8)	12 (1–38)
Number of SEND Courses attended (n = 160)	3 (3)	2 (0–15)
Total number of pupils in the school (n = 160)	444 (300)	400 (12–1400)
No. of pupils in PE class (n = 160)	25 (5)	27 (6–35)
No. of SEND children in PE class (n = 160)	7 (7)	5 (0–30)
No. of different type of SEND in PE class (n = 160)	3 (1)	3 (0–6)

Analysis of the predictors on the outcome variables

Predictors of teachers’ attitude

There was a small-to-moderate positive correlation (r = 0.35; 95%CIs = 0.14–0.56; p = .001) between the number of different types of disabilities in class and PE teachers’ attitudes. Trivial-to-large differences in attitudes were observed between special school and primary (d = −1.32; 95%CIs = −2.65–0.00; p = .050) and secondary (d = −1.41; 95%CIs = 2.76 to −0.05; p = .042) school teachers, with special school teachers holding more favourable attitudes. Finally, teachers who reported having obtained a qualification in teaching children with SEND profiles had possibly more favourable attitudes; however, the uncertainty in the estimate (given by the p value and 95%CI) was too large to be definitive (d = 0.38; 95%CIs = −0.04–0.80; p = .08).

Predictors of teachers’ self-efficacy

There was a small-to-moderate-sized difference between secondary school PE teachers’ and primary school teachers’ self-efficacy levels (d = 0.71; 95%CIs = 0.28–1.13; p = .001), with the former having higher levels of self-efficacy. Other possible differences include the difference in self-efficacy between: primary school and special school teachers’, with special school teachers having higher efficacy beliefs (d = −1.17; 95%CIs = −2.44–0.10; p = .07); and teachers who had different frequencies of contact engagement with people with disabilities in their personal life, with more frequent engagement being associated with higher efficacy beliefs. (d = 0.77; 95%CIs = −0.09–1.22; p = .09). There was also a possible association between experience in teaching PE and perceived self-efficacy, with more years teaching PE possibly associated with higher perceived self-efficacy (r = 0.93; 95%CIs = −0.04–1.89; p = .06).

Predictors of the perceived inclusive context within the school

There was a small-to-moderate-size difference between teachers who had not had contact with SEND people over the year prior to the competition of the survey and those who had ‘just saw them around’ (d = 0.71; 95%CIs = 0.09–1.32; p = .025). Trivial-to-large differences were found between teachers with infrequent personal contact with SEND and teachers who ‘just saw them around’ (d = 0.47; 95%CIs = 0.12–0.83; p = .009), with the former holding more positive perception about inclusivity within the school context. Trivial-to-large differences were also found between teachers who had had frequent personal contact with SEND people and those who had had no contact at all (d = 0.56; 95%CIs = −0.07–1.18; p = .08). A trivial-to-moderate difference was found between subjects who had infrequent contact with SEND people and those who had frequent personal contact with people living with a disability (d = −0.33; 95%CIs = −0.70–0.05; p = .08), with the latter having more positive perception of the inclusive levels within the school.

Although non-statistically significant, a trivial-to-moderate difference was found between male and female teachers, with the former having higher positive perceptions about the extent to which inclusion was supported within the school (d = −0.28; 95%CIs = −0.62–0.06; p = .11). Non-significant trivial-to-moderate differences were found between teachers who held an undergraduate degree and their colleagues who obtained a postgraduate degree (d = −0.31; 95%CIs = −0.64–0.02; p = .07), with the latter having more positive perceptions about the inclusive environment within the school in which they worked. A non-statistically significant trivial-to-small negative correlation was found between teachers’ perceptions about the school context and the total number of pupils attending the school (r = −0.16; 95%CIs = −0.34–0.02; p = .07). Another non-statistically significant trivial-to-small negative correlation was found between teachers’ perception of the school context and the number of students with SEND profiles in their PE classes (r = −0.14; 95%CIs = −0.30–0.02; p = .08). Finally, a trivial-to-moderate negative correlation was found between teachers’ perceptions about the school context and the number of different type of disabilities attending the PE classes (r = 0.13; 95%CIs = −0.06–0.32; p = .19).

The remaining associations between the predictors variables and the outcome variables that were either unclear or trivial are summarised in Table 2.

Table 2. Effect sizes.

Categorical predictors
Effect of the predictor on attitude	Cohen's d	Lower CI	Upper CI	p value	Interpretation
Gender
Females – Males	0.08	−0.29	0.46	0.66	Unclear
Academic Qualification
Postgraduate – Undergraduate	0.19	−0.17	0.56	0.30	Trivial to moderate
Qualification in teaching SEND
Yes – No	0.38	−0.04	0.80	0.08	Trivial to moderate
Type of School
Primary – Special	−1.32	−2.65	0.00	0.050	Trivial to large
Secondary – Special	−1.41	−2.76	−0.05	0.042	Trivial to large
Secondary – Primary	−0.08	−0.53	0.36	0.71	Unclear
Personal contact with SEND
Frequent Contact – No Contact	−0.12	−0.81	0.56	0.73	Unclear
Infrequent Contact – No Contact	−0.28	−0.94	0.37	0.39	Unclear
See them Around – No Contact	−0.23	−0.91	0.44	0.50	Unclear
Infrequent Contact – Frequent Contact	−0.16	−0.57	0.25	0.43	Unclear
See them Around – Frequent Contact	−0.11	−0.56	0.34	0.63	Unclear
See them around – Infrequent Contact	0.05	−0.33	0.44	0.79	Unclear
Effect of the predictor on School Context
Gender
Females – Males	−0.28	−0.62	0.06	0.11	Trivial to moderate
Academic Qualification
Postgraduate – Undergraduate	−0.31	−0.64	0.02	0.07	Trivial to moderate
Qualification in teaching SEND
Yes – No	−0.01	−0.39	0.38	0.96	Unclear
Type of School
Primary – Special	0.29	−0.91	1.50	0.63	Unclear
Secondary – Special	0.49	−0.74	1.73	0.43	Unclear
Secondary – Primary	0.20	−0.20	0.61	0.33	Unclear
Personal contact with SEND
Frequent Contact – No Contact	0.56	−0.07	1.18	0.08	Trivial to large
Infrequent Contact – No Contact	0.23	−0.37	0.83	0.45	Unclear
See them Around – No Contact	0.71	0.09	1.32	0.025	Trivial to large
Infrequent Contact – Frequent Contact	−0.33	−0.70	0.05	0.08	Trivial to moderate
See them Around – Frequent Contact	0.15	−0.26	0.56	0.48	Unclear
See them around – Infrequent Contact	0.47	0.12	0.83	0.009	Trivial to large
Effect of the predictor on Self-Efficacy
Gender
Females – Males	−0.38	−0.74	−0.03	0.036	Trivial to moderate
Academic Qualification
Postgraduate – Undergraduate	0.02	−0.33	0.37	0.90	Unclear
Qualification in teaching SEND
Yes – No	0.15	−0.26	0.55	0.48	Unclear
Type of School
Primary – Special	−1.17	−2.44	0.10	0.07	Trivial to large
Secondary – Special	−0.46	−1.76	0.84	0.49	Unclear
Secondary – Primary	0.71	0.28	1.13	0.001	Small to large
Personal contact with SEND
Frequent Contact – No Contact	0.57	−0.09	1.22	0.09	Trivial to large
Infrequent Contact – No Contact	0.37	−0.26	1.00	0.25	Unclear
See them Around – No Contact	0.35	−0.30	1.00	0.29	Unclear
Infrequent Contact – Frequent Contact	−0.19	−0.59	0.20	0.33	Unclear
See them Around – Frequent Contact	−0.22	−0.65	0.21	0.32	Unclear
See them around – Infrequent Contact	−0.02	−0.39	0.35	0.90	Unclear
Continuous Predictors
Effect of the predictor on Attitude	r	Lower CI	Upper CI	p value	Interpretation
Different types of disabilities in PE class	0.35	0.14	0.56	0.001	Small to moderate
Years teaching	0.25	−0.83	1.34	0.64	Unclear
Years teaching PE	−0.31	−1.31	0.70	0.55	Unclear
Total pupil in the school	−0.02	−0.22	0.18	0.85	Unclear
Number of pupil in PE class	0.04	−0.18	0.25	0.74	Unclear
SEND courses attended	0.01	−0.03	0.05	0.67	Trivial
Number of SEND in PE class	−0.05	−0.22	0.13	0.58	Unclear
Years teaching SEND	0.09	−0.21	0.40	0.54	Unclear
Age	−0.25	−0.63	0.14	0.21	Unclear
Effect of the predictor on School Context
Different types of disabilities in PE class	0.13	−0.06	0.32	0.19	Trivial to moderate
Years teaching	0.14	−0.85	1.13	0.77	Unclear
Years teaching PE	−0.05	−0.97	0.87	0.91	Unclear
Total pupil in the school	−0.16	−0.34	0.02	0.07	Trivial to moderate
Number of pupil in PE class	0.10	−0.10	0.29	0.34	Unclear
SEND courses attended	0.02	−0.02	0.06	0.30	Trivial
Number of SEND in PE class	−0.14	−0.30	0.02	0.08	Trivial to small
Years teaching SEND	0.03	−0.25	0.30	0.83	Unclear
Age	0.06	−0.30	0.41	0.75	Unclear
Effect of the predictor on Self-Efficacy
Different types of disabilities in PE class	−0.08	−0.28	0.12	0.45	Unclear
Years teaching	−0.71	−1.76	0.33	0.18	Unclear
Years teaching PE	0.93	−0.04	1.89	0.06	Trivial to large
Total pupil in the school	−0.03	−0.22	0.16	0.79	Unclear
Number of pupil in PE class	0.01	−0.19	0.22	0.89	Unclear
SEND courses attended	0.03	−0.01	0.07	0.13	Trivial
Number of SEND in PE class	0.02	−0.15	0.19	0.80	Unclear
Years teaching SEND	0.10	−0.19	0.39	0.49	Unclear
Age	−0.20	−0.57	0.17	0.29	Unclear

Discussion

Teachers’ attitudes, self-efficacy levels, and school environment are key elements required to create an inclusive setting in PE classes for students with SEND profiles (Avramidis and Norwich 2002; Folsom-Meek, Nearing, and Kalakian 2000; Hutzler et al. 2019; Kudláček, Sherrill, and Válková 2002; Morley et al. 2005; Tripp and Sherrill 1991; Vickerman 2002). Despite Ireland having enacted policy to support inclusive education over the past two decades, challenges for teachers still exist within PE classes (Shevlin, Winter, and Flynn 2013). Previous studies of Irish PE teachers concluded that further future investigations were required to further elaborate on the correlates of (i.e. potential reasons for) teachers’ attitudes (Meegan and MacPhail 2006). Therefore, the main purpose of the present paper was to empirically investigate the factors that are associated with Irish PE teachers’ (i) attitudes vis-à-vis inclusion; (ii) self-efficacy in delivering inclusive pedagogies; and (iii) perceptions about the school context in which they worked.

Perhaps the most surprising finding of this study was the positive correlation found between different types of disabilities in class and PE teachers’ attitudes. Previous research has primarily focused on the impact of the type and degree of the disability on teachers’ attitudes and self-efficacy (Casebolt and Hodge 2010; Hersman and Hodge 2010; S. Hodge et al. 2004; Meegan and MacPhail 2006; Morley et al. 2005; Sato et al. 2007; Sato and Hodge 2009), and a recent systematic review and meta-analysis has concluded that, overall, teachers’ attitudes were more favourable towards teaching students with mild disabilities (Tarantino, Makopoulou, and Neville 2022). However, our findings showed, for the first time in published research, that teachers who taught students with a wider range of SEND held more positive attitudes toward the inclusion of such children in PE classes. These results can be interpreted as saying that exposure to different types of disabilities can lead to more favourable attitudes vis-à-vis inclusion. Such findings are novel to PE but align well with previous research results that showed that general education teachers who were exposed to different types of disability had more positive attitudes toward inclusion (Goddard and Evans 2018; Keilbaugh 1977; Wall 2002).

The findings of this study also show that previous personal contact with people with SEND (i.e. outside of the classroom) influences teachers’ perceptions about inclusion. Our findings revealed that higher levels of contact with people with SEND were associated with more positive perceptions about the extent to which inclusion was supported within the school in which teachers work. Prior research has reported the school context to be one of the most important factors influencing and shaping teachers’ attitudes, self-efficacy, and behaviours (Caprara et al. 2006; Hutzler, Zach, and Gafni 2005; Obrusnikova 2008; Özer et al. 2013; Rizzo and Vispoel 1991; Tripp and Sherrill 1991). However, our results now suggest that this association might be reciprocal. For example, it is quite possible that teachers with more frequent contact with SEND in their personal life know more about and are therefore more attuned to the ways in which the needs of a person with a SEND profile can be and are met within a given school. It is also quite possible that teachers with more frequent contact with SEND in their personal life positively influence other school staff and thereby shape the beliefs and values that promote inclusion within a given school context. Such findings and interpretations align well with previous theoretical models; for example, social cognitive theory (Bandura 1986), which proposes a bi-directional relationship between environmental and personal characteristics, and previous research in general education has begun to explore in more depth (Thibaut et al. 2018).

In terms of teachers’ attitudes toward inclusion, our findings revealed significant differences between special school teachers and teachers teaching in both primary and secondary, with the former reporting more favourable attitudes toward inclusion. Given that special school teachers work in a setting where the specific goal of inclusion for all based on individual needs and specific abilities is fundamental to teacher practice, these findings are perhaps unsurprising. However, it is potentially surprising that no differences were found between primary and secondary school teachers’ attitudes. Though, a lack of evidence for differences in attitudes towards inclusion between teachers in primary and secondary school settings has been signalled in previous research. In their comprehensive literature review on attitudes towards inclusions, Avramidis and Norwich (2002) reported conflicting evidence from across numerous empirical studies: i.e. some found that secondary school teachers showed more favourable attitudes towards inclusion than counterparts working in elementary schools, whereas others show the opposite. Although such an analysis was not part of the present study, it is possible that differences in attitudes towards inclusion between teachers working in different settings are mediated by the nature and characteristics of the school context within which teachers work.

Another interesting, yet surprising, finding was the differences in self-efficacy levels between primary and secondary school PE teachers. Contrary to what reported by Baker (2005) in the American context, our results showed that secondary school teachers had significantly higher levels of self-efficacy in providing inclusive pedagogies when compared to their primary school counterparts. Although teachers’ self-efficacy has been widely investigated in the Irish context (Hosford and O'Sullivan 2016; Ryan and Mathews 2022; Tindall, Culhane, and Foley 2016), research evidence has been lacking regarding the differences between primary and secondary school teachers, especially within the context of inclusive PE. A possible explanation for the difference in self-efficacy among primary and secondary school teachers could be found in the difference between the curricula and subsequently the provision of PE in primary and secondary schools in Ireland. Although previous evidence has not been reported in the context of inclusive PE, an explanation can be drawn up by looking at general education. Irish primary teachers teach PE alongside other subjects, whereas secondary school PE teachers teach PE only. This may result in a lack of practical confidence and perceived competence among primary teachers as they cannot specialise and simply cannot spend as much time on task learning about or implementing PE activities as a specific subject (among many others). A recent qualitative study conducted by Ní Chróinín and O’Brien (2019) reported that primary teachers did not tend to have a background in sport and/or structured physical activity and that they were also ‘positively disposed’ towards the provision of PE activities by external educators (331). Ní Chróinín and O’Brien (2019) reported that, despite enjoying teaching PE, primary teachers preferred having an external educator teaching PE, who was deemed having more expertise and management skills in teaching PE (Ní Chróinín and O’Brien 2019). In this study, one participant even voiced that having someone with ‘a more fine-tuned skills set’ was pivotal for children’s enjoyment and learning (Ní Chróinín and O’Brien 2019, 331). Another obvious difference between primary and secondary school teachers is that the former (primary school teachers) do not need a specific qualification to teach PE in their practice. On the other hand, secondary PE teachers not only need to obtain such a higher-education qualification for teaching PE, but research shows that secondary school PE teachers are also more likely to have a sporting background and thereby feel much more at teaching and facilitating learning with a physical activity component. MacPhail et al. (2005) reported that more than ninety per cent of secondary school principals requested a recognised PE qualification when employing a PE teacher, as non-qualified PE teachers undermined the quality of PE provision (MacPhail et al. 2005). Having obtained a qualification in teaching PE may have also exposed secondary school teachers to a variety of courses on inclusive PE, which has been highlighted as one of the predictors that positive influence attitudes (Anunah and Hodge 2005; Tarantino, Makopoulou, and Neville 2022; Vickerman and Coates 2009).

Finally, findings showed that there was considerable uncertainty (i.e. wide confidence interval) in our estimate of differences in attitudes between teachers with a formal SEND qualification and those without such a qualification. The observed (i.e. mean) difference was still positive, however. This provides some evidence of a link between having a SEND qualification and having more favourable attitudes towards the inclusion of children with SEND in PE. This finding complements the research of Boyle, Topping, and Jindal-Snape (2013), who reported that teachers with a special education qualification had more positive attitudes towards inclusive education than those with a general education qualification alone (Boyle, Topping, and Jindal-Snape 2013). Research by Hsien, Brown, and Bortoli (2009) showed that teachers with a qualification in special education perceived inclusion as feasible within the educational system (M. Hsien, Brown, and Bortoli 2009). Such findings are promising, but also stand to reason: teachers with formal qualifications in special education should in principle have a higher level of professional readiness for working with this population group and have a better understanding of their needs. Building on previous studies on general education, therefore, our findings lend further support to the view that teacher training may be crucial in enabling teachers to successfully apply inclusive pedagogies within mainstream settings.

Strength and limitations

This study has a number of strengths. The study provides comprehensive and novel empirical evidence regarding the factors associated with Irish PE teachers’ attitudes vis-à-vis inclusion, self-efficacy to deliver inclusive pedagogies, and perceptions about whether and how inclusion was supported within their school. The only other published study of Irish PE teachers’ attitudes towards inclusion was conducted in 2006 (Meegan and MacPhail 2006) and it investigated a more limited number of factors associated with inclusion attitudes. As such, this study provides both an update to an extension of inclusion research that is particular to and can directly inform the Irish educational context.

Despite the novelty and strengths of this study, some limitations need to be acknowledged. The cross-sectional nature of this study is an obvious limitation, within our findings represent only a snapshot (though an updated and extended snapshot at that) of the factors associated with Irish PE teachers attitudes towards inclusion. It is also worth noting that these cross-sectional data were collected during the first lockdown of the COVID-19 pandemic. During this period, schools were closed, and teaching was largely being delivered remotely. As such, given the new pedagogical challenges that teachers were experiencing in the context of social and physical distancing restrictions, the findings of study might also reflect some COVID-related attitudes, self-efficacy, and perceptions about the school context. For example, teachers might have had lower efficacy beliefs at different points of the COVID-19 pandemic as they quickly needed to transition into a new mode of delivery. Equally, teachers’ attitudes towards inclusion might have been more positive because, during COVID, they became more attituded to the challenges of remotely including children with differing needs and even from different backgrounds. The recruitment of participants for this study required participants to opt into the study on a voluntary basis and to complete an online questionnaire. As such, the findings of this study could be open to some biases: namely, response bias (i.e. it is possible that only teachers who are knowledgeable about and interested in inclusion responded) and social desirability bias (i.e. teachers potentially responded in a manner that reflects what they think they ought to say about inclusion, rather than what they actually perceive or believe). It should be noted here again that we used well-validated questionnaires during the conduct of this study, which is a known methodological tactic to reduce these types of response biases. The importance of replication is also worth noting here – we believe that replication of the present findings in different (potentially different international) contexts surveying independents sample of PE teachers would further corroborate and improve the generalisability of our findings.

Conclusion

In summary, our results showed that: (i) having different types of SEND in class; (ii) being frequently exposed to people with SEND; and (iii) obtaining a qualification in teaching PE may positively influence the inclusion of children with SEND in PE classes. Despite the limitations of our study design and sampling strategy, which requires that we proceed with caution when generalising these findings across schools in Ireland, there are a number of practical implications worthy of consideration and future research. The findings of this study highlight the importance of teachers’ exposure to routine and consistent working with children with SEND profiles, which influences not only the attitudes and efficacy beliefs of the people working directly with them, but it also influences the broader school context within which teachers work. Our findings, also potentially highlight the importance of delivering inclusion-related professional learning opportunities to primary school teachers. Such professional learning opportunities could include collaboration among primary and secondary school teachers on inclusion challenges and approaches, as well as practical workshops to address such challenges and apply new approaches directly with (and with feedback and input from) children with SEND profiles.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study was funded by the Irish Research Council under grant number GOIPG/2018/2996.

Notes on contributors

Giampiero Tarantino

Giampiero Tarantino was a PhD candidate at the School of Public Health, Physiotherapy and Sports Science, University College Ireland. He completed a BSc and a MSc in Sports Science at the University of Pisa, Italy (2008–2016). His PhD research primarily focuses on inclusion of children with disabilities and special educational needs in physical activity and physical education. He has also published various systematic reviews and meta-analyses in other scientific fields in peer reviewed journals.

Ross D. Neville

Ross D. Neville, is an assistant professor in the School of Public Health, Physiotherapy and Sports Science at University College Dublin in the Republic of Ireland. His PhD was in the field of recreation and leisure studies, which he received from Technology University Dublin. His current research focuses on teacher- and child-level interventions that promote physical activity and facilitate physical education for children with special educational needs and disabilities.