1 Introduction

Self-efficacy corresponds to the beliefs that the individual has about his/her abilities, with the purpose of organizing and performing the actions necessary for the production of certain achievements (^{Bandura, 1997}). In this sense, motivation, affective state and human actions depend more on what one believes to be real than what it really is, so that one’s perceptions alter the judgment and use of one’s knowledge and skills (^{Bandura, 1997}; ^{Pajares & Olaz, 2008}).

This means that the individual’s expectations about his/her personal efficacy determine whether a given behavior will be initiated, the level of effort that will be spent, how long he/she will face obstacles and aversive experiences, the amount of stress he/she will experience when coping with environmental demands, and even his/her vulnerability to depression (^{Bandura, 1978}, ²⁰⁰⁸). Strong self-efficacy beliefs allow a serene state to be maintained in the face of difficult activities, while weakened beliefs tend to lead to the perception that the task offers an unrealistic level of complexity, generating anxiety and stress (^{Pajares & Olaz, 2008}).

Self-efficacy acts on cognitive, motivational, affective and selective processes (^{Andreu, Pallarés & Juárez, 2004}; ^{Bandura, 1997}; ^{Navarro, 2007}). These will be presented separately for didactic purposes only, as they regulate human functioning in an integrated manner. In the cognitive aspect, self-efficacy will influence the positivity or negativity of the thoughts that precede actions, leading to the establishment of goals and commitment to them, in order to favor or inhibit a given conduct. Thus, high beliefs are related to ambitious goals and great dedication to achieve them (^{Andreu, Pallarés, & Juárez, 2004}; ^{Bandura, 1997}; ^{Navarro, 2007}). Similarly, goals considered unattainable are not established (^{Martínez, 2004}); therefore, the role of cognition in the acquisition and regulation of behavior is emphasized, highlighting that fleeting experiences can cause lasting effects, as they are codified and retained in memory (^{Bandura, 1978}). By suggesting the disposition of thought, efficacy beliefs affect an individual’s understanding of his/her existential circumstances, since, with high self-efficacy, the focus of his/her attention is on current life opportunities (^{Bandura, 2008}, ²⁰⁰⁹). According to the author:

Through ingenuity and perseverance they figure out ways of exercising some measure of control even in environments of limited opportunities and many constraints. Those beset with self-doubts dwell on impediments which they view as obstacles over which they can exert little control. They easily convince themselves of the futility of effort so they achieve limited success even in environments that provide many opportunities. (^{Bandura, 2009}, p. 181).

Considering that motivation tends to result from cognition, consequently it will be affected by the anticipation of the results of conduct, acting on the will and the effort made to achieve success (^{Andreu, Pallarés, & Juárez, 2004}; ^{Bandura, 1997}; ^{Navarro, 2007}). Thus, levels of self-efficacy can increase or compromise motivation, as people who have doubts about themselves find it difficult to motivate themselves (^{Schwarzer & Schmitz, 2004}), as, when there is no belief that their actions can promote the desired results and avoid the unwanted, they lack the incentive and perseverance to act and face obstacles, so that any motivational factor carries at its core the sense that the conduct can promote the intended change/outcome (^{Bandura, 2008}, ²⁰⁰⁹). ^{Bandura (1997)} states that the best skills can be invalidated by personal doubts about one’s own ability to perform and, therefore, different people with similar skills or the same person in different circumstances may exhibit extremely varied performances. Similarly, highly talented individuals may have accomplishments that fall short of their abilities when their self-efficacy beliefs are deteriorated.

To the extent that self-efficacy has influence on human behavior, teaching practice could not be exempt from its effects, since self-efficacy plays a mediating role in this practice, because it acts on the cognitive, motivational and behavioral aspects of educators, influencing their choices, persistence in the face of obstacles, motivation for certain behaviors, and feelings about the choices made. “For example, teachers with high self-efficacy for using movies as a lesson strategy will persist more in discussing complex movies than teachers with low selfefficacy” (^{Azzi, Polydoro, & Bzuneck, 2006}, p. 151). In addition to being related to the didactic options and the way to conduct the educational process, teacher self-efficacy is associated with the students’ academic success (^{Bandura & Pajares, 2007}), their aspirations and conceptions about themselves (^{Bandura, 1997}). This is one of the most impactful teaching characteristics on student achievement (^{Navarro, 2007}).

^{Tschannen-Moran and Woolfolk Hoy (2001}) understand teacher self-efficacy as the teacher’s assessment of their own ability to teach and motivate all students, regardless of the difficulties they may have. Their sources are the same as those of general self-efficacy: domain experiences, vicarious experiences, verbal or social persuasion, and physiological and affective states. The importance of their strengthening stems from the implications of teachers’ beliefs on their teaching skills on the conduct of the teaching-learning process.

The difficulties present in the teaching-learning process are understood by teachers with robust beliefs, in their teaching skills, as surmountable obstacles, starting from ingenuity and additional effort, instead of placing the responsibility for student learning failure (^{Bandura, 1997}). Among the many challenges that educators face, we emphasize school inclusion, largely supported by our legislation (Law no. 9,394 of December 20, 1996); however, far from being effective with quality and guarantee of learning and development for all (^{Carvalho, 2016}). Although inclusion has been a commitment for more than two decades, it is still viewed with some unfamiliarity as student diversity and the break with school standardization are processually being integrated into the social imaginary. As a result, they challenge teachers to deconstruct conceptions and practices that are not always supported by favorable conditions in terms of physical structure, student numbers, available human and material resources, continuing education, and so forth.

Considering the difficulties of our educational system in achieving the inclusion of the public of Special Education (^{Carvalho, 2016}), as well as the implications of this process for teachers (^{Vitaliano & Valente, 2010}), we emphasize the relevance of the instruments capable of measuring teacher self-efficacy for inclusive education, so that they can guide the necessary interventions, especially the formative ones, because, according to ^{Bzuneck (1996}), the proper understanding of the behavioral and motivational aspects of teachers requires the identification of teachers’ self-efficacy beliefs, especially when it is intended to promote changes in their pedagogical practices. In this sense, the objective of this research is to analyze the psychometric properties of the Brazilian version of the Teacher Efficacy for Inclusive Practices (TEIP) Scale (^{Sharma, Loreman, & Forlin, 2012}).

2 Method

Initially, we contacted the first author of TEIP via e-mail to request permission to translate it and adapt it to the Brazilian reality. The authorization was immediate and accompanied by a request for a feedback at the end of the research. The translation and crosscultural adaptation⁴ of the instrument followed the steps suggested by ^{Guillemin, Bombardier and Beaton (1993}).

The validation process involved 40 schools, divided into state and municipal schools and private schools, of Early Childhood Education (final grades, preschool), Elementary School I and II, from the cities of Corumbá and Ladário, in the state of Mato Grosso do Sul, Brazil.

The Portuguese version, called Escala de Eficácia Docente para Práticas Inclusivas (EEDPI) - Teacher Efficacy for Inclusive Practices Scale, was answered by 346 teachers. This is a convenience sample, not a random one. We discarded 38 completed scales because they were incomplete, resulting in a total of 308 instruments analyzed. The distribution of the variables was examined with the Kolmogorov-Smirnov test, which revealed the non-normality of the distributions (p ˂ 0.001).

3 Results

The distribution of the results in the scale items was adequate for the factorial analysis, presenting KMO of 0.90 and Bartlett’s sphericity test of X² = 3363.13 (gl = 153, p ˂ 0.001), considered highly significant.

The option was to analyze the main components, differing the factors to be isolated based on eigenvalue equal to or greater than the unit. Such analysis concluded by three factors, all with values above 1, corroborating the findings of ^{Sharma, Loreman e Forlin (2012}), in the construction of its original version, but with distinct organization. The relevance of these three factors can be found in the Scree Plot, as shown in Figure 1.

Source: Elaborated by the authors from the SPSS software.

Figure 1 Declivity with main components of the Teacher Effi cacy for Inclusive Practices Scale. 

The three factors obtained in EFA explained 62.27% of the total variance. By analyzing the content of the items grouped in the first factor, we found that they all related to the teacher’s behaviors in terms of classroom conduct (avoiding inappropriate behavior, assessing student comprehension, proposing appropriate challenges, etc.), explaining 46.96% of the variance. The second factor, which grouped the items related to students with specific demands (aggressive, confused, disabled students), explained 8.82% of the variance. Factor 3, whose content referred to the relationship with families, explained 6.50% of the total variance.

Although the number of factors obtained is identical to that of the original scale, the typology of the items in each of them and the significance we assigned to the three dimensions differ substantially. At the same time, one of the factors points to two items (Factor 3), yet there is item 18 (providing examples and alternative explanations when students are confused), which extracted the same value in two factors, similarly to item 9 (ability to involve parents in the school activities of their disabled children), with double saturation.

Despite the relevance of the third factor being indicated by the declivity graph (Figure 1), ^{Tabachnick and Fidell (2013}) recommend caution in interpreting factors defined by only two items because of their questionable reliability. In fact, it would be meaningless to maintain a dimension with so few items. At the same time, the interaction between school and family is of great importance to the teaching and learning process, especially in school inclusion, which would not justify the exclusion of these items. Thus, we undertook a new analysis, fixing the extraction of two factors.

From the analysis of the factor weights with the delimitation of the extraction of two factors, we obtained the explanation of 55.78% of the variance. The saturation of the items ranged from 0.45 to 0.82, while the commonality of items 1, 3 and 4 was less than 50% (0.36, 0.33 and 0.35), indicating little common variance in these variables, but within the acceptable level (^{Tabachnick & Fidell, 2013}).

The restructuring of the scale, based on the fixation of two factors, resolved the double saturation of item 9. However, item 18 remained with high saturation in factors 1 and 2 and another item was born loaded in both: number 14. Item 18 concerns offering other explanations or examples when students are confused, so we find it appropriate to exclude them. Item 14, which concerns joint work among students, was considered indispensable, both for the importance of the other in learning and for the existence of positive social interaction as one of the indicators of successful school inclusion.

Next, the content and the relationship between the items of each factor were analyzed. Therefore, there is the first factor with high factor weights in the statements regarding the management of behaviors and teaching strategies, being called “Classroom management”. The second factor demonstrates high factor loadings in the ability to work together with other professionals and plan activities, called “Planning and collaboration”. At that time, we noticed the inadequacy of item 17, which referred to ways of dealing with physically aggressive students. Moreover, the creators of the original version of the instrument certainly linked the possibility of aggression to the particularities of certain audiences belonging to the Special Education. However, Brazilian education is marked by the unfortunate situations in which teachers are victims of verbal and physical aggression by their students (^{Soares & Machado, 2014}), which makes the total agreement of educators with the statement “I feel confident to deal with physically aggressive students”. Given the little coherence of this statement with the remaining items of the dimension, we decided to eliminate it.

The appropriateness of item 15 was also questioned, as it is related to the use of diversified assessment strategies, fitting in the dimension of “Classroom management”. However, a diversified assessment goes beyond the application of tests and direct observation of behavior, and may involve families and other professionals, especially regarding the public student of Special Education. In addition to requiring planning, reflection and decision making, these are processes that are not restricted to the classroom space, justifying their approach to the other items of Factor 2, related to teaching activities performed outside the classroom (planning, relationship with parents, collaborative work with other professionals, knowledge of inclusive legislation).

In the 16-item version of the EFA, we maintained the option to extract the main components, with two-factor fixation and Varimax rotation. Based on this analysis, 56.95% of the total variance is explained. Table 2 systematizes the results obtained.

Factor 1 represents 47.07% of the total variance and carries eight items linked to “Classroom management”. Factor 2 explains 9.88% and includes items associated with “Planning and collaboration”. Cronbach’s alpha of the 16-item scale was 0.92, with 0.89 for Factor 1 and 0.88 for Factor 2, showing high internal consistency. If any of the items in this resolution were excluded, Cronbach’s Alpha would remain at 0.92. Thus, it is proposed to use this version of the Teacher Efficacy for Inclusive Practices Scale with 16 items, which shows appropriate psychometric property.

Next, we descriptively analyze the frequencies of the data obtained in completing the two scales.

4 Final considerations

Teacher self-efficacy plays an important role in regulating teacher behavior when influencing student dedication, performance, and encouragement while conducting the teaching and learning process. Because it is a specific concept, a teacher may feel able to perform one task and inefficient for another; therefore, believing in his/her own ability to teach Mathematics does not mean that his/her beliefs related to arts teaching are positive. Transposing the argument into the context of inclusive education, a teacher with high selfefficacy for pedagogical work may generally feel unable to teach a student belonging to the Special Education audience. Therefore, it is important to measure teacher self-efficacy for inclusive practices and develop timely actions in view of the finding of deficient beliefs. From this perspective, it is proposed to use the Teacher Efficacy for Inclusive Practices Scale, a version translated and adapted to the Brazilian reality, whose psychometric qualities are adequate. It is suggested that this scale be applied in different contexts and populations in order to contribute to its validation.

In fact, it is important to highlight the contributions of this research in relation to the advance of the state of the art in epistemological terms, since it broadens the discussions of the Special Education area in Brazil by highlighting the influences of teachers’ self-efficacy beliefs in the effectiveness of school inclusion and introducing a scale that consists of a new source of knowledge, with validity and reliability, which represents the first instrument of teacher self-efficacy for inclusive practices rigorously validated at the national level, which gives it the character of unprecedentedness.