Validation of a Perception Scale for Knowledge Acquired in Emotional Education During Initial Teacher Training

Abstract

Emotional education is essential in teacher training processes, but historically it has been neglected in the training system. The purpose of this study was to design and psychometrically validate the EEITT Scale, an instrument that assesses the perception of knowledge acquired about emotional education in student teachers and practicing teachers. A quantitative, descriptive and comparative approach was used with 548 participants, applying confirmatory factor analysis and invariance analysis to evaluate the model. Confirmatory factor analysis showed that the model fit well and had high reliability scores, which backs up the validity and internal consistency of the EEITT for measuring perceptions about emotional education training. Statistically significant differences were identified between groups, with students reporting greater social-emotional learning in the four factors evaluated. Likewise, a negative and significant effect of age on the perception of emotional education training was observed. These findings highlight the importance of emotional education in teacher training and point to the need for educational policies that integrate holistic and continuous approaches throughout the teaching career. Despite its limitations, this instrument provides relevant tools for future research and for guiding the design and improvement of teacher training practices.

1. Introduction

Within the school system, emotional education understood as the educational process that seeks the development of the emotional dimension under the recognition, management and emotional regulation, as a necessary aspect for the development of identity and personality [1], is insufficiently addressed in the regular curriculum, which affects children and young people’s emotions [2], even though emotions are considered fundamental in the teaching and learning processes [3,4]. The importance lies in the fact that the school environment is the main socialization space after the family [5] where it is possible to educate the processes of emotional regulation [6]. For teachers, competencies regarding emotional regulation are directly related to learning outcomes, teacher effectiveness in the classroom, their well-being and health, and emotions and motivations on the part of students [7]. However, teachers have chosen to neglect or suppress their emotions because these could threaten the goals of the school system with respect to teachers’ pedagogical practices [8].

The increasing levels of mental illness associated with the teaching profession have resulted in more teachers leaving the profession [9], being identified as the professional group most likely to experience burnout [10]. This is because the teaching profession is a demanding activity that involves constant emotional changes. The above is associated with the high prevalence of stress, sustained physical and mental health problems, translating also in the deterioration of the performance and health of children and youth [11,12]. However, the process of considering the emotional dimension in educational institutions at different levels has been slow, in response to the devaluation of the emotional as opposed to the rational [6,13,14]. This is even more relevant during initial teacher training, characterized as one of the stages where academic and professional changes linked to the manifestation of stress are developed [15,16], so that emotional regulation can become a modulating factor.

Emotional regulation is understood as the processes by which people monitor, evaluate, and express emotions in order to adapt to the objectives and demands of the context in which they find themselves [4,17]. This is in line with Gross’s process model, which states that emotional regulation includes the selection and modification of situations, and these involve interventions in the environment to respond according to cognitive reevaluation and the expression of emotions [17]. In this sense, emotional regulation is a key skill for those who are learning to be teachers [18], because awareness and knowledge regarding emotions, the recognition of one’s own emotions and how they impact on others, can come to exert influences on the well-being of the teacher and students [3,4,19,20]. Likewise, they affect learning engagement, academic performance, interpersonal relationships, course satisfaction, psychosocial development, and successful classroom and/or professional practice experiences [20]. Thus, the ability to regulate emotions may come to function as a protective factor against the effects of stress on the health of trainee teachers [16]. In this sense, emotions are at the center of teaching, i.e., it is not just a matter of knowing pedagogy and curriculum, since good teachers are emotional beings who connect with their students [21]. Thus, teaching practice and good teaching is accompanied by interest, pleasure, challenge, creativity and other emotional factors [22].

One study showed the emotional changes that trainee teachers have during their professional practice, revealing that they experienced emotions such as longing and anxiety at the beginning, shock and shame after teaching, and guilt and regret at the end of the practice or internship process [22,23]. Also, it was found that supervising teachers of trainees tend to neglect the role of emotions in support towards learning, affecting their professional identity development [24,25]. Though the literature indicates that people with greater capacity for emotional regulation could come to provide support on how to manage emotions to people with less developed emotional competencies [26]. However, there are few studies conducted in contexts of professional experience, understood as the process where trainee teachers experience a real classroom environment, with emotional aspects [27]. According to this author, emotions are fundamental in teacher training and development, highlighting the connection between emotion and cognition, driving through this the need to find strategies to express and manage emotional experiences, mainly in transition processes, as is the case of the change from being a student in training to being a teacher.

In this regard, there are several instruments that measure emotional aspects in individuals. For example, the Cognitive Emotion Regulation Questionnaire (CERQ) by Garnefski et al. [28], with emphasis on the study of risk factors and individual protection associated with emotional problems. The Emotional Regulation Others and Self (EROS) questionnaire by Niven et al. [29] which is responsible for measuring how affective regulation can influence well-being, performance and relationships. The Emotion Regulation Questionnaire for Children and Adjustments (ERQ-CA) by Gullone and Tafe [30], which aims to measure emotional regulation during middle childhood and adolescence. The Wong–Law Emotional Intelligence Scale (WLEIS) by Wong and Law [31] to measure emotional intelligence in studies linked to management and leadership. The Trait Emotional Intelligence Questionnaire by Cooper and Petrides [32], which is responsible for operationalizing emotional intelligence according to the subjective nature and information of emotional experiences as a personality trait. The TMMS of Salovey et al. [33], which is responsible for assessing the meta-cognition of emotional states. Finally, the DERS scale by Gratz and Roemer [34] that seeks to identify difficulties in emotional regulation, among other instruments and/or versions that contribute to the consideration of emotional aspects.

Furthermore, most of these instruments have been contextualized to the educational field, being applied in studies with teachers and students at different levels. Likewise, recently published systematic reviews have identified various approaches and tools for assessing emotional competencies [19,34], highlighting the advances and challenges of constructing contextualized and specific instruments for a given population and stages during professional teacher development. The synthesis of the described instruments allows us to realize that, although there are several instruments to evaluate emotional regulation, emotional well-being or emotional competencies, there is no specific instrument to measure the perceptions regarding emotional education acquired in teacher training. This would make it possible to identify strengths and needs to guide the improvement of educational policies and teacher training programs.

Given the above, it is possible to argue that emotional education in future teachers is crucial to improve the practices of socialization of emotions [35]. This research is conducted in response to the lack of instruments that measure perceived emotional education, and considering the lack of knowledge about how emotions are part of the initial teacher education. This would enable us to identify strengths and needs to guide the improvement of educational policies and teacher training programs, considering the different stages of a teacher’s career, which is a fundamental aspect for evaluating the evolution of perceptions of emotional training over time. In addition, it allows us to verify the sensitivity of the instrument in detecting gaps and providing comparative data between groups. Therefore, the objective of the study is to validate a measurement model that captures the perception of teachers and teacher trainees about their training in emotional education. Thus, the study proposes an approach to the object of study that considers teachers in training and practicing teachers in order to know the differences in socioemotional training and the perception of this from both educational actors.

Along with this objective, and considering that the study included both education students and practicing teachers, the following hypotheses were proposed:

• H1: There are statistically significant differences between education students and teachers in relation to the received socio-emotional training. This is based on the idea that emotional education is not distributed equally among groups, as in the Chilean education system emotional skills education has only recently been incorporated into teacher training curricula. Consequently, teachers who were trained decades ago may not have received this preparation, which implies variation in perceptions of the obtained training [2,4,9].

• H2: Education students report a greater perception of socio-emotional training than practicing teachers. This reflects the current context in which emotional education has been normalized as a component of professional preparation [16,21].

• H3: There is a statistically significant influence of age on the perception of the participants. Older teachers are less likely to have been exposed to emotional education during their initial preparation, whereas younger teachers have had experience accessing updated programs where emotions are part of the training content [8,22].

2. Materials and Methods

In accordance with the methodological guidelines set forth by McMillan and Schumacher [36], and León and Montero [37], the present study is framed as quantitative research, with a descriptive, cross-sectional and comparative design.

2.1. Participants

A total of 548 people participated; 83.4% were teachers in the Chilean school system and 15.7% were teacher trainees. In this regard, both in-service teachers and teachers in training were included to analyze possible differences at different stages of the teaching career. This allows us to observe how training in emotional competencies evolves from the initial stage to professional practice and whether there are gaps in preparation that could be addressed in continuing professional development. It should be noted that the proportional difference between practicing teachers and teachers in training is based on logistical and access factors. Despite this, the sample imbalance does not invalidate the analyses carried out, but it is recognized as a relevant consideration in the interpretation of the research findings. The age of the participants ranged from 18 to 76 years (M = 35.23; SD = 10.89) and with respect to gender 83.4% were female, 15.7% male, 0.5% non-binary and 0.4% preferred not to respond. The 99.5% are Chilean and 0.5% declared to be of another nationality. Regarding ethnic ancestry, 81.6% do not belong to any ethnic group, 14.1% self-identify as Mapuche, 2.4% say they belong to another ethnic group and 2% do not know if they have ethnic ancestry.

2.2. Instrument

The perception scale for knowledge acquired in emotional education during initial teacher training (EEITT), an instrument designed for this study, was applied. This is a measure that seeks to know the perception of teachers and student teachers regarding the learning of the referred knowledge during their professional training process. It is a 6-point Likert scale whose response options range from 1 = Strongly disagree to 6 = Strongly agree with the items raised. The measurement model presents the following structure (

Table 1. Structure of the EEITT questionnaire.

Factors	Range of Items	Total Items
F1 Mood (M)	X1–X10	10
F2 Emotional Regulation (ER)	X11–X18	8
F3 Psychosocial Well-being (PSW)	X19–X28	10
F4 Identity (ID)	X29–X49	21
Total		49

Source: Prepared by authors.

).

As shown in Table 1, EEITT consists of four factors. Mood (M) is considered to be the general disposition and state of how the subject feels, which can influence how we perceive experiences and how we react to them; Emotional Regulation (ER) refers to the intrinsic and extrinsic processes that contribute to the monitoring and evaluation of emotional experiences; Psychosocial Well-being (PSW) is the emotional balance that individuals need to maintain healthy personal and social functioning; and Identity (ID) focuses on contextual and behavioral aspects, considering behavioral changes in physical and social spaces (classrooms, school climate, natural spaces, among others) that allow for a reduction in states of dysregulation and promote healthy emotional states. So, these factors are interrelated and allow regulatory processes to be linked to the use and perception of the context in which the subjects operate.

In

Table 2. Sample EEITT scale items.

Item Content (Correlative Order According to Table 1)
I learned that maintaining a positive state of mind favors learning processes.
I was taught about the influences of emotional regulation on the teaching-learning process.
I was taught that self-concept and self-image is a component that influences students’ learning development.
I learned that I must create environments of dialogue that contribute to the reduction of discriminatory attitudes in the classroom.

Source: Prepared by authors. Note: the complete scale is available in the dataset by Fuentes-Vilugrón et al. [38].

we present some of the indicators that make up the instrument.

2.3. Analysis Plan

Content validity was evaluated by ten experts in emotional education with experience in measurement model development through two complementary procedures. First, Aiken’s V [39] (Aiken, 1980) was calculated to assess relevance, clarity, and wording of each item on a 1–5 scale, establishing V ≥ 0.70 as the retention criterion. Second, a categorical judgment of factorial belonging was implemented where experts assigned each item to the most appropriate factor, establishing ≥ 70% agreement as the acceptable criterion [40] (Lynn, 1986). Items meeting both criteria were retained for subsequent confirmatory factor analysis and renumbered consecutively starting from X1 while maintaining their original factorial distribution.

The normality in the distribution of the data in all indicators is evaluated using the Kolmogórov–Smirnov test, in order to choose the subsequent methods of analysis (parametric or robust methods). It is confirmed that the variables do not follow a normal distribution of the data if the statistical significance shows values less than 0.05. According to Maydeu-Olivares [41], a confirmatory factor analysis (CFA) is performed by estimating the parameters by maximum likelihood with mean and variance adjusted (MLMV). This analysis looks for evidence in favor of construct validity.

To establish that evidence in favor of construct validity is found, for the CFA, goodness-of-fit indices such as the Chi-square statistic and the ratio between it and its degrees of freedom, which should be less than 3:1 to indicate a good fit, are calculated [42,43]. Also, another three fit indexes were calculated: 1. The root mean square error of approximation (RMSEA), which determines an acceptable fit of the model to data, when their values are less than 0.08 and excellent if the value is less than 0.05 [44]. 2. The comparative goodness of fit index (CFI) whose values greater than 0.90 show an acceptable fit and those greater than 0.95 show an optimal fit [45]. Finally, the Tucker–Lewis index (TLI) with acceptable values greater than 0.90 and ideal values being those greater than 0.95 [44].

The convergent validity of the instrument is evaluated according to Hair et al., which contemplates for each dimension the following elements: (1) the standardized loadings whose values should be greater than 0.50 and their level of statistical significance with a p-value less than 0.05; (2) the composite reliability where the values should be greater than 0.70; and (3) the average variance extracted (AVE), requiring values greater than 0.50. Subsequently, discriminant validity is analyzed according to the criteria of Fornell and Larcker [46], who state that a construct has discriminant validity if its AVE is greater than the squared correlations between that construct and the others present in the model. This indicates that the construct is distinctively different from the others.

To estimate reliability, composite reliability is used through McDonald’s [47] omega coefficient, where values between 0.70 and 0.90 are acceptable and greater than 0.90 are considered excellent [48]. If discriminant validity is not achieved, or if achieved and it is found that the correlations between factors are greater than 0.50, a model in which the latent variables are measured by a second-order factor will be proposed [49], considering that, in the second-order CFA the assumptions about the measurement model remain unchanged with respect to the first-level CFA [50].

In order to establish the instruments are measuring adequately the perceptions of teachers and student teachers, so the scores are comparable, first, the structure of the scale is studied through CFAs for each group, and then a measurement invariance analysis of the category containing them is performed, using the configural, metric and scalar models. The configural model is achieved if the goodness-of-fit indicators are met according to the criteria already referred to for a CFA. Both metric and scalar invariance are considered achieved if one or more of the required variance criteria are met [51,52]. In this sense, a variation in RMSEA less than 0.015 and/or CFI less than 0.01 is expected, considering that for groups smaller than 300 the variation in CFI should be less than 0.005. The above gives way to review the existence of statistically significant differences in perception between teachers and student teachers, by means of the Mann–Whitney U hypothesis test. This is performed using the Mean Ranks because of their robustness against outliers and their descriptive capacity of the central dispersion. These allow us to reliably compare the variability between the two groups, which is essential to identify statistically significant differences in the distribution of the data [37].

In order to investigate the influence of age on the perception of the participants, an analysis is estimated by means of structural equation modeling (SEM), where the fulfillment of the criteria of a CFA, the relationship between variables and their statistical significance are sought [53]. Finally, in order to propose a reference for interpreting the results in future applications of the instrument, cut-off points were calculated for each factor using K-means cluster analysis to establish three levels of perception (low, medium, and high). The cut-off points were defined using the minimum and maximum values of the medium-level cluster.

Normality tests, hypothesis testing, descriptive statistics and K-Means cluster analysis were performed with SPSS v.27 [54]. CFA, measurement invariance analysis and SEM analysis were performed with MPlus v.8 [55]. AVE, reliability, convergent and discriminant validity calculations were performed with Excel v.16 [56]. For the measurement of effect size and statistical power, G*Power V.3 software was used [57].

2.4. Ethical Considerations

This research was conducted in compliance with the declarations of Helsinki [58], of Singapore [59] and the current regulations in Chile regarding scientific research [60]. In this sense, the research project underlying the present study was reviewed and authorized by removed for peer review.

The participants agreed to an informed consent and confidentiality notice, being informed about the characteristics of the research, the instrument and the approximate time to answer it. The above emphasizes the voluntary nature of participation and the anonymity of the data, with respect to the handling of information and the publication of results. All participants answered the questionnaire in web format, from computers or smartphones in approximately 15 min.

3. Results

The retained items showed Aiken’s V values between 0.72 and 0.87, exceeding the established criterion (V ≥ 0.70). The percentage of agreement in factorial belonging ranged from 83.3% to 100%, meeting the minimum criterion of 70%. Of the total 49 initial items distributed across four factors, 23 items met both content validity criteria. Factor F1 ‘Mood’ retained 6 items from the original 10, F2 ‘Emotional Regulation’ preserved 5 items from the initial 8, F3 ‘Psychosocial Well-being’ maintained 6 items from the proposed 10, and F4 ‘Identity’ kept 6 items from the original 21. The retained items were renumbered consecutively from X1 to X23 for subsequent analysis while maintaining their original factorial distribution, as shown in Figure 1.

The Kolmogorov–Smirnov test is statistically significant (p < 0.001) for the 23 items selected, which implies that the data do not follow a normal distribution. The CFA evidenced a good fit of the proposed model to the data (X2 = 490.304; DF = 224; p < 0.001; RMSEA = 0.045; CFI = 0.943; TLI = 0.935). Figure 1 shows the model with first order factors.

It is possible to sustain that the model presents convergent validity given the high saturations of the variables towards their respective factor (Figure 1), by the values observed in the AVE and by the composite reliability in each dimension (

Table 3. AVE, saturations and reliability.

Factors	Loadings Min–Max	Mean Variance Extracted	Composite Reliability	McDonald’s Omega
F1 (M)	0.654–0.800	0.538	0.874	0.9
F2 (ER)	0.771–0.875	0.692	0.918	0.9
F3 (PSW)	0.753–0.920	0.752	0.948	0.9
F4 (RWE)	0.809–0.869	0.708	0.936	0.9

Source: Prepared by authors. Note: the saturations of all indicators are statistically significant at the 0.001 level. F1 = Mood (M); F2 = Emotional Regulation (ER); F3 = Psychosocial Well-being (PSW); F4 = Identity (ID).

). In addition, the instrument shows high reliability indices. Discriminant validity is evident, except between factors 3 and 4 (

Table 4. Discriminant validity.

Factors	AVE	F1	F2	F3	F4
F1 (M)	0.538		0.377	0.407	0.415
F2 (ER)	0.692	0.614		0.539	0.389
F3 (PSW)	0.752	0.638	0.734		0.524
F4 (RWE)	0.708	0.644	0.624	0.724

Source: prepared by the authors. Note: below the diagonal are the correlations between factors. Above the diagonal are the correlations between factors squared.

).

Considering the correlations between the latent variables in the first-order model (Figure 1), and specifically the absence of discriminant validity between factors F3 and F4 (Table 4), a model incorporating a second-order factor (Figure 2). The estimation of this model showed a good fit to the data (X2 = 495.597; DF = 226; p < 0.001; RMSEA = 0.045; CFI = 0.942; TLI = 0.935). This will also simplify the model used to subsequently measure the influence of age on the construct.

The CFA showed a good fit of the proposed model to the data for the Teachers group (X2 = 440.709; DF = 224; p < 0.001; RMSEA = 0.046; CFI = 0.940; TLI = 0.932) and fair fit for the Student Teachers group (X2 = 282.847; DF = 224; p < 0.05; RMSEA = 0.043; CFI = 0.890; TLI = 0.876). Although a CFI and TLI greater than 0.90 were not reached for this group, the invariance analysis was continued, which allowed us to reach scalar invariance in this category (

Table 5. Measurement invariance analysis.

Model	X2	DF	RMSEA	CFI	ΔRMSEA	ΔCFI
Configural	689,910	448	0.043	0.928	–	–
Metric	704.346	467	0.041	0.930	0.002	0.002
Scalar	737.208	486	0.042	0.926	0.005	0.004

Source: Prepared by authors. Note: In all three models the Chi-square statistic showed a p-value with statistical significance at the 0.001 level.

).

The model proposed to measure the direct effect of Age on the perception of training in emotional education (Figure 3) presents a good fit to the data (X2 = 536.626; DF = 248; p < 0.001; RMSEA = 0.044; CFI = 0.939; TLI = 0.933). The effect of Age on participants’ perception is of low magnitude, statistically significant (γ = −0.209; p < 0.001) and shows a good confidence interval (range = −0.209; L = −0.316; U = −0.101).

With respect to group comparisons as can be seen in

Table 6. Comparison of average ranges.

Factors	In-Service Teacher’s N = 412 Mean Rank	Pre-Service Teacher’s N = 136 Mean Rank	Z	U	p
F1 (M)	253.98	336.66	−5.303	19,562.500	<0.001
F2 (ER)	243.25	369.17	−8.052	15,140.500	<0.001
F3 (PSW)	258.15	324.04	−4.220	21,279.000	<0.001
F4 (RWE)	254.15	336.16	−5.294	19,630.000	<0.001

Source: Prepared by authors. F1 = Mood (M); F2 = Emotional Regulation (ER); F3 = Psychosocial Well-being (PSW); F4 = Identity (ID).

, teacher trainees perceive more socioemotional learning than teachers. The differences are statistically significant at the 0.001 level.

Considering that statistically significant differences in perception were found between In-service Teachers and Pre-service Teachers, cut-off points were calculated separately for each group to establish 3 perception levels (Low, Medium and High), which are shown in

Table 7. Cut-off points for first and second order factors scores.

	In-Service Teachers			Pre-Service Teachers
Scales	Low	Medium	High	Low	Medium	High
F1 (M)	<3.50	3.50–4.83	>4.83	<4.00	4.00–5.17	>5.17
F2 (ER)	<2.80	2.80–4.40	>4.40	<3.60	3.60–4.60	>4.60
F3 (PSW)	<2.83	2.83–4.50	>4.50	<3.67	3.67–4.83	>4.83
F4 (RWE)	<3.00	3.00–4.67	>4.67	<3.83	3.83–5.17	>5.17
G1 EEITT	<3.13	3.13–4.57	>4.57	<4.13	4.13–5.17	>5.17

Source: Prepared by authors.

.

In

Table 8. Participants at each level.

	In-Service Teachers			Pre-Service Teachers
Scales	% Low	% Medium	% High	% Low	% Medium	% High
F1 (M)	10.2	27.9	61.9	2.9	26.5	70.6
F2 (ER)	29.4	42.5	28.2	9.6	37.5	52.9
F3 (PSW)	15.3	28.6	56.1	6.6	29.4	64.0
F4 (RWE)	7.0	27.4	65.5	2.2	32.4	65.4
G1 EEITT	10.2	36.2	53.6	4.4	42.6	52.9

Source: Prepared by authors.

we show the distribution of participants in the low, medium and high levels of perception for each of the dimensions.

4. Discussion

The research fulfills the objective of validate an instrument that measures, from the perspective of students in training and practicing teachers, the perception of the knowledge acquired about emotional education. Based on the obtained results, it is possible to argue that the EEITT scale is a valid and reliable instrument that allows understanding and analyzing the phenomenon of emotional education during teacher training. In this way, the scale contributes to evidencing a series of challenges and needs that exist in the higher education system, whose academicist characteristics cause an omission of emotional aspects in the learning and teaching processes [6,61,62]. In this context, this research proved that the measurement model designed for the evaluation of the perception of emotional education training among teachers and teacher trainees has shown a robust effect of the collected empirical data. Moreover, the comprehensive description in perspective of both teachers and teacher trainees in initial teacher education provides validated psychometric data that address each of the dimensions jointly, which adds significant value to the field of education and emotional regulation.

The results obtained in this research are consistent with previous research that has highlighted the importance of emotional education in teacher education and its positive impact on the professional development of education professionals [63,64,65]. Likewise, the findings may vary according to different sociodemographic groups, as is the case of the differences observed between teachers and teacher trainees. These data contribute to the collection of proven information to provide teachers in training with technical and emotional competencies for personal well-being and work success [66]. This allows novice teachers to have the tools to fulfill their role as a guiding and mediating figure in the transmission of knowledge, in the socialization processes and in the emotional education of students [67,68]. The generation gap between practicing teachers and teachers in training can be interpreted from a life cycle perspective, where subjective aspects such as experiences or a lack of emotional education during professional training can lead to a lower perception of emotional competencies. Therefore, these findings suggest further investigation into how professional trajectories can influence teachers’ emotional perception and how this can be a fundamental aspect throughout their professional career.

The structural analysis reveals that age has a negative and statistically significant effect on the perception of training in emotional education (γ = 0.209; p <0.001), indicating that younger participants tend to have a more positive perception regarding their training in this area. This becomes relevant, considering that it provides information on how perceptions may vary with age, which could have implications for curriculum design and also for the implementation of teacher training programs.

Regarding the hypotheses posed, it should be noted that

• H1: There are statistically significant differences between education students and teachers in relation to the received socio-emotional training. It is confirmed, given that the statistical tests showed a p-value at the 0.001 level in all factors.

• H2: Education students report a greater perception of socio-emotional training than practicing teachers. This is confirmed, in that the mean ranks are higher in the students for each factor, which could be related to the greater focus on socio-emotional education in current university degrees.

• H3: There is a statistically significant influence of age on the perception of the participants. This is confirmed, as a statistically significant effect of low magnitude was found. The younger the age, the greater the perception, which is consistent with Hypothesis 2, suggesting that education students perceive greater socio-emotional training than practicing teachers. This may also extend to recently graduated teachers, as they could share a similar perception to education students.

In addition to the statistical comparisons, cut-off points of scores were calculated for the first and second-order factors, allowing for the classification of participants into low, medium and high levels of perception regarding their emotional education training. This analysis revealed interesting findings about the distribution of participants across these levels. Notably, 68.4% of participants reported a high perception of training in the ‘Mood’ factor, while 61.3% showed a low perception in ‘Psychosocial Well-being,’ and 70% expressed a low perception in the ‘Identity’ factor. These findings highlight the varying levels of perceived emotional education and underscore the need for targeted interventions in teacher training programs to address these gaps.

Despite the novelty of the findings, there were limitations associated with the sample size, so the results cannot be generalized to the Chilean population, nor to other populations or sociocultural contexts, which could limit their applicability to results at the international level. In particular, the difference in size between the trainee teachers and the practicing teachers who participated in the study emerges as one of the main limitations. Although it does not affect the internal validity of the analyses, it could limit the generalization of the findings, specifically in comparisons between the two groups. This suggests that future research could replicate the application of the scale to a larger number of trainee teachers.

5. Conclusions

Based on the obtained results, it became evident that students in initial teacher training perceive greater knowledge about emotions than practicing teachers. This responds to a generational change that reflects the historical omission of the emotional dimension in the training system. However, currently, albeit slowly, education has begun to incorporate holistic elements in professional training processes. This is manifested in the emergence of public policies and regulations related to emotional regulation in educational contexts, from preschool to adulthood. Thus, the results coincide with the growing attention that educational policies are giving to emotional education, which has been demonstrated through the incorporation of various standards of coexistence and well-being in teacher training, thereby advancing toward more inclusive learning and teaching processes. In the case of Chile, there are already laws such as Law 20.911 on school coexistence, as well as guidelines for the development of emotional well-being from the Ministry of Education, which reflect a concern for emotional development as an emerging but equally necessary change. Although there is a wide range of regulations, significant challenges persist. In particular, when emotional regulation is associated with teacher mental health, if high levels of emotional dysregulation are in place, in some cases, they lead to attrition within the first five years.

This study validated and verified the reliability of the instrument proposed to assess the perception of knowledge acquired on emotional education in initial teacher training. However, although the instrument is adequate for teachers, the same does not occur with teacher trainees, showing significant differences between the two groups. Even so, scalar invariance was evidenced, allowing the development of unbiased comparisons between the compared groups.

This research represents a first approach that highlights the relevance of considering sociodemographic characteristics in the perception of emotional education. In addition, the findings offer a theoretically and empirically grounded instrument, useful for future research and the design of educational policies that promote more effective training in emotional education at all educational levels. However, more studies are needed to delve deeper into the differences observed between teachers and teacher trainees. It is also key to explore variables such as age and gender in these perceptions, as well as to consider other elements, such as years of experience or educational context. Finally, understanding how these perceptions impact teaching practice and teachers’ emotional well-being could inform the development of effective interventions in the field of emotional education.