Perceptions of Compulsory Education Teachers About Cultural Diversity: A Study in the City of Messina

Ensuring inclusive education, which is one of the priorities of the 2030 Agenda for Sustainable Development, is not possible without teachers. Teachers’ perceptions influence the way they organise and develop their professional activity and the commitment they ultimately give to cultural diversity in educational practice. In this paper, we study the perceptions about cultural diversity which are held by teachers working in compulsory education with Roma students in the Italian city of Messina. The study involved 182 teachers from four educational centres for primary education and the first grade of secondary education. It is a quantitative descriptive study in which a questionnaire has been used as the instrument to collect information. The results obtained highlight, among other issues, the fact that the teachers who were involved in the study favour theoretical and practical perspectives which encourage cultural diversity. In addition, they reveal that there are significant differences in teachers’ perceptions according to the educational level they teach.


INTRODUCTION
Ensuring inclusive education and lifelong learning for all is one of the objectives set out in the 2030 Agenda for Sustainable Development (PNUD, 2015), which is supported by UNESCO and other organisations (2016).However, despite the efforts made by different nations, including Italy, progress towards quality education for all students is a goal which has yet to be reached (Echeita, 2017).
The inexorable arrival of immigrants and refugees to European territory -and growing social inequality, accentuated after the last economic crisis -have emphasized the need to promote inclusive and intercultural educational processes aimed at fighting inequality and discrimination (European Commission, 2015;Malusà & Tarozzi, 2017).
Responding to this challenge, the case of the Roma community is particularly complex, since its migrant and minority character adds cultural diversity to its status (Liégeois, 2004).
As pointed out by the European Commission (2015), this community has the worst living conditions, the lowest levels of learning and the highest unemployment rates in the European Union.In this sense, Roma make up the most rejected and marginalised community in Europe (Liégeois, 2004).To alleviate this situation, the European Commission (2011) has proposed, to EU governments, the need to move forward in a significant way, in order to include the Roma community in national strategic plans.
In 2012, Italy, "country of nomad camps", in which approximately 180,000 Roma live, entrusted to the Ufficio Nazionale Antidiscriminazioni Razziali the creation of the "Strategia Nazionale d'inclusione dei Rom, dei Sinti e dei Camminanti 2012/2020" (UNAR, 2012).Thus, the Italian Ministry of Education, in the "Indicazioni per il curriculo 2012", declared that diversity of cultures and languages is a reality in schools, and pointed to the need for intercultural education.
The intercultural model is today the model that allows all children to recognise each other's identity.School successfully embraces the universal challenge of openness to the world and the practice of equality in the recognition of differences.MIUR, 2012, p. 6 However, the results obtained in the last "Diversi da chi" Report (ISMU-MIUR, 2016) reveal that the presence of Roma students in schools has not only declined, but that the impact of the initiatives undertaken has been limited.As Curcic, Miskovic, Plaut, and Ceobanu (2005), recognise, throughout these years there was no reduction in the inequalities existing between Roma and non-Roma students.
This limited impact of inclusion policies in Italy not only calls into question the effectiveness of the actions undertaken, but also invites us to observe other factors that may be decisive in improving these results.One of the key elements is, undoubtedly, teachers.Promoting, generalising and embracing the inclusion of culturally diverse students requires commitment on a collective level.These factors are essential in raising awareness, modifying professional and institutional cultures and improving educational practices.But none of this is possible without the commitment of teachers (Dusi, 2017;Grant & Gibson, 2011).
The dissatisfaction shown regarding the progress experienced in many of the inclusive and intercultural initiatives undertaken (Malusà & Tarozzi, 2017;Sleeter, 2009) is closely related to the lowly status given to the teaching profession and to the training of teachers in this area (López & Hinojosa, 2016;Nieto & Mcdonough, 2011).A recent international review, carried out by the latter, shows that numerous studies highlight the negative perception that teachers of cultural diversity tend to have.They comment on their low expectations, their poor training and a lack of commitment to an educational practice which is sensitive to the cultural diversity of students.
In the case of Italy, Law 107/2015, promoted by the Ministero dell'Istruzione, dell'Università e della Ricerca (2016), points to the need to "Guarantee, in every school, the presence of a core of teachers capable of dealing with the processes of reception, literacy and cultural mediation, in areas with strong immigration" (p.44).This measure, however, is clearly restrictive, since it is limited to teachers who work in certain contexts.It does not apply to all teachers, as stipulated by international organisations and experts (Gay, 2013;Hinojosa & López, 2018;UNESCO., 2016).
Moving towards an education that perceives cultural diversity as mutually enriching requires paying more attention to teachers as agents of change, and to their training.Reflective and committed teachers are needed; they should be teachers who are willing to explore the training potential offered by cultural diversity, to assess their contribution to personal and collective development, and to reconsider their vision of cultural diversity in a responsible and ethical way (Hinojosa & López, 2018;Tarozzi, 2014).In this line, Shulman (2005) and Markic and Eilks (2013) have underlined the correspondence which exists between teachers' perceptions and the way in which they organise and develop their professional activity and, consequently, in the commitment they finally give to an intercultural and inclusive educational practice.
In this paper, being aware of the low impact of the initiatives undertaken in line with Italian educational policy in favour of the inclusion of Roma students (ISMU. & MIUR., 2016), the lack of attention given to teachers and their training, and the delicate situation existing in the country, following the new immigration policy proposed by Minister Salvini, led us to consider it necessary to reaffirm the commitment of educational research to interculturality and inclusion, as stipulated by UNESCO and other organisations (2016), as well as to emphasise teachers as a key factor in improvement.This is a novel initiative, since it allows us to understand the teachers' perspective on cultural diversity within the Italian context.This follows trials already underway in other countries (Carrasco & Coronel, 2017;Nelson & Guerra, 2013) but still little explored in Italy.It means identifying difficulties and challenges, offering data which helps teachers and educational leaders to reconsider the decisions taken and confronting the issue of the inclusion of Roma students in compulsory education schools with greater guarantees of success.

OBJECTIVES
The purpose of this study is to analyse the perceptions of compulsory education teachers working with Roma students in the city of Messina about cultural diversity.This is broken down into the following specific objectives: 1. Describing the perceptions about cultural diversity held by Messina's compulsory education teachers.
2. Determining if there are significant differences relating to the educational level they teach (primary and first grade of secondary).

METHOD
This is a quantitative investigation which responds to an expost-facto study of a descriptive type.It can be considered a comparative-causal design (Cohen, Manion, & Morrison, 2010), since it not only aims to describe real situations and experiences, but also to iden-tify cause-effect relationships between variables not experimentally manipulated.The study thus differentiates the independent variable of an attributive (not manipulable) characterin this case it is the educational level teachers teach (primary and first grade of secondary)from dependent (non-experimental) variables, which correspond to the set of items that make up the questionnaire used to gather information.The research hypotheses are: • H1.Compulsory education teachers, working with Roma students in the city of Messina, tend to perceive cultural diversity in a positive way.
• H2.There are no statistically significant differences -at the level of arithmetic meansbetween the cultural diversity perceptions of primary and secondary teachers in the total of the test, in dimensions and items.

Context and participants
The investigation was carried out in the city of Messina, because it is one of the thirteen Italian cities participating in the Progetto Nazionale per l'inclusione e l'integrazione dei bambini rom, sinti e caminanti -PON "Inclusione" 2014-2020 Ministero del lavoro e delle politiche social (2014).This involved 182 compulsory education teachers, from a total of 308, belonging to four educational centres which were selected through intentional sampling.This nonprobabilistic sampling procedure was chosen because these four schools are the only ones in the city of Messina that host Roma students.To determine a representative sample size, a degree of confidence of 0.95 (1-α) was used, with a sample error of approximately 4.5% and unknown proportions p=q=0.5.
33% of the teachers surveyed came from the IC La Pira Gentiluomo, 24.2% were from the IC Mazzini-Gallo, 18.7% from the IC Vittorini and 24.2% from IC Paradiso.13.2% were male while 86.8% were female.The percentage of educators teaching in primary education was 68.2%, while 31.8% were first grade secondary teachers.

Collection and analysis of information
For collecting information, the Inventory of Future Teachers' Beliefs About Cultural Diversity, validated in the Spanish context by López and Hinojosa (2016), was used.
This questionnaire is a Likert scale consisting of 71 items split into four dimensions: • Dimension 1: Approach to cultural diversity as a problem.This includes statements in which cultural diversity is negatively valued (items 6, 8, 12, 13, 14, 16, 18 and 21).
The scale used includes four response options: 1) Strongly disagree, 2) Disagree, 3) Agree and 4) Strongly agree.Before being applied, the questionnaire was translated, adapted and validated in the Italian context.It is not just a literal translation of the original instrument, but also an adaptation which includes cultural, idiomatic, linguistic and contextual aspects in order to preserve its validity and reliability (Hambleton & Zenisky, 2011).
Four experts participated in the content validation process.They were affiliated to various departments of theUniversità degli Studi Messina (COSPECS: Dipartimento di Scienze cogni-tive, Psicologiche, Pedagogiche e degli Studi culturali y DICAM: Dipartimento civiltà antiche e moderne).These educators valued the totality of the items according to their clarity, relevance and pertinence, responding to the following scale: 1=Nothing, 2=Little, 3=Enough, 4=Very.The judgements given by the experts showed that the arithmetic averages obtained in clarity, relevance and pertinence in each of the items were greater than 3 without exception.Therefore, they were higher than the Archimedean point 2.5 (1+2+3+4/4) set as a minimum acceptable.This confirms their high degree of excellence.Likewise, in order to evaluate the homogeneity versus heterogeneity of the scores given by the experts in each item, and to determine the degree of agreement granted to each item, the interquartile range was calculated (difference between Q3-Q1).The ambiguity coefficient values obtained were, in all cases, equal to or less than 1.So, according to Barbero, Vila, and Suarez (2010), the items should be as formulated.
To determine the concurrent criterion-related validity, the corrected item-total correlation was calculated.According to Martínez-Arias, Hernández-Lloreda, and Hernández-Lloreda ( 2006), the corrected item-global values higher than ±0.30(r>±0.30)must be considered.Thus, some items were obtained r<±0.30,although the scale, as a whole, presents most items with r>±0.30.This means that the items individually measure the same dimensions as the total scale and, therefore, they guarantee the unidimensionality of the scale.
In adapting the questionnaire to the Italian context, the type of reliability considered was understood as internal consistency, given that the instrument for the collection of information was administered only once.In this case, the reference coefficient for the calculation of reliability was Cronbach's alpha coefficient.The values obtained are (Table 1): As can be seen, Cronbach's alpha value reaches values greater than 0.6 in all cases.According to Zumbo, Gadermann, and Zeisser (2007), minimum values of 0.6 are usually considered sufficient in exploratory research such as this one.
Once the validity and reliability of the questionnaire was determined, it was applied to the teachers of the four selected schools.The information collected underwent two types of analysis: univariate analysis and bivariate analysis.The univariate analysis was based on measurement of the dependent variables with measures of central tendency (arithmetic mean) and dispersion measures (standard deviation).This analysis allows us to describe, in a general way, the perceptions of participants relating to cultural diversity (objective 1).The bivariate analysis, on the other hand, helped to cover the two objectives of the study (determining if there are significant differences depending on the educational level the teachers teach).For this, a comparison of differences of means was carried out.This allowed us to determine the degree of association of one variable (total test/dimension/item) over the average of another variable (educational level taught by teachers participating in the study).In this case, due to the characteristics of the data, the Mann-Whitney U test (p<.05) was applied for non-parametric data.
For the descriptive and inferential analysis of the data obtained, the statistical package SPSS v.24 was used.

RESULTS
The results obtained, in accordance with the objectives set out in the study, are presented below:

Results related to objective 1
To describe the perceptions of Messina's compulsory education teachers about cultural diversity, presented firstly are the values of arithmetic means and the standard deviations obtained in each of the dimensions and in the total scale.Subsequently, these parameters are shown in the items that reached the most relevant scores (Table 2).As can be seen, the highest average scores correspond to the dimensions referring to a positive approach to cultural diversity (M=3.25) and practices that favour the inclusion of culturally diverse students (M=2.82).These results can be translated into "Agree" because they show values close to 3. The dimension that presents cultural diversity as a prob-lem (M=2.32) and that other dimension referring to practices that do not favour inclusion (M=2.11)approximate, to a greater extent, the average 2 ("Disagree").
The low values of the standard deviations in all dimensions also show a strong convergence and homogeneity in the perceptions of the teachers participating in the study.
Regarding the total average score of the scale (M=2.63) and its standard deviation (SD=.20), it can be seen that the teachers participating in the study tend to favour cultural diversity.This is because the average is closer to the response level 3, with a clear homogeneity in the answers given, since the total score of the standard deviation is near to 0.
In Table 3 the average values and the standard deviations corresponding to the items that reached the most outstanding (high and low) scores are presented.The means with the highest scores correspond to items 4, 10, 23 and 27, and all of them belong to dimensions 2 and 3 (cultural diversity as a mutually enriching factor and practices favourable to the inclusion of culturally heterogeneous students).Conversely, the lowest average values are found in items 31, 33, 54 and 37, all of them being in dimension 4 (educational practices that do not favour the inclusion of culturally diverse students).
As for the standard deviation scores, we can see a greater convergence in items 1, 2, 54, 42, 23, 31, 27 and 10.The majority of these items belong to the positive dimensions of the scale, except items 31 and 54.These two items are located in dimension 4, with especially low average scores.
On the other hand, the highest scores in standard deviations and, therefore, the greatest dispersion of responses can be seen in items 59, 16, 8, 20, 30, 21, 13, 28 and 14.All of them are part of dimensions 1 and 4, which refer to a negative perception of cultural diversity and actions that do not contribute to the inclusion of cultural diversity in educational practice.In this regard, it is worth highlighting the dispersion of opinions existing in these aspects, which differs from the homogeneity and convergence found in those items that extol cultural diversity as a positive factor together with the need to promote inclusive practices.
Items 23, 27, 10, 31 and 54 deserve special attention.The first three have markedly low values in the standard deviation (s=.623, s=.634 and s=.640 respectively) and average values that are certainly high (M=3.59 in the case of items 23 and 10, and M=3.56 in item 27).Items 31 and 54, meanwhile, stand out owing to their low scores, both in standard deviation (s=.629 and s=.606) and in arithmetic mean (M=1.39 and M=1.41).
These data confirm the first of the hypotheses formulated in this study, as they show the tendency of compulsory education teachers, working with Roma students in the city of Messina, to perceive cultural diversity in a positive way, rather than see it as a problem.

Results related to objective 2
To address objective 2 (To find out if there are significant differences depending on the educational level educators teach), the educational level variable has been cross-checked with the dimensions of the questionnaire and the total test.Secondly, this same variable has been cross-checked with each of the items that make up the scale.Table 4 shows the results corresponding to the first crossing.As can be seen, there are no statistically significant differences in any of the dimensions, or in the total of the test regarding the level in which the educators teach.According to the average scores obtained in both collectives, it can be observed that both groups are more in accordance with the approach which presents cultural diversity as an enrichment factor (averages greater than 3) than with one which considers it problematic (M=2.32 y M=2.35).They are also more favourable to practices that advocate the inclusion of cultural diversity (M=2.83y M=2.83) than those that hinder it (M=2.09y M=2.14); although, as highlighted above, these differences are not significant.We can affirm, then, that hypothesis two of this study is fulfilled in the total test and at the level of dimensions.
Finally, we present the results related to the cross-checking between the educational level variable and each of the items of the questionnaire to determine whether or not there are significant differences in any particular aspect (Table 5).The cross-checking demonstrates that there are only five items showing statistically significant differences.This indicates that hypothesis two is confirmed in sixty-six of the items that make up the scale.
In Table 6 taking as a reference the value of the means obtained, we link the five items that show significant differences to their dimensions for clarity of presentation of results.These results show that the average of secondary school teachers is lower than that of primary school teachers in item 4.This item corresponds to the dimension presenting cultural diversity as an enrichment factor, and expressly refers to the equal value of all cultures.
However, when it comes to the educational considerations and practices which have a negative vision of cultural diversity, primary school teachers disagree more than secondary school teachers, who have higher means.So, secondary school teachers are more in agreement than primary school teachers with regard to the following: 1. Students' religious education is the responsibility of school (item 21), 2. the inclusion of all students is only possible with extracurricular activities (item 46), 3. teachers' beliefs about cultural diversity are not modified by initial training (item 51), and 4. the presence of culturally diverse students does not involve changes in the programming, planning, development and assessment of teaching (item 68).
According to these data, we can affirm that these five items do not confirm hypothesis two of this research, since there are significant differences between primary and secondary school teachers.These differences indicate that primary school teachers are more favourable to cultural diversity than secondary school teachers.

DISCUSSION
The approach of compulsory education teachers who work with Roma students in Messina to perspectives (theoretical and practical) favourable to cultural diversity confirms the issues emphasized in the works of Matencio, Molina, and Miralles (2015), and Chiner, Cardona, and Gómez-Puerta (2015), However, at the same time, it contradicts the results obtained in other studies (Carrasco & Coronel, 2017).
The teachers who participated in this study positively value the use of dialogue in the process of social organisation and believe that schools should favour the learning of shared ethical values and that they should be based on participatory forms of functioning.Thus, they call for the strengthening of teachers' ethical commitment (Tarozzi, 2014) and the development of more democratic and inclusive forms of organisation and functioning in schools (Cardno, Handjani, & Howse, 2018).In line with this, they refuse to classify students according to their culture, and do not hold the view that schools that host culturally diverse students are less socially prestigious.All this shows their commitment to social justice and equality, and their rejection of segregation on cultural grounds, as Gorski (2009); Nieto and Mcdonough (2011) and Malusà and Tarozzi (2017) defend.
These perceptions collide, however, with the dispersion existing among the teachers participating in this study.For example, this can be seen when considering that there are cultures that are not integrable in a democratic society, or when the school system has to guarantee only the teaching of the professed majority religion of the country.The existing lack of consensus on aspects like these suggests that "teachers are still far from having internalised an intercultural model of education" (Borrero & Blázquez, 2018, p. 88).
The results of this study confirm, as Silverman Silverman (2010) and Hinojosa and López (2018), show, that teachers are more in agreement with theoretical principles (in favour, or not in favour, of cultural diversity) than with ways of putting these principles into practice.This reflects the challenge they face when undertaking truly inclusive actions in multicultural contexts (Cardno et al., 2018).For Carrasco y Coronel, this difficulty can be attributed to their lack of intercultural competency: "…the lack of intercultural competency is very likely to appear as a key indicator when explaining teachers' perceptions of cultural diversity and its impact on classroom activity" (Carrasco & Coronel, 2017, p. 92).However, for Chiner et al. (2015), this problem stems from the result of the conflict between the personal and professional beliefs of active teachers, since "they show lower levels of sensitivity in the educational field, while their personal beliefs remain positive" (p.23).
If we focus on the second of the objectives of the study, we find that, in general, there are no significant differences between primary and first grade secondary school teachers, except with regard to five issues.In them, primary school teachers tend to be more in favour of cultural diversity and more critical of those specific aspects of theoretical and practical perspectives that do not favour cultural diversity.These findings confirm, if only in specific aspects, the results obtained by Bagant, Jiménez, Ferrer, and Ruíz (2005) which these authors attribute to the different pedagogical training of both professional groups.In the Italian context, first grade secondary education represents one of the weaknesses of the school system (Malusà & Tarozzi, 2017) and poor pedagogical training is frequently noted as one of the factors which influence this the most (Dusi, 2017).This could be one of the reasons that account for the few differences because, in this specific case, none of these professional groups have received specific training in how to manage the cultural diversity of their classrooms from an intercultural approach.In this sense, it would be interesting to extend studies focused on analysing the impact that those training programmes, committed to cultural diversity inclusion, have on teachers and intercultural educational practice in other countries.Regarding this, there are more optimistic perspectives, which rely on the transformative power of teacher training (Bodur, 2012;Hinojosa & López, 2016) but there are others which are not so optimistic (Nelson & Guerra, 2013;Reiter & Shannon, 2011).It is possible that the key to this controversy is not only about having access to initial and permanent teacher training, but also about analysing the type of teacher training offered and the approaches adopted.In their early research.Hinojosa and López (2016), analysed the impact of different training initiatives aimed at qualifying teachers in the management of cultural diversity.Following this, they concluded that, although the results of the reviewed investigations are inconclusive: we can affirm that the formative proposals which are concerned with linking theory and practice, and which incorporate critical reflection, dialogue, exchange with colleagues, case studies and collaborative work, usually have a more positive impact on the attitudes, behaviours, beliefs and perspectives of future teachers towards cultural diversity Hinojosa & López, 2016, p. 102 The significant differences found between teachers of both educational levels show a greater degree of agreement among primary school teachers, compared to their first-grade secondary counterparts, in which all cultures are perceived as having the same value.This is a view which Silverman (2010) and Tarozzi (2014) criticise as superficial and uncritical.Primary school teachers are also more willing to adapt their teaching to the cultural diversity of students, incorporating changes in the programming, development and assessment of their teaching.This is possibly because, as Díez (2014), argues, secondary school teachers have incorporated the concept of intercultural education into educational discourse, even if they continue to be reluctant to incorporate it into their professional practice.This may derive from the belief that they are already offering an inclusive response and want to end the idea -also questioned by Chiner et al. (2015), that the best way to manage cultural diversity is by giving a differentiated or compensatory response.However, according to Carrasco and Coronel (2017), this attitude is due more to an attempt to make cultural diversity invisible.The limitations imposed by the characteristics of our study do not allow for clarification of this point, so it would be interesting to broaden the inquiry with more qualitative investigations that help to explore this issue more deeply.

CONCLUSIONS
The results obtained have allowed us to respond to the two objectives of the study and conclude that: • The perceptions of the teachers participating in this study are closer to theoretical and practical perspectives favourable to cultural diversity, regardless of the educational level which they teach.
• The compulsory education teachers show a highly favourable attitude to the use of dialogue between cultures as well as to the configuration of a school open to participation and committed to the adoption of shared values.They do not believe that the schools which host culturally diverse students are schools with less social prestige.In addition, they are opposed to the grouping of students according to their culture.
• There are, however, teachers who have a negative vision of cultural diversity and are resistant to the development of educational practices that favour the inclusion of culturally heterogeneous students.
• The teachers participating in this research find it difficult to transfer their theoretical, generally positive, perspectives about cultural diversity to professional practice.
• There are some differences in perceptions about diversity between primary school teachers and first-grade secondary school teachers.These differences indicate that primary school teachers tend to value the cultural diversity of students in a more positive light and are more committed to educational practices which favour their inclusion.
• However, the quantitative nature of this work has not allowed us to investigate, in depth, some of the issues revealed in the study, issues that deserve greater attention.
For this reason, we consider it necessary to broaden the scope of inquiry by adopting new, more qualitative approaches.In addition, there is a need to incorporate new factors which help to better understand teachers' perceptions about cultural diversity, the differences existing between primary and secondary school teachers, their commitment to intercultural education and the inclusion of all students.In this sense, studies which focus on teacher training (Hinojosa & López, 2016), inclusive leadership (Cardno et al., 2018), the institutional culture or the educational policy (Szeto & Cheng, 2018) are likely to be very revealing.

Table 1
Total and by dimensions values of Cronbach alpha coefficient.

Table 2
Means and standard deviations of the scale dimensions

Table 3
Means and standard deviations relating to the most relevant items of the scale In plural societies the best form of social organisation is that which comes from the dialogue between the different existing cultures.The entry of people from other countries has to be regulated, taking into account the contribution they can give to the general welfare of the host nation.

Table 4
Arithmetic means of the dimensions and total scale according to educational level

Table 5
Arithmetic means of the items that present significant differences, by educational level

Table 6
Items where the educational level variable (relating to the level educators teach) offer statistically significant differences.