Influencing Factors on In-Service Teachers’ Competence in Planning CLIL

In spite of the multiple competencies necessary to design and deliver proper Content and Language Integrated Learning (CLIL) lessons, the teachers’ linguistic proficiency is the primary competence considered in the accreditation of CLIL teachers in Spain. However, teachers’ competence in planning CLIL lessons is key to bilingual education. This article explores this competence and reports on the several factors that influence the level of integration of CLIL methodological principles in the lesson planning style of in-service teachers at primary and secondary education. The factors studied are the stage of education, curricular subjects, teacher education, status at school, years teaching CLIL and type of school in a sample of 383 in-service CLIL teachers. Results show statistically significant differences in all the factors studied except in the stage of education (primary-secondary) and the type of school (state-semiprivate). These results reveal a high heterogeneity in the sample, which allows the description of the CLIL teacher profiles according to the factors that influence their competence in planning and delivering CLIL lessons. It is concluded that education and training in CLIL competencies, such as planning CLIL lessons, is also necessary in order to achieve a more homogeneous competence profile of the teachers. Therefore, a review of the academic programs and of the accreditation model so as to guarantee that they include prescriptive education and training in CLIL is recommended, since the quality and sustainability of bilingual programs also depends on teachers’ competence in CLIL and not only on the students’ academic results.


Introduction
In a political context of international cooperation in the field of education, important methodological and didactic changes are taking place in classrooms. Content and Language Integrated Learning (CLIL), which emerged in Europe in the 90s as an innovative approach to integrate content and language learning (Marsh, 1994), has already extended beyond the borders of the European Union (Ruiz de Zarobe, 2013), awakening the interest of practitioners and education authorities in Latin America, Japan, and Southeast Asia.
The increased implementation of CLIL in schools across Europe since the 1990s responds to an international strategy of the European Union aimed at making European citizens acquire a knowledge and command of two or more foreign languages in addition to their mother tongue (European Commission, 1995). Therefore, Spain, as a European country, has regulated bilingual education -mainly in Spanish and English -in all the educational systems of the 17 autonomous communities and the two autonomous cities of Ceuta and Melilla. A majority of these bilingual programs have opted for the CLIL approach, "a dual-focused educational initiative which advocates the learning of academic content and a foreign language simultaneously" as defined in Pavón-Vázquez and Ellison (2013, p. 68). One of the reasons that have highly contributed to the popularity of CLIL in Spain is the fact that the learning of curricular content in a foreign language (L2) allows increasing the number of the L2 learning hours without making any changes in the official curricula.  2016b, 2017a, 2017b). In addition, other national and regional studies confirm that the levels of competence in the foreign language are higher in students who go to bilingual schools than to non-bilingual schools (Comunidad de Madrid, 2016Madrid, , 2017Madrid, , 2018Gisbert, Martínez de Lis, & Gil, 2015;Shepherd & Ainsworth, 2017).
Furthermore, it seems that students' performance in the areas of mathematics, language, and science is not reduced due to this approach to teaching and learning (Comunidad de Madrid, 2016Madrid, , 2017Madrid, , 2018Ministerio de Educación, Cultura y Deportes, 2016b, 2017bRuiz, 2016;Tamariz & Blasi, 2016).
On the other hand, one of the pillars that sustains these programs are the CLIL teachers and their different roles in CLIL programs widely described by Marsh, Mehisto, Wolff, and Frigols (2010), Salaberri (2010), Pavón-Vázquez and Ellison (2013) or Ball, Kelly, and Clegg (2015). However, improvements in students' language proficiency and academic performance have not been accompanied by improvements in the training of the teachers who teach CLIL to these students. The impact that CLIL has meant for those teachers, who face the challenge of teaching curricular content in an L2, has not been measured with the same intensity or with the same reliability (Pérez-Cañado, 2018).
Teacher education and teacher training are essential for quality bilingual education. As Pavón-Vázquez and Ellison (2013) put it: "What is understood across most contexts is that CLIL is demanding for teachers in terms of adjusting practice and developing competences, and that prior training is essential" (p. 69). However, Morton (2016) points out that "one of the most challenging issues in all types of content-based instruction is teachers' perceived lack of pedagogical content knowledge necessary for effective content and language integration" (p. 144). This situation is the result of a deficient provision for teacher preparation in these programs (Banegas, 2012(Banegas, , 2015Coyle, Hood, & Marsh, 2010;Fernández-Cézar, Aguirre-Pérez, & Harris, 2013;Jover, Fleta, & González, 2016;Lancaster, 2016;Mehisto, Marsh, & Frigols, 2008;Pavón-Vázquez & Ellison, 2013;Pérez-Cañado, 2012, which generally consider linguistic competence as the only criterion to accredit teachers to teach CLIL.

CLIL teacher competencies
Quality CLIL teaching practice requires multi-skilled practitioners.
Content teachers and language teachers who deliver CLIL lessons have to put into practice a wide variety of competencies ranging from pedagogical demands (content and language knowledge and awareness, CLIL fundamentals and management, and learning resources and environments) to personal and professional skills (personal reflection, interpersonal and collaborative competence, developmental and research competence) (Bertaux, Coonan, Frigols, & Mehisto, 2010;Madrid-Manrique & Madrid-Fernández, 2014;Marsh et al., 2010;Pavón-Vázquez & Ellison, 2013;Pérez-Cañado, 2017.
In order to offer adequate education and training to CLIL teachers, it is, first of all, necessary to decipher their specific needs in practical terms, based on the competencies required to integrate language and content in a single lesson (Brüning & Purrman, 2014). Thus, as many researchers agree (Mohan; Nunan; Snow, Met, & Genesee;van Lier, as cited in Coyle et al., 2010;Marsh, 2012), this study also considers the CLIL lesson plan as the tool to carry out this integration and as a determining factor in the quality of the teaching and learning processes that take place in the classroom. The majority of authors who have defined the competencies of the CLIL teacher agree on the importance of providing these teachers with the appropriate tools to develop an adequate lesson plan (Bertaux et al., 2010;López-Hernández, 2016;Llull, Fernández, Johnson, & Peñafiel, 2016;Madrid-Manrique & Madrid-Fernández, 2014;Marsh et al., 2010;Pavón-Vázquez & Ellison, 2013;Pérez-Cañado, 2017. Therefore, the next step is to determine what the core elements of a CLIL lesson plan are.

The CLIL lesson plan
The cognitive and linguistic demand that CLIL implies, not only for students but also for teachers, requires the mastery of basic methodological principles that must be present in all types of CLIL lesson plans.
These principles should be known by all CLIL teachers and serve them as a reference for the design and delivery of their lesson plans. In spite of that, a majority of bilingual programs in Spain only consider linguistic competence to accredit the teachers to impart CLIL. In the Commu-UNIVERSIDAD DE LA SABANA DEPARTMENT OF FOREIGN LANGUAGES AND CULTURES nity of Madrid, where this research was carried out, CLIL training is not mandatory at any stage of the accreditation process. Since 2010, when the accreditation process was regulated as a linguistic certification, the training on pedagogical content knowledge of bilingual education was relegated to the volunteerism of teachers, which has generated a very heterogeneous competence profile of CLIL teachers (Custodio-Espinar, 2019;Morton, 2016;Herrero-Rámila, 2015;Pérez-Cañado, 2018).
For this reason, an ad hoc scale was created to measure the level of integration of the CLIL methodological principles in the teaching style of accredited teachers who teach CLIL in bilingual schools in order to identify those profiles and their training needs. The tool was designed following a set of methodological principles that must be present in a CLIL lesson plan model (Custodio-Espinar, 2012, 2019): • The CLIL lesson plan must include the four CLIL components defined by Coyle (1999): content, cognition, communication, and culture, known as Coyle's 4Cs.
• Content needs to be seen through a conceptual, linguistic, and procedural dimension with each of these dimensions representing planning tools and priority objectives for any didactic and pedagogical activities or tasks (Ball et al., 2015).
• The language must be approached from content and cognitive processes necessary to study in the foreign language, such as cognitive academic language proficiency (CALP), and from the interaction processes involved in a real communicative context, such as basic interpersonal communication skills (BICS) (Cummins, 1999). In CLIL, language is used to learn and communicate at the same time.
• The content determines the linguistic demands that teachers must analyze and support according to their students' linguistic competence. The Common European Framework of Reference (CEFR) is a useful tool to determine the level of proficiency and to develop language assessment strategies adapted to that level (Council of Europe, 2001).
• Therefore, it is necessary to develop content-obligatory language and content-compatible language objectives for the lesson in order to guarantee that students learn the key content concepts, as well as expand their language learning beyond academic forms and functions (Snow, Met, & Genesee, 1992). • The language demands analysis must guarantee that classroom interaction is supported by strategies that are adapted to the language proficiency of the students (Coyle, 1999(Coyle, , 2007Coyle et al., 2010). This analysis allows teachers to develop reception, transformation and production scaffolding (Dale & Tanner, 2012) likely to ensure a more interactive learning in the L2.
• Attention to diversity through the use of cognitive taxonomies, such as Bloom's Taxonomy (Bloom, 1956), is necessary to adapt the cognitive demand of the activities to the level of cognitive development of each student and ensure a progression from lower-order thinking skills (LOTS) to higher-order thinking skills (HOTS) (Anderson & Krathwohl, 2001).
• A coherent integration of theories and methodologies that underlie CLIL, such as student-centered, and autonomous, flexible learning should be used to promote the development of the key competencies described in the Spanish educational law, LOMCE (Ley orgánica para la majora de la calidad educativa) (Ley Orgánica 8, 2013).
• The activities must connect with curricular objectives, from an approach to content and meaning. They must be open and flexible, connected to students' interests, realistic and motivating, and aimed at the completion of tasks and final products. They should allow evaluating content, process and language in an integrated form (Pérez-Torres, 2015).
• "CLIL-specific learning materials support the creation of enriched learning environments where students can simultaneously learn both content and language whilst becoming more adept learners of both" (Mehisto, 2012, p. 17). In particular, information and communication technologies (ICT) resources can greatly contribute to the development of activities and tasks that promote the type of learning required by CLIL and help students to progress autonomously at their own pace, thus facilitating the development of strategies of academic-scientific learning (Custodio-Espinar & Caballero-García, 2016).
• CLIL assessment should cover language and content progress by combining different strategies to assess students individually and in groups with oral and written tasks that provide opportunities for co-evaluation and self-evaluation. Formative assessment strat-UNIVERSIDAD DE LA SABANA DEPARTMENT OF FOREIGN LANGUAGES AND CULTURES egies should be included to optimize the learning process, as well as summative strategies to measure the quality of that learning (Assessment Reform Group, 2012;Llinares, Morton, & Whittaker, 2012;Lofft-Basse, 2016;Wewer, 2014;Wylie & Lyon, 2015).
• The organization of timetables must respect the principles of intensity and reiteration. Additionally, the classroom's arrangement must show a positive, affective environment through the use of visual resources, the organization of furniture or any other element that promotes interaction, motivation and trust. This can help to decrease anxiety and the threshold of students' affective filter (Krashen, 1985).
This theoretical framework was the basis of the questionnaire created to measure the level of integration of the CLIL methodological principles in the planning styles of teachers and, thus, their competence in planning CLIL lessons. The construct CLIL lesson plan, which corresponds to the structure of the questionnaire, was organized around five major dimensions: core CLIL components, methodology, activities and resources, evaluation, and organization. This study intended to analyze the differences in the level of competence to plan CLIL lessons of accredited teachers by measuring the level of integration of the CLIL methodological principles in their planning styles according to the factors stage of education, curricular subjects, academic background, status at school, experience teaching CLIL and type of school where they work.

Method, objectives and variables
This study had a non-experimental design. It was an ex-post-facto descriptive and differential study whose general objective was to obtain information that can contribute to the consolidation of the bilingual programs from the perspective of the teacher competence to plan CLIL lessons. The specific objective was to analyze the differences in the degree of integration of the CLIL methodological principles in the planning style, according to six factors: educational stage, curricular areas, academic background, status at school, experience delivering CLIL and type of school of the CLIL teachers. The research questions were as follows: • RQ1 Are there differences in the level of integration of CLIL principles of accredited teachers according to their stage of education?
• RQ2 Are there differences in the level of integration of CLIL principles of accredited teachers according to their curricular subjects?
• RQ3 Are there differences in the level of integration of CLIL principles of accredited teachers according to their academic background?
• RQ4 Are there differences in the level of integration of CLIL principles of accredited teachers according to their status at school?
• RQ5 Are there differences in the level of integration of CLIL principles of accredited teachers according to their experience teaching CLIL?
• RQ6 Are there differences in the level of integration of CLIL principles of accredited teachers according to their type of school?
There are six dependent variables in the study. They are the global score in the level of integration of CLIL principles in the sum of the five CLIL lesson planning dimensions and in each of these dimensions: D1 Core CLIL components, D2 Methodology, D3 Activities and resources, D4 Evaluation, and D5 Organization.

Population and sample
In The software Ene3.0 was used to calculate the sample using a stratified and proportional random sampling technique (precision level of 0.7, confidence interval of 95% and a standard deviation of 3).
The result was a sample of 74 schools, which were distributed proportionally in five strata. Each stratum corresponds to one of the educational districts, known as DAT (dirección de área territorial), in which the Community of Madrid is organized for educational purposes. Teacher participation was estimated at five teachers per school, 370 teachers in total (see Table 1). A comparison of the distribution of the ideal and the real sample in the five strata is included in Table 2. 1 indicates "never" and 6 indicates "always," was used to measure the 31 variable-items, with which the teachers evaluated their own classroom planning practices according to how often they took into account the CLIL principles in their actual planning practice. These variables were organized in five dimensions. Each of these dimensions included a series of sub-dimensions that allowed the identification of the core components of the approach and its basic methodological principles described in the theoretical framework (see Table 3).

Procedure and analysis
Data was collected in 2017 following a strict on-site application protocol, in which the researcher visited every school of the sample. The statistical analysis of the data was developed using IBM SPSS 20. The analysis consisted of descriptive studies of all the variables and differential analyses calculated for the whole scale (the 31 study variables), as well as for each of the five dimensions. The procedure for contrasting the hypotheses was the analysis of variance, by means of ANOVA (with Tukey b for subsequent contrasts), when the factor had three or more groups. When the factor was dichotomous (two groups), the Student's "t" was used. Significance levels were set at the 5% level and the effect size was estimated using η2.

Instrument reliability
The total reliability of the scale is high and satisfactory with a Cron-

Descriptive analysis
The   A majority of CLIL primary teachers in the sample teach Natural Science, Social Science and Arts & Crafts at the same time, which is one of the main differences with their colleagues from secondary school who only teach the subjects related to their academic degree. In secondary school Advanced English teachers, who also teach English Literature, are a majority.
Concerning teachers, education, only 11.7% of the sample hold a four-year-degree. Despite the fact that a majority of teachers work in primary school, 35.5% of the sample hold a Licenciatura. This means that a low percentage of primary-school teachers hold more than one university degree, since the former Diplomatura in Education or the current Degree in Primary Education is compulsory to work as a primary-school teacher in Spain.
Almost two thirds of the sample have been teaching CLIL for more than four years (see Figure 4). Regarding the status of the teachers in their schools, more than half are permanent teachers (220), 183 of whom are civil servants (47.8%), and 37 have an indefinite contract in a semiprivate school (9.6%) (see Figure 5).
It is remarkable that almost one third of the sample are interim teachers working in public schools. Finally, the majority of the teachers (87.5%) work in public schools.
With regard to the dependent variables, the sum of total scores in the variables of the five dimensions has a normal distribution in the sample (see Figure 6).  The maximum score in the sample is 179 points out of 186 and the minimum is 88 points. This makes a difference of 91 points, which means that the sample is very heterogeneous with respect to the level of integration of the CLIL methodological principles.
The means in the level of integration of the CLIL methodological principles are shown in Table 4, including the means, the standard deviation, and the number of teachers in each group of the variable.  Source: Own Elaboration.

Differential analysis of the competence to plan CLIL
The study analyzed the level of integration of CLIL principles in the planning style of the sample according to the six independent variables described in order to identify the factors influencing teachers' competence to plan CLIL lessons. No differences in the level of integration of the CLIL methodological principles were expected in any of the 42 hypotheses (Curricular Subjects is divided into primary school and secondary school, thus IV 2a and IV 2b). First, the results of the seven hypotheses that measure the effect of the six independent variables on the main dependent variable of the study are shown in Table 5. Statistically significant differences were found in all the independent variables except the stage of education. It is also remarkable that the curricular areas taught by primary teachers do not show differences whereas they show differences in the secondary teachers. This is coherent with the differences found in the experience teaching CLIL, since primary teachers started their participation in the program six years before their secondary counterparts and, those primary teachers who did it before 2010, received specific training on CLIL called formación de entrada before their accreditation. The results by dimensions are summarized in Table 6.  Despite the fact that no differences were expected in any of these 35 hypotheses, statistically significant differences were found in 18 of them. The differences found in the Stage IV of education are not relevant since dimension 5 showed a very poor reliability, particularly comparing primary and secondary-school teaching scenarios. These results confirm the heterogeneity of the sample in the level of integration of the CLIL principles and, therefore, in the competence to plan and deliver CLIL lessons.

Discussion
Regarding RQ1, the results show that the stage of education is the only factor that does not influence the level of integration of CLIL principles.
As to the factors that show influence, the results of RQ2 show that, in primary school, a majority of teachers obtain similar means, although there are differences in the first and fifth dimensions in favor of Science On the other hand, the curricular subjects show differences in the secondary-school teachers in favor of the group of Advanced English teachers. It is important to point out that they hold a C2 in English and some of them are the coordinators of the bilingual program at their schools, which involves training and mastery in CLIL. The size of the effect of these results -39% in D1 Core CLIL components and 44% in D4 Evaluation -may be due to the fact that, at this stage, the vast majority of teachers have been qualified according to their language proficiency, with at least C1 level of proficiency; therefore, their knowledge of CLIL is not guaranteed. These differences in favor of the Advanced English teachers, who hold a C2, also confirm the direct relationship between language proficiency and mastery of theoretical aspects of CLIL pointed out by Pérez-Cañado (2016). explain that the new programs must be revised in order to improve the training of generalist teachers who do not study the specialty in a foreign language but can also become CLIL teachers.
With respect to the fourth RQ, a determining factor in the level of integration of CLIL principles is the status of the teacher at school. The results of these analyses revealed that the group of interim teachers shows statistically significant differences with the groups with indefinite-term contracts, in favor of the latter. This fact is very remarkable, because it points out that staff mobility is a factor of negative influence on teachers' competence in CLIL.
Concerning the influence of the number of years teaching CLIL, RQ5, the groups showing a higher level of integration of the CLIL principles are the 7-9 and 10-12 groups. These differences are statistically UNIVERSIDAD DE LA SABANA DEPARTMENT OF FOREIGN LANGUAGES AND CULTURES significant with the 1-3 group, which puts into value the knowledge that teachers develop from practice in the classroom. Although this experience is not usually recognized by research, as noted by Perines (2018), it seems to have a positive impact on teachers' competence to plan CLIL lessons.
Finally, regarding RQ6, it is necessary to say that, although the 4 th factor, type of school, showed significant differences in favor of the private school, this result has a relative value because of the limited sample of this group, which is represented by one private school. It is also interesting to note that the means of state and semiprivate schools are similar, and no significant differences were found between these two groups.
A thorough analysis of these results allows the identification of the following profiles in the sample: • These are the teachers who show a higher level of integration of the CLIL methodological principles in their planning style: -Teachers accredited with prior linguistic and CLIL training (Diplomados who received linguistic and CLIL training prior to their accreditation).
-Teachers with a C2 level of English. These differences are not found between levels B2 and C1, probably because B2 teachers are primary-school teachers accredited with prior linguistic and CLIL training, whereas C1 teachers are primary-and secondary-school teachers accredited after 2010 when the accreditation process was regulated as a linguistic accreditation; thus, CLIL training is not guaranteed.
-Trained CLIL teachers who manifest desire to still be trained in CLIL.
-Primary-school teachers who teach Social Sciences, Natural Sciences and Arts & Crafts simultaneously -a majority in the primary-school stage.
-Secondary-school teachers who teach Advanced English (including Literature), a majority in the secondary-school stage.
They hold a C2.
-Public servants with a permanent position at one school or temporally working at one school for particular reasons. They are usually very skilled teachers who stand out in important aspects for the school.
-Teachers with 7 to 12 years of experience teaching CLIL.
• These are the teachers who show a lower level of integration of the CLIL methodological principles in their planning style: -Teachers accredited without previous CLIL training, after 2010 (70% of the sample) and without post-accreditation training (50% of the sample).
-Teachers with a four-year degree, after Bologna 2003.
-Interim teachers and Physical Education teachers.
-Teachers having 1 to 3 years of experience teaching CLIL.

Conclusions and recommendations
From the analysis of the factors studied, it is concluded that the only factors that do not influence in-service teachers' competence in planning CLIL lessons are the stage of education (RQ1) and the type of school (state or semiprivate) (RQ6). Regarding the other factors, it is necessary to offer specific training on the core CLIL components (dimension 1) to primary-school teachers of Physical Education and to all secondary-school teachers except Advanced English teachers.
The secondary-school teachers also need training related to the formative and summative evaluation of students in CLIL contexts (dimension 4) (RQ2).
Furthermore, initial teacher education should include more ef- □ Lingüística y metodológica antes de la habilitación.