EFL teachers ’ beliefs and practices regarding writing activities and class organization

This study aimed to find out the relationship among EFL university teachers’ code-based and meaning-based beliefs in writing instruction, classroom writing activities and classroom organization. To this end, 120 university teachers, who were teaching writing to EFL learners at different universities in Iran, participated in the study. For the data collection, a valid and reliable questionnaire on writing was used. Descriptive statistics and Pearson Product-Moment correlation were carried out for analyzing the data. The results showed that the correlations between both teachers’ meaning-based and code-based beliefs in writing and writing activities exist. It was also found that the correlations between meaningbased beliefs in writing and whole class activity and group work as ways of doing writing activities were significantly positive. However, meaning-based beliefs in writing and individual writing were not correlated significantly. Moreover, correlations between code-based beliefs in writing and individual writing and whole class were positively significant and there was a positive but not significant correlation between code-based beliefs in writing and group work. The results can be useful for

Abstract: This study aimed to find out the relationship among EFL university teachers' code-based and meaning-based beliefs in writing instruction, classroom writing activities and classroom organization. To this end, 120 university teachers, who were teaching writing to EFL learners at different universities in Iran, participated in the study. For the data collection, a valid and reliable questionnaire on writing was used. Descriptive statistics and Pearson Product-Moment correlation were carried out for analyzing the data. The results showed that the correlations between both teachers' meaning-based and code-based beliefs in writing and writing activities exist. It was also found that the correlations between meaningbased beliefs in writing and whole class activity and group work as ways of doing writing activities were significantly positive. However, meaning-based beliefs in writing and individual writing were not correlated significantly. Moreover, correlations between code-based beliefs in writing and individual writing and whole class were positively significant and there was a positive but not significant correlation between code-based beliefs in writing and group work. The results can be useful for ABOUT THE AUTHOR Touran Ahour is an assistant professor in TEFL at Islamic Azad University, Tabriz branch, Iran. She has authored several books and published many articles in scholarly journals and presented papers in national and international conferences. Her research interests include materials evaluation, reading-writing connection, assessment, teaching skills, and other ELT issues. Farhad Golpour is a Ph.D. candidate in TEFL at Islamic Azad University, Tabriz branch. He is teaching at Teachers' university. His research interests include language teaching, writing, material development and teachers' beliefs. He has published some books and articles on some international journals and has presented some papers in both international and domestic conferences. Saeideh Ahangari is an assistant professor in TEFL at Islamic Azad University/ Tabriz Branch. Her main interests are Task-based language Teaching, CALL and their interface with the issues in language testing. She has published many articles and participated in many national and international conferences.

PUBLIC INTEREST STATEMENT
The way that teachers teach writing at schools or universities can be affected by their beliefs. In this article, we have investigated the role of university teachers' meaning-based and code-based beliefs in their preferred kinds of writing activities and class organization for doing these activities. The results revealed that teachers who had meaningbased beliefs highly used decoding and spelling and grammar activities and those who had codebased beliefs highly preferred product activities. It was also found that teachers with meaningbased beliefs preferred their students do writing activities in groups or as whole-class and those with code-based beliefs asked their students to do writing activities as a whole-class or individually. The findings might be useful for teachers to match their beliefs with practices and use the useful types of activities and organize them in a way that suits the students' level and interests to encourage them to write. Bauersfeld (1979) reported that before 1980s, the learners' amount of learning was investigated in order to reform and evaluate the educational content; however, the influential roles of teachers in the success of learners were ignored. In other words, in this period, teachers and their beliefs regarding how they can make a difference in the classroom that can result in the progress of the learners were not considered as important. It was believed that because prospective teachers spend lots of hours in the classroom, their beliefs are shaped from the time they receive formal training, therefore; later they would try to use the same latent beliefs in their own classrooms (Zeichner & Tabachnick, 1981). In this regard, Pajares (1993) found that "beliefs are the best predictors of individual behavior" (p. 45) and so we can claim that "teachers' knowledge and beliefs about teaching and subject matter are major determinants of what they do in classrooms" (Borko & Putnam, 1996, p. 675). Skott (2009) declared that belief can influence practice and play an important role in teaching practice. Therefore, teachers' beliefs have an essential role in their own teaching and influence their teaching (Grossman, Wilson, & Shulman, 1989). In other words, belief and practice have reciprocal influence on teachers. Richardson (1996) believed that teachers' beliefs can influence their practices and the reverse, that is, their practices can influence their beliefs. Gaitas and Martins (2014) identified two kinds of beliefs as code-based and meaning-based beliefs: teachers who focus on teaching grammatical points and help learners to write correctly have code-based beliefs, and those who focus on writing for developing writing skill instead of teaching how to write have meaning-based beliefs. Our study is the reflection of Gaitas and Martins' conceptualization of teachers' code-based and meaning-based beliefs but at the university setting.

Introduction
Some studies have shown the importance of teachers' beliefs in pedagogical practice (e.g. Borg, 2003;Farrell & Kun, 2008;Johnson, 2011;Johnson & Johnson, 1974;Pajares, 1992). Schoenfeld (2014) indicated that there is a direct connection between teachers' beliefs and classroom practice. Some other researchers such as Fives and Buehl (2012), Lederman (1992), Wilson and Cooney (2002) mentioned that classroom practice is formed by teachers' beliefs and preferences. Clark and Peterson (1986) stated that values, principles and beliefs can control and make sense of the teachers' behavior. According to Buehl and Beck (2015), in some quantitative studies (e.g. Thoonen, Sleegers, Peetsma, & Oort, 2011;Wilkins, 2008) through self-reported measures a relationship ranging from weak to medium was found between teachers' beliefs and practices. There have also been some studies such as Poulson, Avramidis, Fox, Medwell, and Wray (2001), Tolchinsky, Bigas, and Barragan (2012), and Gaitas and Martins (2014) that found positive and significant relationships between teachers' beliefs and different writing activities.
One factor that may affect learners' writing is class organization, that is, the way teachers organize their classes to do the exercises. To do this, teachers may ask learners to do their activities either in pairs and small groups, individually, or as a whole class. Some researchers including Porto (2001) and Gaitas and Martins (2014) have studied the relationship between teachers' beliefs and class organization and have found a positive relationship. The study done in Spain by Gaitas and Martins (2014) showed that the correlations between beliefs and activities and beliefs and classroom organization, supported by meaning-based and codebased beliefs were significant.
Generally, teachers have different beliefs and views toward teaching writing activities and methods of teaching writing. Some teachers focus on new theories that encourage meaning-focused writing; while, many teachers do not believe in this model and their focus is more on correcting learners' errors and form of writing. Birjandi and Malmir (2009) stated that in spite of all of the developments in teaching of writing in different EFL contexts, the traditional approach, that is, "product approach" is still used in teaching writing to learners in Iranian universities and colleges. In addition, qualitative research on writing in general and teaching of writing in particular is very poor in Iran. In this regard, the purpose of this study was to find the degree of relationship between university teachers' beliefs, kinds of writing activities and class organization they employ in their writing classes in order to determine whether there are matches or mismatches of their beliefs and practices, something that is missing among the studies in this area in the EFL context of Iran.

Theoretical views
In this part, some theoretical views concerning the variables of the study are presented.

Teachers' beliefs
The construct of belief has been defined differently in different educational contexts. For example, belief is conceptualized as an "individual's judgment of the truth or falsity of a proposition" (Pajares, 1992, p. 316). For Pajares (1992), beliefs usually overlap with close psychological terms such as knowledge, attitude, opinion and ideology. In his perspective, a definition for beliefs is "at best a game of player's choice" (p. 309). Kuzborska (2011) suggested that all teachers hold beliefs about their work, their subject matter, their students, and their roles and responsibilities. Yero (2002) indicated that a teacher's beliefs about school are drawn from his/her experiences as a student. Porter and Freeman (as cited in Pajares, 1992) called belief as a category of orientation concept in teaching. Borg (2001) defined belief as an evaluative proposition that people have consciously or unconsciously and is accepted as a fact by individuals. Pajares (1992) stated that for planning a lesson, the kinds of decision-making, and for general classroom practice the influences of the beliefs are more than knowledge, and he added that the actual teachers' behaviors toward students are drawn from teachers' beliefs. Teachers can adjust their behaviors and educational decisions based on learners capabilities if they can recognize learners' abilities. Kennedy (1997) indicated that the source of the teachers' beliefs is not clear, it may be the result of their training, socialization at schools or the reflection of their life experiences. Tao and Juliang (1999) uncovered that teachers' beliefs are drawn from cultural interaction, selfconstruction and social history and culture. Davis and Andrzejewski (2009) pointed out that for the interpretation of ambiguous situations, teachers use the ways that are in line with their beliefs. Moreover, teachers use their beliefs for designing educational goals and setting standards, and that beliefs help teachers to omit unnecessary matters and make clear what should be examined and emphasized. In this study, belief refers to the code-based and meaning-based teachers' beliefs in the way of using writing activities and classroom organization, which are conceptualized in the following sections.

Code-based and meaning-based approaches and beliefs
In literary instruction, teachers' writing practices have been divided into two important approaches and a great deal of research has been conducted on this object . According to Ehri, Nunes, Stahl, and Willows (2001), code-based approach, also called phonics or skill-based approach, states that students should focus on letter and sound relationships and the best way for recognition and writing correctly is repetition and practice. This approach also considered as bottom up processing, that is, language knowledge is built little by little to higher levels. Moreover, by correspondence between graphemes and phonemes, children turn unfamiliar written words into familiar spoken forms. In this study, according to Gaitas and Martins (2014) teachers who used this method believe that it is the best method of teaching writing and doing writing activities; therefore, teacher with such beliefs considered to have code-based beliefs. For example, they believe that before beginning a written expression activity, the teacher should prepare students to write correctly or teaching specific grammar classes are necessary in order to write correctly.
The second approach, i.e. meaning-based approach, also called whole language states that from the beginning students should be immersed in written language by reading books and writing their own story books. According to Goodman (1992), Dahl and Freppon (1995) and Teberosky and Colomer (2003), reading and writing are communicative activities. They mentioned that learners make hypotheses about words they encounter and find enough information for testing their hypotheses. In this top-down process approach if learners are put into an environment, that is full of print, and ask them to investigate, they will become literate. According to Goodman (1992), meaning-based activities include using context for understanding the new words, writing personal experiences and reading stories to the children. In this study, according to Gaitas and Martins (2014), teachers who used this method of teaching writing believe that this way of teaching is the best method of teaching writing and doing writing activities; therefore, they are considered to have "meaning-based beliefs". Meaning-based beliefs allows greater concentration, and individual writing is the best way to develop writing skills; moreover, in a writing session instead of telling students how words are written, the teacher should encourage them to do it alone.

Writing activities
Writing is one important skill in language learning and teaching and teachers use different activities to teach this skill. In this study, writing activities refer to different kinds of activities including product, process, decoding, and autonomous activities that are conceptually explained below.
Product activity is "a traditional activity in which students are encouraged to mimic a model text, usually is presented and analyzed at an early stage" (Gabrielatos, 2002, p.5). For example, in this approach students are given a sample and are asked to write a piece of writing by using the given sample and standards. Teachers expect their students write a composition that is grammatically errorless and is rich in rhetorical conventions. Product-oriented view toward writing emphasizes mastery of correct grammar and vocabulary (Kroll, 2003). Likewise, Hyland (2003) contended that this view emphasizes the achievement of the final result by students rather than the improvement of the skill of writing itself. The final result refers to the similar production of a model text with the correct use of grammar and vocabulary.
Process activity is a complex activity in which the writer should pay attention to how to create a piece of writing and teacher must discover how it is created. Based on process approach to writing, learners go through different stages (e.g. planning, writing, revising, and editing) to create a text (Hyland, 2003;Seow, 2002;Weigle, 2014), which is beneficial for them in that they can develop writing and cognitive skills.
Decoding activities are useful in developing writing skills. One such activity is dictation and Davis and Rinvolucri (1988) indicated the value of decoding and recoding of sounds in dictation as one important task in language learning. Dictation permits work on different language features such as spelling, vocabulary, and grammar. It can be a very effective tool for improving all four skills in both aspects of language: accuracy and fluency. Actually, dictation helps in developing students' abilities to listen for gist and very specific information, give a practice in listening, speaking and writing, and develops learners' awareness of word stress in speaking.
Autonomous activities are activities that encourage the involvement of the individual learners, such as writing everyday reports, recipes, stories, descriptive reports, messages, etc. The common phenomenon among learners is a passive role that they assume in the process of learning; they rely on teachers too much and are reluctant to develop a sense of responsibility for the outcome of their learning. Littlewood (1999) defined an autonomous person as "one who has an independent capacity to make and carry out the choices which govern his or her actions" (p. 428). The definition highlights two important aspects of autonomy: (1) learners' abilities to take charge of their own learning; and (2) their independence in decision-making, that is, they are able to regulate their learning without relying on others, for example, teachers. Weinstein, Curran, and Tomlinson-Clarke (2003) defined class organization as organization, especially the social organization that includes how students communicate and interact with each other and the teacher, also the ways that teachers structure the classroom interactions and activities to promote learning including communication, relationships, time, and the arrangement of the physical environment. According to Cohen (1994), group work means when students are working together, the group is so small that they can participate in doing an assigned task. Cohen, Manion, and Morrison (2000) stated that when learners doing group work, different members of the group bring different knowledge, the more members involve in a group the more knowledge is available. Moreover, when learners doing an activity individually it means every person does an activity based on his personal and individual abilities. Moreover, whole class activity means whole class doing the same activity. This is what we mean by class organization in our study.

Empirical studies
There have been some studies investigated the relationship between code or product-based and meaning or process-based beliefs in writing instruction and different writing activities. For example, Graham, Harris, MacArthur, and Fink (2002) conducted a survey on 150 primary grade teachers and found a positive correlation between belief in process writing instruction and whole language, and writing activities including invented spelling, students helping each other, students sharing writing with peers, student selection of writing topics, teacher mini-lessons, and student/teacher conferences. They also uncovered a positive correlation between belief in correctness in writing and activities such as spelling instruction, handwriting and grammar instruction, that is, skill instruction. Graham et al. (2002) discovered that teachers mostly spent their times on teaching mechanics of writing and grammar, and they spent less than one hour a day on teaching writing. Indeed, within a month most of the time was devoted to the skill teaching instead of teaching writing processing such as planning and revising. The survey of Cutler and Graham (2008) among random samples of primary grade teachers (N = 178) is similar to Graham et al. (2002) in that they paid more attention to code-based writing. More specifically, the typical teacher placed considerable emphasis on teaching skills such as spelling, grammar, capitalization, punctuation skills, handwriting and sentence construction skills. They also reported that most of the teachers in their study in the United States used an eclectic method in teaching writing, that is, a combination of process and skill-based methods. Poulson et al. (2001) by investigating 225 primary grade teachers' responses found a positive correlation between teachers' beliefs in correcting learners' handwriting and writing and regular use of spelling lists and copying by children. They reported that the beliefs that young children should select their own writing topics were negatively correlated with these activities. Regarding spontaneous writing activities, they were positively correlated with the beliefs that children should write for audiences other than teacher and that they should choose their own writing topics. These activities were negatively correlated with the belief in handwriting and correctness. As we can see, most of the mentioned studies have been conducted among the teachers teaching to the primary grade learners with various results.
Some studies indicate the mismatch of belief and practice among writing teachers. For instance, Ferede, Melese, and Tefera (2012) conducted a survey (using questionnaire and observation) of preparatory school teachers (n = 19) and students (n = 295) in Jimmy Zone. Their results showed that perception of teachers about writing and their practice of teaching writing were loosely related. In this regard, although most of the teachers preferred teaching writing based on process approach, a majority of them indicated that guided writing is better than free writing. Even, while some of them contended that expecting error-free writings from the preparatory school students is not reasonable, some other posited that tolerating their errors can prevent learners from accurate writing later. Tolchinsky et al. (2012) conducted a research to provide a detailed characterization of teachers' practices in primary schools when teaching writing. A 30-item questionnaire was used, and among the items were those related to activities and class organization, which were the concerns of our study. Teachers stated a group of activities such as letter recognition, autonomous writing, strategic reading activities, explicit analysis of word-sounds and explicit teaching directed at letterto-sound correspondences. According to the responses, three practices were reported: explicit teaching practices (concentrated on learning outcomes and code-based activities), situational practices (preferred strategic reading and occasional learning and spontaneous writing), and both autonomous writing and explicit instruction. Regarding the class organization, the results showed that teachers who have code-based beliefs use individual writing activities; while, those who support meaning-based beliefs employ pair, group, and whole-class writing activities.
In another study, Khanalizadeh and Allami (2012) aimed to find out Iranian EFL teachers' beliefs with regard to writing instruction. Their participants were 122 teachers of private language institutes to whom the belief questionnaire indicating teachers' orientation such as form, process, and social-based ones were distributed. The results indicated that a majority of teachers had eclectic orientation with the form-based orientation as the dominant one. Similarly, in Shahvand and Rezvani's (2016) study on 90 EFL teachers of different language institutes in the cities of Tehran and Isfahan in Iran, no significant relationship was found between teachers' beliefs and practices regarding effective teaching.
Regarding the relationship between teachers' beliefs, writing activities, and class organization, Gaitas and Martins (2014) surveyed the code-based and meaning-based beliefs of 255 Portuguese primary school teachers in using different forms of writing activities (e.g. process writing such as planning texts and revising texts; decoding such as copying and dictation; autonomous writing such as texts, everyday reports, informational texts, messages, recipes/instructions/rules, stories, descriptive reports, and wordlists; writing as a product activity such as themed compositions and free compositions; and spelling and grammar worksheet activities), and classroom organization (pairs or small groups; individual; and whole classroom). The results revealed the positive and significant use of autonomous and process activities with small to medium effect size by the teachers with meaningbased beliefs, and positive and significant use of decoding and product activities with medium effect size by the teachers with code-based beliefs. They also concluded that meaning-based beliefs are correlated with small group and whole classroom writings, while, code-based beliefs are associated with individual writing. However, there were different associations, which revealed that teachers used multidimensional aspects in their writing instruction theory and practice. Tagle et al. (2017) conducted a study on 37 EFL teachers in two Chilean universities to find their beliefs about teaching writing. The results indicated that the teachers believed in teaching writing production based on correct grammar and vocabulary as well as the replication of text types that are the embodiments of code-based and product approach in teaching writing. In addition, the teachers believed in the use of successive stages at the moment of production, which embodies the process approach to teaching writing.
In some studies, the use of writing strategies in improving the writing skill of the learners was investigated. In this respect, the aim of Bidabadian and Tabatabaei (2015) study was to find the relationship between teachers' beliefs and instructional practices regarding writing strategies. The results revealed their positive beliefs towards most of the writing strategies out of which compensational and social aspects of writing strategies were more considered in their actual practices. The other strategies such as metacognitive, cognitive, and affective ones were not among their actual instructional preferences.
As the literature showed, there were some controversies in the results of studies related to the teachers' beliefs, especially meaning-based and code-based, kind of writing activities and class organization. In addition, few studies conducted in the EFL context of Iran investigated general orientation of teachers in writing instruction in the language institutes. Therefore, in order to fill the gap and contribute more to the field of writing, this study aimed to find out the relationship between EFL university teachers' code-based and meaning-based beliefs and their classroom organization and classroom writing activities. In this regard, the following specific research questions and the related null hypotheses were formulated: 1. Is there any significant relationship between teachers' meaning-based and code-based beliefs and writing activities done in writing classes?
Ho: There is no significant relationship between teachers' meaning-based and code-based beliefs and writing activities done in writing classes.

Is there any significant relationship between teachers' meaning-based and code-based beliefs and classroom organization in doing writing activities?
Ho: There is no significant relationship between teachers' meaning-based and code-based beliefs and classroom organization in doing writing activities.

Participants
The study was conducted by the cooperation of accessible teachers teaching writing to EFL learners at different universities in Iran. Teachers were selected based on convenience sampling (Best & Kahn, 2006). In this regard, 120 teachers participated in the study. These teachers were males and females with different academic degrees and various years of experiences. Regarding their experience, teachers with less than 10 years as indicated by Fernandez-Garcia et al. (2019) considered as less experienced ones and those with ten years and above as experienced ones. Table 1 indicates more information about the participant teachers.

Questionnaire of teachers' beliefs
Teachers were asked to complete a survey questionnaire (Gaitas & Martins, 2014) comprising four distinct parts (see Appendix), which are explained separately below.

A. First part
The first part gathers demographic information including gender, years of experience, level of education, and field of study.
Teachers were asked to express their degree of agreement or disagreement in relation to each item. Twelve items weighted on the first factor (i.e. meaning-based beliefs) had the internal consistency of .94, and the 12 items weighted on the second factor (i.e. code-based beliefs) had the internal consistency of .92 as reported by Gaitas and Martins (2014). Examples of meaningbased beliefs include: "because it allows greater concentration, individual writing is the best way to develop writing skills; and in a writing session, instead of telling students how words are written, the teacher should encourage them to try to do it alone. Examples of code-based beliefs are: Before beginning a written expression activity, the teacher should prepare students to write correctly; and teaching specific grammar classes are necessary in order to write correctly".

C. Third part (writing activities)
The third part asks how often teachers, or their students, engage in specific writing activities. This part consists of 15 items evaluated on a 6-point scale from 1 (never) to 6 (always). Teachers indicated how often each activity occurs. The activities were generally grouped as suggested by Tolchinsky et al. (2012), that is, process writing activities (planning texts and revising texts); decoding writing activities (copying and dictation); autonomous writing activities (texts, everyday reports, informational texts, messages, recipes/instructions/rules, stories, descriptive reports and wordlists); writing as a product activity (themed compositions and free compositions); and worksheets (spelling and grammar). The internal consistency of .92 was reported by Gaitas and Martins (2014) for this part of questionnaire.

D. Fourth Part (classroom organization)
The forth part allows us to assess how often teachers use specific classroom organization procedures to develop writing activities. It consists of 20 items evaluated on a 6-point scale ranging from 1 (never) to 6 (always). Six items which were related to the writing in pairs or small groups had the internal consistency of .88. Seven items related to the writing individually had the reliability of .73. In addition, seven items related to the whole classroom writing had the internal consistency of .80, as reported by Gaitas and Martins (2014).
Teachers expressed how often they use each procedure. For example, for writing in pairs or small groups, one item indicates, "For text revision purposes, students produce works with peers or small groups". Similarly, for the individual writing, one item indicates, "Organizes and plans written work in such a way as to teach how to write individually". Or, for identifying teachers' beliefs in whole classroom writing, an item states, "With the participation of students, collectively revises their texts to improve their productions".

Procedure
To make it more convenient for the respondents, the questionnaire (Gaitas & Martins, 2014) was changed into the digital type in the net and sent to 250 Iranian university teachers via emails. Their emails were obtained by sending request letters to the officials of different universities and explaining the objectives of the study. Because the participants were teachers at the university level, the researchers used the same English version of the questionnaire but changed it to the digital format. Out of these teachers, 150 filled out the questionnaires and returned them to the researchers. After checking, those questionnaires that were not completed fully were removed and the remaining 120 questionnaires were considered in the data analysis. Because the questionnaire was converted to a digital format and it was sent through the e-mail to the teachers, they could answer it if they liked and there was not any obligation for completing the questionnaire. In this case, the researchers provided some guidelines to the teachers on the aim of the study and how they can respond to the questions and that their responses will be considered as confidential.

Data analysis
The collected data were entered into the SPSS 20 for further statistical analysis. Descriptive statistics including mean, standard deviation, frequency and percentage were estimated in order to answer the research questions descriptively. Because the independent variables (code-based and meaning-based beliefs) and the dependent variables (kinds of writing activities and class organization) are considered continuous variables in the study, Pearson Product-Moment correlation was used as an appropriate test to answer the research questions inferentially. For conducting the Pearson correlation, the related assumptions of normality (through Skewness, Q-Q normality plot, and Kolmogorov-Smirnov test) and linearity (through scatter plot) were checked and the results showed no violation of these assumptions, which legitimized the use of this analysis.

The relationship between teachers' meaning-based and code-based beliefs and writing activities
In order to answer the first research question and find the relationship between teachers' beliefs (meaning-based and code-based) and writing activities including: process writing activities (planning texts and revising texts); decoding writing activities (copying and dictation); autonomous writing activities (texts, everyday reports, informational texts, messages, recipes/instructions/rules,  stories, descriptive reports and wordlists); writing as a product activity (themed compositions and free compositions); and worksheets (spelling and grammar), Pearson Product-Moment correlation was carried out. Table 2 shows the descriptive statistics such as mean (M) and standard deviation (SD) of the selected items (independent variables). Table 3 shows the descriptive statistics such as mean (M) and standard deviation (SD) of the selected items (dependent variables). Table 4 indicates the main results of Pearson correlation between the variables of the study concerning the beliefs of the teachers and the kinds of writing activities they prefer.
The results, as shown in Table 3, indicated that there was a significant and positive correlation between meaning-based beliefs of teachers and writing activities such as decoding, spelling and grammar. While, this is the only significant relationship between code-based beliefs and writing activities came out to be the product activity. In other words, the teachers who had meaningbased beliefs preferred decoding activities such as copying and dictation as well as worksheet activities for spelling and grammar; whereas, those who had code-based beliefs mainly preferred product types of activities such as themed compositions and free composition. However, compared to the Cohen's (1988) criteria (r = .1-.29, small; r = .3-.49, medium; r = .5 and above, large), the effect sizes of these results (r decoding = .24; r spelling = .26; r product = .28) are small.
The other activities were not significantly and highly preferred by the teachers with both kinds of beliefs, though the results showed direct or positive relationships for all except for the process activity with the negative and non-significant correlation with code-based beliefs of the teachers. This shows that teachers who had code-based beliefs in writing instruction did not pay more attention to the process writing activities, which are mainly related to the meaning-based beliefs. In general, the response to the first research question is positive and, therefore, the related null hypothesis is rejected.
The results revealed that the beliefs and practice of the teachers do not correspond. Logically those who have meaning-based beliefs have a tendency to use process writing activities and those who have code-based beliefs tend to apply product types of activities (Gaitas & Martins, 2014), which is the reverse of our results. Even the autonomous activities are not preferred by the teachers who have meaning-based beliefs. They mainly use worksheet activities practicing spelling and grammar with the students and ask them to work on copying sample texts and writing dictation.

The relationship between teachers' meaning-based and code-based beliefs and classroom organization
In order to answer the second research question and find the teachers' meaning-based and codebased beliefs and their preferred use of classroom organization (group work, whole class and individual writing), Pearson Product-Moment correlation was used. Table 5 shows the descriptive statistics such as mean (M) and standard deviation (SD) of the selected items (independent variables). Table 6 shows the descriptive statistics such as mean (M) and standard deviation (SD) of the selected items (dependent variables).
The main results of Pearson correlation are demonstrated in Table 7 for the teachers with codebased and meaning-based beliefs and their preferred classroom organization.
As Table 7 reflects, there were positive and significant relationships between meaning-based beliefs of the teachers and the group work and whole-class as ways of organizing the class for writing activities. There were also positive and significant correlations between code-based beliefs and whole-class and individual writings as ways of class organization. The results indicated a small and large effect sizes for the teachers with the meaning-based beliefs in the use of group-work (r = .26) and whole-class activities (r = . 65), respectively. The effect sizes for the teachers with the code-based beliefs preferring the whole-class and individual writing activities were large (r = .58) and medium (r = .31), respectively. Based on the results, the answer to the second research question is positive and, thus, the related null hypothesis is rejected.

Discussion and conclusion
The purpose of this study was to find the relationship between EFL teachers code-based and meaning-based beliefs and writing activities (product, process, autonomous, decoding, and spelling activities) and class organization (group work, pair work, and individual activity).
In our study, positive and significant correlations with small effect sizes were found between meaning-based beliefs of teachers and decoding (copying and dictation) and spelling and grammar worksheet activities, which is the reverse of Gaitas and Martins (2014) findings. In the study of Gaitas and Martins, positive and significant correlations with small to medium effect sizes were found between meaning-based beliefs and autonomous activities (texts, everyday reports, informational texts, messages, recipes/instructions/rules, stories, descriptive reports and wordlists) and process activities (planning and revising texts). In contrast, negative and significant relationships with small effect sizes were found between meaning-based beliefs and decoding (copying and dictation) activities. Their findings reflect a logical line of relationship and correspondence between their teachers' beliefs and practices; while, the results of our study indicate a mismatch between our teachers beliefs and practices. Our study is similar to Gaitas and Martins' in that both   investigated the beliefs of teachers with regard to their writing activities and class organization. The difference is that they used primary grade school teachers, while, we recruited university teachers. The mismatches between our finding and results of Gaitas and Martins can be explaining by consideration of the levels that the researches have been done. At university level, teachers may not pay attention to the specific methods of teaching writing and they have more freedom to select the way of teaching writing and choosing different kinds of writing activities. Moreover, Birjandi and Malmir (2009) stated that despite the developments in the teaching of writing in different EFL contexts, the traditional approach, i.e. "product approach" is still used in teaching writing to learners in Iranian universities and colleges. This may be another reason that even teachers with meaning-based beliefs teach writing by focusing on product-based writing activities.
Our results also indicated that the correlation between code-based beliefs and writing activities such as product was significant and positive with a small effect size and correlations between codebased beliefs and writing activities such as autonomous and spelling and grammar activities were non-significant but positive. However, the correlation between code-based beliefs and process activity was negative. In contrast to our study, Gaitas and Martins (2014) study showed positive and significant correlations with medium effect size between code-based beliefs and decoding and product (theme composition) activities and negative and significant correlations with small to medium effect sizes with some of the autonomous and process activities. Considering these results, the matches are between beliefs and practices in Gaitas and Martins' study, which is comparable to our results for the correlation between code-based beliefs of teachers and product writing activities.
It seems that teachers with code-based beliefs try to focus on skill-based writing activities more than teachers with meaning-based beliefs, because they did not emphasize the process activities, which need more creativity and writing based on successive stages of planning, writing, revising, and editing (Weigle, 2014). These results verify the findings of Graham et al. (2002) and Cutler and Graham (2008) in that their teachers emphasized mechanics of writing and grammar and spelling as well as capitalization, punctuation skills, handwriting and sentence construction skills.
Logically there should not be a positive correlation between meaning-based beliefs and activities related to spelling, grammar and writing mechanics, but some studies on writing instruction have found that teachers in teaching writing combine two approaches of meaning-based and code or skillbased ones and have a multidimensional view in teaching writing (Gaitas & Martins, 2014;Tagle et al., 2017;Tolchinsky et al., 2012). Some studies, similar to our study with regard to the meaning-based beliefs, have indicated the mismatch of beliefs of teachers and actual practice in writing instruction (Ferede et al., 2012;Khanalizadeh & Allami, 2012;Shahvand & Rezvani, 2016). In spite of this result, teaching writing has encountered opposing views toward using product approach (Ferede et al., 2012;Matsuda & Silva, 2010), which is based on form-focused teaching principles that only considers the correct final product. These views suggest that process approach has more advantages and develops the skill of writing in learners through successive stages of planning, writing, revising, and editing, which is a recursive process rather than being a linear one, and helps the learners to acquire different strategies while correcting their errors (Manchón, de Larios, & Murphy, 2009;Weigle, 2014). Considering these perspectives, in our study, both types of teachers who have code-based and meaning-based beliefs in writing instruction do not significantly have a tendency to use process approach. However, the results showed its positive relation with the meaning-based and negative relation with the code-based beliefs of the teachers, which reveals the matches of their beliefs with their actual practice in the use and not use of the process approach.
Some scholars consider both the product and process approaches as complementary (Badger & White, 2000;Kim & Kim, 2005) because one approach may not suit all educational levels and settings (Kumaravadivelu, 2003). It has also been proposed that the use of process approach in teaching writing to the students with low level of language competency would be problematic because it requires learners to use their autonomy in active production of writing (Hasan & Akhand, 2010;Ho, 2006). As the literature showed, most of the studies in the area of belief and practice in writing have been done on the primary level students (Cutler & Graham, 2008;Ferede et al., 2012;Gaitas & Martins, 2014;Graham et al., 2002;Poulson et al., 2001;Tolchinsky et al., 2012). Very few studies have been conducted in the university (Tagle et al., 2017) and language institutes (Khanalizadeh & Allami, 2012;Shahvand & Rezvani, 2016). A majority of the teachers in these settings believed in the use of eclectic approach to teaching writing. This view may not be applicable to our study with the university teachers teaching writing to the intermediate level students in the EFL setting of Iran. Although the results showed that teachers with meaning-based and code-based beliefs, to some extent, use a variety of writing activities but they are not all significant and the size of the effect for the significant ones is small.
To answer the question about the teachers' meaning-based and code-based beliefs and classroom organization (group work, whole class, and individual writing) as ways of doing writing activities, the results indicated that the correlations between meaning-based writing beliefs and whole class activity and group work were significantly positive. In contrast, the correlation between the meaning-based writing beliefs and individual writing was not significant. It seems that teachers who have meaning-based beliefs ask their students to do their writing activities in pairs or in groups and they do not ask them to do their writing activities individually.
The findings also showed that the correlation between code-based belief in teaching writing and group work was positive but not significant. It seems that teachers who emphasize code-based writing prefer not to organize their classes for group work. On the contrary, correlations between code-based beliefs and individual and whole class writing were significantly positive. The finding of our study is in line with the results of Gaitas and Martins (2014) who found that code-based beliefs are significantly associated with individual writing and meaning-based beliefs are significantly correlated with writing in pairs or small groups and whole-class writing. Likewise, there is a correspondence between our results and Tolchinsky et al.'s (2012) in that the teachers who had code-based beliefs asked their students to do writing activities individually, and those who had meaning-based beliefs requested their students to do pair, group, and whole-class writing activities. Moreover, the findings of Graham et al. (2002) and Poulson et al. (2001) are compatible with our study findings; both of the studies found a direct correlation between practice and beliefs, i.e. between meaning-based beliefs and actual practices in the classroom. Their students shared writing with peers and helped each other.
The results showed that both types of teachers with meaning-based and code-based beliefs consider group work as useful way of doing activities and use it in their actual practices in the classroom. In this regard, Rivers (1984) indicated the important role of teachers in motivating learners to do writing activities in group work since it provides some opportunities to work together and learn from each other how to interpret content and conceptualize ideas.
This study have some implication for those who are involved in education especially university and school policy and decision-makers to provide conditions for teachers to put their beliefs into practice. Moreover, this study can be useful for teachers (especially university teachers as a generalization) to consider their beliefs and match them with their practices and use the activities that are more beneficial in their classes and organize their classes in the ways that stimulate the learners' interest in writing. Some researchers (e.g. Brinkley, 1993;Kamman, 1990;Pajares, 1993) believe that the success of a teacher in using writing-to-learn activities in the classroom depends on his beliefs and attitudes towards writing and his capability in developing instructional activities. According to Mehrpour and Moghaddam (2018), if matches and mismatches of teachers' beliefs and practices are found, in-service and prospective teachers as well as teacher trainers can comprehend each other's viewpoints and as a result can cooperate to unite the beliefs and practices of teachers.
This study examined the beliefs and practices of university teachers with regard to the writing activities and classroom organization. Further study can be done with the high school teachers to compare their opinions with those of university teachers. In this study, teachers' age, teaching experience, gender, and level of education were not considered as variables to be studied; future study can focus on these factors and compare teachers' beliefs and practices in terms of age, less and more experienced teachers, male and female teachers, and those with MA, and PhD degrees teaching writing to the undergraduate or postgraduate university students. This may provide more information to the field of writing and by finding the matches and mismatches of teachers' beliefs and practices in selecting writing activities and class organization; we may find the ways of making writing instruction motivating and stimulating learners to improve their writing quality.