Examination of Relationships between Factors Affecting on Oral Participation of ELT Students and Language Development : A Structural Equation Modeling Approach

Class participation is considered as a way that accordingly, the students appeared actively into the educational process and to help in strengthening our teaching and bringing liveliness to the classroom. Oral participation (OP) is important for students of English Language Training (ELT). This study attempts to determine which factors students find most influential in their oral participation in a foreign language class and its relations with English language development (ELD). Structural equation modeling using LISREL software was used to analyze data. According to the derived constructs and the evaluation criteria for goodness-of-fit, the results approved the validity of the projected construct. The interpretation of the results obtained from SEM and the results of hypothesis testing showed that there are significant relationships between factors affecting on oral participation and also relationships between oral participation and language development. SEM results show that final model based on ELT have proved that ELD was controlled with OP by 65%. Therefore, proposed model of this research can increase the success of ELD studies in L2. Therefore final model has proved that ELD was controlled by educational factors (EF) more than social factors (SOF) and student factors (SF). The structure of the general model presented should be applicable to students of ELT and second language learning environment. Findings suggest this case study fits the unique criteria of a ‘second language’ learner.


Introduction
The advantages of participation in the class discussions have been researched extremely during the last years (Mustapha et al., 2010).One of the scarce cases that create roughly homogeneous consensus in the universities is student of OP (Cole, 1996;Finn & Voelkl, 1993;Stanton-Salazar, 1997).Students' participation in the classroom exists inside a hierarchy anticipated upon students' increasing interaction with school (Patchen, 2006).Also, active classroom participation has a crucial role in the success of learning and students' personal expansion in the future (Tatar, 2005).Students, who are actively involved, reported higher satisfaction and higher persistence rates (Astin, 1999).
Exploring classroom participation from students' perspective is important because it provides a firsthand account and insight into their feelings and perceptions.The students' perceptions are their own realities in experiencing classroom participation.The different requirements to be made between evaluation of participation in class dialogues and evaluation of performance in a structured task such as a seminar presentation.Academic success in a foreign language classroom is largely dependent on OP in the class.Motivating students to participate orally is a struggle for many foreign language teachers (Gardner, Masgoret, Tennant, & Mihic, 2004).
There have been a variety of reasons identified in prior studies as having influences in encouraging or discouraging students' OP.Factors specific to the students like the role of gender in a complex array of everyday social practices (Wenger, 1998) or gender differences (Howe, 1997); particularly ethnic (Patchen, 2006); age, students' willingness to talk and motivation to talk (Peng, 2006) , course level, student preparation and student emotions like confidence or fear, pedagogical and education like faculty authority, class size and classroom instructor communication variables and also In a foreign language classroom, relations between the teacher' and the students' perspectives are a particularly important variable that affects oral participation, a vital element in academic achievement (Consolo, 2006) or the same teacher-learner interaction (TLI) (Tuan & Nhu, 2010;Stewart, 2007), the students' English proficiency (EF), differences in values of social-cultural and training performances between two cultures and socio-cultural factors (SO-CUF) increase the learning opportunities in a second language learning environment (Anton, 1999;Mitchell & Myles, 2004;Lantolf, 2006), students differences together with teaching space were correlated, and each factor separately and together affected the students' OP in class discussions and students skilled in daily life, gives worthwhile viewpoints into understanding the mutual between students (Nomnian, 2007).
Studies have confirmed that foreign language anxiety was experienced by most of students in language learning, negatively impacted their practices in that language (Liu, 2007), and also environmental factors including lecturer traits, classmate traits, engaging class content and conducive physical settings (Mustapha et al., 2010); pair work and group work as learner-learner interaction (LLI) pattern can provide language models for the creation of interact with each other (Erten, 2000) and facilitate language development.Language classrooms can be seen as sociolinguistic environments (Cazden, 1988) and discourse communities (Hall & Verplaetse, 2000) in which interaction is believed to contribute to learners' language development.Therefore, language is as a wonderful present that enables people to make new ways of speaking with others about various issues.Language is a rule-governed, meaningful communication system (Honig, 2007).
It should be mentioned that oral language is the main medium of contact for young students, even when their literacy skills is shaped, it is most significant linkage with environment and life place (Monsalve1 & Correal, 2006) and also one of the most drastic issues that ELT students confront in their study is weakness to contact and handle English to participation in class or to classmates in university.Despite the importance often assigned to participation in classroom, it has been repeatedly reported that most students remain passive in class (Mustapha et al., 2010).Taken all of the preceding discussion into account, the importance of oral participation in promoting students' second language acquisition seems to be crucial.
The purposes of the study are to explore the ELT students' perceptions regarding OP and discover relation between OP in class and ELD that in conjunction with this subject, the research questions to be expressed: Is there a relationship between OP in classroom and ELD of students?
Will EF have a considerable effect on OP in class?
Will SOF have a significant impact on oral participation in class?
Will SF have a significant impact on oral participation in class?Therefore, in current study, we examine factors influencing on OP in between student ELT and its relations with LD to finally show in a final model the associations among OP and influential factors on it with ELD by SEM.

Theoretical Background and Development of the Conceptual Model
In this part, theoretical views related to our research topic will be discussed.One conceptual model in a systematic approach can be represented of a system that is made to deal special questions.The conceptual model can be improved by adding levels of predictability and understanding what is already acquired by the earlier research, that is to say, the realization of relationships and correlations among variables has already been validated and are already known.Investigations conducted by researchers such as Sibat (2005) and Jacob, Huui and Ing (2006) show that the use of English language skills is regarded as essential basis in hiring the graduates.Considering that the connection between students, class participation and their academic success is indisputable (Wudong, 1994).Studies have indicated that activities of associated with class participation is very important in beneficial learning (Gomez, Arai, & Lowe, 1995;Tsou, 2005).
Ferguson-Hessler and de Jong have proved that students participate in class actively, attitude to have academic success compared to students who are passive in class.In general, OP as observable behavior (Ellis, 1999;Spada, 1986) has a special importance, in confirmation of that, from point of Swain (1993) language manufacturing provides a meaningful opportunity to exercise of one's linguistic sources that leads to expressiveness and following generate and using the language as frequently as plausible.
Class participation can be studied in two dimensions consist of oral/non-oral participation.In oral participation, students speak in class, then answer and ask questions, make comments, and taking part in discussions (Lee, 2005).Students who do not participate in the ways mentioned above are often considered to be passive and are generally penalized when participation is graded (Larsen-Freeman, 1986).
Successful second language acquisition for students is based on to obtain feedback on how to use a second language, thereby taking advantage of teacher's and peers' to improve their language through the reform of language errors and oral feedback, learners can modify and improve their oral production.Likewise, with regard to error correction, learners can incorporate new language elements of a correct response to the teacher.In the classroom, the teacher often asks questions to learners and learners answer the questions and vice versa; or the teacher participates in learning activities (Tuan & Nhu, 2010).A recent study on Chinese EFL (English as a Foreign Language) students found that students were least likely to participate in oral activities in class that required them to respond to teachers' questioning.They were much more likely to participate if it involved working in a pair with a peer (Liu & Jackson, 2009).According to Chaudron (1988), teacher talk takes up the largest proportion of classroom talk.It represents approximately two-thirds of the discourse in classroom.
The findings of a study of teacher-student interaction conducted by Musumeci (1996) showed that the teacher talk time occupies about 66% or 72%.Having a supportive relationship with the teacher influences the students' sense of belonging thereby affecting student effort, achievement, self-efficacy and even long-term goals (Walker & Greene, 2009); (Anderman, 2003).Other type of relation is learner-learner interaction that occurs among learners.In this form of interaction, the teacher plays a role as a monitor and learners are the main participants (Gillies, 2006).Therefore, it can be inferred that cooperate exercise should make easy language development.Harmer (2001) proposed that couple practice increases the amount of talking time available to every learner in classroom.
Various studies pointed to limited EP as a factor that limits students' OP in foreign language class, specifically an insecurity and self-consciousness about a low ability to speak the target language in front of one's peers (Lee, 2009).In classrooms where there is a large cluster of students with behavior problems, such that the classroom environment is not viewed as safe, whether physically, socially or emotionally, there is a negative effect on student motivation (Walker & Greene, 2009;Koth, Bradshaw, & Leaf, 2008).Also, researchers have proven that learners tendency to participate in class arguments depend on many factors (Tatar, 2005).Fessinger (1995) quoted by Gomez, Arai and Lowe (1995) categorizes the factors into three major traits consist of 1) class traits, 2) student traits and 3) teacher traits.Class traits include interaction norms and emotional climate.Student traits are divided into three components: confidence, preparation, and intimidation.Fear of appearing unintelligent to peers or instructor, lack of organization skills, communication apprehension, fear of offending, and intimidation are all considered as confidence trait (Gomez, Arai & Lowe, 1995).The third traits-teacher traits, meanwhile, are linked to approachability and supportiveness of the teacher as well as whether the teacher welcomes discussion.Also, Fawzia (2002) divides the factors influencing students' OP into three broad categories; learner/student factors (SF), social factors (SOF), and educational/pedagogical (EF/PF).Learners' tendencies, views, learning styles, their background and personal affective factors are examples of student factors whereas social factors include the gender of students and nature community feeling in a group.Factors relates to educational/pedagogical index consist of the lecturer, the course of training, the subject and nature of query.In this factor, OP is a part of evaluation, teacher's encouragement, class size, peer help and support, and also the way the lesson is conducted.The most of research have demonstrated that the approach a teacher takes in the classroom affects students' participation.Studies has displayed that learners are more willing to make easy and thus contribute more in group dialogues compared to in whole-class discussions.It is found that student-centered class encourages more participation as compared to teacher centered classroom (Barry, King, & Burke, 2000).In study Persaud and Salter (2003) have shown that female learners are reluctant to participate for fear of being scolded and criticized by their teacher and/or peers.In addition, when student participates, she risks treated with disrespect in front of the class.
In next classified by Liu's (2001), socio-cultural factors concern with the participants' values, beliefs and moral judgments that are strongly affected by their cultural backgrounds and educational experiences in their home countries.For example, Japanese cultural values of self-restraint and respect for seniority restrain Japanese students from speaking freely (Shimuzu, 2006) and also according to Lim, Korean students are used to viewing teachers as authority figures in class.Therefore, leader a teacher in class and guide with talk can be helpful for students while listen and talk notes.Korean students believe that a teacher is a person who knows everything.They tend to value quietness, be less opinioned and believe that they from elders and wiser persons who are usually represented by a teacher in class (Warschauer, 1996).It should be noted that Krashen (1982) hypothesized that anxiety contributed negatively to an "affective filter", which made an individual less responsive to language input and language anxiety is more associated with oral participation.Eventually mentioned factors imply to the students' linguistic capabilities and communicative competence.For further understanding of the issue, there are many students that are reluctant to participate in classroom discussions because of their poor command of the language, in compared with it, there are students who have good speaking abilities tend to participate in class discussions with ease. Figure 1.Conceptual model-SEM With appropriate theoretical support and done studies, the main objective of conceptual model is to determine to what extent variation of oral participation is responsible for the differences in language development and also to explore general model for improving language.The conceptual framework presented as in figure 1 is drawn from the SEM approach.In analytical model, each latent variable includes a number of constructs.EF is represented by the following three constructs: TE, TLI, and LLI.Each LISREL model is comprised of two sub-model measurement and structural model.In the current study the effects of three latent variables for OP, ''EF'', ''SOF'' and ''SF'' were assessed and also effect of one latent variable for ELD, ''OP'' was assessed.Each of these latent variables can be linked with observed variables that SEM is based on structural relationships between observed variables and hypothesized relations, as in traditional path analysis (Markland, 2006).Accordance with the model, four hypotheses can be constructed: H 1 : There is a relationship between OP and ELD in between ELT students.
H 2 : There is a relationship between EF and OP in between ELT students.H 3 : There is a relationship between SOF and OP in between ELT students.H 4 : There is a relationship between SF and OP in between ELT students.

Research Methodology
A self-report survey method was used in this study and the research variables have been studied in it through questionnaires and participants of this study, who were 85 of M.A students of ELT at Islamic Azad Universities (Ahar & Ardabil branch).The sample was drawn by the Cochran formula and a stratified random sampling procedure was used.A questionnaire was developed to collect related data, as well.Face validity was created by a panel of experts comprising of university professors and experts will approve.In current study, Cronbach's coefficient alpha was used to evaluate internal consistency reliability of the scale by using SPSS software, by which reliability for the overall instrument was estimated at 0.85.
In order to form relationship between factors affecting on OP and also relationship between OP and ELD; (I) first, a conceptual model of causal relationships was created between the variables of sets, (II) then, SEM was performed by LISREL 8.5 (Joreskog & Sorbom, 1999).EF, SOF and SF were observed as predictors to OP and then OP was viewed as predictors to explain ELD.The structural relationships between the exogenous variables and endogenous variables (Vieira, 2011), by SEM analysis were explored.In SEM, CFA (confirmatory factor analysis) is performed and the causal relationships are assessed among latent variables (Anderson & Gerbing, 1988) and the conceptual model is examined by using LISREL 8.50 (Joreskog & Sorbom, 1999)

Data Analysis
The quantitative phase of the study involved a combination of descriptive and inferential statistics.Descriptive statistics including CV, and KS values were used to show the distributions and numerical characteristics of the Educational, social and student factors, OP and ELD.Table 2 displays the results of the main statistical descriptors of the data sets analyzed.The variability in the listening in ELD is less than others, as it is characterized by CV = 0.22.But the other measured components varied to a greater extent, as revealed by the coefficient of variation (CV) (

Hypotheses Testing by SEM and Discussion
In SEM, hypotheses are assessed with regard to means, variances and covariance of observed data in terms of traits characterized by an underlying model (Kaplan, 2000).In this type of analyze, it is possible that is provided simultaneous evaluation of multiple dependent variables and both direct and indirect impacts of one variable on another (Wright, 1934).SEM permits the estimation of latent variables instead of only measured variables and thereby omits random error.Moreover, SEM has the privilege of efficient indices of general fit of hypothesized models to the data.In current study, after studying research hypotheses and evaluating the condition of initial structural model, the most important variables that could be useful for SEM, that is a total of five latent variables, including two internal latent variables as the OP and ELD, shown respectively by Figure 2 shows that four hypotheses indicate causal relationships between (1) EF, (2) SOF and (3) SF with OP, and also (4) OP with ELD, therefore research hypotheses are approved.In fact, EF, SOF and SF have significant positive effects on OP and then OP has significant positive effect on ELD.According to the estimation of standardized coefficients of research structural model and the obtained significance level ( 05 .0 


), the effects of EF, SOF and SF have been assessed on OP with positive and direct coefficient of 071, 063 and 0.51, respectively and the impact of OP has been assessed on ELD with positive and direct coefficient of 0.65.The fit of the model was assessed with various measures (Bentler, Model fit was evaluated using a variety of fit measures outlined by Bentler (1990).The results displayed that the hypothesized model had acceptable fit indices.Table 3 shows that the theoretical model adequately constructed the observed covariance of the data with all of the mentioned above indicators of good model fit the minimum specifications.The X 2 for this model was 214.02 (d.f = 94) and the X 2 /d.f ratio of 2.28 is below the threshold of 3.0 as prescribed by Wheaton et al. (1977).All of the fit indexes (NFI, NNFI, CFI, & GFI) are above the 0.90 minimum prescribed by Bentler (1990) for well fitting models.The RMSEA is one of the most criteria in covariance structure modeling.It takes into account the error of approximation and ask the question "how well would the model, with unknown but optimally chosen parameter values, fit the covariance matrix if it were available?"(Brwne & Cudeck, 1993), values less than 0.08 display good fit.Turning to Table 3, the RMSEA values for the models are 0.05 for WYC model; thus, it can be concluded that the models fit the data well by Raykov and Marcoulides (2006).

Conclusion
The research backgrounds and offers display the values of participation in the college classes.ELD changes with OP is positively correlated with its.SEM provided an adequate explanation of the variance of ELD in the taken samples.
In fact, the EF was the most important parameters in determining OP variability.Based on data obtained, OP can be foreseen with acceptable precision.The final model reveals a promising result, which can be useful for forecasting situations with ELD.In addition, the findings show that EF is better indicator of OP than did SOF and SF.The discussion and estimation of different modeling forms has therefore gained much attention in teaching and ELT literature.Therefore final model has proved that ELD was controlled by educational factors (EF) more than social factors (SOF) and student factors (SF).The structure of the general model presented should be applicable to students of ELT and second language learning environment.Findings suggest this case study fits the unique criteria of a 'second language' learner.
Table 2).Meanwhile, factors distributions of the study area are normal as demonstrated by the Kolmogorov-Smirnov (KS) Test (p-values KS > 0.05).

Table 3 .
Results of structural model tests-goodness of fit summary