English Teachers' Self-Efficacy Beliefs and Students Learning Approaches: the Role of Classroom Structure Perception

This study aimed at investigating the direct and indirect effects that teachers' self-efficacy beliefs exert on students' learning approaches via affecting their perceptions of classroom structure. The sample included 40 English teachers and 240 first-grade female students from high schools in Iran. To collect data, three questionnaires were applied: (a) Self-Efficacy Beliefs Questionnaire was answered by the teachers, and (b) Study Process Questionnaire and Survey of Classroom Structure Goals were given to the students. Path analysis revealed that, via Motivating Tasks, Mastery Evaluation, and Autonomy Support, teachers' self-efficacy beliefs had an indirect and positive effect on students' deep learning approaches but an indirect and negative effect on their surface learning approaches. Also, teachers' self-efficacy beliefs affected students' deep learning approaches directly and positively but their surface learning approaches directly and negatively. Moreover, it was found that Motivating Tasks, Mastery Evaluation, and Autonomy Support had direct and positive effects on students' deep learning approaches but direct and negative effects on their surface learning approaches. All the relationships between model variables were statistically significant. The results tend to verify that students' perception of classroom structure plays a mediating role between teachers' self-efficacy beliefs and students' learning approaches.

he term learning approaches has been widely used since 1970. According to Biggs (2003), learning approaches are methods that students use when they do learning tasks with regard to learning results. He makes a distinction between deep and surface learning approaches. Learners with deep learning approaches focus on understanding, associating, and relating the ideas or concepts in a learning task. When such learners study, they put newly learnt materials into more comprehensive and coherent conceptual frameworks or structures. Learners with surface approaches, on the other hand, tend to memorize facts and reproduce them later, without any focus on the coherence and logic existing within them or any attempt to create or discover new relations in what they have learnt (Biggs, 2003). There are almost three views about whether the approaches which students adopt are stable or not. Some researchers hold that they are essentially stable in all learning situations (Eley, 1992). According to the second group, learning approaches are flexible depending on learning environments and contexts (Entwistle & Peterson, 2004). Finally, there are some others who argue that learning approaches are both stable and variable (Peterson, Rayner, & Armstrong, 2009). Curry (2002) states a variety of constructs that researchers have turned to their in study students' learning (e.g., instructional preferences, learning style, and cognitive style) can be conceptualized like the layers of an onion. Learning strategies are the outer layers of the onion, implying that they are most influenced by the environment. This also implies that compared to other constructs, learning strategies are most adaptable to change.
A key question for researchers is to understand how students learning can change in particular contexts. The answer to this question would let them make generalizations of such learning experiences and better understand students' learning strategies. That different students employ different learning methods is to state the clear. It is already known that some students are highly motivated and eager to learn and understand whereas some others seek to only pass the course through minimal learning possible (Biggs, 2003). It is also agreed that some teachers foster their students' interest to learn while others do not (Sadlo & Richardson, 2003). Students' learning T  Learning Approaches motives and their perceptions of learning environment are just two of the factors that affect their learning (Prosser & Trigwell, 1999). Biggs (2003) notes that students' learning is affected by what he calls a complex ecosystem which brings about changes in their learning process. The ecosystem consists of several variables, one of which is learning context and environment which plays an important role in learning (Biggs, 2003).
Research studies that focus on classroom and school-level environments have produced promising findings leading to an enhancement of the learning and teaching process. According to Fraser (1998), learning environment refers to psychological, pedagogical, and social context in which learning takes place and which influences students' achievement and attitudes. In their learning environment studies, several researchers (e.g., den Brok, Brekelmans, & Wubbels, 2004) have demonstrated that teachers' and students' perceptions of the classroom environment influence cognitive and affective outcomes. They have also shown that there is a strong relation between students' outcome and their perceptions about their learning environment.
Research findings suggest that students' positive perceptions of the learning environment can affect their cognitive outcomes (Wubbles & Brekelmans, 2005), classroom attitudes (Kim, Fisher, & Fraser, 2000), and satisfaction (Strayer, 2012). How students perceive the classroom structure is highly significant since such perceptions affect their motivation and performance considerably. Blackburn (1998) discusses three measures of classroom perceptions: motivating tasks, autonomy support, and mastery evaluation. The first measure deals with the extent to which students find classroom tasks to be meaningful, relevant, and interesting to them. The second is concerned with whether students think the teacher supports their autonomy through providing opportunities to choose and by encouraging responsibility for self-regulated learning. Finally, the third measure establishes the extent to which students find that the evaluation and recognition practices are fair, focus on learning, and de-emphasize social comparisons and competition.
It is believed that students' perceptions of learning environment influence their learning approaches. Researchers such as Ramsden (1992) argue that students' perceptions of learning environment are more important than the learning environment itself since such perceptions determine their learning approaches. He believes that to change students' learning approaches we do not try to change the learners rather we seek to change their experiences or perceptions of their learning environment. Learning environments oriented to problem-solving (Mergendoller, Maxwell, & Bellisimo, 2000) encourage deep approaches. The students, however, are likely to adopt surface learning approaches when they perceive that the assessment tasks ask no more than reproducing the learnt materials (Entwistle & Ramsden, 1983) In other words, students' perception of the assessment procedure affects their learning approaches too. Case and Gunstone (2003) demonstrated that when students perceived a supporting role from their teachers, they adopted deep approaches. Furthermore, students' perception of the assessment goals seemed to play a role; when they believed assessment is intended to help them learn better, they turned to deep approaches again.
It is important to note that classroom structure is based on teachers' goals and values. Educational theory suggests that teachers themselves are one of the most important determinants of whether a classroom exhibits higher versus lower quality of instruction (Desimone, Smith, & Fris-vold, 2007;Mashburn et al., 2008). There is substantial research evidence that teachers have great potential to affect students' educational outcomes (Anderson, 2004). The teachers' role is not limited to knowledge transmission. It includes teaching learners how to learn and encompasses boosting their confidence, motivating, enhancing self-esteem and organizing an appropriate learning environment (Williams & Burden, 2000).
There is a great emphasis on teachers' behaviors, views, perceptions, beliefs, theories, and motivational levels in education. Teachers' self-efficacy beliefs play a key role in determining how they organize their teaching. The construct of self-efficacy has evolved from Bandura's social cognitive theory. Bandura (1997) defined self-efficacy as the "belief in one's capabilities to organize and execute the courses of action required to produce given attainments" (p.3). It is believed that these perceived capabilities influence behavior (Czerniak & Chiar-elott, 1990) in that when a person holds a belief that his or her behavior can lead to a desired outcome, he or she executes the behavior to achieve that outcome. As the concept of self-Learning Approaches efficacy is applied to teaching and teachers, it is defined as the belief about the role of one's capabilities to bring about desirables changes in students' behaviors and achievements.
There is evidence that teachers' perceptions of their self-efficacy play an important role in students' educational outcomes. Evidences show that there is a relationship between teachers' self-efficacy beliefs and students' achievement and motivation. Teachers' self-efficacy beliefs also affect their teaching activities, commitment, and behaviors. Pajares (1992) found a strong relationship between teachers' educational beliefs and their lesson planning, instructional decisions, classroom practices, and subsequent teaching behaviors. According to Dembo and Gibson (1985), teachers who do not have a strong sense of self-efficacy, such that they do not believe they are capable to affect student performance positively, may not accept responsibility for motivating students or take the necessary steps to do so. Teachers with a low sense of self-efficacy are more likely to attribute difficulties in teaching to student failure and make fewer, more tentative, innovations to ameliorate the difficulties. Goddard and Goddard (2001) concluded that teacher self-efficacy was related to improved planning and organization (Allinder, 1994), studentcentered learning (Czerniak & Schriver, 1994), the use of activity-based learning (Enochs, Scharmann, & Riggs, 1995), and a more humanistic approach to student control (Woolfolk & Hoy, 1990).
In order to determine how teachers' efficacy affects student achievement, Ross (1994) analyzed 88 teacher efficacy studies and concluded that teachers who have a higher sense of efficacy are more likely to: (1) use new approaches and strategies for teaching, (2) use management techniques which enhance and reinforce student autonomy and diminish student control, (3) provide special assistance to low-achieving students, (4) build students' self-perceptions of their academic skills, (5) set achievable goals, and (6) persist if their students fail (cited in Woolfolk, Hoy, & Spero, 2000).
Studying affective characteristics among teachers is, therefore, a promising area of research that has the potential to shed light on what constitutes effective teaching. There is little known about the relationship between teachers' particular characteristics such as self-efficacy beliefs and students' perceptions of classroom structure. Most studies are focused on students' self-efficacy beliefs. For example, in study of Green et al (2004) path analysis was used to test predictions of a model explaining the impact of students perceptions of classroom structures (tasks, autonomy support and mastery and evaluation) on their self-efficacy, perceptions of the instrumentality of class work, and their achievement goals. While in order to determine how teachers' self-efficacy beliefs affects students perceptions of classroom structures and student learning approaches further research is needed.
The proposed model in this study is based on socio-cognitive, constructivist, and learning approaches and earlier related studies (Blackburn, 1998;Pajares, 1992;Green et al., 2004;Yilmaz, 2011). In this model, teachers' self-efficacy beliefs affects students' learning approaches both directly and indirectly-through students' perceptions of classroom structure. The model is an indication of the fact that teachers' self-efficacy beliefs play a key role in building learning environments for the learners. More importantly, it is the learners' perceptions of these environments that lead them to adopt either deep or surface learning approaches. In earlier studies, the direct and indirect effects and relationships of these three variables have been given little attention. More specifically, earlier studies examine the relationships between two variables and how one affects the other. This study, however, attempts to provide a more comprehensive picture through discussing a mediating variable-learners' perceptions of classroom structure-and how these variables affect each other directly as well as indirectly. The model consists of three variables: Teachers' selfefficacy beliefs as the endogenous variable and learners' perceptions of classroom structure and their learning approaches as the endogenous variable.

Purpose of Study
The purpose of the study is to determine how well this theoretical model fits the data from a sample of high school English students in Iran. English  Learning Approaches involves various components (e.g., reading, writing, oral communication, grammar skills, creative expression, etc.) and English classes provide an intriguing context to study variables such as students' perception of classroom structure and English teachers' self-efficacy beliefs and the relationship between them. Therefore, this study is an attempt to examine how teachers' self-efficacy beliefs directly and indirectly influence Iranian high school students' learning approaches in English classes. Broadly speaking, the study and its findings can help with a better understanding of factors affecting students' learning approaches in English classes.

Significance and Justification of the Study
As went before, the relationships between these variables have been investigated in previous studies. But the present study does so in the framework of a tentative model and goes beyond a 'one-to-one' approach to variable investigation. In other words, the complexities and intricacies inherent in classroom realities are reflected more since the role of a mediating variable is highlighted. As a result the study is a step toward filling the gap in prior studies in which the direct and indirect effects of variables on each other is paid little attention.
Besides, the findings of this study would demonstrate if Iranian students' perceptions of classroom structure affect their learning approaches-hence enabling us to examine the relationship in Iranian context. The results would help us better understand what influences Iranian students' learning approaches in English classes.
The model (figure 1) is based on Bandura's social cognitive theory and constructivist view of the learning process is shared by social cognitive theorists (e.g., Schunk & Zimmerman, 1996) and learning approach theorists (e.g., Biggs, 2003).
The three variables in the model are: (1) teacher's self-efficacy as the exogenous variable, (2) students' perceptions of classroom structure, and (3) students' learning approaches as the endogenous variable that also is based on research studies in literature (e.g., Pajares, 1992;Greene et al., 2004;Blackburn, 1998;Yilmaz, 2011).

Research Hypotheses
Hypothesis one: Teacher's self-efficacy has direct and indirect effects on students' learning approaches Hypothesis two: Teacher's self-efficacy has a direct effect on students' perception of classroom structure Hypothesis three: Students' perception of classroom structure affects their learning approaches directly.

English Education in Iran
Nowadays, English plays a pivotal role in knowledge and information transmission globally (Wang, 2008). Therefore, English education enjoys a high status in national curricula of many countries and is seriously discussed by educational policymakers and curricula planners throughout the world. In Iran, several foreign languages such as English, French, Russian, Chinese, and German are formally included in the national curriculum; however, it is the English language that is known as the prime and the only foreign language in practice. Iranian students begin learning English formally when they are twelve or thirteen. Due to general inefficiency of English education in public sector (Mazlum, 2013), private language schools have recently increased in number. In general, due to problems pertaining to textbooks, teacher and student factors, the late start of the course…, English education in Iran's public schools is encountered with challenges and problems reflected in several local studies (Riazi, 2005;Hayati & Mashhadi, 2010;Atai & Mazlum;2012).

Participants
Through random sampling, 40 female English teachers and 240 first-grade high school students were selected from public schools in Yazd, Iran. The participating students were taught by the participating teachers. The population consists of all female English teachers and first-grade students in Yazd city. Yazd has two districts; therefore, equal number of participants was randomly selected from each district for both groups (i.e. teachers and students). The average age of students and teachers was 16 and 29 respectively. Teachers' teaching experience varied from 3 to 14 years.

Instruments
Study Process Questionnaire: This questionnaire can help with the identification of possible problem areas in the way students study. The revised Study Process Questionnaire (SPQ) has been developed by Biggs et al. (2001) for the evaluation of students' learning approaches. It is a 20 item, five-scale Likert questionnaire that is intended to evaluate deep and surface approaches only (while each approach has a motivation and a strategy dimension). For the reliability, Cronbach's Alpha was applied in this study. It turned to be 0.58 for the deep approach and 0.68 for the surface.

Survey of Classroom Goals
Structure: This is used to measure students' perceptions of class structure. Their perceptions reflect their understanding of the learning environment, performance goals (getting a good score or giving the right answer) as well as mastery goals (motivation tasks, mastery evaluation, and autonomy support). Using confirmatory factor analysis, Green et al. (2004) revised the factor structure (loading) of Blackburn's Survey of Classroom Goals Structure (SCGS). Three independent factors (i.e., sub-scales) were identified: (1) Motivation Tasks with 11 items, (2) Autonomy Support with 6 items and, (3) Mastery Evaluation with 11 items. In their study, Cronbach Alpha Coefficients for Motivation Task, Autonomy Support, and Mastery Evaluation were 0.85, 0.65, and 0.80 respectively. In this study, the coefficients turned out to be 0.75 for Motivation Tasks, 0.58 for Autonomy Support, and 0.64 for Mastery Evaluation.
Teacher's Self-efficacy Beliefs Questionnaire: Developed by Schwarzer, Schmitz, and Daytner in 1999, this questionnaire is a 10 item measure that identifies job skills and groups them into four major areas: (a) job accomplishment, (b) skill development on the job, (c) social interaction with students, parents, and colleagues, and (d) coping with job stress. The measure was constructed following Bandura's social cognitive theory. The questionnaire is a four-scale Likert one and includes ten items. The scores range from 10 to 40. For the reliability, Cronbach's Alpha was applied in this study. It turned to be 0.72.
The psychometric properties of these instruments have been investigated in earlier local studies and in Iranian context (Yamini, 2008).
With the official permission of the Organization of Education Office in Yazd, first, several districts were randomly selected followed by the random selection of some high schools. Forty English teachers and 240 students (6  Learning Approaches students for each teacher) were randomly selected from these high schools. The research objectives were made clear to the participants and they answered the questionnaires with consent and individually. Data suggest that there is a positive and significant relationship between teacher's self-efficacy beliefs and motivation tasks (0.14) and deep learning approaches (0.18). A negative and significant relationship is observed between surface approach and motivation tasks (-0.39), autonomy support (-0.23), and mastery evaluation (-0.29). Also, a positive and significant relationship is present between deep learning approaches and motivation tasks (0.17) and mastery evaluation (0.17).

Results
The present study sought to investigate the mediating role of perceptions concerning motivation tasks, autonomy support, and mastery evaluation among self-efficacy beliefs and deep and surface learning approaches. To predict deep and surface learning approaches, path analysis was applied to examine the suggested model. Figure 2 shows the path coefficients of the suggested model.  Figure 2 shows that all paths are significant. Compared with all the other variables of the study, the direct effects of teachers' self-efficacy beliefs on motivation perception and through motivation perception on deep approaches have been more-which is 0.58 for the first and 0.43 for the latter. Teachers' self-efficacy beliefs (i.e., the exogenous variable) affect students' surface and deep learning approaches through motivation perception, mastery evaluation, and autonomy support. The effect procedure Learning Approaches is as follows: self-efficacy beliefs have direct effects on both deep approaches (0.19) and surface approaches (-0.14). They also have an indirect effect on deep approaches (0.32) and surface approaches (-0.23). In Table 2, direct and indirect coefficients, all research variables along with their significance levels are presented.  Table 3 shows the model fit indexes. The model fit is considered to be appropriate provided that 2 is not statistically significant but in larger samples the index is usually significant and therefore is not an appropriate index to fit models. Furthermore, if 2 /df is above 3, it would not lead to an acceptable fit. For AGFI, GFI, and CFI indexes, above 0.90 and for RMSEA, less than 0.06 is an indication of appropriate and acceptable fit. Above 0.80 is an acceptable fit for CFI, GFI, and AGFI indexes and below 0.08 for RMSEA (Hooper et el., 2008). It is observed that model fit indexes, GFI, AGFI, and RMSEA are at appropriate levels and therefore the model fits the data adequately.

Discussion and Conclusion
The purpose of this study was to investigate the direct and indirect effects that English teachers' self-efficacy beliefs exert on students' learning approaches via affecting their perceptions of classroom structure. The results of this study revealed that teachers' self-efficacy beliefs have direct effects on students' both deep and surface approaches, but the effect is positive for the first and negative for the second approach. To explain this finding, it can be argued that belief in perceived capabilities affects behavior as reflected in earlier studies in literature (e.g. Czerniak & Chiar-elott, 1990). Thus, a person who believes he or she is capable of achieving a desired goal or outcome is more likely to follow the necessary behaviors for the attainment of that goal or outcome. Similarly, teachers who have high self-efficacy perceptions take better advantage of classroom time and spend it more effectively, criticize their students less for their incorrect and wrong answers, and guide them to right answers by asking questions. Teachers with low selfefficacy beliefs, however, spend more time on irrelevant activities and employ ineffective techniques and strategies to guide their students (Yilmaz, 2011). Teachers with high self-efficacy perception rely on their students' learning capacity more compared to those with low levels of self-efficacy, and they endeavor to create an effective educational life using a variety of strategies, methods, and techniques in the classroom (Alderman, 1999).
Teachers who do not have high self-efficacy perception (i.e., do not believe they are capable of affecting their students' behaviors positively) do  Learning Approaches not feel responsible for motivating their students (Dembo & Gibson, 1985). Teachers' self-efficacy beliefs lead to an increased perception of learning efficacy in students (Anderson et al., 1988), facilitate their involvement in classroom activities, and increase their efforts to solve problems (Ross et al, 2001). Therefore, the fact that teachers' self-efficacy beliefs have direct and positive effects on students' deep learning approaches-the findings of this study-is rooted in behaviors of a teacher with high self-efficacy. The behaviors, activities, and thoughts of such teachers can influence the learning approach students adopt.
The results of the study also revealed that students' perception of classroom structure (motivating tasks, mastery evaluation, and autonomy support) affects their learning approaches (surface and deep) directly and significantly. To explain this finding, it might be said that students' motivation and goals develop within the broader social and psychological context in which they learn. If students experience threat, anxiety, and discrimination in their learning environment and if their teacher is an unfeeling and demotivated one who has a negative attitude towards teaching and his or her learners, students will adopt surface learning approaches since such a learning environment does not entail in itself the necessary motivational and emotional conditions for the development of deep approaches. This coincides with the findings of Greene et al. (2004). In their study, they found a positive relationship between autonomy support perception and deep strategies and mastery goals.
When students believe that the teacher focuses on mastery in learning or on deep understanding, they tend to develop a similar attitude too. In other words, when students feel that the teacher values and merits competence and awards better performance, they internalize such values (Ryan et al., 1998). Teachers who use more individual assessment (compared to group assessment) and consider their students' errors as a natural part of learning process decrease the effects of social comparisons and fear from failure in their students (Snow & Jackson, 1994). If the assessment goal is social comparison rather than mastery, most students would only try to get the right answer and a higher score. As a result, they would not be interested in understanding concepts but memorizing them.
Task design is regarded as a component of classroom structure perception. The findings of the study suggest that students who view tasks as potentially meaningful and motivating tend to adopt mastery goals. Thus, in line with arguments and suggestions in previous studies (e.g., Green, 2004), teachers are suggested to design and use tasks that have functional values and are interesting. This will motivate students intrinsically partially because doing such tasks is more enjoyable (Boekaerts, 1999). Overlooking the role of valuable, motivating and interesting tasks and too much reliance on textbooks might undermine the importance of students' active learning. This, in turn, might lead students to develop a passive attitude towards learning and adopt surface approaches to learning (Kember & Wong, 2000).
One more finding of this study relates to the mediating effects of classroom structure perception. It was found that teachers' self-efficacy beliefs through classroom structures (motivating tasks, mastery evaluation, and autonomy support) affects students' surface and deep learning approaches. To explain the finding, it can be maintained that teachers' selfefficacy beliefs are not independent from other social and psychological determinants like classroom structure perception that affects performance and motivation. They affect teachers' teaching activities and behaviors (Skaalvik & Skaalvik, 2007). Research findings confirm that teachers' capability in managing classroom and organizing learning are the key factors. Compared to teachers with low levels of self-efficacy perception, teachers with high self-efficacy perception are more likely to be instructionally creative and to use management and teaching methods that support students' autonomy. These teachers assign responsibilities according to learner needs (Jordan, et al., 1993) manage classroom problems (Chacon, 2005) and keep students focused on task (Podell & Soodak, 1993).
Teachers who believe in their capabilities are more likely to employ a model of strategies that reduces negative effects and enhances class expectations which are formed on warm interpersonal relations and academic endeavor (Woolfolk, 1998). Also, teachers with a great sense of self-efficacy tend to be humanistic rather than custodial. In other words, the more efficacious the teacher, the less custodial to control students and the more likely he or she is to support student autonomy and responsibility.  Now that students' motivation and learning behaviors are affected by their perceptions of psychological-social context of the classroom, teaching methods, pedagogical tasks, etc. teachers need to rely on their capabilities in order to provide a satisfactory learning environment and, as a result of this, make their students' perceptions of learning environment positive. All this would enhance their students' learning outcomes because in providing an effective learning environment the key is the teacher's belief in his capability to manage the class and organize learning. If teachers really believe that they can affect their students' learning positively, they would make any attempts to create the required environment. Students find such an environment a positive one and their positive perception of the learning environment affects their learning outcomes positively.