the second ten questions, the mode of three questions was 5, the mode of six questions was 4 and the mode of one question was 3. Therefore the most answers to these questions were 4 which was the sign of disagreement about them.

Considering the pyramid, pie chart, bar chart and frequency tables which were elaborated in chapter 4, the frequency of the means of the third ten questions above 3 more than the frequency of the means under 3, which means the participants who are in fact the teachers mostly agree with the options 4 and 5. It indicates that most of the teachers disagree with the questions 21-30 and these questions are the ones which support the perceptions about Metalinguistic Feedback.

In addition to the results of this study, a number of descriptive studies suggested this type of feedback should be incorporated in the instruction of grammar, so that further elaboration on the types of structures being introduced could enhance the visual saliency and hence more efficient noticing and encoding of information in memory. So these studies also indicated that Metalinguistic Feedback was not very effective in correcting Pronunciation errors as well as Recast.

According to the first research question which was: What are the teachers’ perceptions about the effects of Recast on Elementary learners’ pronunciation? , the first hypothesis was formulated: Teachers don’t have any significant perceptions about the effects of Recast on Elementary learners’ pronunciation. Through investigating the descriptive statistics of the second ten questions and their frequency data, the null hypothesis was rejected. Therefore Teachers have significant perceptions about the effects of Recast on Elementary learners’ pronunciation and according to the statistics analysis they have positive perceptions about it and in this ground they consider it very important.

According to the second research question which was: What are the teachers’ perceptions about the effects of Metalinguistic Feedback on Elementary learners’ pronunciation?, the second hypothesis is formulated : Teachers don’t have any significant perceptions about the effects of Metalinguistic Feedback on Elementary learers’ pronunciation. Through investigating the descriptive statistics of the third ten questions and their frequency data, the null hypothesis was rejected. Therefore Teachers have significant perceptions about the effects of Metalinguistic Feedback on Elementary learners’ pronunciation and according to the statistics analysis they have negative perceptions about it and in this ground they don`t consider it very important.

The Friedman test is a nonparametric alternative to the one-way ANOVA with repeated measures. It means that if the data are normal, ANOVA is used for scores mean comparison but because the second ten questions` data was not normally distributed, the nonparametric alternative to ANOVA which is Friedman test was used. As indicated in the Friedman test the mean rank of the second ten questions was orderly less than the first and the third ten questions, which really means the answers of the second ten questions led to options 1 and 2 and the answers of the third ten questions led to options 4 and 5. In Friedman test, the null hypothesis states that: there is no significant difference among the three groups` scores means. Since the Sig Value was .000 and it was less than .05, the null hypothesis was rejected and there were significant differences among the three groups` scores means

Because the Recast data was not normal, the researcher used a nonparametric test. The Wilcoxon signed-rank is the nonparametric test equivalent to t-test. Regarding the obtained results, with due attention to rank and compare the data, the researcher applied Wilcoxon Signed Ranks Test.

In this test, while Mean2 (mean of the questions 21-30) was more than Mean1 (mean of the questions 11-20), Numbers, Mean Rank and Sum of Ranks reach maximum. In this ground, the positive ranks which indicated Mean2 was more than Mean1 and equaled to 89, were more than negative ranks which indicated Mean2 was less than Mean1 and equaled to 8.If the significance level was equal or less than .05 then the researcher can conclude that the difference between the two scores was statistically significant. In this example the Sig. value was .000 (which really means less than .05). Therefore the researcher can conclude that the two sets of scores were significantly different.

According to the third research question which was: Is there any significant difference between teachers’ perceptions about the effects of Recast and Metalinguistic Feedback on Elementary learners’ pronunciation? ,the third hypothesis was formulated: There is no significant difference between teachers’ perceptions about the effects of Recast and Metalinguistic Feedback on Elementary learners’ pronunciation.

In this test the Sig. value was .000 (which really means less than .05). Therefore the hypothesis was rejected and the researcher concluded that the two sets of scores were significantly different and there was significant difference between teachers’ perceptions about the effects of Recast and Metalinguistic Feedback on Elementary learners’ pronunciation. To sum up, considering all the statistics steps, tables and figures the teachers’ perceptions about the first ten questions which were about the pronunciation error corrections was approximately positive, the teachers’ perceptions about the second ten questions which were Recasts’ effects on pronunciation error correction was mostly positive and the teachers’ perceptions about the third ten questions was mostly negative. It can be inferred that according to the teachers’ perceptions Recast’s effects on pronunciation is very important in encouraging the learners to overcome their pronunciation errors and it gives teachers a chance to step back and think how and when they can correct the learners’ pronunciation errors and how they can help their learners to overcome this kind of error.

5.3. Implications

From theoretical point of view, a successful teacher requires higher skill in using Recast and more knowledge about Metalinguistic Feedback on pronunciation error correction. This study presented Recast and Metalinguistic Feedback as pronunciation error correction feedbacks for language teaching achievement.

From pedagogical point of view, the findings of this study showed that the teachers agreed that pronunciation error correction especially Recast is necessary for L2 improvement. In general there were more significant differences between the teachers regarding Recast than those teachers who considered Metalinguistic Feedback. Teachers should correct learner pronunciation errors instead of ignoring them as if there were no errors in their utterances.

Given the fact that teachers’ immediate pronunciation error correction decreases learners’ motivation to speak (Allwright & Bailey, 1991), pronunciation errors should be treated after learners finish speaking. Teachers should also consider learners’ anxiety when making a decision on the degree of explicitness. Teachers can build learners’ confidence and self-esteem in their foreign language ability via encouragement and positive reinforcement. In this respect, teachers should be sensitive when correcting their learners’ errors and should remind them that it is natural for language learners to make pronunciation errors in the process of acquiring the target language. The teachers’ responses to learners’ errors may play the most important role in helping them alter their speaking for the better. Teachers, however, cannot and should not correct all the pronunciation errors made by their learners.

The findings provided an awareness of the feedback practices adopted in the classroom and the significant effects on learners’ pronunciation learning and their opportunities to use language,

Although learners want to receive pronunciation error treatment as much as possible, in reality, constant corrective feedback from the teacher can discourage learners from participating in activities in class and increase anxiety. As a result, learners feel uncomfortable and lose motivation to practice their pronunciation in class. To have a good pronunciation, learners need an environment that makes them feel encouraged to speak. They can learn by trial and error, by taking risks, and thus improve their speaking. Also, language learners need both time and opportunity for repair in the class. By trying to understand and acknowledge learners’ beliefs, teachers can minimize conflicts that may contribute to learner frustration, anxiety, and lack of motivation (e.g., Schulz, 1996). Since this study presented a description about the teachers’ perceptions and evaluation of them, it can be helpful and effective for teachers in the outset and meanwhile service courses. They can employ these feedbacks in their teachings methods. Teacher-learner rapports also mediate learners’ attention to and understanding of feedback. Generally by knowing the result, learners can learn faster and teachers also can use strategies and aids in their teaching that enhances their teaching power.

5.4. Recommendation

According to the findings of this study, teachers have significant positive perception on pronunciation error corrections, but in fact pronunciation is not the only language item which learners make errors in it and there are other language factors such as speaking, listening, reading and writing which may need error corrections. Therefore some further studies must be done about the effects of Recast and Metalinguistic Feedback on other language skills.

It is also true when it comes to education, when learners continue their education to higher levels, they may receive the best feedback on their pronunciation errors from other correction feedbacks than Recast, as a result of mentioned issues, further researches is recommended in investigating the learners’ levels and also the kind of feedback that teachers apply in order to correct learners’ pronunciation errors.

A need is also felt to investigate the usage of Recast and Metalinguistic Feedback by the teachers that played key role in answering the questionnaire and see to what extent they themselves use these feedbacks in their real classes. A study in this case would be helpful for both teachers and learners. If the teachers could understand how and when to use Recast and Metalinguistic Feedback, they would be able to teach more effectively and at the same time they could help learners to learn better.

Appendices

Appendix A: BELIEF QUESTIONNAIRE

Thank you for agreeing to complete this questionnaire .The information you provide will be very useful in helping us understand your views on language teaching.

Language Teaching Experience:

Indicate how well you agree with each of the following statements about your teaching experience.

Please circle your

- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
- 22
- 23
- 24
- 25
- 26
- 27
- 28
- 29
- 30
- 31
- 32
- 33
- 34
- 35
- 36
- 37
- 38