The Effect of Peer-Assisted Mediation vs . Tutor-Intervention within Dynamic Assessment Framework on Writing Development of Iranian Intermediate EFL Learners

Dynamic assessment originates in the Zone of Proximal Development (ZPD). Practicing dynamic assessment necessarily requires the development of ZPD. This study aimed to investigate the effect of peer-assisted mediation vs. tutor-intervention within dynamic assessment framework on writing development and the attitude of Iranian intermediate EFL learners. To do so, a quasi-experimental design and a questionnaire survey were carried out. After conducting a pilot study, a language proficiency test was administered to homogenize the two intact groups with 30 students in each. The writing part of the proficiency test was also considered as pre test. Then, the two groups were asked to fill out the pre-test attitude questionnaire. While in peer-assisted mediation group, the writing assignments were assessed by peers followed by feedback provided by them, in tutor-intervention group, the tutor assessed and provided feedback on the assignments during the instruction. At the end both the post test of writing and the questionnaire were administered. Comparing the post tests indicated that the peer-assisted mediation group outperformed tutor-intervention group. Although, peer-assisted mediation group had significant effect on the writing of the learners, for tutor-intervention, this effect was not significant comparing the pre and the post tests within groups. The study also revealed that both peer-assisted mediation and tutor-intervention significantly changed the attitude of the learners towards writing development between the pre and the post test of questionnaires, though the difference between their effects on the learners’ attitude was not significant comparing the post tests of questionnaires.


Problem and Purpose
Today, the role of assessment in teaching and even language learning becomes crucial in all stages.In past decades, the assessment from traditional views shifted to a new one called dynamic assessment.As the name represents it has some alternation over the traditional one.Poehner (2008) cites that "dynamic assessment posits a qualitatively different way of thinking about assessment of how it is traditionally understood by classroom teachers and researchers" (p. 1).Dynamic assessment is based on Zone of Proximal Development (ZPD).A difference between what a learner achieves by herself and what she achieves by assistance of others directly refers to ZPD.Development in learning process is along with sociocultural theory which is noted by Swain, Kinnear and Steinman (2010).Sociocultural theory considers theory of mind and cultural interactions.Someone like Peterman (2005) believes that sociocultural theory assumes; learning happens when an individual participates in cultural context and he is supported initially by more knowledgeable person.language, or education that emphasis on the ability of the learner to respond to intervention and mediation (Haywood & Lidz, 2006).Adopting interactionaist approach of DA, this study was an attempt to investigate the effect of peer-assisted mediation and tutor-intervention on writing development.It also sought to discover the effect of these types of assessment on the attitude of the learners towards their writing development.The present study highlighted the role of dynamic assessment in writing skill learning.Also, it signified the role of assistant and mediator (peer or tutor) in learning process, besides it revealed the effect of providing feedback in writing process.

Research Questions
1) Does peer-assisted mediation have any significant effect on writing development of Iranian intermediate EFL learners?
2) Does tutor-intervention have any significant effect on writing development of Iranian intermediate EFL learners?
3) Is there any significant difference between the effect of peer-assisted mediation and tutor-intervention on writing development of Iranian intermediate EFL learners?4) Does peer-assisted mediation and tutor-intervention have any significant effect on the attitude of Iranian intermediate EFL learners towards writing?5) Does tutor-intervention have any significant effect on the attitude of Iranian intermediate EFL learners towards writing?6) Is there any significant difference between the effect of peer-assisted mediation and tutor-intervention on the attitude of Iranian intermediate EFL learners towards writing?

Review of the Related Literature
Traditionally, assessment is described as an information-gathering activity (Bailey, 1996).For instance, McNamara (2004) explains that assessment is to find understandings of the pupil's knowledge or their learning ability.Based on this viewpoint, it is not possible to discover why teachers, including second language ones, often refer to the assessment as a necessary part of teaching and learning process.One teacher might think that the data achieved through the assessment procedures would be eagerly welcomed, and viewed as a vital constituent of better teaching.
Dynamic assessment considers mediation, provides constant feedback during the process of learning and the responses to these feedback, so feedback is a very important factor in powerful student learning.The benefits of successful feedback set in the context of learning outcomes are many.Successful feedback will be beneficial in the following issues: It builds self-assurance in the students, stimulates students to improve their learning; provides students with presentation improvement information, corrects errors and recognizes strengths and weaknesses.
Dynamic assessment includes presupposes giving feedback, and the responses to the feedback, so feedback is crucial in successful learning.Dockrell (2001) believes assessment should provide feedback to students on their progress towards the achievement of learning outcomes.Feedback enables students to realize where they have done well and indicates what they could improve on.It also justifies the grade or mark of comprehensive assessments.
It is important that feedback is timely.Cheng (2005) cites that if you provide feedback too soon, it may disrupt the student's reflective process.However, it is far more common that feedback is provided too late when it is no longer salient to the student.Feedback should not be held off until the end of a year or semester, as the student is unlikely to benefit from it once the task is complete and they have moved on to a new one.The benefits of successful feedback set in the context of learning outcomes are many.Successful feedback will be beneficial in the following issues: It builds self-assurance in the students, stimulates students to improve their learning; corrects errors and recognizes strengths and weaknesses.
Trends in the teaching of writing in ESL in past decades differentiated.Teachers learned more and more about how to teach fluency not accuracy, how to use authentic texts and contexts in the classroom.Process writing is one of the modern issues related to writing skill.Process writing helps writers to understand their own composing process, gives students time to write and rewrite, lets students discover what they want to say and write, gives students feedback throughout the composing process, and encourages feedback from both teacher and peers (Brown, 2007).
The effect of peer mediation with young children on autonomy behavior of children mediated by trained peers was conducted by Shamir and Steven (2005).The results indicated that children who received instruction in peer mediation with young children outperformed children who received general preparation for peer-assisted learning.Also, the higher level of mediational techniques and higher cognitive modifiability was associated with autonomy.
Another study investigated the improving oral reading fluency with a peer-mediated intervention.It examined the effects of an experimentally derived, peer-delivered reading intervention on the oral reading fluency of a first-grade student who had been referred for poor reading fluency.Results indicated that reading improvements were obtained through appropriate and efficient peer intervention as mediators of the learning reading comprehension (Duke & Daly, 2011).
Peer-assisted learning strategies on promoting word recognition, fluency, and reading comprehension in young children were investigated by Douglas and Lynn (2005) who summarized a good portion of the comprehension research program on reading in the early grades.First, they described investigations conducted in kindergarten, where their focus was on the development of decoding and word recognition.Then, they discussed studies conducted in first grade, where they continued to emphasize decoding and word recognition but expanded their focus to include fluency and comprehension.The Findings showed peer-assisted learning strategies are useful in fluency and reading comprehension.
The use of tutor mediation within a DA framework to support business students in the context of open and distance education investigated by Shrestha, Prithvi and Coffin (2012).The study explored the value of tutor mediation in the context of academic writing development among undergraduate business students studied in open and distance learning, following the DA.The analyses of the interactions suggested that DA could help to identify and respond to the areas that students need the most support in the study.Finally, they argued that a learning theory-driven approach such as DA could contribute to undergraduate students' academic writing development.Also, results showed that traditional assessment methods were unable to sufficiently support students.DA's focus on learning and development, on the other hand, helped to identify participants' evolving writing abilities.
A simple process and framework for teaching English writing to Iranian EFL intermediate learners based on the principle of Dynamic Assessment DA was introduced by Azarizad and Ghahremani (2013).Reflections and results of the research reiterated that the dialogic way of teaching is of great help in enhancing learners' writing interest and improving their writing competence.
A case study was done on introducing DA and producing a simple framework (or a process) for English writing instruction based on the principle of DA by Xiaoxiao and Yan (2010).The results of the study proved that the dialogic way of teaching was of great help in enhancing learners' writing interest and improving their writing competence.
The regulatory scale offered by Aljaafreh and Lantolf (1994) was applied to Iranian EFL learners' writing ability by Isavi (2012).In the study, the learners responded differently to the same type of errors they made in the pretest stage after introduction of mediation by the teacher.The regulatory scale which was applied in the intervention stage uncovered the fact that the individual learners had different developmental levels.The result of the study showed that a DA approach to EFL learners' writing ability could be useful.
All conducted studies have been in some sense meditational but there are other approaches to assessment that include intervention and response to intervention which are not meditational.These would fit within the broad definition of DA.However, in this study, both teacher intervention and peer assisted were mediational with focus on moving the learners beyond their ZPD in their writing ability.

Method
While peer-assisted mediation and tutor-intervention were considered as independent variables, writing skill and the attitude of the learners were dependent variables of the study.

Participants
Participants of the main study comprised of 60 female learners studying English as a foreign language at intermediate level, based on a proficiency test taken from American English File test pack.The sample was not randomly assigned to groups rather the intact classes were used.Therefore, the sample was assigned to conditions using nonrandomized ways as permitting researcher to choose conditions based on presumed needs.

Language Proficiency Test
A language proficiency test was administered to test the homogeneity of the participants.The test consisted of different parts namely grammar, vocabulary, pronunciation, reading, writing and listening aimed at evaluating the homogeneity of the participants and its writing section was also used as pretest.The language proficiency test was taken from American English file Test Pack by the institutes.Its grammar section had 20 items, vocabulary section had 20 items, pronunciation section had 10 items, reading section had 10 items, listening section had 10 items and writing section included paragraph writing.The test scored out of 100, each item got 1 point.Normally, proficiency tests do not include pronunciation section, but since the test was packed by the institutes and the validity of the test had been reported in advance, this section was not eliminated.The writing section of the test was used as a pretest.The pretest of writing included writing a cover letter to apply for a job based on an advertisement which was provided to them.The students had to write 120-150 words.

Attitude Questionnaire
The questionnaire was adapted from writing skill questionnaire by Community of Writers (Elbow & Belanoff, 2002).The questionnaire was related to the variables of the present study in order to measure the outlooks of participants towards learning and development of writing through treatment.It had 5 parts including attitude towards general writing, attitude towards generating ideas, attitude towards mediation, attitude towards feedback, and attitude towards collaboration.It consisted of 24 items that participants answered with yes, no, and sometimes.

Rating Scale of Writing
A rating scale used in this study to assess the writing assignments of the learners at each session.It belongs to DA as RECIPROCITY rating scale.It was devised by Van der Aalsvoort and Lidz (2002) as it is cited in Poehner (2008).It focuses on bidirectional interaction between mediator and learner.It also signifies the role of documentation which will be revealed by comments and written feedback.Ten scores were assigned for taught points which were repeated in subsequent assignments since this kind of rating scale as a part of DA focuses on removing the errors during a course of study.

Post Test of Writing
The post test of writing was quite different from the pre-test.It consisted of a paragraph writing around 120 to 150 words taken from the topic of American English File series.The students had to write a post card to a friend they hadn't seen or spoken to for a long time.They had to write based on the provided instruction.The writing taught points in treatment sessions regarding organization of the paragraph, punctuation, capitalization, descriptive paragraphs, using linking words and adverbs in narratives and connecting sentences with relative pronouns were all considered.The scores of posttests of writing were calculated out of 10 the same as pre -tests were scored.

Design
This quantitative research enjoyed a quasi-experimental and a survey questionnaire design.

Procedure
At the outset of the study, a pilot study was conducted in a small scale of 10 participants with the same characteristics of the main study in three sessions.While KR-21 showed 0.86 for the reliability of the proficiency test with 100 items, Cronbach alpha index was estimated 0.85 for reliability of the questionnaire with 24 items.The treatments, feedback provisions, and assistance were piloted as well.
The writing points were taught followed by paragraph writing.Then, the participants were taught how to provide feedback on the peers' writing at the end of each session.Also, the rating scale was presented to all participants as checklist to assess the paragraphs.
In all three sessions, in part with the main study, participants were divided in pairs.The pairs switched their assignments.At the first stage one reads the other's paragraph once without correcting.Second, the peer spied two important points.Then, the peer gave comments in full sentences on the margin; the rating was based on 1-10.When it was finished, the assignments were switched back.They talked and negotiated the problems.On the other hand; the tutor read the paragraphs once and then she found the errors by providing feedback, writing comments, using abbreviations and phrases in order to highlight the problems.Then the tutor returned the assignments to the participants, also tutor and participants talked about problems and solutions.The tutor then rated the assignments out of 1-10 basis.
In the main study, first, a language proficiency test and an attitude questionnaire were administered to two groups.Then, the treatments sessions started.The following writing points were taught in treatment sessions in each group.

Taught Writing Points 1) Capitalization and Organization
2) Punctuation: Period (.), Question Mark (?), Comma (,) 3) Linking ideas in narrative: (and, but, so, because) 4) Writing letter based on the presented format.5) Writing a postcard based on the given format.6) Using adverbs in narratives: For example: Suddenly, therefore, at last, at the end, then, now, soon The participants in both groups were asked to write a paragraph based on the given topics and apply the taught writing points.

Topics for Writing Assignments
1) Write a paragraph on the first day of school.
2) Write how movies or television influence people's behavior 3) Write a paragraph about your first trip and use at least three linking words.4) Write a letter to a friend and invite him or her to an occasion.In session, the researcher gave the rating scale in printed copy to the peer participants; explained and trained them how to rate the writing paragraphs based on the topics and how to provide feedback on their assignments.Then, in peer-assisted mediation group, participants were divided in pairs.The pairs switched their assignments.At the first stage, a peer as a mediator read the other's paragraph once without correcting.Then, the peer as a mediator spied two important errors.After that, the peer gave comments in full sentences on the margin and wanted the others to correct them all.Finally, the assignments were switched back.They talked and negotiated the mistakes based on provided feedback.

Rating
On the other hand, in tutor -intervention group, the tutor read the paragraphs as a mediator once and then she corrected the errors by providing feedback, writing comments, using abbreviations and phrases in order to highlight the problems.At last, the tutor rated the assignments in the same way.When the tutor returned the assignments to the participants, both the tutor and the participants talked about problems based on provided feedback.
Negotiating made the participants aware of their errors with feedback provided by tutor or peers as mediators.
They tried to come over their mistakes and avoid repeating them in subsequent assignments.Since ZPD is the essence of dynamic assessment, accordingly, a peer who was not able to find writing problems independently they did it through interaction with mediators (peer or tutor).In fact, by mediation, assistance provided by tutor or peers through interaction, rechecking the previous problems and negotiation; actual level of participants developed.This was evident through checking their previous errors in following assignments which moved the students went beyond their ZDP.After treatment sessions, both the post test of writing and the questionnaire of attitude were administered in two groups to test the research hypotheses.
As DA is a team up approach to add-on assessments within the domains of psychology, it focuses on the learner development to respond to intervention, mediation, assistance, reaction or feedback.Based on this, during the present study, feedback provided by the peer or tutor towards writing difficulties of the learners seemed to be helpful enough to remove their problems and to move them beyond their ZDP.
Rating scale used for writing posttest belonged to DA called RECIPROCITY rating scale.It was devised by Van der Aalsvoort and Lidz (2002, see Poehner, 2008).This rating scale is on bidirectional interaction between mediator and learner.It also signifies the role of documentation which revealed by comments and written feedback.Because, this rating scale is to some extent qualitative, so to make it quantitative and to assess clearly; It was considered 10 scores for the posttest too.Since this rating scale as a part of DA focuses on removing the errors during a course of study from pre-test to post-test each score of writing points was repeated in the other sessions, e.g.punctuation was repeated from the first session to the last session even in posttest, At the end, all posttest were rated and scored by two raters.

Results
As displayed in Table 1, the K-R21 Reliability Index for General Language Proficiency was .86.An independent t-test was run to compare two groups' mean scores on the pretest of general language proficiency test in order to prove that both groups enjoyed the same level of general language proficiency prior to the administration of the treatments.As displayed in Table 2, the peer-assisted mediation (M = 90.23,SD = 5.48) and tutor intervention (M = 89.16,SD = 8.06) groups showed almost the same means on the pre test of general language proficiency.The results of the independent t-test (t(58) = .526,P > .01,R = .069it represents a weak effect size) (Table 3) indicated that there was not any significant difference between the two groups' mean scores on the pre test of general language proficiency test.Thus it can be concluded that two groups were homogeneous.It should be noted that the assumption of homogeneity of variances was met (Levene's F = .000,P > .01).That is why the first row of Table 3, i.e. "Equal variances not assumed" was reported.

Hypothesis One
An independent t-test was run between pre and post test of writing to see the effect of peer-assisted mediation on writing development of Iranian intermediate EFL learners.Table 4 shows the peer-assisted mediation group in post test (M = 8.81, SD = 0.341) and pre test (M = 7.70, SD = 0.174).As displayed in table 5, the probability associated with t-observed value (.000) was lower than the significant level of .05.Therefore, peer-assisted mediation had a significant effect on writing development of Iranian intermediate EFL learners.

Hypothesis Two
To investigate the effect of tutor-intervention on writing development of Iranian intermediate EFL learners, an independent t-test was run between pre and post tests of writing.Table 6 demonstrates the tutor-intervention group in post test (M = 6.77,SD = 0.52) and pretest (M = 6.14, SD = 0.22).As displayed in Table 7, the probability associated with t-observed value (.142) is higher than the significant level of .05.

Hypothesis Three
An independent t-test was run on writing posttests of two groups to test significant difference between peer-assisted mediation and tutor-intervention on writing development.As displayed in Table 8, the peer-assisted mediation (M = 8.81, SD = 1.87) outperformed the tutor-intervention (M = 6.77,SD = 2.85) groups on the post test of writing.The results of the independent t-test (t(58) = 3.27, P < .01,R = .582it represents a large effect size) (Table 9) indicated that there was a significant difference between the peer -assisted mediation and tutor -intervention groups' mean scores on the post test of writing.It should be noted that the assumption of homogeneity of variances was met (Levene's F = 3.406, P > .01).That is why the first row of Table 9, i.e. "equal variances not assumed" was reported.
A multivariate analysis of variances (MANOVA) was run to compare the two groups' means on the five sections of the pretest of questionnaire in order to see their homogeneity in attitude towards writing.Before discussing the results it should be mentioned that the assumption of equality of covariance matrices and homogeneity of variances were met.Based on the results in table 10 (F = P > .01) it can be concluded that the homogeneity of covariances matrices was met..135 The assumption of homogeneity of variances which tested through the Levene's test assumes that groups did not differ significantly in terms of their variances.As displayed in Table 11, the Levene's F-values of attitudinal questionnaire were non-significant, i.e. (P > .01),which showed homogeneity of variances of the groups on pre tests of attitude.Because the assumption of normality was me, none-parametric data was changed to parametric ones.Based on the results displayed in Table 12 (F(5, 54) = 1.59,P > .01,Partial η 2 = .129it represents a moderate to large effect size) it can be concluded that there were not any significant difference between the means of the two groups on attitudinal questionnaire before treatment.

Hypothesis Four
To test that peer-assisted mediation does not have any effect on the attitude of Iranian intermediate EFL learners, the mean scores of pre and posttest questionnaires were compared within each group.Based on results the peer-assisted mediation group had a higher positive effect on the attitude of the learners towards writing in post test (M = 13.92)than pre test (M = 11.82)(Table 15) (t(29) = 2.91, P < .05,R = .47 it represents an almost large effect size) (Table 16).It can be concluded that peer-assisted mediation had significant effect on the attitude of the learners and positively changed their attitude towards writing skill.

Hypothesis Five
To test the effect of tutor-intervention on the attitude of Iranian intermediate EFL learners, the mean scores of pre and post test of questionnaires were compared within the group.
Based on results, the tutor-intervention group performed significantly better on the post test (M = 14.42) than pre test (M = 12.28) (Table 17) (t(29) = 3.87, P < .05,R = .58it represents a large effect size) (Table 18).It can be concluded that tutor-intervention had significant effect on the attitude of the learners and positively changed their attitude towards writing skill.

Hypothesis Six
To test the significant difference between peer-assisted mediation and tutor-intervention on Iranian intermediate EFL learners' attitude towards writing, a multivariate analysis of variances (MANOVA) was run to compare the peer-assisted mediation and tutor -intervention groups' means on the five sections of the post test of questionnaire.
Before discussing the results it should be mentioned that the assumption of equality of covariance matrices and homogeneity of variances were met.Based on table 19 (F = 1.45,P > .01) it can be concluded that the homogeneity of covariances matrices was met..113 The assumption of homogeneity of variances which tested through the Levene's test assumes that groups did not differ significantly in terms of their variances.As displayed in Table 20, the Levene's F-values for pre tests of attitude were non-significant, i.e. (P > .01),which showed the homogeneity of variances of the groups on post tests of attitude questionnaires.it can be concluded that there was not any significant difference between the means of the two groups on post tests of questionnaire.Although, both peer-assisted and tutor-mediation had significant effect on the attitude of the learners toward writing but the difference between their effect was not significant.To show the nature of dynamic assessment more clearly the assessed ten writing assignments in two groups during the courses were also compared.By comparing mean of the means of 10 writing assignments, t-value 2.65 preceded t critical 2.00 (α = 0.05).It showed that the peer-assisted group outperformed the tutor-intervention group on writing assignments during the courses.

Discussion and Conclusion
This study concluded that peer-assisted mediation proved to be more effective on the learners' writing development.It also concluded that although both peer-assisted mediation and tutor-intervention affected positively on the attitude of the learners but the difference between their effects was not significant.The applied dynamic assessment both in a form of peer-assisted mediation and tutor-intervention led to positive development on writing skill during treatment sessions, however the writing development of peer-assisted mediation group was higher.
As DA is a team up approach to add-on assessments within the domains of psychology, it focuses on the learner development to respond to intervention, mediation, assistance, reaction or feedback.In present study, feedback provided by the peer or tutor towards writing difficulties of the learners seemed helpful enough to remove their problems to a great extent.Therefore, the study concluded that both peer-mediation and tutor-intervention were effective on the learners writing development during the course of instruction; however, peer-assisted mediation was more efficient on post-test.
Learner attitude has been the essential area of inquiry in language acquisition and it is related to the internal behaviors that can affected by external ones which supports the conclusion of the study that both peer-assisted mediation and tutor-intervention affected positively on the learners' attitude toward writing.Since, attitude is usually affected positively by some external factors; in the present study these factors were peer-assistance, tutor-assistance, and mediation in line with feedback provision which led to a positive change in the learners' attitude toward writing.
The findings of the study were in line with Shamir and Steven (2005) who found that mediators and learners received significantly higher scores on autonomy behavior criteria which displayed the significant role of peer mediation.The findings were compatible with the results found by Azarizad and Ghahremani (2013) both emphasized the influential role of dynamic assessment on writing development.
On the attitude of the learners, the findings were somehow aligned with the result obtained by Johnson and Douglas (1976) that cooperative, compared to individualized, learning resulted in greater ability to take the affective perspective of others, more altruism, more positive attitude towards classroom life, and higher achievement.
The conclusion of the study supported the studies which have found traditional assessment methods unable to sufficiently support the students.DA with focusing on learning and development helps the assessors identify the participants' evolving writing abilities.All the studies concluded that feedback was of great help in enhancing learners' writing interest to improve their writing ability.The study also concluded that DA approach to EFL learners' writing ability could prove to be useful, and more appropriately designed mediation played a significant role in promoting learners' writing ability in developing their learning potential in ZPD.
The findings once again were supported by the idea that dynamic assessment can unify instruction with assessment to provide learners with mediation in order to promote their reserved learning potential during the assessment.The conclusion of the study also signified the role of attitude in enhancing learning process and it was supported by other studies and the idea that if learners are reluctant to learn or they do not have a positive attitude, they do not produce any result and language learning is stimulated by the attitude.
EFL Learners and tutors can benefit from the results of this study.Peers can mediate the learning process of writing through negotiation on writing assignments, participation in pairs or groups, providing comments and feedback on writing assignments of the peers.Moreover, the learners' attitude toward writing could positively change using peer and tutor assistance while the course is going on.Attitude can be defined as a set of beliefs developed in the course of time in a sociocultural setting and having a positive attitude certainly facilitates learning.Therefore, by positively changing the learner's attitude towards writing, those who are even reluctant to learn would be keen on learning.
EFL tutors may promote techniques of dynamic assessment through peer-assisted mediation, making the peers as mediators via teaching them how to provide feedback to remove the errors.This may also establish a friendly and challenging atmosphere which facilitates learning process and in turn enhances cooperation and collaboration learning.Teachers should assist the learners and help them to think and generate ideas by providing timely feedback.The peer-assisted mediation can adjust the tutors' responsibility in some cases, so that they will be able to manage the class more efficiently.Peer-assisted mediation leads to a decrease in complications at educational settings, enhancement of learners' self-esteem, improvement of their attendance, and encouragement of the learners in problem-solving situations to find more novel solutions.Accordingly, tutors can detect the learners' writing difficulties and their timely interventions would display that they not only help them write but also encourage them think.Similarly the tutors make available the ongoing feedback on writing learning process to support the learners at each stage.
Dynamic assessment needs instruction, intervention, and feedback to promote L2 development by the means of ZPD.How exactly Vygotsky's ZPD triggers a change from dependent performances to the process of maturing and performing independently in all aspects of L2 learning via dynamic assessment still needs more investigations.

References
7) Conjoining Sentences by which, who, where, … 8) Descriptive Paragraphs 9) Writing a paragraph with examples 10) Coherence 5) Write a postcard to a friend and tell him about a beautiful place.6) Tell a story and use at least five adverbs in it.7) Write a paragraph about what you need for a trip.What and why.Use at least 3 connectors.8) Describe your dream house 9) Write some important qualities of a good boss 10) Is homework harmful or helpful?

Table 2 .
Descriptive statistics for pretest of general language proficiency of two groups

Table 3 .
Independent t-test between the means of two groups on general language proficiency tests

Table 4 .
Paired sample statistics for peer-assisted mediation group between pre-and post-tests of writing

Table 6 .
Paired sample statistics for tutor-intervention group in pre and post tests of writing It can be concluded that tutor-intervention had no significant effect on writing development of Iranian intermediate EFL learners.Although the students outperformed in post test but the difference with pre test was not significant.

Table 8 .
Descriptive statistics of writing for two groups on post tests

Table 9 .
Independent t-test of writing posttests for two groups

Table 10 .
Assumption of equality of covariance matrices of attitude questionnaire before treatment

Table 11 .
Homogeneity of variances assumption; pretests of attitude

Table 12 .
Multivariate tests; pre tests of attitude questionnaire of two groups Table13displays the descriptive statistics for the two groups on the pre tests of attitude.The largest difference lied between their attitude towards general writing (MD = 2.67) and the smallest difference belongs to generating ideas (MD = .07).

Table 13 .
Descriptive statistics for pre test of attitude questionnaires of two groupsAs displayed in Table14, the Cronbach Alpha reliability indices for the pre and post tests of attitudinal questionnaire towards writing were .75 and .77respectively.

Table 14 .
Cronbach alpha reliability indices, pretest and posttest of attitude questionnaire toward writing

Table 15 .
Descriptive statistics of pre and post test of attitude questionnaire of peer-assisted mediation group

Table 17 .
Descriptive statistics of pre-and post-test attitude questionnaire of tutor-intervention group

Table 18 .
Paired samples t-test for tutor-intervention group between pre-and post-tests of attitude questionnaire

Table 19 .
Assumption of equality of covariance matrices, posttests of attitude questionnaires of two groups

Table 21 .
Multivariate tests for post tests attitude questionnaires for two groups

Table 22 .
T-value of 10 writing assignments of two groups Figure 1.Development of groups in 10 treatment sessions assignments