The Effect of Dynamic versus Static Assessment on Syntactic Development of Iranian College Preparatory EFL Learners

This study investigated the effect of dynamic assessment on learners’ syntactic knowledge. The main concern of this study was significant difference between dynamic and static assessment and the possible role of these two forms of assessment on the syntactic development of Iranian EFL college preparatory learners. To fulfill the purpose of the study, NELSON was administered to 70 pre university male learners. Those learners with scores one standard deviation above and below the mean were selected for the purpose of the study. There were 60 students who were then divided into experimental and control groups. Next, the pre-test (a teacher-made, multiple-choice test) was administered to 20 subjects of parallel group. Then, twelve multiple-choice grammar tests were administered to the subjects of the experimental group. The students received mediation in dynamic assessment model which involved some strategies like looking for clues, eliminating the answers that do not fit and comparison strategies. The mediation process was designed to enable teachers to mediate each of the items in an interactive way. The mediation was divided into two stages: a) manipulation of grammatical and structural point through the strategies mentioned above and b) information paper which was given to the students at the end of each session to work on them at home. The control group received deductive grammatical rules during twelve sessions. Finally, the pre-test was re-administered as the post-test and t-test was applied. The results showed that at 0.05 level of significance the mean of experimental group was more than the mean of control group for different scores and consequently the null-hypothesis was rejected. The study proved that dynamic assessment outperformed in improving syntactic knowledge of the learners.


Introduction
A test is a method of measuring a person's ability or knowledge in a given area.A test is first a method and that method generally requires some performances or activities on the part of either the testee or the tester or both.Testing is an inextricable part of the instructional process.If a test is to provide meaningful information on which teachers and administrators can base their decisions, then many variables and concerns must be considered.No EFL program can deny the significance of testing for evaluating learner's acquisition of the target language.Testing is used for two purposes.On the one hand, it is a means to reinforce learning and to motivate students and on the other hand, it is a device to assess learners' performance in the language.In the former, the testing is geared to the teaching; while in the latter case, teaching is geared to the testing.There seems to be a certain inherent contradiction between the goals of student assessment and its means.The goal is usually to evaluate the students' learning ability and to gain information useful for more effective instruction.The means, however, are often limited to measuring the students' current performance level.According to Vygotsky these functions belong to the Zone of Proximal Development (ZPD) in counter-distinction to fully developed functions that belong to the zone of actual development.While the results of the static assessment show us the already existent abilities of the student, the analysis of ZPD allows us to evaluate the ability of the student to learn from the interaction with a teacher or a more competent peer.Grammar may be considered in several different and sometimes overlapping ways: 1) as a synonym for syntax, 2) to refer to the entire body of elements and units that constitutes a language (including not only syntax but also phonology and word formation), 3) to label the description of a language prepared by linguist, 4) to describe the body of linguistic knowledge that resides in the minds of the speaker of language (Falk, 1978).
In all academic settings, assessment is viewed as closely related to instruction.assessment is needed to help teachers and administrators make decisions about student's linguistic abilities, their placement in appropriate levels, and their achievement, also help in evaluating the suitability and effectiveness of the curriculum, and the instructional materials.Language teachers are often faced with the responsibility of selecting or developing the language tests for their classrooms and programs.However, deciding which testing alternatives are the most appropriate for a particular language education context can be daunting, especially given the increasing variety of instruments, procedures and practices available for language testing.Most teachers spend a considerable amount of classroom time on assessment, giving and checking tests, and 'going over' the tests with their students.Of all classroom activities, testing is possibly the activity that pupils and teachers take most seriously.'Teaching to the test' is in sharp contrast with the priority attached to the As early as 1934 the Swiss psychologist Andre Rey proposed basing the evaluation of students' abilities on directly observable learning processes.The concept of learning potential assessment was further developed by Vygotsky (1934Vygotsky ( -1986) ) and Feuerstein in the field of psychology to assess cognitive functions.(See also Minick 1987, andKozulin, 1998).While the results of static testing (assessing current performance levels (show us the already existent abilities of the student, DA allows us to evaluate the ability of the student to learn from the interaction with a teacher.This learning ability may serve as a better predictor of the students' educational needs than the static scores. It is only recently that the DA concept has been adopted (and adapted) for use in domains that depend on the use of cognitive strategies (thus far, physics, astronomy, mathematics and reading comprehension).How do we know it works?Well, there has been some promising research.A recent study (Vollmeyer & Rheinbreg, 2002) showed not only that it works, but also produced an additional surprising result.Using the DA format to test learning in physics, the researchers predicted that mediated feedback would affect both motivation and performance.What they found instead was that learners who were told that they would receive explicit feedback on the use of strategies, used better and more systematic strategies (compared to a previous static assessment) even before the mediation stage had begun.In other words, the mere expectation of strategy feedback led to deeper processing of the learning material.
Vygotsky believed that the normal learning situation for a student is a socially meaningful cooperative activity.New cognitive functions and learning abilities originate within this interpersonal interaction and only later are they internalized and transformed becoming the student's inner cognitive processes.Thus under conditions of collaborative or assisted performance students may reveal certain emergent functions that have not yet been internalized yet.According to Vygotsky these functions belong to the Zone of Proximal Development (ZPD) in counter-distinction to fully developed functions that belong to the zone of actual development.

The Study
Foreign language learners terminate their studies with an almost sophisticated knowledge of conventional grammar rules.For instance, in our country high-school students are dealing with a strong background of formal grammar rules whereas their performance on some of the tests shows their lack of familiarity with the language.There seems to be a certain inherent contradiction between the goals of student assessment and its means.The goal is usually to evaluate the students' learning ability and to gain information useful for more effective instruction.The majority of learners may be acquainted with grammar rules in isolation and know the structures and formulas very well but they always have problems with syntactic rules in context and also EFL teachers know from experience that their students often have difficulties with both at sentence and discourse level of grammar, a problem that cannot be ignored.Another problem is that a plenty of teachers feel that teaching through a test is an inappropriate and useless job while some teachers believe that teaching through testing can be completely communicative and enjoyable for learners even if achieving this can sometimes be quite demanding of our creativity as teachers.The purpose of this study is to determine the effect of dynamic assessment versus static assessment on syntactic development of Iranian college preparatory learners.Vygotsky (1978) believed that the early development of understanding occurs through interaction with others.In this period greater achievement is possible when a child learns through collaboration with a more experienced or informed guide.This principle usually underlies therapeutic interventions but is not the way in which cognitive or language assessments are traditionally conducted.This difference between a student's own performance and his/her achievement when guided by a teacher or in collaboration with a more experienced peer, reflects the student's development potential referred to by Vygotsky (1978) as the "zone of proximal development" (commonly ZPD).The term "dynamic assessment includes a range of methods and materials to assess this potential for learning, rather than a static level of achievement assessed by conventional tests.Its aim is to reveal an individual maximum performance, by teaching or mediating within the assessment and evaluating the enhanced performance that results.Feuerstein et al. (1987) noted that it is used in contrast to the notion of "static assessment" assessed by conventional psychometric approaches.The case for DA of cognitive ability has been described by Feuerstein et al (2002), but there remains void in the field of speech and language therapy.The limitation of static assessment are considerable in the field of SLT where the multidimensional nature of language "does not easily lend itself to single unitary measures" (Dockrell,2001: 75).screening measures used to detect language problems have been shown to be less than accurate (Law et al, 1998) and little better than the opinions of parents (Laing et al., 2000).Furthermore, Dockrell (2001) suggests that diagnostic tests which target specific aspects of language systems are consistently inadequate for determining whether a learner is developing typically or is experiencing a delay.The diversity of learner's problems requires a range of interventions, but test batteries fail to specify an intervention likely to be useful, or to give a prognosis.
There seems to be a certain inherent contradiction between the goals of student assessment and its means.The goal is usually to evaluate the students' learning ability and to gain useful information for more effective instruction.The means, however, are often limited to measuring the students' current performance level.This contradiction was identified as early as 1934 by Vygotsky (1934Vygotsky ( /1986; see also Minick 1987, Kozulin, 1998).Vygotsky believed that the normal learning situation for a student is a socially meaningful cooperative activity.New cognitive functions and learning abilities originate within this interpersonal interaction and only later are they internalized and transformed becoming the student's inner cognitive processes.Thus under conditions of collaborative or assisted performance students may reveal certain emergent functions that have not yet been internalized yet.
According to Vygotsky these functions belong to the Zone of Proximal Development (ZPD) in counter-distinction to fully developed functions that belong to the zone of actual development.While the results of the static assessment show us the already existent abilities of the student, the analysis of ZPD allows us to evaluate the ability of the student to learn from the interaction with a teacher or a more competent peer.This learning ability may serve as a better predictor of the students' educational needs than the static scores.Vygotsky mentioned the whole range of possible interactive interventions to be used during ZPD assessment, such as asking leading questions, modeling, starting to solve the tasks and asking students to continue, and so on, but he produced no standardized procedure for the ZPD assessment.Vygotsky also made no particular distinction between ZPD assessment of general cognitive functions and ZPD assessment in content-based learning areas.

Participants
To accomplish the task, 60 students participated in this study.All of them were Iranian college preparatory high school students.Their major was mathematics.They enrolled in an extracurricular program that was held as a grammar course in fall 2009 in Eghbal Lahouri educational center.That is a high school in Garmsar.The age limit of subjects was 17 to 19 students of Upper and lower grade were not accessible for the researcher.In this study, all subjects were male students.To do random assignment of subjects, the students were numbered from 1 to 60. Odd numbers were assigned to one group and even numbers were assigned to the other one.One group determined as the control group (C group) and the other one was regarded as the experimental group (E group).Each of two classes was held twice a week on Saturday and Tuesday mornings.Each session lasted 90 minutes.

Procedure
To accomplish the purpose of this study, the following procedure was applied: Fowler and Norman Coe) was administered to select the homogeneous subjects .70subjects took the test.After selecting subjects with scores one standard deviation above and below the mean, they were divided randomly into two parallel groups.The t-test was applied and it proved the homogeneity of two groups.Then the researcher prepared two different tests.First, a pre-test (it was the same as the post-test) which was a multiple -choice teacher-made test was made .Then twelve multiple-choice tests were given to experimental group that worked as a mediation for the experimental group.

PHASE TWO: Pre-test piloting
To make sure about the appropriateness of the pre-test (a teacher-made test) for the subjects under investigation, the test was piloted with parallel group 20 college preparatory EFL learners.Then, item analysis was done; i. e., item facility, item difficulty, item discrimination, and choice distribution were determined.10items out of 30 items had to be deleted, for they were malfunction items.After estimating reliability and validity, 20 acceptable items of this test were given to the subjects under the investigation, both subjects of the control and experimental groups.Considering the scores of the pilot group.The reliability of pre-test (or post-test) was estimated through KR-21 formula (Table 1).On the other hand, to ensure the validity of this test, the scores of the pilot group were correlated with the scores of the structure and written expression section of NELSON utilizing Pearson Product Moment correlation coefficient (Table 3).Checking acceptable reliability and validity, 20-item pre-test (post-test) could serve for the purpose of the study.It was given to 30 subjects of experimental group and 30subjects of the control group.At the end of the term, it was used as post-test.To make sure about the appropriateness of twelve multiple-choice (teacher-made tests) for the subjects under investigation, the tests were piloted with a parallel group of 20 college preparatory subjects.Then, item analysis was done, i. e., item facility, item difficulty, item discrimination, and choice distribution were determined.10items out of 30 items had to be deleted, for they were malfunction items.Considering the scores of pilot group the reliability was estimated through KR-21formula (Table 2).On the other hand, to ensure the validity of tests the score of the pilot groups were correlated with each other (Table 4) and they were correlated with the Structure and Written expression part of Nelson utilizing Pearson Product Moment Correlation Coefficient (Table 3)

PHASE THREE: Test Administration
In the main phase of the study, the first test to be administered was given to 70 subjects.The rational for administering NELSON was to select the subjects with nearly the same proficiency level and choosing the sample population randomly.After NELSON administration, subjects were selected with scores one standard deviation above and below the mean.Reliability of NELSON was estimated through KR-21 formula.Then, subjects were randomly divided in to two groups.Application of t-test proved that two groups were homogeneous.Next, pre-test was given to two groups.Then, twelve multiple-choice grammar tests were administered to the subjects of the experimental group.At the heart of the process the students received mediation in dynamic assessment model which involves some strategies like looking for clues, eliminating the answers that do not fit and comparison strategies.
The mediation process was designed to enable teachers to mediate each of the items in an interactive way.The mediation was divided in to two stages: a) Manipulation of grammatical and structural point through the strategies mentioned above and b) Information paper which was given to the students at the end of each session to work on them at home.The control group receives deductive grammatical rules during twelve sessions.The post-test was finally administered to both groups.

PHASE FOUR: Data Analysis
The data analysis stage included the following steps: first, the basic descriptive statistics; that is, mean, median, variance ,standard deviation, and standard error for NELSON, multiple-choice grammar test, pre -test, post-test were calculated.Second , the pilot study for all tests (multiple -choice grammar tests, pre-test post-test) were carried out, and then the reliability of each test after scoring each one was calculated through KR-21 (Table 3).Furthermore, to estimate the validity of pre-test, and multiple-choice grammar tests were correlated with NELSON test through Pearson Product Moment Correlation Coefficient (Table 3).Next, through the application of the t-test, the mean scores of the subjects in the experimental group were compared with the mean scores of the subjects in the control group.It was done to determine any significant difference between two groups in terms of subjects' syntactic development.3, Table 4).According to the Table 3 the correlation coefficient of pre-test and NELSON is 0.96 and probability level is 0.01, so we come to conclusion that pre-test has been an acceptable test and has enjoyed criterion validity.Before administering the multiple-choice grammar tests in the experimental group, in order to estimate the reliability of the tests, they were given to other groups (not in the main groups) and the number of students in each group was 20.Table 6 shows the descriptive statistics for pilot study of these tests and there is no significance difference between both experimental and control groups in pre-test which is 0.31.It means that there is no significant difference between pre-test experimental and control groups.This table shows that there were 30 students in both post-test experimental and control groups so there were no absentees among the subjects of the two groups ,In other words, All of the subjects took part in the post-test.The indexes of central tendency (mean, standard deviation and median) in the experimental and control groups divulge that the performances of subjects in the experimental group was much better .Meanwhile the spread of scores in the experimental group and control groups has not been similar.In fact the spread of scores in the control group was wider in comparison to experimental group.The subjects of the experimental group could get the mean score of (17.737) which is significantly higher than the control group which was (13.83).Nevertheless, it deserves mentioning that it deserves mentioning that the difference between the performances of the two groups has been calculated through applying a t-test that will be presented in (Table 7) In order to be sure of accuracy of results, three assumptions were considered: Data have been estimated through interval or rational scale.
The distribution of data is normal.
If we compare two groups (experimental and control), there should be homogeneity of variances.

Results and Discussion
The present paper aimed at investigating the effect of dynamic assessment on learners' syntactic knowledge.It focuses on the significant difference between dynamic and static assessment and the possible role of these two forms of assessment on the syntactic development of Iranian EFL college preparatory learners.
The concept of learning potential assessment was further developed by Vygotsky (1934Vygotsky ( -1986) ) and Feuerstein in the field of psychology to assess cognitive functions.(See also Minick 1987, andKozulin, 1998).While the results of static testing (assessing current performance levels (show us the already existent abilities of the student, Dynamic Assessment allows us to evaluate the ability of the student to learn from the interaction with a teacher.The premise of this approach is that This learning ability may serve as a better predictor of the students' educational needs than the static scores.A recent study (Vollmeyer & Rheinbreg, 2002) showed not only that it works, but also produced an additional surprising result.The term "dynamic assessment includes a range of methods and materials to assess this potential for learning, rather than a static level of achievement assessed by conventional tests.Its aim is to reveal an individual maximum performance, by teaching or mediating within the assessment and evaluating the enhanced performance that results.Feuerstein et al. (1987) noted that it is used in contrast to the notion of "static assessment" assessed by conventional psychometric approaches.The case for DA of cognitive ability has been described by Feuerstein et al (2002), but there remains void in the field of speech and language therapy.
According to Vygotsky these functions belong to the Zone of Proximal Development (ZPD) in counter-distinction to fully developed functions that belong to the zone of actual development.While the results of the static assessment show us the already existent abilities of the student, the analysis of ZPD allows us to evaluate the ability of the student to learn from the interaction with a teacher or a more competent peer.This learning ability may serve as a better predictor of the students' educational needs than the static scores.
The effectiveness of DA approach was evaluated by conducting a study with 60 Iranian students , all studying in a college preparatory high school and assigned randomly into experimental and control groups.Each of the two classes was held twice a week and each session lasted 90 minutes.Next, the pre-test (a teacher-made, multiple-choice test) was administered to 20 subjects of parallel group.Then, twelve multiple-choice grammar tests were administered to the subjects of the experimental group.The students received mediation in dynamic assessment model which involved some strategies like looking for clues, eliminating the answers that do not fit and comparison strategies.The mediation process was designed to enable teachers to mediate each of the items in an interactive way.The mediation was divided into two stages: a) manipulation of grammatical and structural point through the strategies mentioned above and b) information paper which was given to the students at the end of each session to work on them at home.The control group received deductive grammatical rules during twelve sessions.Finally, the pre-test was re-administered as the post-test and t-test was applied.Specifically the post test results showed that at 0.05 level of significance the mean of experimental group was more than the mean of control group for different scores and consequently the null-hypothesis was rejected.The study proved that dynamic assessment outperformed in improving syntactic knowledge of the learners.

Conclusion
The main concern of this study was to test this presupposition whether or not using dynamic assessment can make any effect on syntactic development of Iranian college preparatory EFL learners, and whether this effect would be equal on static assessment.To assure and determine if any significant change appeared in the syntactic knowledge of our groups of subjects, after receiving at the treatment, the results of performance of each group of pre-test was compared with the results of its performance at the post-test stage applying t-test.Performing t-test revealed that there is a significant increase in the performance of subjects in experimental group; this means that the subjects in experimental group benefited significantly from subjects in control group.So null-hypothesis of this study was rejected at 0.99% of confidence level.The findings of this research are of a great value for learners in a sense that developing grammatical knowledge can help learners perceive a relationship between grammatical and three other dimensions of language.These dimensions are: 1. Social function 2. Semantics (meaning) 3. Discourse pragmatics.Foreign language learners terminate their studies with an almost sophisticated knowledge of conventional grammar rules.For instance, in our country high-school students are dealing with a strong background of formal grammar rules whereas their performance on some of the tests shows their lack of familiarity with the language.In this study the most useful technique is dynamic model which provides learners some strategies like deletion, clues and comparison for the purpose of grammar development.Moreover, the findings of this study may help teachers to change their method of teaching and testing, so they would be able to fulfill the learners' needs to be the professional language learners.Considering techniques in dynamic model, teachers can encourage students to pay more attention to grammatical points in test instruction.Dynamic assessment can function as a teaching device for all that it measures, and it can test learners' reading comprehension and grammar knowledge.This research, may, also, help the test developers to construct appropriate, valid, and reliable grammar tests and test writers will be able to develop tests which are more effective assessments of students' grammatical knowledge.
Phase one-test preparation Phase two-Pre-test piloting Phase three-test administration Phase four-data analysis PHASE ONE: Test preparation NELSON (test 050 A which was extracted from Nelson English Language Tests, book 1 Elementary written by W.S.
exist.The smallest value is shown.

Table 1 .
Reliability estimate of pre-test and NELSON

Table 3 .
Reliability Estimate of 12 Multiple-choice Grammar Tests

Table 4 .
Descriptive Statistics for Pilot Study of 12 Multiple-choice Grammar TestsBased on this table, the reliability of 12 Multiple-choice grammar tests is evaluated through Cronbach's alpha for battery, it means, for the whole ten tests.Item total correlation column shows that the relationship between the scores of one tests and the total score and alpha if deleted column represents that when one test is deleted, if the reliability increases, they will not be good tests and vice versa.So it might be concluded that multiple choice tests have internal consistency because in the case of deleting one test the reliability decreases.The reliability estimate of these tests is 0.9589 (Table2, Table

Table 5 .
Descriptive Statistics for Pilot Study of 12 Multiple-choice Grammar Tests

Table 6 .
Descriptive statistics for post-test Experimental and Control group Table 7 reveals through the application of t-test, with 58 degree of freedom at 0.01% of significance.And the mean of experimental group (17.73) was more than the mean of control group (13.83).So null-hypothesis of this study was rejected at 0.99% of confidence level.

Table 7 .
The t-test compared the results of two groups