The Effect of Mock Tests on Iranian EFL learners ’ Test Scores

The effect of using tests in test preparation courses has been subject to debate. While some scholars such as Yang and Badger (2015) believe it is a cause of positive washback effect, others argue that this issue is tentative and context-bound (Green, 2007). Therefore, this study investigated the effect of using Mock tests in International English Language Testing System (IELTS) preparation courses on students’ overall IELTS scores. Fifty one IELTS students were selected non-randomly through the quota sampling approach out of 76 students at Mahan Language Institute in Birjand, Iran. These participants were distributed into Group 1 (n=25) and Group 2 (n=26). A complete IELTS test was administered to ensure that the Groups were homogeneous and to serve as pretest. After 10 sessions of intervention, a different IELTS test was administered as posttest. The results of between subject analysis through independent samples t-test revealed that using Mock tests in the IELTS preparation courses can positively affect the participants scores on IELTS exam. Pedagogical implications are discussed.


INTRODUCTION
Regulations in many institutes of higher education in countries where English is the medium of instruction require students to sit for high-stakes language tests as an entry requirement.As a result, students' success on such tests is a determinant of their academic success.Recently focus has been accorded to these tests in terms of validity, reliability, fairness, etc.In case of China, for example, Jin (2011) reports that College English Test (CET) is among the tests that students need to get through to enter the higher educational system.In Malaysia, Malaysia University Entrance Test (MUET) was the subject of study by Kuen and Embi (2012).
In Iran, International English Language Testing System (IELTS) has been determined as a popular test and has been subject to a few recent studies (Amirian, 2016;Mohammadi, 2016).One question left open, in line with this universal trend, is how one can increase language learners' (LLs) performance on high-stakes tests.This question motivated the researchers to conduct a study and investigate the effect of Mock tests on Iranian IELTS candidates' overall performance on the IELTS exam.
Recent studies conducted in the EFL context of Iran on high-stakes tests such as the IELTS indicate insufficiencies in terms of both research and learner's performance.First of all, most of the studies have their focus accorded on specific aspects of IELTS test such as the writing skill (e.g., Mohammadi, 2016;Panahi, 2015); and the washback effect (Salehi & Yunus, 2012), or learner's attitude towards the exam (Rasti, 2009).The studies that investigate IELTS candidates overall performance are scant.In addition, although Mock tests are used in many language centers across the country, research on their effect is sparse.Hence, there is no evidence whether or not the effect of Mock tests on IELTS candidates' performance is statistically significant.

Purpose of Study
The prime purpose of study was to find out whether Mock IELTS exam could have any effect on IELTS candidates overall score on the IELTS exam.This can determine whether or not such preparatory tests are useful for IELTS test takers.This knowledge can, in turn, motivate both IELTS candidates and instructors to either make use of such tests or find an alternative.

Research Question
In order to delve into the effect of Mock IELTS tests on IELTS candidates overall performance on IELTS, the following research question was proposed: • What is the effect of Mock tests on IELTS candidates overall performance on IELTS exam?

Hypothesis
The following research hypothesis was proposed: H0: Mock IELTS exams do not have any effect on IELTS candidates' overall performance on IELTS exam.

Significance of the Study
The findings of this study signify the role of IELTS preparatory courses and materials, and may be useful to guide language teachers, material developers, and curriculum designers.In the first place, curriculum designers who set educational objectives can use the findings of this study to realize whether or not Mock tests should be a part of the course.Based on this decision, material developers can revise the type of material used for IELTS preparation courses.Language teachers, on the other hand, find out whether it is worth administering such tests to the learners.

LITERATURE REVIEW
IELTS exam is taken by non-native English speakers to enter Anglophone countries for the purpose of work or study (Phakiti, Hirsh, & Woodrow, 2013).As a world-wide test, IELTS is taken by many candidates annually.Only in 2011, 1.7 million candidates took the test in different parts of the World (IELTS, 2013).IELTS is a prerequisite to enter many institutes of higher education at both undergraduate and postgraduate levels; therefore, it has been subject to numerous studies.Zhenhua (2008), Ghamarian, Motallebzadeh, Fatemi (2014), and Rashidi and Javanmardi, (2011) studied the IELTS exam in terms of washback and Construct validity.Moore and Morton (2005) focused on the IELTS Writing test and attempted to find out if the test meets the needs for university students.Lewthwaite (2007) focused on IELTS candidates' attitudes towards the test.Some other scholars such as Golchi (2012) studied students' anxiety and its relationship to their scores on the IELTS exam.
Tests in language classes and exam preparation courses have been a controversial issue.Chen (2011) calls the use of tests a threat to language teaching and learning due to negative washback effect.Ghamarian, Motallebzadeh, and Fatemi (2014) also stated that there is no relationship between constructs used in the IELTS exam and what language learners face in real life.Therefore, using tests to enhance EFL learners test skills may be questionable.On the other hand, there are scholars in the field who support the use of tests in test preparation courses.For example, Lumley and Stoneman (2000) observed that tests have more effect than teaching language use to language learners in IELTS preparation courses.
Lack of consensus on the conducting test preparation courses has led to a number of studies in different contexts.Cheng and Watanabe (2004) investigated different aspects of this issue, i.e., negative and positive washback effect.By considering a series of studies conducted in Seri Lanka, Hong Kong, Uganda and Japan with regard to high-stake tests such as IELTS and TOEFL, they concluded that there are two sides to this issue.One is the negative washback which focuses language learners' attention to the test solely; therefore, they learn language for the sake of good score.The other side is the positive effect which prepares the learners for the test and eventually brings about good results for them.Although the authors made suggestions to control negative effect on using tests in preparation courses, they still believed that using tests and testing strategies is an inseparable part of such courses.Green (2007) also noted that knowledge of the test demand is a significant issue to master, which is not covered in normal language classes unless it is deliberately added to the curriculum.Thus, one should be aware of such demands to be able to gain the desired score.Finally, Alderson and Wall (1993) posited that understanding of the concept of washback in different context is not usually supported by empirical evidence.Indeed, what teachers observe account for this phenomenon; therefore, more research is required to investigate this issue.
The researchers were motivated by lack of consensus on the effect of using tests to teach how to tackle tests.In more technical words, based on Green (2007), and Alderson and Wall (1993) who stated that researcher in terms of washback and high stake tests is scant, this study was conducted.

Research Design
This study has a quasi-experimental design, as we made use of a non-random sampling procedure.Quota sampling method was used as we did not have access to all the participants in the context of Iran.There were two main cohorts of participants in this study.The results gained from these participants (e.g., Group 1 and Group 2) was compared to answer the research question.
The main philosophical stance considered in this study is positivism; therefore, all conclusions are drawn based on researchers' observation and observable data.Other issues such as participants' perception were not taken into account.

Participants, Sampling Procedure and Setting
This study was conducted at Mahan Language center in the city of Birjand.In order to avoid the practice effect, the participants who had not attended IELTS course before were used in this study.
The researchers made use of quota sampling procedure believing that the participants in this language center have common features with the participants in the EFL context of Iran.As two cohorts of participants were required for this study, i.e., Group 1 (with Mock test) and Group 2 (without Mock test), and considering that the researchers were seeking large effect size, power analysis was conducted (Cohen d=.7).The results revealed that 23 learners are required in each group.Due to the possibility of the attrition effect, the researchers began the study with 25 participants in Group 1 and 26 participants in Group 2. These participants were selected out of the population of 76 IELTS students in the center.The participants were all upper intermediate and advanced language learners who aimed at taking an IELTS preparation course.These participants came from various backgrounds; therefore, socio economic background of the learners was not considered as a variable.These participants were selected by administering IELTS (academic module) and by considering 1 standard deviation above and below the mean score.Demographics of the participants are shown in Table 1.

Instruments and Materials
The main instruments used in this study as homogeneity test/ pretest and posttest were two different versions of IELTS academic module administered by the British Council in 2016.In addition to that, Mock tests were used in Group 1.

Procedure
Consent forms were administered at the beginning of study.Next, academic IELTS exam was administered as both homogeneity test and pretest.Considering 1 standard deviation above and below the mean score, 51 IELTS students were selected as the main participants.These participants were distributed into 2 different groups, i.e., Group 1(n=25), and Group 2 (n=26) with no statistical difference in their scores.
The main phase of the study lasted for 10 sessions during 3 months.Classes were held once a week and were both conducted by the same teacher (the second author).In Group 1, the participants were given a Mock test every session.The four different sections of the test were administered and the results were emailed to the learners prior to the next session.Therefore, no particular teaching was done in Group 1 and no discussions were held after the test to discuss the test issues.In Group 2, the conventional approach to IELTS classes in the institute was used.The participants were taught writing, reading, listening, and speaking.The test procedure was explained to them and individual tasks regarding the exam subsections were assigned to them.The procedure in Group 2 was deductive and rule-based.The participants were briefed on how to answer a particular type of question in IELTS tests.For example, to practice the speaking skill part 1, they were told about the procedure and type of ques-tions first, and next, they were asked to take the roles of the candidate and an examiner and practice the conversations.
The rubrics of the Common European Framework of References (CEFR) were used to score the speaking skill and the writing skill.The reading and listening sections were scored using the answer key.Each student received 4 scores for 4 different skills which were then turned into 1 score (average of 4 scores) to be used in the analysis.The researchers made use of the participants' gained scores instead of band scores.

Exploratory Data Analysis Results
At the beginning of data analysis Shapiro-Wilk test was run to ensure normal distribution of data.
As is depicted in Table 2, the results of Shapiro-Wilk test (p≥.01)indicate normal distribution of all scores; therefore, parametric data analysis could be conducted.
Having distributed the participants whose scores fell 1 standard deviation above and below the mean score on the homogeneity test, the two groups' scores (considered as pretest) were compared through independent samples test to make sure the groups did not have a statistical difference prior to the study.It should also be mentioned that total scores used to score the IELTS exam in this study were between 0 and 36.

Effect of Mock Tests
In order to test the null hypothesis, "Mock tests do not have any effect on IELTS candidates overall performance on IELTS exam", the participants' posttest scores in both groups were compared through independent samples t-test.
As can be seen in Table 4, the results of independent samples t-test (t(49)=-3.91,p=.002< α=0.05) reveal that the difference between Group 1 and Group 2 on the posttest is sig- As both groups' mean scores were improved paired samples t-test was also run.
The results (t(24)==2.11,p=001<α=005) indicate that the difference between the pretest mean score and posttest mean score is significant for Group 1.
The results (t(25)=2.67,p=.003<α=005) indicate that the difference between the pretest scores and posttest scores for Group 2 were also significant.

DISCUSSION
The findings of the study revealed that Mock tests can have positive effect on preparing IELTS candidates in the EFL context of Iran, as Group 1 outperformed the Group 2 in this study.The researchers attempted to aid Group 1 participants to experience real test situations rather than lecturing about the test.In Iran, Erfani (2012) observed this phenomenon between the IELTS exam and TOEFL iBT test and concluded such tests which motivated both language teachers and learners to use them as preparation tests in the class.However, test-based language classes have not always been favored.Watanabe (2004) posited that using tests in language classes and practicing the test taking strategies are signs of negative washback effect.However, the results of this study revealed that practicing test taking strategies is more effective than teaching course content in terms of a high stakes test such as IELTS.However, more detailed studies are required to find out about the extent of wash back effect in these situations.It may be due to the fact that many Iranian EFL learners, as stated by Mohammadi (2016) suffer from lack of test taking strategies, due to which they lose the test time.
In a recent study, Yang and Badger (2017) concluded that IELTS preparation courses that make students test-wise give them a sense of security, as most students want to learn how to tackle the test and gain high scores.Therefore, the authors recommend administering mock tests in test preparation courses.In congruence with their remarks, Lumley and Stoneman (2000) concluded that tests were integral parts of IELTS preparation courses and show more positive effect than focusing on linguistic features of language in such courses.In line with Lumley and Stoneman (2000), in this study, it was observed that actual exam experience has more effect than lecturing about the exam to the language learners in IELTS preparation courses.Naseri, Maghsoudi, and Rajabi (2014) observed that practicing speed reading strategies with IELTS candidates can increase their reading comprehension in the IELTS exam.Speed reading strategies as state by Chung and Nation (2006) include techniques such as skimming and scanning.As stated by Bell (2001) such techniques are best conceived if they are practiced in the test situation; therefore.Practicing test-taking techniques can increase test takers reading comprehension.Although the current study was not a study on the reading skill solely, such findings are congruent with Bell's (2001).

CONCLUSION
The findings of this study revealed that using Mock tests in IELTS preparation courses has positive effect on Iranian IELTS candidates overall IELTS score as the null hypothesis stating that this effect is not different was rejected.It was also discussed that the results of this study, congruent with previous studies, advocate the use of Mock Tests in IELTS preparation courses.Therefore, there is a need to add tests to IELTS preparation courses by curriculum designers.Fu-  Alderson and Wall (1993) is a context-bound issue.One may also wish to look at the issue of implementing mock test from a different perspective; for example, language learners or language teachers' perception.

Table 1 .
Demographics of the participants

Table 2 .
Test of normal distribution

Table 3 .
Comparison of the groups' mean scores on pretest

Table 6 .
Comparison of pretest and posttest scores for Group 2

Table 4 .
Comparison of the mean scores on the posttest

Table 5 .
Comparison of pretest and posttest scores for Group 1 researchers can study other high stakes tests such as the TOEFL exam in terms of washback, since washback effect as stated by ture