DISERTASI dan TESIS Program Pascasarjana UM, 2016

Ukuran Huruf:  Kecil  Sedang  Besar

EFL Students’Ability in Performing Autonomous Learning and their Writing Proficiency across Cognitive Styles

Dyah Dewi Masita



EFL Students’Ability in Performing Autonomous Learning and their Writing Proficiency

across Cognitive Styles

Dyah Dewi Masita

Graduate Program in English Language Teaching-Universitas Negeri Malang-Jl. Semarang 5, Malang 65145

diadewi_masita@yahoo.com. 087859349491

Abstract :This study is intended do investigate the relationship between students’ autonomy and their writing proficiency across cognitive styles. Correlational research design was fit to understand the relationship between two continuous variables. To obtain the required data, Writing Autonomy Questionnaire, and writing test was administered to 155 students of sixth semester from two private colleges at Malang. The result presented that there is significant and positive correlation between those variables (r = .437). Meanwhile variables pressented gave contribution to the other variable due to its correlationship.

Key Words : students’ autonomy, writing proficiency, cognitive styles

Abstrak : Penelitian ini bertujuan untuk mengetahui hubungan antara autonomy mahasiswa dan kemampuan menulis mereka berdasarkan gaya berpikir. Metode penelitian yang sesuai adalah korelasi untiuk memahami hubungan dua variabel. Writing Autonomy Questionnaire, dan tes menulis disebarkan kepada 155 siswa semester 6 untuk memperoleh data yang diperlukan. Hasil peneliatian ini menunjukkan bahwa ada hubungan positif dan signifikan antara kedua variabel tersebut (r = .437). Sementara itu variabel yang teridentifikasi tersebut saling berkontribusi dikarenakan hasil analisa yang menunjukkan hubungan positif dan signifikan.

Kata Kunci : belajar mandiri, kemampuan menulis, gaya berpikir


Over the past few decades, Communicative Language Teaching (CLT) become more prominent in Indonesian education especially English at formal institution. CLT has been trend in Asian EFL context including Indonesia recently and documented in international surveys such as by Jarvis (2004), Butler (2011),  Ho & Wong (2004),  and Nunan (2003). Along with the decree published by Indonesian government No. 12 Year 2012 Article 1 that the purpose of education is for developing students’ potential consciously to have character, self management, as well as skills which will be essential in the society, the application of CLT approach in Indonesian education is seen to be an effective learning situation in which focusing more on meaning rather than form (English served as communication, not merely about sentence structure and grammar). In line with this approach, government expects all the teachers to nurture their students to have an ability to maintain and manage their own learning. Thus, autonomy appears to be recent issue nowadays in higher education. This kind of ability is called autonomy in which it is one of three basic psychological needs beside need for competence and need for relatedness (Deci & Vansteenkiste, 2004). Each individuals requires the ability to be aware and responsible toward his action whether it is initiated or spontanously. In academic setting, there have been a number of studies reveal that the higher degree of autonomy that students have, the better their academic achievement will be.  Dafei (2007) and Myartawan, et al. (2013) have conducted similar study regarding autonomy in language learning, which was investigating the relationship between learner autonomy and English proficiency in higher education. The first one was conducted in China, while the other was in Indonesia. Both studies reveal that there is significant correlation between students’ autonomy and English proficiency. Moreover, a study done by Lowe (2009) found that there was a positive, significant correlation between the total Score on the LAP-SF which measures individual’s autonomy in learning and Grade Point Average (GPA).

Proficiency may be thought of as skill, adequacy, sufficiency fora defined purpose, or capability. In connection with speaking and writing which are productive skills in language learning, those become the predictor of someone’s proficiency somehow when he/she perform English. Focusing in this study, proficiency in writing can be explored when a person have already mastered all micro and macro skills in writing. Hyland (2003) explained that writing as the visual channel and the productive mode of language is a vital skill for the L2 students to develop their language knowledge and the teaching of this skill has become central in second language classrooms. Further, students in higher education needs to write a good composition by themselves so that their proficiency can be reflected.  Thus, writing autonomy directs to students’ independency in writing and knowing what and how they should do monitored and consciously.

A number of existing studies, Lowe (2009); Dafei (2007); and Myartawan, et al. (2013) who put interest in similar areas, autonomous learning, investigate its relationship and English academic performance in general which was represented by GPA. Positive and significant correlation revealed in those previous studies. Besides, Kulsirisawad (2012) conducted experimental design about developing writing autonomy in EFL classrooms via peer feedback and found that it gives students more oportunities to take charage of theie learning process especially in writing. However, there is still lack of research which investigate the relationship between autonomous learning and English skill. Meanwhile students in higher education mostly majoring English and in the process of composing thesis, they are required to collect sources as many as posible and explore arguments to defense their ideas. This study was conducted at the Kanjuruhan University and Islamic University of Malang to the learners who have passed all paragraph writing involving genres and academic writing which expose argumentative essay. The consideration of choosing both of them was having similarity as the private university in Malang where the English Language Teaching Department is  accredited B. Meanwhile, the sixth semester learners were apropriate for this study since they are required to apply what they have been learnt on the previous semester and they are preparing for teaching practicum that requires them to learn and perform independently. In addition, they have been being ready for writing their own thesis as the requirement to obtain Bahelor’s degree.  According to the information gathered from pilot study which lecturer and few learners were involved, learners who participated in this study mostly have learned English for more than 10 years since they were in kindergarten. Meanwhile, they were used to write Argumentative essay with browsing sources at first rather that baesd on their own logical reasoning. Besides, they were still need leads from the lecturer to write a composition well and more time at least a day until a week.Thus, autonomy serves as the important point for students in college level.

An effective language learner is characterized by any kind of personal characteristics, learning style and learning strategies. Rubin (1975), Stern (1975),  Naimah, et al. (1978) noted that there are some descriptions of an effective language learner as well as considering his/her personal characteristics and awareness of learning style and strategies. Dealing with autonomy, Nunan, (1991) and Benson (2001) agree that those who are good and effective language learners will display a high degree of autonomy. The result of an investigation held by Nunan and Wang, (2011) show that the dominant style of more effective language learners was communicative, it is characterized by Field Independent and active. On the other hand, the dominant style for less effective language learners was authority-oriented. These learners exhibit characteristics of Field Dependent and passitivity since they are used to do better in “traditional classroom”. In addition, Nosratina, et al. (2014) noted that there is significant relationship between learning style and learning strategies.  Thus an effective language learner will be able to choose what learning strategies are suitable for them, and it will lead to higher degree of autonomy.

Another study conducted by Altun and Cakan (2006),  Ahmadzade and  Shojae (2013) which put interest in the relationship between cognitive style and academic achievement showing the different results. The first one noted that there is no relationship between cognitive style and academic achievement. Meanwhile, the other pointed out that there is significant relationship between cognitive style and academic achievement. The difference result is likely because of both studies conducted in two areas of study : computer and English.

Eventhough a number of existing studies have conducted the relationship between autonomous learning and academic achievement in general, this study is different from those previous studies. First, it concerns in investigating the relationship between learners’ writing autonomy and their achievement across cognitive styles in college learners. Besides that, the present study also involves cognitive styles as one of the variables. Its purpose is to find out how cognitive styles have impact to boost learners’ autonomy in writing proficiency which is considerably effective.  All previous studies do not include cognitive styles as one of the variables. Further, the different setting of place and subject also distinguish this study compared to the previous ones. Because there are still lack of study which study about writing autonomy as well as the impact of cognitive style, this study is expected to support some theories related to that.

From background of the study mentioned earlier, the researcher formulates research question as follows :

1.         Is there any relationship between learners’ autonomy and their writing proficiency ?

2.         Is there any significant difference between autonomy scores of FI and FD cognitive styles ?

3.         Is there any significant difference between writing proficiency scores of FI and FD cognitive styles ?


The idea oflearner autonomy was first developed out of practice-that of teacher-researchers at the Centre de Recherches at d’Applications Pedagogiques en Langues (CRAPEL), University of Nancy, France in the early 1970s. According to the former Director of CRAPEL Henri Holec (personal communication), the need for a term to describe people’s ability to take charge of their own learning (for this is how he and his colleagues came to conceptualize ‘learner autonomy’) (Smith, 2008).  Defining learner autonomy, Holec (1981), Long (1989) and Little (1990) agreed that it is in line with psycological aspect and ability to take charge on one’s own learning. According to Little (1991) autonomy help students to set their aims, decide on the content and process of their own learning, and assess their improvement and success.

Benson and Voller (1997) pointed out the term autonomy has been used in at least five ways : (1) easy accesibilty for students to develop themselves (supportive environment and situation in which they are able to study entirely on their own), (2) a set of skills which can be learned and applied in institutional education. Getting academic knowledge and skill in school (institutional education) are not the only important skill to face the future, there is also soft skill which involves expected character such as being responsible, persistence, and others. Those are also provided in institutional education beside academic skill. (3) An inborn capacity which is surpressed by institutional education. Autonomy can be developed within students naturally (since they were born), though it needs to be improved in supportive educational situation. (4) The exercise of students’ responsibility for their own learning. The learnng method that is available for students to nurture autonomy such as problem solving. Its purpose is to raise students’ awareness and responsibility for their own learning. After all the characteristics of students’ autonomy, they will have (5) the right to determine the direction of their own learning. 

Some definitions regarding to writing such as writing as the visual channel and the productive mode of language is a vital skill for the L2 learners to develop their language knowledge and the teaching of this skill has become central in second language classrooms (Hyland, 2003). Furthermore, Hayes (1996 cited in Weigle, 2002 : 17) stated that writing is not an individual project but it has social value in which what, how, and who to write are within social interaction. Moreover, writing cannot be separated from the readers. In order to make them comprehend what the writing composition is about, it should be clear and containing self explanatory. In terms of EFL classes, writing is an essential tool for learning and helping learners improve their academic achievement in English language learnning. Due to the excessive importance of writing, it becomes the predictor of individual’s success in English language learning. However, as writing is seen to be the most complicated skill since it expresses learners’ thoughts in the form of written mode and using correct sentence structure.Foroutan, et al. (2013) asserts that socialization is important in learning language skill, more especially writing skill, they also added the usefulness of letting students collaborate with each other in improving language skill. The results pointed out that online interaction which occurs through CMC (Computer Mediated Communication), in this case using weblog that emphasizes the role of communication and social contacts in the form of commenting, exchanging ideas, collaborating, or adding some materials are important factors in developing autonomy among language students. Meanwhile, Khodadady and Khodabakhshzade (2012), Kulsirisawad (2012), Bagheri and Aeen (2011), and Hyland (2000) agree that autonomy in writing is build through activities which lead them to cooperate with others such as collaborative learning, direct feedback from teacher, and peer feedback with supportive tool such as portfolio and journal.  By documenting growth over time through a systematic collection of their work, portfolios enable students to see possibilities for reflection, redirection, and confirmation of their own learning efforts (O’Malley & Pierce, 1996). Moreover, it has been argued  that self-assessment as an effective language learning strategy to promote autonomous language learning because it encourages language students to assess their learning progress and in turn helps them to stay focused on their own learning (Hung, 2009). It helps students to be judge of their own strength and weakness in order to develop self-directed learning and becoming autonomous students.

Similar to autonomy which concerns on psychological aspects, cognitive styles refer to individual’s preferred way of processing, i.e. of perceiving, conceptualizing, organizing, and recalling information  (Saville-Troike, 2006 : 87).  The statement means the process of cognition-how information is being processed. Due to the individual differences, cognitive styles are divided into Field Independent (FI) and Field Dependent (FD). This division is based on whether the learners tend to separate details from the general background or see things more global. Witkin et al. (1954) explored that FI learners perceive things as distinct from their background field, while FD learners tend to be influenced by any embedding context. In other word, FI learners tend to be independent in doing the tasks, analytic and systematic, while FD learners tend to be more holistic, seeing everything through their surroundings, focusing more in meaning rather than form.

The first construct of Field Dependent and Field Independent done by Witkin (1977) shows the advantages of each preferences in achieving success in language learning.  Some researchers (Chapele & Roberts, 1986 cited in Robertson) argued Field Independent learners have correlated to language learning success. On the other hand, Johnson, Prior, and Artuso (2000) concluded that Field Dependence learning process has right advantage in social skills and dependency to the teacher. To mediate the report of those researchers, Brown, (2007 : 123) has suggested both position process a certain degree of validity. Field Independent learners perform better in language learning, they can find their own ways to learn, while Field Dependent learners could theoretically do the same with their social skills.


            Explanatory correlational research was used in this study since it explores the correlationhip between two or more continuous variables (students’ autonomy and writing proficiency). Here, students’ autonomy (X) is predictor variable, while writing proficiency (Y) becomes criterion variable. The goal of correlational research is to predict other variable as the result of correlation and make the consistency of variables being investigated. Creswell (2012) asserts that although it is not as rigorous as experiment, correlational research can be best in making connection between variables and being best predictor for further result. Thus, it was appropriate for answering the research questions formulated in this study because it aimed at studying and investigating the relationship between two or more continuous variables (learners’ autonomy and writing proficiency).

This study was done at Kanjuruhan University and Malang Islamic University which are located in Malang, East Java Province. Kanjuruhan University of Malang provides multicultural environment since the learners come from the Indonesia’s ethnics groups, and Malang Islamic University has the Islamic religious ambience. Both of them are known of having accreditation B of English Department among other private colleges in Malang. They were chosen as the target population due to accessibility, suitability, and availability reason. Kanjuruhan University of Malang provides 8 faculties with 11.497 learners in academic year 2015/2016 and Islamic Malang University consisted of 10.568 learners within 10 faculties. The research subjects of the study were the people who fit and and represent the conceptual construct for this study. They were undergraduate learners who were in sixth sesmester of English Department at both colleges and already taken Writing IV or Advanced Academic Writing course. The consideration of choosing sixth semester learners because of their readiness in teaching practicum for next semester. Moreover, they were getting content course, instead of skill course in this semester, and they have enough knowledge from the Structure, Vocabulary, and Writing course series.Then, the appropariate participants involved in this study were 155 students in total, from two classes at Kanjuruhan University of Malang and other five classes from Malang Islamic University within average 23 students in each class, considering their ability in writing argumentative essay and being ready for applying all their knowledge and skills in that level. Considering consistency in giving response to tbe questionnaire, only 108 of them were valid.


GEFT (Group Embedded Figure Test)

Group Embedded Figure Test (GEFT) is a test used to determine individual’s cognitive style whether he/she is Field Independent or Field Dependent style. It contains some complex figures/pictures and the participants are asked to simplify them. First introduced by Witkin (1954), it became main attention in describing cognitive style because of its simplicity and validity. According to Witkin et.al (1971), GEFT has been commonly applied to determine the FI and FD cognitive style. The researcher use this kind of test to determine the cognitive style because this test is valid and it has been investigated in many researches across cultural setting and accepted as quite reasonable (Altun & Cakan, 2006 : 290) and it has high reliability which is .82 (Witkin, et al, 1971). Moreover, a number of psychological tests use visual and spatial mode such as simplifying picture or rearrange jumbled picture into correct order to know the characteristics of test takers. In addition, it enables learners to analyze and find a hidden figure so that their critical thinking would be activated. Thus, it has been applied by many researchers on any levels of participants only for the subject classification

The test required each individual to find a specified simple form provided that was embedded within a complex design. The test consists of 3 sections which should be completed in 12 minutes. The first section is given for practice purposes and included 7 items to be done in 2 minutes. The real tests begin in both the second and third section. Those sections contains 9 items which each of sections have to be done in 5 minutes. The subject score will be based on these two sections. The correct answer will be scored 1 point while the wrong answer will have 0 point. The score will range from 0 (the most FD) to 18 (the most FI). Although Witkin et al (1971) do not specify a clear cut off score for determining field dependent and independent individuals, the researcher followed the conventional division system. The learners who scored 0-9 were categorized as FD, meanwhile those whose scores range from 10-18 were categorized FI. Further, during the administration of the GEFT, the exact procedures set out in the technical manual (Witkin, et al., 1971) regarding time limits and directions were closely followed.

Writing Autonomy Questionnaire

To measure the learners’ writing autonomy the researcher adapted Learner Autonomy Profile (LAP) which was developed by Confessore and Park (2004) based on the four pillars of autonomy introduced  by Meyer (2001),

Carr(1999), Ponton (1999), and Derrick (2001) which developed measurements for each of them, ILD (Inventory of Learner Desire), ILR (Inventory of Learner Resourcefulness), ILI (Inventory of Learner Initiative), and ILP (Inventory of Learner Persistence).The questionnaire was constructed to three preconative factors and in general learning. Further the present questionnaire was modified into writing skill consisting writing process, reflection, and autonomy characteristics (supportive learning situation, conscious learning, and self learning management). In addition, learners’ belief and self efficacy were added as the conative factors of autonomous learning which concerns more in learners’ perception and what they believe in learning English. Considering that LAP-SF was valid analyzed by many researchers, reflected individual’s intentional act to perform autonomous learning, this questionnaire was chosen to be adapted.

            It then was tried out to the 31 learners of sixth semester who had the same characteristics with the target population. The result then showed that its reliability presesnted by Cronbach Alpha was .732. The 30 items then were analyzed and found that item number 1, 3, 10, 13, 21, 23, 27, and 30 were invalid because r counted (correlation) on two-tailed of those items were below r table (significance .05 or 5%). Consequently, the 22 items were used due to their validity.

Writing Test

In this study, writing test was administered to see learners’ writing proficiency. Homogenity of learners’ writing proficiency in target population for this study was required to determine learners’ writing autonomy. In order to make them homogenous, they were taken from sixth semester according to the consideration they have passed all writing genres course. Besides, to make sure that their writing proficiency were similar, the researcher conducted test which ask learners to write their own essay based on certain topics. The required essay was hortatory explanatory text in the form of argumentative. It was chosen because argumentative essay promotes learners’ critical thinking in exploring their arguments in certain ways. On the subject matter and theme in the blueprint, the researcher considered  to provide 10 familiar topics, which then learners were able to choose their preferences. The next stage, the researcher rated 3 most learners’ preferred topics and added those into the writing prompt.

In giving score of productive skills such as writing and speaking, scorer’s subjectivity usually appears. Cohen et al., (2007 : 117) describes reliability as the consistency of measurement. It is concerned with precision and accuracy. In this study, assessing learners’ essay, the researcher used a writing scoring rubric which includes content, organization, accuracy, and mechanics. To avoid the appearance of inconsistency score due to the subjectivity of raters, the researcher applied inter-rater reliability.

In order to achieve the agreement between raters in understanding scoring rubric, they have to apply the same standards to their rating process. Raters were selected by considering their ability and experiences in teaching English, especially in writing. To ensure that both raters were applying the same standard to their scoring, they had a test previously. It was held held by giving them model of essays then score them based on the standard argumentative essay scoring rubric. When they were consistent in giving score for some times the tests were conducted, they could be raters for this study. After the analysis was done among raters, they had reliability in giving essay score was .989 (r = .989)

Data Collection

Data were collected by distributing GEFT at first to classify subjects based on their cognitive style which takes only 15 minutes. Then, participants were required to give response to Writing Autonomy Questionnaire by ticking the statements according to their appropriateness to their situation in learning English. At last, they were asked to write an argumentative essay at least three paragraphs (250-350 words) within time constraints (70 minutes) in which they have chosen the given topics previously.

Collection of the data was done firstly at Kanjuruhan University of Malang one week after the try out finished, while at Malang Islamic Unversity, questionnaire and writing test were undergone on the first week of semester. The participants involved in this study were taken from sixth semester students of English Department at Kanjuruhan University of Malang and Malang Islamic University. They were 155 students in total from two classes at Kanjuruhan University of Malang and other five classes from Malang Islamic University within average 23 students in each class. Collection of the data was done firstly at Kanjuruhan University of Malang one week after the try out finished in which class A on Tuesday, February, 23rd at 2 p.m while class B was on Thursday, February, 25th at 7 a.m. In addition, at Malang Islamic University, the distribution of questionnaire, GEFT, and writing test were held on March, 2nd 2016 for three classes, E, C, and D. The learners from last two classess, A and B were  having those test on the day after.

Data Analysis

The primary data in this current research were in the form of Writing Autonomy Questionnaire result and Writing test in the form of Argumentative essay. The questionnaire was used to measure learners’ autonomy in writing.  After the learners finish doing the questionnaires, the researcher scored them. According to the score obtained from questionnaire, conventional classification was made to see the degree of their autonomy. For those who scored 22-44 were classified having low autonomy, 45-66 score was belong to moderate and learners who obtain score ranges from 67-88 would be highly autonomous.

To answer the research questions, there were two stages employed by the researcher in analyzing the data. The first was concerned with the descriptive statistics which purpose was to reveal the characteristics of the data, such as maximum score, minimum score, and standard deviation. Followed by testing normality and linearity of data, Kolomogrov-Smirnov and scatterplot diagram were employed to see that the gathered data were having normal distribution and linear relationship. The second stage deals with inferential statistics, to test the hypothesis. Testing hypothesis to answer the research questions was by using statistical computation. It aimed at investigating whether or not there is positive significant correlation between the variables (autonomy and writing proficiency) across learners’ cognitive styles. The significant correlation was converted into null hypothesis to find the hypothesis significance as a basic of the null hypothesis acceptance or rejection. The criteria of acceptance or rejection of the hypothesis was a level of 0.05 (Level of Significance).

After getting the score from raters, the results were correlated to see the degree of their reliability. Further, the researcher computed the final score from both raters of the writing test and the questionnaire to examine the consistency by employing Pearson Product Moment Formula. The computation was done using SPSS 16 software. Writing Autonomy Questionnaire served as the X variables and the Y variable was represented by the Writing test score. When the positive correlation appeared between learners’ autonomy and writing proficiency, there would be post hoc analysis which gain more possibilities within the relationship. Following the attribute variable involved in this study, the different score in autonomy and writing proficiency revealed. To answer the second and third research question dealing with which learners are more autonomous and proficient in writing between FI and FD cognitive styles, the researcher compared mean from the data by using independent sample t-test. At first, the researcher made equal number of learners from two groups, FI and FD. Then, the scores of autonomy questionnaire and writing proficiency from those learners computed into indepedent sample t-test. Lastly, mean obtained from the variables was being compared to see which one was higher or lower. 


4.1       Characteristics of Data

Descriptive statistics to show the characteristics of data is the first step in analyzing the data Table 1 reveals mean, range, both minimum and maximum score, and Standard Deviation from autonomy and writing proficiency score.

Table 1. Descriptive Statistics of Students’ Autonomy and Writing proficiency

            N         Range  Minimum         Maximum        Mean   Std. Deviation

Autonomy       108      33        53        86        65.45   6.781

Writing            108      58        32        90        66.63   10.937

Valid N (listwise)        108                                                     


As shown in Table 3.1, the mean of writing proficiemcy which is 66.63, while the mean of the students’ autonomy score obtained from questionnaire was 65.45, but then the range of writing score was quite wide which is 58, while autonomy score was only 33. Standard Deviation (SD) of students’ autonomy was 6.781, on the other hand, writing proficiency scores having 10.937. 

Before analyzing the relationship between variables (students’ autonomy and writing proficiency), it is necessary to display the normal distribution of the data as well as their linearity. Its purpose is to ensure that those variables which are going to be analyzed, distributed normally to be put into correlation computation. Kolmogorov-Smirnov technique was used to investigate the normality test. The result presented in Table 3.2

Table 2. Normality Test of Students’ Autonomy and Writing Proficiency

One-Sample Kolmogorov-Smirnov Test

            Autonomy       Writing


Normal Parameters                 Mean

                                                Std. Deviation

Most Extreme Differences     Absolute



Kolmogorov-Smirnov Z

Asymp. Sig (2-tailed)  108







.437     108








a.         Test distribution is Normal    

From the Table 3.2, it was obvious that the data were distributed normally in which  p-value from Kologorov-Smirnov analysiswas above level of significance .05. From the total number of subject 108, the p-value from autonomy questionnaire was .437 and for writing proficiency was .284. The result showed that obtained p-value was higher than the level of significant, thus it fulfill the requirements to become normally distributed.

            Testing the linearity of the data is necessary to understand that the variables relates to each other before getting into correlation analysis. Linearity is shown by using Analysis of Variance (ANOVA) table, which aims at understanding the sigificant coefficient of linearity, and scatterplot diagram to illustrate it. Table 3.3 presents the ANOVA table of linearity.


Table 3. Linearity Test of Students’ Autonomy and Writing Proficiency using ANOVA Table



            between groups           Within groups Total

            (Combined)     Linearity          Deviation from Linearity                   

Sum of squares


Mean square


Sig.      4838.403




.033     2.447E3




.000     2391.239




.637     7960.782


100.769           1.280E4



The results of analysis showed that obtained p-value from Table 3.3 was above .05 level of significance (.637 > .05) Based on those results linearity of the data were exist within variables. Considering to illustrate the linearity, scatterplot diagram was displayed. When score from varriable X (students’ autonomy) increases, will be followed by increasing score from variable Y (writing proficiency). The result of linearity test analaysis can be seen from the Figure 3.1



Figure 1. Scatterplot Diagram of Linearity Test from Students’ Autonomy and Writing Proficiency

            The diagram above presented the distribution of data which was linear as shown by the scatterplot which spread increasing normally when score from variable X (students’ autonomy) was high, so was the score from writing proficiency. Eventhough some of data were not linear, it was obvious that majority of data were linear to be analyzed into correlation.

4.2 Relationship between Learners’ Autonomy and Writing Proficiency

Conventional classification of the degree of students’ autonomy in writing was decided, based on the minimum and maximum score they got from the questionnaire. The minimum score was 22 while 88 was the maximum score, thus it can be classified in which for those who achieved 22 – 44 (low autonomy), 45 – 66 (moderate autonomy) and 67 – 88 (high autonomy). Based on the result, it was found that there was none of them were having low autonomy. 58% of all students (63 in total) was high autonomy, while for moderate one was obtained by 45 students or 42% of all students.To investigate how students’ autonomy have relationship with their writing proficiency, the researcher analyzed those variables using Pearson Product Moment with two-tailed direction within .05 level of significance (95%), and then a Table 3.4 provided below showing the degree of correlation between those variables.

Table 4. Correlation between Students’ Autonomy and

Writing Proficiency

Autonomy       Writing

Autonomy   Pearson Correlation

                    Sig. (2-tailed)

                    N 1

108      .437**



Writing        Pearson Correlation

                    Sig. (2-tailed)

                    N .437**


108      1



** Correlation is significant at the 0.01 level (2-tailed)


From the output table above, it was shown that there is relationship between students’ autonomy and their writing proficiency and it was positive. From 108 participants, the degree of correlation was .437 which was classified into moderate correlation. Significance of its correlation can be seen at two tailed within .05 level which was .00. Investigating whether there is significant correlation, .00