Wang Yingxue

Jiamusi University,Jiamusi,China

Busan University of Foreign Studies,Busan,South Korea

Email:wonderfulcristina@163.com

[Abstract] This research focused on the effect and efficacy of interaction through their computer-mediated interaction and L2 writing.In terms of data collection and analysis,45 participants'791 English writing assignments from Pigainet were collected to check whether there is a correlation among the interaction of the use of computer-mediated groups.The research found out that this computer-mediated L2 writing method was more effective in SLA writing classes.Furthermore,this research explored that L2 learners' autonomy that positively influenced the frequency of learners'computer-mediated interaction improved their SLA writing skills.

[Keywords] computer-mediated interaction;SLA writing;CALL;autonomy

Introduction

As it rolls on,ubiquitous high-tech teaching and learning equipment has been widely applied to L2 education.This phenomenon accelerates to find out whether the approach could be a springboard to success in L2 education.To identify its efficacy,particularly in the improvement of learners' SLA writing skills in computer-mediated classes,this research explores the effect and efficacy of L2 writing improvement through the computer-mediated class context in which interaction theory,computer-mediated language learning,and autonomy of learning are applied.With the development of science-technology,"human-computer" interaction is increasingly frequent and even essential of language learning,as well as second language learning(Amory,2014).Therefore,it is significant to do some research on the learner-computer interaction in learning.

Interaction theory

Educators and researchers have slightly different perspectives on the definition of interaction.This research introduces a few which are main and representative literature.Wilikins (1972) proposed that the communicative approach and the core of it would be interaction.Long (1983) put forward the "interaction hypothesis" theory,who believe that the interaction is critical to gain comprehensible input and the opportunity to obtain language output.So,interaction is a possible prerequisite for the success of second language learning,requiring three types:instructorslearners,peers,and learners-content(Moore,2016).A lot of research on the first and second types of interaction(Patricia et al.,2000)has been administrated,named as"human-human"interaction.

Computer mediated language learning

CALL is an abbreviation of Computer-Assisted Language Learning,which can be defined as learning a language on a computer as a direct activity through structural lessons or as an activity that is unrelated to language study but promotes language awareness and acquisition (Kern &Warschauer,2000).With rubric of modes of Computer-Mediated Communication，the growing importance of computer-mediated language learning(CMLL)has significantly impacted writing instruction and writing study recently as the result of the increasing growth of computer technologies and the popular use of computers among linguistic researches.The utilization of writing software applications in learners'compositions may be positively correlated with the quality of the compositions of L2 proficiency(Littlewood,2002).The goal of automated essay scoring software as CMLL was to relieve teachers of the burden of grading huge quantities of student essays.

Autonomy in Language Learning

Autonomy is the ability to take charge of one's learning(Holec,1981)for all the decisions concerning all aspects of this learning,i.e.,determining the objectives;defining the contents and progressions;and selecting methods and techniques to be used;monitoring the procedure of acquisition properly speaking(rhythm,time,place,etc.);and evaluating what has been acquired.The term has been used in at least five different ways:(1) situations where learners could study totally on their own;(2)skills that can be learned and practiced in self-directed learning;(3) natural capacities that are suppressed by institutional education;(4) exercises that learners could have the responsibility to do for their own learning purpose;(5)learning right that learners could use to determine the direction of their own learning.

There has been increasing emphasis on autonomy language learning in recent years(Littlewood,2002).Interestingly,Asian students have similar learning patterns for autonomy (Littlewood,2001).Some teachers have reported that their attempts to promote autonomy in Asian learners were successful(Ohno et al.,2018).

Research Questions

This research aims to investigate the effect of learner-computer interaction in English writing and the effect factors such as proficiency level and autonomy of language learning.Therefore,participants were randomly grouped into four levels by stratification according to pretest scores.Following research questions,read as below.

(a)Are there any differences between proficiency level groups in terms of improvement scores of English writing in pre-posttest,frequency of learners'computer-mediated interaction，and autonomy of language learning？

(b)What is the correlation between improvement of English writing,the frequency of learner-computer interaction,the autonomy of language learning,and proficiency level?

Previous Research

Previous literature conducted in Taiwan has discussed the effectiveness of MyAccess in the writing classroom(Chen &Cheng,2006;Yang,2004).Some students felt the computerized feedback more facilitate,while some other students were not satisfied with it for the computerized feedback could not offer specific feedback on the essay content(Chen&Cheng,2006).However,some researchers have focused on learner autonomy of language learning could be enhanced by using computer programs,particularly in the field of ESL/EFL writing(Williamson,2009).If the use of the computer software is carefully modeled,it can offer students with both assistance and autonomy in the writing process.(Williamson,2009).There were differences in the degree of autonomy among different proficiency groups:Good performers showed higher autonomy(Shien&Akiko,2019).

Methodology

Participants

Forty-five college freshmen joined both pre-test and post-test,whose ten tasks were assigned to submit their compositions to Pigainet,with 791 drafts including 450 first drafts,251final drafts,45 pre-test,and 45 post-test drafts.According to the distribution of pre-test scores,they were grouped into four-level groups.

Instruments and Materials

All materials of tests and tasks were excerpted from the topics of CET4 Bank.They were required to complete the tests and tasks by submitting the compositions to Pigainet,including revision drafts.Pigainet could record the revision history and frequency that would be collected for analysis in this research.

Pigainet is a kind of Computer-Mediated Error Analysis instruments that provide holistic comment and feedback information in grammar,vocabulary,and expression method according to every sentence.Participants submitted their tests and tasks into Pigainet,which would give them feedback,including the holistic score and the detailed information.They can revise their compositions according to the feedback information as many times as they want.

The link of Wenjuannet,which is a kind of software specialized for designing Questionnaires,was used to send the questionnaire to participants through WeChat.And the data of the questionnaire can be collected from the Wenjuannet after participants' answering.The questionnaire of the autonomy of language learning,which followed the"Questionnaire on Learner Autonomy"(Sakai et al.,2009).

Procedure

Forty-five collegians voluntarily took part in the pre-test on 11 Dec.2020,who were required to submit their compositions with the topic of Selling the used book only once time within 30 minutes in the classroom under the invigilation.

The pre-test was designed to get the distribution of proficiency of participants'English so as to divide them into groups for comparison of their divergence and similarity.Post-test is designed to check whether there is an improvement of scores compared with pre-test and what difference and similarity in terms of error types and number of each type of errors compared with the pre-test.Both pre-test and post-test were taken in the classroom for 30 minutes.

Ten tasks were assigned to finish in 5 weeks,namely twice a week.There would be a topic and direction posted on the Pigainet,and they are supposed to finish it without time limit,without any scoreline,and without invigilation.They could revise their compositions as many times as they wish according to revision feedback from Pigainet and submit their finals.In terms of the questionnaire,it was designed to examine participants' autonomy of language learning for research whether there is a correlation among the frequency of revision,the improvement of scores,proficiency of English,and autonomy of language learning.

Data Analysis and Results

Data analysis

When data collection was completed,SPSS was employed for analysis:The first was pre-test scores and posttest scores of 45 participants,which would be calculated to be improvement scores automatically by calculating variables in the transfer column of SPSS.Descriptive statistics such as Mean,SD,Maxim,and Minimum of each level group improvement scores were shown after running ANOVA for comparison of the improvement scores in pre-posttest between groups.Post Hoc was also conducted when the p-value is lower than 0.05 in the ANOVA test.Because the number of each level group is different,Bonferroni and Games-Howell were used to Post Hoc Multiple Comparisons.The second was the frequency of revision by using Pigainet of 45 participants who finished throughout ten tasks on the basis of calculating the sum of revision times of 10 tasks totally and then could get the descriptive statistics of Mean,Standard Deviation,Maxim,Minimum of each level group frequency of revision by using Pigainet.ANOVA was run to compare the frequency between groups.The third was the distribution of 45 participants' questionnaires,including strategy of language learning,the autonomy of language learning,and perception of Pigainet on the basis of coding and calculation.As for the questionnaire,the Chi-square test was conducted to investigate the relationship between autonomy of language of learning,and multiple comparisons of ANOVA were carried out to comparing the difference in terms of the autonomy of language learning between proficiency level groups in detail.

Results

RQ1.Is there any difference between proficiency level groups in terms of improvement scores of English writing in pre-and post-test,frequency of learners'computer-mediated interaction，and autonomy of language learning？

Firstly,table1 shows that the improvement scores participants got in pre-and post-test were significantly different between the 4 proficiency groups(F=4.705,p=.007).

In order to observe the difference in details between groups,the Post Hoc was needed to run.As the number of each level group is different,Bonferroni and Games-well of Multiple comparisons need to be conducted.It is necessary to only focus on the numbers with an asterisk,as the p-value is lower than 0.05.Post hoc shows that level1 is statistically different from level 3(MD=7.8,P=.039),and level 1 is statistically different from level4 with(MD=10.95,P=.007).The results imply the improvement scores are statistically different between Level1and level3 and between level1 and level4.It may be the result of the furthest distance between proficiency level1 and level4 in the pretest and then further distance between proficiency level1 and level3 in the pre-test.Therefore,the statistical difference of improvement scores existed between level1 and level 4 and between level1 and level3.

On the basis of calculating the sum of the frequency of revision by using Pigainet in 10 tasks,ANOVA for examining the difference in terms of the frequency of revision by using Pigainet,as table2 shows that there was not a significant difference in terms of frequency of learner-computer interaction between the four proficiency level groups(F=.733,p=0.539).

It is necessary to analyze data of the Liker-scale questionnaire about the autonomy of language learning with the Chi-square test.According to the result of the chi-square test of independence,there is a significant relationship between participants' proficiency level and their choice of strategy of language learning (χ2=234.838,df=12 and p<0.05) (Table 4.8).Given strongly disagree,disagree,neutral,agree,or strongly agree,126 out of 200 selected agree and 71 out of 200 selected strongly agree within level1;19 out of 260 selected disagree and 129 out of 260 selected neutral within level2;81 out of 280 selected neutral and 58 out of 280 selected strongly agree within level3;and 17 out of 160 selected disagree and 89 out of 160 selected agree within level4.The result implies that participants of proficiency level1 group have better autonomy of language learning through agreeing and strongly agree with the items of strategy in the questionnaire;participants of proficiency level2 have a neutral or negative attitude to the autonomy of language learning;participants of proficiency level 3 have a neutral and positive attitude to the autonomy of language learning;participants of proficiency level4 have a negative or positive attitude to the strategy of language learning.It is undoubtedly that the higher the proficiency level,the higher autonomy of language learning.While the participants in proficiency level lower may also have autonomy of language learning at present or in the future because they have a desire to improve their proficiency,at least it is real for participants as volunteers in this study.

Table 3.Chi-Square Tests for Autonomy

The results of the ANOVA test shows that there was a significant difference between groups in an (F=5.886,p=.002),and post hoc shows that level1 is significantly different from level2(MD=1.00269,p=.001),level1 is significantly different from level3(MD=.95925,p=.001),level1 is significantly different from level4(F=89500,p=.005).Level 1 is significantly different from other levels.

RQ2.What is the correlation between improvement of English writing,the frequency of learner-computer interaction,the autonomy of language learning,and proficiency level?

Table4shows that there is a significantly high positive relationship between improvement scores in pre-and posttest and proficiency levels(r=.499,p<.05);frequency of learner-computer interaction is significantly positively related with the autonomy of language learning(r=.487,p=.001),proficiency level has significant negative relation with the autonomy of language learning(r=-.335,p<.05),while positive relation with improved scores in pre- and post-test.Here the notes of proficiency level need to be taken as proficiency level1 are higher of scores in the pre-test,namely the number of level is bigger,the proficiency level is lower.Therefore,the result should be interpreted and transferred as proficiency level has a significant positive relation with the autonomy of language learning and negative relation with improved scores in pre-and post-test.In sum,the results imply that autonomy of language learning and frequency are positively statistically related with each other;different proficiency level participants revise their composition frequently;improvement scores in pre-and post-test are negatively related to proficiency level.

Table 4.Correlation between improvement scores,frequency of interaction,autonomy of language learning and proficiency level

Discussion and Conclusion

This research explored the effect and efficacy of L2 learners'computer-mediated interaction English writing improvement.The result from the analysis implied that there was no significant correlation between the frequency of learners'computer-mediated interaction and improvement scores in pre-and post-test(Bi&Bing,2020).Obviously,this research is different from previous studies in comparing divergences and correlations in terms of frequency of learners' computer-mediated interaction and improvement in pre-and post-test between proficiency levels:The result of ANOVA implied that the lower proficiency level groups were,the more frequent learners'computer-mediated interaction happened;and the lower proficiency level group was,the more improvement scores got.It is because of the causes of the more errors that need to be corrected in their compositions,and participants could get more instruction in writing by learners'computer-mediated interaction.The result of the chi-square test of analysis autonomy of language learning was concordant with previous studies that the higher proficiency level group was,the higher autonomy of language learning had (Littlewood,2002).In addition,the autonomy of language learning would influence the frequency of learners'computer-mediated interaction.Namely,the higher autonomy of language learning was,the higher frequent learners'computer-mediated interaction happened.

The implication of this result put forth that second language learners could improve their English writing through computer-mediated interaction,especially for proficiency level groups whose learning style was a bit shy to interact with instructors or peers.To motivate their confidence,SLA teachers could encourage them to practice their writing and revise their compositions by frequent interaction with the computer as a method of self-learning.This research has limitations:collected samples are small and were from a small group of students in a university.This research would go further in some qualitative aspects.Further qualitative research and analysis relevantly with a larger and multi-majors sample is necessary.