What happens to ESL students’ writing after three years of study at an English medium university?

Available online at www.sciencedirect.com

ScienceDirect Journal of Second Language Writing 28 (2015) 39–52

What happens to ESL students’ writing after three years of study at an English medium university? Ute Knoch a

a,*

, Amir Rouhshad a,1, Su Ping Oon b, Neomy Storch a,2

School of Languages and Linguistics, University of Melbourne, Parkville, Victoria 3010, Australia b Nanyang Junior College, 128 Serangoon Avenue 3, Singapore 556111, Singapore

Abstract The number of international students coming to study in Australian universities has increased exponentially in the past decade, mirroring trends in other English speaking countries. For the majority of these students, English is an additional language. The motivation for studying in an English medium university is presumably not only to gain a degree but also to improve English language proficiency. Studying in an English-medium university provides these students with a rich immersion environment. However, research on whether these students’ English proficiency does improve has to date been scarce, and has generally examined students’ gains after a relatively short duration (e.g., 12 weeks). The current study examined 31 undergraduate students’ L2 (ESL) writing proficiency following a three-year degree study in an Australian university. The study used a test–retest design which required participants to write a 30-minute argumentative essay on the same topic at the commencement and at the end of their degree program. A range of measures was used to assess writing, including global and discourse measures. All participants were also interviewed about their writing experiences at the university. The study found that students’ writing improved after three years of study but only in terms of fluency; changes in accuracy, grammatical and lexical complexity, as well as global scores of writing fell short of significance. The students reported in the interviews that they were not required to do much writing in their degree studies and when they did, their lecturers almost exclusively commented on the content of their writing. We discuss our findings in relation to the students’ L2 proficiency, the nature of the test, and the nature of their immersion experience. # 2015 Elsevier Inc. All rights reserved. Keywords: Academic writing; Writing development; International students; Second language writing

Introduction The number of international students coming to Australian universities to pursue their degree has increased greatly in the past decade. In 2013 international students formed a substantial proportion (almost 20%) of the 1.1 million higher education students in Australia (Australian Government Department of Education, 2014a). These students come from a diverse range of countries, but predominantly from Asian countries (Australian Government Department

* Corresponding author. Tel.: +61 3 83445206. E-mail addresses: [email protected] (U. Knoch), [email protected] (A. Rouhshad), [email protected] (S.P. Oon), [email protected] (N. Storch). 1 Tel.: +61 3 94793049. 2 Tel.: +61 3 83445208. http://dx.doi.org/10.1016/j.jslw.2015.02.005 1060-3743/# 2015 Elsevier Inc. All rights reserved.

40

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

of Education, 2014b). For the majority of these students English is an additional language (EAL). Thus studying in Australia presents for these students not only an opportunity to gain a degree but to concurrently attain higher proficiency in English. Hyland (2013, p. 54) suggests that English language skills are ‘‘becoming less a language than a basic academic requirement for many users around the world.’’ The attainment of high levels of language skills is indeed what universities expect from all their students. The published graduate attributes of leading universities in Australia state explicitly that all students are expected to develop excellent communication skills in addition to acquiring disciplinary knowledge when studying for a degree course. These attributes often single out written communication. For example, at the University of Melbourne, the document ‘‘Melbourne graduates’’ (University of Melbourne, n.d.) states that Melbourne University graduates are ‘‘effective oral and written communicators.’’ However, research on whether students’ second language writing develops as a result of studying in L2-medium universities is scant. What research exists suggests that these learners’ writing does not necessarily improve whilst studying for their degree courses (e.g., Humphreys et al., 2012; Storch, 2007, 2009). The findings of such studies bear theoretical and pedagogical implications for the field, as well as political implications for the educational policy makers at universities who need to examine the extent to which this particular graduate attribute (i.e., written communication skill) has been achieved. However, most of these studies have examined writing development following usually one semester of study (i.e., 12 weeks) which is a period of time perhaps not long enough to record improvement. The present study set out to investigate second language writing development of EAL students following three years of studying an undergraduate degree course in an English-medium university. In addition, the students’ perceptions and reflections on their writing development were examined through individual interviews at the end of the third year. We begin with a review of the available studies on L2 writing development before describing our study. Research on L2 writing development There is a small but growing body of research on second and foreign language writing development. Studies on foreign language writing development have been conducted either with young learners (in high schools) receiving limited language instruction (e.g., Gunnarsson, 2012; Verspoor & Smiskova, 2012) or adult learners who were beginners (e.g., Vyatkina, 2012). Of more relevance to our research are studies on second language writing development with adult learners of high intermediate proficiency in immersion environments. These studies have been conducted following either a specific course of instruction in the L2 or an immersion experience in the L2-medium university without formal language instruction. All these studies have tended to employ a test–retest design and writing development is gauged by using composition/band scores (e.g., Sasaki, 2007, 2009, 2011; Storch, 2009) and/or a range of discourse-analytic measures (e.g., Larsen-Freeman, 2006). These discourse-analytic measures attempt to assess improvement in terms of fluency, accuracy, and complexity, the latter distinguishing between grammatical/ syntactic and lexical complexity. Our discussion of these studies is organized under three sections. The first two sections distinguish between research based on the length of time between test and retest. We start with studies whose duration was shorter than one year. This is followed with studies that were one year or longer in duration. In our third section we review the studies that examined the factors that could explain writing development or lack thereof. L2 writing development after less than one year The bulk of research on second language writing development has focused on development following one semester of study in the L2 medium university. In some of these studies, the students participated in a course of language instruction (e.g., Elder & O’Loughlin, 2003; Humphreys et al., 2012; Larsen-Freeman, 2006; Storch & Tapper, 2009) with only scant research (e.g., Storch, 2007, 2009) available on writing development of students without formal L2 instruction. The studies have employed a range of measures to gauge development in L2 writing, and their results suggest some improvement, albeit depending on the measures used. Available research using composition/band scores of writing suggests that the writing of second language learners improves following three to six months of instruction (Green, 2005; Hu, 2007), or even following 8–12 weeks of instruction (e.g., Archibald, 2001; Elder & O’Loughlin, 2003). Sasaki (2007, 2009) showed that study abroad (ranging

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

41

from 1.5 to 11 months) significantly increases composition scores of students when combined with L2 writing instruction. The increase in band/composition scores in the cited studies (e.g., Hu, 2007; Sasaki, 2011) is regardless of context (ESL vs. EFL; pre-university vs. university) and participants’ L2 proficiency level. Among the few exceptions is the study by Tsang and Wong (2000) which reported that band scores of TESL undergraduates did not increase significantly following 14 weeks of instruction. However, this was a relatively small-scale study, with only 11 participants. Unlike composition/band scores, studies using discourse-analytic measures (e.g., fluency, accuracy, complexity) to gauge writing development have shown mixed findings. Two measures of fluency (word count and ratio of words to Tunits) employed in the study by Larsen-Freeman (2006), for example, showed an increase following six months of language instruction; however, the results were not validated statistically. Similarly, Tsang and Wong (2000) found that the fluency of their EAL undergraduates (word count) increased significantly following 14 weeks of language instruction in a university in Hong Kong. However, Storch and Tapper (2009) showed that EAL students did not increase significantly in their fluency (measured via word count and ratio of words to T-units) following 10 weeks of EAP instruction in a university in Australia. Similar mixed findings are evident about accuracy and complexity. In terms of grammatical accuracy, Tsang and Wong (2000) and Larsen-Freeman (2006), who operationalized accuracy as the ratio of error-free T-units (EFT/T), showed that it increased following three to six months respectively of language instruction. However, in Storch and Tapper (2009) EFT/T did not increase significantly following 10 weeks of EAP instruction. Other measures of accuracy (e.g., ratios of error-free clauses, errors to words, and errors to T-units) also fell short of significance in a number of studies (Deng, Lee, Varaprasad, & Lim, 2010; Shaw & Liu, 1998; Storch & Tapper, 2009), regardless of the context of study (ESL vs. EFL). Grammatical complexity, when operationalized as the ratio of clauses to T-units and the length of T-units, was found by some studies (e.g., Larsen-Freeman, 2006; Tsang & Wong, 2000) to improve following instruction. However, other studies that operationalized grammatical complexity differently reported no improvement following instruction. For example, Shaw and Liu (1998) operationalized complexity as the use of nominalization, use of subordination or change in pattern of subordination, and reduction of clauses to prepositional or participial phrases. The authors showed that complexity did not change significantly following two to three months of EAP for university-bound students in England. Results concerning lexical complexity have also been mixed. Storch and Tapper (2009) operationalized lexical complexity as the percentage of words which appeared in Coxhead’s (2000) Academic Word List (AWL). The authors found a significant increase in lexical complexity following 10 weeks of EAP instruction; however, in Deng et al.’s (2010) study, which operationalized lexical complexity similarly, the difference fell short of significance following a similar period of instruction with university students in Singapore. Taken together, available research on L2 writing development following formal instruction suggests that such instruction results in a significant increase in band/composition scores. However, discourse measures showed mixed results following instruction. Perhaps, the duration of these studies was not long enough to result in a significant improvement. Ortega (2003), for example, claims that 12 months of language instruction are needed to show improvement in complexity. It is important to note that because in these studies the students participated in formal language instruction, the studies investigated whether instruction made a difference to writing proficiency, rather than immersion per se. The studies by Storch and Hill (2008) and Storch (2009) are among the few that investigated the writing proficiency of EAL students following a semester (12 weeks) of degree study that did not include language instruction. Both of these studies showed that the discourse measures of fluency (number of words), accuracy (ratio of error-free T-units and clauses), and grammatical complexity (ratio of clauses to T-units and dependent clauses to clauses) did not change over 12 weeks of degree study. However, band scores of organization and style, content, form, and average writing score improved significantly in both studies over the same period. Clearly the dearth of studies calls for further investigation of this issue. L2 writing development after one year or longer Studies investigating L2 learners’ writing after immersion in the L2 environment of one year or longer are rare. These include the studies by Craven (2012), O’Loughlin and Arkoudis (2009) and Knoch, Rouhshad, and Storch (2014) investigating the impact of degree study on the writing of ESL students and by Serrano, Tragant, and Llanes

42

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

(2012) who investigated the impact of the study abroad experience on the writing of EFL students. Their findings are more uniform particularly in terms of some discourse measures. In terms of composition/band scores, the study by Knoch et al. (2014) found no improvement in composition/band scores after one year of study; similar results were reported by Craven (2012) after 18–36 months of study. O’Loughlin and Arkoudis (2009) reported some improvement but only for those with initial low band scores. O’Loughlin and Arkoudis examined development in English language proficiency of 66 EAL students who were completing undergraduate or graduate programs at a large Australian university. Proficiency development was assessed using IELTS scores after 1.5 or 3 years, depending on the participants’ degree program. Participants presented the original copy of their IELTS results used for entry to the university and also sat for an official IELTS test toward the end of their final semester. The results showed that writing skill showed the least improvement (.20) among the language skills, and that participants with a lower writing score at the outset showed the greatest improvement, a finding that is consistent with Elder and O’Loughlin (2003). In terms of discourse measures, fluency has been reported to improve after one year of immersion experience in an L2 medium university by Knoch et al. (2014) measured by words produced, and by Serrano et al. (2012) measured via the ratio of words to T-units. Both studies found that accuracy (measured as ratios of error-free T-units, error-free clauses, and errors to words) fell short of significance. There are mixed findings for syntactic complexity. Whereas Knoch et al. (2014) observed no improvement, Serrano et al. (2012) reported significant improvement over one year. As far as lexical complexity is concerned, a range of measures was employed by Knoch et al. (2014) but none recorded significant change. In Serrano et al.’s (2012) study, lexical richness (type/token ratio) improved but the results did not reach significance. Thus the few longitudinal studies investigating L2 writing development suggest little improvement over time, with the exception of fluency. The findings concerning grammatical and lexical complexity are mixed. However, given the small number of studies to date that have investigated writing development beyond one year, it is difficult to draw firm conclusions about the impact of immersion in an L2 medium university on L2 writing without further investigation. Variables which account for improvement or lack thereof in L2 writing Few studies (e.g., Elder & O’Loughlin, 2003; Green & Weir, 2003; Storch & Hill, 2008) also administered questionnaires to identify factors that influence second language writing development. Elder and O’Loughlin (2003), for example, showed that participants’ perceptions of the importance of academic writing for future studies were a predictor of writing gain. Storch and Hill (2008) also found a positive correlation between accuracy in writing and participants’ integration (operationalized as having Australian friends). The authors also conducted interviews with a small number of their participants and concluded that lack of opportunity to produce extended writing and to receive feedback on writing during the course of study could be possible reasons for lack of improvement in L2 writing despite immersion in the L2 university environment. Serrano et al. (2012) examined the attitudes of study abroad students toward English people and English language. The authors showed that students who rated English people as more sociable and humble showed greater improvement in their accuracy than students who rated them as less sociable or snobbish. The authors explained that positive attitudes toward the L2 language community may encourage learners to interact with native speakers and thus provide more opportunities for receiving comprehensible input and producing output. In addition, students who were living with English-speaking people showed more improvement in lexical richness (type/token ratio) than students who were living with one or more shared L1 speakers. However, most of the studies employed questionnaires to account for their findings rather than interviews, which yield richer data. Clearly factors which may explain learners’ L2 writing development or lack thereof is an area which requires further in-depth investigation. In summary, our review of the literature showed a dearth of research on writing development, particularly tracking development beyond one year of immersion in an L2 university context. What research exists seems to have investigated writing development mainly after one semester of study and where students participated in a formal language subject concomitantly with their degree course. The results of this body of research are mixed. Research investigating the reasons for improvement or lack thereof has relied on questionnaires, rather than interviews with more in-depth analysis of the learners’ experiences. Our study set out to address some of these research gaps.

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

43

The present study The present study sets out to investigate the possible impact of three years of undergraduate degree study in an Englishmedium university on the English writing development of EAL students. We employed a test–retest research design. Participants completed two Diagnostic English Language Assessment (DELA; described in more detail below) essays three years apart (Time 1 and Time 2). Writing development was gauged using rater-based band scores and a wide range of discourse measures. Participants were also interviewed at the end of the study to explore their perceptions on their writing development over the three years. Therefore, the current study investigated the following questions: 1. Does the writing of L2 students improve after three years of study in a degree course in an L2-medium university? 2. What factors may explain L2 learners’ improvement or lack thereof? Participants Data was collected from 31 EAL students over three years of their degree study at a large university in Australia. At entry all participants had satisfied the university’s English language proficiency requirement in one of several different ways (e.g., having an IELTS overall score of 6.5, completing secondary education in an Australian school, etc.). Twenty participants (64.15%) were female and 11 male (35.48%). Nineteen (61.29%) were either native speakers of Chinese or Malaysian. The rest were from countries such as Vietnam, South Korea, Germany, and Fiji. Seventeen participants (54.83%) received their secondary education in Australia and the rest did in their home countries. In addition, 20 participants (64.51%) were studying an undergraduate degree in the Faculty of Business and Economics or the Faculty of Medicine, Dentistry and Health sciences. These faculties tend to attract a large number of international students. The rest were mainly from the Faculties of Land and Environment, Arts, and Science. Pseudonyms are used throughout this study to protect the identity of the participants. Instruments Diagnostic English Language Assessment (DELA) Although all students at the university are required to meet minimum English language proficiency requirement at entry, those whose proficiency is at the minimum level required (e.g., IELTS 6.5 or equivalent) are strongly encouraged to sit the Diagnostic English Language Assessment (DELA). DELA was designed and validated by the university’s Language Testing Research Centre. The rationale for this policy is that even students who meet the proficiency requirement can benefit from further English language development and that it is important to diagnose students early who may be at risk of failing due to inadequate L2 proficiency. DELA comprises sub-tests of reading, listening, and writing. The writing component requires participants to write an argumentative essay in 30 minutes. Some relevant ideas are provided to help students generate ideas. The writing sub-test is assessed based on the three criteria of (a) Organization and style, (b) Content, and (c) Form. Each of these criteria is assessed using a 6-point scale and then averaged to compute the average writing score. All other sub-tests are also assessed on a 6-point scale, with 6 representing advanced proficiency. Based on their DELA scores, participants are divided into three categories which delineate the degree of language support they may need. Participants with an average DELA score of 3.3 or below are classified as at-risk and are required to take a credit-bearing English for Academic Purposes (EAP) subject in their first semester. Those whose score ranges from 3.4 to 4 are classified as borderline students and are recommended to attend language support workshops run by the university. Finally, those who score above 4 are classified as proficient students. It is worth noting that a DELA score of 4.5 is approximately equivalent to IELTS 7 (although no formal linking study has been undertaken); therefore, even students who are classified as proficient are likely to benefit from some language support programs. Interview schedule A semi-structured interview was conducted with each participant at Time 2. The interview questions (see Appendix A) were developed based on our reading of the literature (e.g., Storch & Hill, 2008) and extensive

44

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

discussions among the researchers. We enquired about the participants’ academic experiences during their study at the university but focused mainly on their L2 writing opportunities as well as how much feedback they received on their writing. Thus, the participants were asked about writing requirements of the subjects they undertook throughout their three years, the type of assessment tasks, and about the nature of feedback they received on their written assignments. We also asked the participants how they assessed their writing skill compared to when they commenced their degree program. Background questionnaire Participants were also asked to complete a brief background questionnaire designed to collect basic bio data information as well as information on previous language learning and study experiences and standardized English language test scores. Data collection procedures Data collection began in February 2009 (Time 1), just prior to the commencement of the academic year. For the writing component, students were required to write an argumentative essay on the topic of corporate advertising. They also completed a questionnaire which elicited background information. A year later all who sat for DELA in 2009 were invited to take part in a writing development study. One hundred and one students (101) consented to participate (see Knoch et al., 2014). Two years later in November 2011 (Time 2) an email was sent out to invite all those 101 students for a further data collection. Thirty-one participants consented to partake and formed the participants of the present study. At Time 2 participants completed the writing component of the DELA and wrote about the same prompt as at Time 1. The topic was held constant to control for the possible effect of topic. Individual interviews (see Appendix A) with the participant were conducted by the first and second authors. Participants were paid for their participation at Time 2. Data coding and analysis All essays completed at Time 1 and 2 were marked using the DELA rating criteria (i.e., content, form, organization and style and an average score was then computed). The essays were marked by an experienced DELA rater who had several years of experience in marking DELA essays. The rater was not aware of the purpose of the study and therefore had no idea that two essays were written by the same person. A subset of essays was rated by a second rater to ensure rater reliability (reported below). In addition, essays were coded for a wide range of discourse-analytic measures (e.g., accuracy, fluency, complexity) either manually or automatically or a combination of both, as shown in Table 1. As the table shows, many of these measures are T-unit or clause based. To code the essays for T-unit and clause boundaries, we followed Cummings et al.’s (2006) coding scheme. In the following we explain the coding of these discourse-analytic measures in greater detail. Accuracy was measured manually via the percentage of error-free Tunits and clauses. Following Knoch et al. (2014) spelling errors were not counted and clauses and T-units were considered as error-free when a highly competent English user judged them as such. During a pilot study we also explored other measures of accuracy, including the number of errors per 100 words. We hoped that such a measure would give us a better picture of subtle changes in accuracy. However, the measure provided no different information than the percentage of error-free T-units and clauses. Fluency was measured automatically by counting the number of words for each essay by a Perl script program designed for that purpose by a qualified software programmer. The Perl script counts contractions and numbers as one word. Other measures of fluency employed in the current study were the number of T-units and T-unit length, i.e., the average number of words per T-unit. Moreover, the essays were coded for grammatical and lexical complexity. We attempted to sample broadly from available measures (see e.g., Bulte´ & Housen, 2012; Norris & Ortega, 2009). Grammatical complexity was measured via the average numbers of words per clause, clauses per T-unit, and the ratio of dependent clauses to all clauses. For lexical complexity, three different measures were used which included percentage of words from the Academic Word List (AWL), lexical sophistication, and average word length. The AWL was developed by Coxhead (2000) and includes 570 word families which comprise 10% of words in a wide range of academic texts. Vocab Profile (Cobb, 2002) was employed to compute the percentage of AWL words in each script. Lexical sophistication was measured by

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

45

Table 1 List of discourse-analytic measures & type of coding. Category

Measures

Type of coding

Accuracy

Error-free Clauses (EFC) Error-free T-units (EFT)

Manual Manual

Fluency

Number of words Number of T-units T-unit length

Automated Manual Automated/manual

Grammatical complexity

Clauses per T-unit Words per clauses Ratio of dependent clauses to clauses

Manual Automated/manual Manual

Lexical complexity

Percentage of Academic Word List Lexical sophistication Average word length

Automated Automated Automated

dividing the number of sophisticated words by the number of content words. The tokens from the Academic Word List and the Off-List words (i.e., words which do not appear on any of the word lists) comprised the list of sophisticated words. The Off-List words were screened to exclude non-words or abbreviations. The third measure of lexical complexity was average word length, which was computed by dividing the total number of characters by the number of words in each script, using the word and character count features of MS Word. To compute the inter-coder reliability scores (for band score), a subset of six essays (19%) was rated by a second rater. Because of the small number of essays, we chose to calculate inter-coder reliability using simple percentage exact agreement scores in each category (organization and style, content, and form). The results were .92, .89, and .95 respectively. For the discourse measures, a subset of 14 essays (comprising 45% of total essays) were randomly chosen and double coded. The inter-coder reliability scores for the discourse measures were computed using simple percentage agreement and were .95, .90, .83 and .85 for T-unit, clause, error-free T-unit, and error-free clause respectively. To examine how the discourse measures and band scores compare between Time 1 and Time 2, first the normality of distribution of the variables (i.e., discourse measures and band scores) were tested at both times using Shapiro–Wilk tests. The results showed mixed findings as some variables had a normal distribution while others did not. Therefore, both parametric (T-test) and non-parametric (Wilcoxon Signed Rank Tests) were run and both showed the same variables as being significant. Thus only the results of non-parametric tests are reported, for the sake of consistency. As we conducted 15 statistical tests on the same data set, Bonferroni correction was applied and the p-value was lowered to 0.003. While the use of the Bonferroni correction can increase the probability of a Type-II error (Larson-Hall, 2010), we preferred being cautious with the interpretation of the data. The results were, however, not changed by this correction. The interviews were transcribed and the emerging themes related to the nature of the participants’ written assignments, feedback, perceived improvement/lack of improvement were analyzed qualitatively. It should be noted that not all participants were asked the same questions due to the semi-structured nature of the interviews. We also elicited and coded the participants’ reasons for improvement/lack thereof. To answer Research Question 2, the amount of writing assignments that participants did throughout the three years, participants’ responses were classified as ‘‘little/no writing,’’ ‘‘fair/small amount,’’ or ‘‘large amount of writing.’’ Participants were classified as students with little/no writing when they had writing assignments for only a few of their subjects and the length of their assignments generally did not exceed 1500 words. Participants who wrote assignments for nearly half of the subjects, and when the length of their assignments ranged between 1000 and 2000 words were coded as students with fair/small amount of writing. Finally participants were classified as students with large amount of writing assignments if they wrote assignments for nearly all subjects and the assignments were 3000–4000 words in length or the total number of words written for all assignments revolved around 20,000 words.

46

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

Results Research Question 1: Writing development over time To answer the first research question, we first compared the band scores for each of the criteria as well as the average writing scores between Time 1 and Time 2. As shown in Table 2, the means of the band scores decreased from Time 1 to Time 2 for all band scores; however, the differences were more noticeable for form and average writing scores. The results of the Wilcoxon Signed Rank Tests showed that for all band scores the differences fell short of significance as they were smaller than the set p-value .003 (see Table 2). Therefore, the band scores of participants did not differ statistically over time. We then compared writing at Time 1 and Time 2 using the discourse measures. Table 3 presents the descriptive statistics and Wilcoxon Signed Rank Test for the fluency measures between Time 1 and 2. As the table shows the number of T-units and T-unit length slightly increased in mean and median at Time 2, but the differences fell short of significance. However, as the table shows, participants generated a larger number of words at Time 2 and the results of the Wilcoxon Signed Rank Test showed that the difference was significant ( p < .003). Therefore, participants were able to write significantly longer essays in the set time of 30 minutes at Time 2. Table 4 presents the descriptive and inferential statistics on accuracy measures over time. As the table shows there were very slight changes across time, and the differences fell short of significance ( p < .003) in both ratios of errorfree T-units and error-free clauses. Thus far the results showed that participants generated more words at Time 2 but the accuracy of their essays did not change over time. We now turn our attention to grammatical complexity measures. Table 5 shows the descriptive and inferential statistics on grammatical complexity measures. As the table shows, the means and medians changed slightly across the time but the differences did not reach significance ( p < .003). Finally, Table 6 presents the descriptive and inferential statistics on lexical complexity measures. As the table shows, means and medians changed slightly and the p-values were larger than .003. In summary, the current study showed that nearly all discourse measures, excluding the number of words, fell short of significance. In other words they did not change after three years of degree study in an L2 medium university. However, participants were more fluent at Time 2 as they produced more words within the same time period. In the next section we report the findings of the interviews. Although not directly related to this research question, we also investigated whether there was a significant difference in the improvement in band scores and discourse measures between participants who completed their secondary education in Australia and those who were newly arrived international students. In the same vein, we also compared the writing development of students who were more proficient and those less proficient at the outset of the study. In both cases, no difference was found between the two groups, and due to space limitations we have decided not to report the full results of this analysis here.

Table 2 Comparing analytic and average DELA scores at Time 1 and Time 2. Time

Mean

SD

Median

Wilcoxon signed rank test Z

p

Organization & style

Time 1 Time 2

4.35 4.09

.87 .83

4.00 4.00

1.71

.087

Content

Time 1 Time 2

4.48 4.19

.85 1.01

4.00 4.00

1.88

.060

Form

Time 1 Time 2

4.16 3.80

.73 .87

4.00 4.00

2.66

.008

Average writing score

Time 1 Time 2

4.33 4.03

.75 .82

4.33 4.00

2.58

.010

Note: The p-value is set at .003 after doing Bonferroni correction

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

47

Table 3 Descriptive statistics on fluency measures at Time 1 and 2. Measure

Time

Mean

SD

Median

Wilcoxon signed rank test Z

p

Number of words

Time 1 Time 2

281.16 325.06

81.12 84.62

275 329

3.35

.001

Number of T-units

Time 1 Time 2

17.65 19.74

5.14 5.56

17 20

2.10

.035

T-unit length

Time 1 Time 2

16.27 16.96

3.40 3.58

15.50 15.83

1.96

.050

Note: The p-value is set at .003 after doing Bonferroni correction

Table 4 Descriptive statistics on accuracy measures at Time 1 and 2. Measure

Time

Mean

SD

Median

Wilcoxon signed rank test Z

p

Error-free T-unit ratio (EFT/T)

Time 1 Time 2

.33 .31

.20 .18

.33 .29

1.01

.300

Error-free clause ratio (EFC/C)

Time 1 Time 2

.43 .41

.22 .18

.41 .40

.99

.310

Table 5 Descriptive statistics on grammatical complexity measures at Time 1 and 2. Measure

Time

Mean

SD

Median

Z

p

Clauses per T-unit ratio (C/T)

Time 1 Time 2

1.47 1.47

.22 .22

1.47 1.40

.47

.640

Average clause length in word (W/C)

Time 1 Time 2

11.08 11.60

1.56 1.94

11.21 11.25

1.76

.800

Ratio of dependent clauses to clauses (DC/C)

Time 1 Time 2

.28 .27

.09 .08

.30 .27

.71

.500

Table 6 Descriptive statistics on lexical complexity measures at Time 1 and 2. Measure

Time

Mean

SD

Median

Z

p

Lexical sophistication

Time 1 Time 2

.50 .50

.07 .06

.50 .49

1.09

.270

Percentage of Academic Word List

Time 1 Time 2

7.67 8.12

2.61 2.46

7.28 7.73

.97

.330

Average word length

Time 1 Time 2

5.33 5.27

.28 .32

5.34 5.32

1.13

.260

48

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

Table 7 Proportion of students reporting on the amount of writing produced over 3 years.

1st year 2nd year 3rd year

Little/no writing

Fair/small amount

Large amount

43.33% 44.44% 38.70%

48.33% 44.44% 41.93%

8.33% 11.12% 19.37%

Research Question 2: Factors influencing writing development To answer the second research question, we asked participants about the length/nature of writing assignments throughout their degree study, the nature of feedback from lecturers, and self-perception of their improvement/lack thereof. We compared their self-perception of improvement with quantitative measures of the discourse features to examine whether they were consistent. The length/nature of assignments & feedback Most of the participants reported that they did little or a small amount of writing in each of the three years. Table 7 summarizes the results, showing what proportion of the students interviewed reported about the length of their assignments. As the table shows, in their first year, 43.33% of participants had no or only little writing. For example when Christina (Faculty of Medicine, Dentistry and Health Sciences) was asked about the amount of writing she did for assignments in her first year of study, she said ‘‘we don’t write that many assignments, is more, we have more examine task, so, I think we only had one assignment in the first year. I think it was about 1000 words.’’ A slightly higher proportion (48.38%) of participants had small/fair amount of writing. For example, Maya (Faculty of Business and Economics) reported that ‘‘for each subject its 10 to 20 percent because its commerce and commerce need to write essays . . . about a 1000 [words] for each one yeah.’’ Similarly Rosa (Faculty of Land and Environment) claimed that ‘‘half of them [subjects] I guess, its sometimes practical sense sometime design projects. . ..’’ She also mentioned that her assignments in general were 1500–2000 words long. Similarly in the second year, 44.44% of respondents mentioned that their assignments involved little or no writing and the same proportion reported a fair/small amount of writing. When asked whether he had to do more writing in the second year, Kami (Faculty of Business and Economics) replied: ‘‘no, it’s more about math, and math, financial math and probability statistics, those kind of things.’’ He claimed that he wrote 4000 words in total for all assignments in his second year. Mat (Faculty of Science) also commented that ‘‘Mostly assignments came in semester 2, second year . . . second year there were more assignments but they didn’t count for much marks . . . you get three assignments for a subject throughout the semester but they were not really long, about 1000-word essays . . . altogether yeah three thousand words.’’ By the third year, about 20% of respondents reported that writing assignments became more substantial in length. For example, Yani (Faculty of Arts) mentioned that she wrote on average 4000 words for each subject in the third year, but in Year 1 and Year 2 wrote only two or three essays during each year, and the assignments were 1500–3000 words in length. Ray (Faculty of Medicine, Dentistry and Health Sciences) wrote a portfolio of 90 pages in Year 3, while in Year 1 and 2 he wrote 1500 words and 6000 words respectively for all assignments in total. Taken together, the results of the interview show that throughout their degree course most participants did not write much for their assignments. A closer examination of the data also revealed that the majority of participants who had little/ no writing were from the Faculty of Medicine, Dentistry and Health Sciences or the Faculty of Business and Economics. Participants were also asked whether lecturers commented on their written assignments. Nearly all respondents (90.32%) said they did not receive comments on the quality of their writing. For example, Misi (Faculty of Business and Economics) commented that: Every written assignment we only get feedback on how the assignment is good or bad, but- general, but the lecturer don’t tell you this word is not right or this sentence is not right, or your or your . . . structure is not right, they don’t talk about that, they only talk about you should write this, you should write that, you don’t write this, you don’t write that . . . yes just focus on the question, you answer the question, no matter how you write, you get marks.

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

49

Self-perception of their improvement or lack thereof/possible reasons & lecturers’ CF Most respondents (87.50%) mentioned that before commencing their degree they presumed that studying at an L2medium university would improve their L2 writing. For example, Maya (Faculty of Business and Economics) said ‘‘Yeah . . . I was hoping so, because three years is, three years experience in university obliviously, I I live here for three years, and the three years would make at least some difference in my writing skills.’’ The majority of respondents (72.41%) believed that their writing skill had improved over the three years of their study, with academic writing style (e.g., paragraph structure) and vocabulary being the most frequently cited areas by 73.68% and 63.15% of respondents respectively. However, only a quarter of respondents (26.31%) were of the opinion that their grammatical accuracy improved during their study. Of the respondents who believed they had improved, half of them (52.38%) attributed their improvement to reading text books, articles, newspapers and/or listening to tutorials, news, etc. For example, Raya, a student from Singapore, (Faculty of Medicine, Dentistry and Health Sciences) believed that her writing improved, that at the end of her course her ‘‘writing is more cohesive form more logical, it flows.’’ She attributed her improvement to extensive reading outside university. She said: The style of writing and the vocabulary that I had, I don’t think they have improved much from the university, if if if I had were to learn more it was from myself, like reading papers or magazines, and improving myself, but I don’t think during university I learned my writing. Another reason mentioned by a number of participants (8) for perceived improvement in writing was the requirement to produce assignments. For example, Parsa (Faculty of Medicine, Dentistry and Health Sciences) explained his improvement in terms of ‘‘just practicing it so being forced to write for the assignment . . . that was about it . . . I started reading a few books just English books, never read much but now I started reading a bit, so that might have helped.’’ The respondents who believed that their writing did not improve (27.58%) attributed this mainly to lack of practice in writing or the dearth of feedback on their language use. Tara, a Chinese student in the Faculty of Medicine, Dentistry and Health Sciences, claimed that her writing skill deteriorated over time due to lack of writing practice. In her interview she said: ‘‘I do feel that when I was in high school we had English course and we had to write a lot. But I guess practice may change, like we don’t have a lot of writing, my writing skill actually degraded.’’ Moreover, she noted that she revised her assignments for spelling and not for grammar as she was of the opinion that lecturers would not care about grammar. Discussion The current study set out to investigate whether the writing (measured via band scores and discourse-analytic measures) of second language learners develops over three years of degree study in an L2-medium university and what factors can explain improvement or lack thereof. Consistent with Knoch et al. (2014) band scores of form, content, organization and style and average writing score fell short of significance suggesting that three years of degree study did not improve these aspects of the students’ writing. These findings are inconsistent with findings reported by others who reported improvements on band scores even after one semester of study (e.g. Storch, 2009; Storch & Hill, 2008). This inconsistency could be explained in relation to proficiency level of the participants in these studies. In studies conducted by Storch (2009) and Storch and Hill (2008), the students’ proficiency was such that they were recommended to seek support. This suggests that in both studies the majority of participants were either at-risk or borderline (as identified by the Diagnostic English Language Assessment); however, in the current study well over half of the participants (17/31 or 54.83%) were grouped as proficient at the outset. The studies by O’Loughlin and Arkoudis (2009) and Elder and O’Loughlin (2003) showed that participants with lower initial writing score are more likely to show improvement. In terms of the discourse measures, our findings regarding fluency are in line with the findings of other longitudinal studies (e.g., Knoch et al., 2014; Serrano et al., 2012). Fluency (measured via word count) increased significantly over three years of degree study, suggesting that participants were able to produce more words within the same allotted time. The increase in fluency in the current study could be attributed to the fact that over the three years, participants had writing opportunities inside and outside the academic context (e.g., using social media) and thus had not only a greater word knowledge to draw on when writing but also increased efficiency in retrieving this vocabulary.

50

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

However, accuracy did not change over time, consistent with previous studies (e.g., Knoch et al., 2014; Serrano et al., 2012; Storch, 2007, 2009) regardless of the interval between the pre-test and post-test. Additionally, grammatical and lexical complexity did not change over time, consistent with Storch (2009) and Knoch et al. (2014). One reason that could explain these findings is the relative short length of the writing task in DELA. In short texts, the number of dependent clauses is likely to be quite finite. As far as lexical complexity is concerned, the lack of improvement could be explained in relation to the assigned topic which allowed for a limited range of vocabulary. Perhaps participants could have shown improvement in lexical complexity had they been asked to write on a discipline-related topic, as was the case in the study conducted by Storch and Tapper (2009). The interviews provide additional explanation for our findings. The most salient finding in the interview data was that most of the participants reported having to produce very little writing over the course of their degree program. As argued by a number of researchers (e.g., Rouhshad, 2014; Storch, 2013; Williams, 2012), writing practice may provide learners with language learning opportunities. When writing, learners have more processing time when compared with speech. Kormos (2012) pointed out that oral production of 100 words by L2 learners takes one minute while writing the same number of words can take as long as 30 minutes to produce. This additional time allows students to focus on form. Cumming’s (1990) study, using think aloud protocols, showed that a third of the decisions made by L2 writers in the process of composing a text were about the content and language use concurrently. Some participants in our study were aware of the importance of extensive writing as in the interviews they attributed their perceived improvement or lack thereof to the amount of writing they were required to do for their assessment tasks throughout their degree. Future studies need to investigate the impact of the writing demands in different degree courses on EAL students’ writing development over time. An additional reason for the lack of improvement was the fact that the majority of participants did not receive feedback on their writing. Ferris (2003) and Leki (2006), among others, argue that feedback on writing may be the single most important element that leads to improved writing. Recent research on corrective feedback (e.g., Bitchener & Knoch, 2009; Ellis, Sheen, Murakami & Takashima, 2008; Van Beuningan, de Jong, & Kuiken, 2012) shows that corrective feedback leads to improved accuracy in writing. A quarter of the respondents who believed that their writing did not improve attributed it to lack of practice in writing and dearth of feedback on their writing. The findings of this study should be interpreted with caution, not only because of the relatively small size of the study but also because of the nature of the writing tasks used. Had we used more discipline-specific topics in the writing tasks, we may have found some improvements in lexical complexity. After three years of degree study we can assume that the students’ discipline-specific vocabulary grew via the extensive reading the students had to do in their degree courses. However, the use of generic tasks did not provide our participants with opportunities to display their newly acquired discipline-specific vocabulary. Students might have also developed their writing skills in areas not captured by the tasks. Engineering students, for example, might have gained skills in technical writing or report writing and generic writing tasks such as the ones employed by DELA are not able to capture these skills. This is also the point made by Norris and Mancho´n (2012) in their review of studies on L2 writing development. The authors note that research on writing development needs to take into consideration the extent to which the tasks used enable learners to display their knowledge and in turn what it is that researchers can observe about writing development. Other aspects of academic writing (e.g., incorporating sources, development of an authorial voice) were also not examined in the current study nor could they be captured by the measures used. Another shortcoming of this study relates to the measure of syntactic complexity employed. The measures we employed were based on clause per T-unit ratios, a commonly used measure in studies on writing development. However, recently a number of scholars (e.g., Biber, Gray, & Poonpon, 2011) have questioned the use of these measures of complexity when applied to writing. Biber et al. (2011), using a corpus-based analysis of written and oral texts, demonstrate that these measures are more appropriate for assessing oral language development rather than written language development. Their analysis reveals that complex phrases (e.g., nominalizations) are more prevalent in formal academic writing and are thus more suitable measures of writing development. Authors such as Bulte´ and Housen (2012) have also argued for a wider construct of complexity than was operationalized in this study. Future studies thus need to consider using a broader range of measures that capture features characteristic of academic writing, including more appropriate measures of syntactic complexity. Nevertheless, the findings of the current study, which suggested that writing of these EAL students showed little improvement over three years of degree study, are of concern. Most of the participants in this study (72.41%), as revealed in the interviews, expected their writing to improve whilst studying in an English-medium university.

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

51

Furthermore, and more importantly, excellent writing skills are a graduate attribute expected of graduates of the university. The findings of this study thus have important pedagogical and policy implications. They suggest that if this graduate attribute is to be achieved, the university may need to review the nature of assessment tasks across degree courses and develop guidelines about the feedback given to students on their writing. Appendix A. (interview questions/themes at Time 2) 1. Please describe your previous English learning experience & preparation for study at English-medium university. 2. Opportunity to use English: What language do you speak at home (in Australia), with friends, % English used. 3. How much writing in different subjects at your time at the university, especially final year; were assignments group essays or individual. 4. What types of writing did you have to do in your three years of study; length of writing; 5. Did you choose courses according to how much writing is required? 6. Did you experience difficulties with writing at university? 7. Did any of the lecturers comment on your writing? If so, what did they say? If so, what did you do about it? 8. Did you attempt to get help with English writing during your time at uni; If so, what was that help? 9. Do you think your English has improved over your time at uni; which aspects improved? 10. What factors helped/hindered improvement? 11. What do you think about most when writing essays for uni?

References Archibald, A. (2001). Targeting L2 writing proficiencies: Instruction and areas of change in students’ writing over time. International Journal of English Studies, 1, 153–174. Australian Government Department of Education. (2014a). Research Snapshot. Onshore higher education international students as a proportion of all onshore students, by university, 2013 Retrieved from https://internationaleducation.gov.au/research/Research-Snapshots/Documents/ International%20students%20in%20Australian%20unis%202013.pdf. Australian Government Department of Education. (2014b). Research Snapshot. International student enrolments by nationality in 2013 Retrieved from https://internationaleducation.gov.au/research/Research-Snapshots/Documents/Enrolments%20by%20Nationality%202013.pdf. Biber, D., Gray, B., & Poonpon, K. (2011). Should we use characteristics of conversation to measure grammatical complexity in L2 writing development? TESOL Quarterly, 45, 5–35. Bitchener, J., & Knoch, U. (2009). The relative effectiveness of different types of direct written corrective feedback. System, 37, 322–329http:// dx.doi.org/10.1016/j.system.2008.12.006. Bulte´, B., & Housen, A. (2012). Defining and operationalising L2 complexity. In A. Housen, F. Kuiken, & I. Vedder (Eds.), Dimensions of L2 performance and proficiency: Complexity, accuracy, and fluency in SLA (pp. 21–46). Amsterdam and Philadelphia: John Benjamins. Cobb, T. (2002). Web Vocabprofile Retrieved from http://www.lextutor.ca/vp/. Coxhead, A. (2000). A new academic word list. TESOL Quarterly, 34, 213–238. Craven, E. (2012). The quest for IELTS Band 7.0: Investigating English language proficiency development of international students at an Australian universities. IELTS Research Reports. Vol. 13 (pp.1–61). Canberra: IELTS Australia. Cumming, A. (1990). Metalinguistic and ideational thinking in second language composing. Written Communication, 7, 482–511. Cumming, A., Kantor, R., Baba, K., Eouanzoui, K., Erdosy, U., & James, M. (2006). Analysis of discourse features and verification of scoring levels for independent and integrated prototype writing tasks for new TOEFL. (TOEFL Monograph No. MS-30) Princeton, NJ: Educational Testing Service. Deng, X., Lee, K. C., Varaprasad, C., & Lim, M. L. (2010). Academic writing development of ESL/EFL graduate students in NUS. Reflections on English Language Teaching, 9(2), 119–138. Elder, C., & O’ Loughlin, K. (2003). Investigating the relationship between intensive English language study and band score gains on IELTS. IELTS Research Reports. Vol. 4 (pp.207–254). Canberra: IELTS Australia. Ellis, R., Sheen, Y., Murakami, M., & Takashima, H. (2008). The effects of focused and unfocused written corrective feedback in an English as a foreign language context. System, 36, 353–371. Ferris, D. R. (2003). Response to student writing: Implications for second language students. Mahwah, NJ: Lawrence Erlbaum Associates. Green, A. (2005). EAP study recommendations and score gains on the IELTS Academic Writing test. Assessing Writing, 10, 44–60. Green, A., & Weir, C. (2003). Monitoring score gain on the IELTS academic writing module in EAP programmes of varying duration. Phase 2 report. Cambridge: UCLES. Gunnarsson, C. (2012). The development of complexity, accuracy, and fluency in the written production of L2 French. In A. Housen, F. Kuiken, & I. Vedder (Eds.), Dimensions of L2 performance and proficiency: Complexity, accuracy, and fluency in SLA (pp. 247–276). Amsterdam and Philadelphia: John Benjamins. Hu, G. (2007). Developing an EAP writing course for Chinese ESL students. RELC Journal, 38, 67–86.

52

U. Knoch et al. / Journal of Second Language Writing 28 (2015) 39–52

Humphreys, P., Haugh, M., Fenton-Smith, B., Lobo, A., Michael, R., & Walkinshaw, I. (2012). Tracking international students’ English proficiency over the first semester of undergraduate study. IELTS Research Report Online Series. Vol. 1 (pp.1–41). Canberra: IELTS Australia. Hyland, K. (2013). Writing in the university: Education, knowledge and reputation. Language Teaching, 46, 53–70. Knoch, U., Rouhshad, A., & Storch, N. (2014). Does the writing of undergraduate ESL students develop after one year of study in an Englishmedium university? Assessing Writing, 21(1), 1–17http://dx.doi.org/10.1016/j.asw.2014.01.001. Kormos, J. (2012). The role of individual differences in L2 writing. Journal of Second Language Writing, 21, 390–403. Larsen-Freeman, D. (2006). The emergence of complexity, fluency, and accuracy in the oral and written production of five Chinese learners of English. Applied Linguistics, 27, 590–619. Larson-Hall, J. (2010). A guide to doing statistics in second language research using SPSS. London: Routledge. Leki, I. (2006). You cannot ignore: L2 graduate students’ response to discipline-based written feedback. In K. Hyland & F. Hyland (Eds.), Feedback in second language writing: Contexts and issues (pp. 266–286). Cambridge: Cambridge University Press. Norris, J. M., & Mancho´n, R. (2012). Investigating L2 writing development from multiple perspectives: Issues in theory and research. In R. Mancho´n (Ed.), L2 writing development: Multiple perspectives (pp. 221–244). Boston: deGruyter. Norris, J., & Ortega, L. (2009). Measurement for understanding: An organic approach to investigating complexity, accuracy, and fluency in SLA. Applied Linguistics, 30(4), 555–578. O’Loughlin, K., & Arkoudis, S. (2009). Investigating IELTS exit score gains in higher education. IELTS Research Reports, 10, 95–180. Ortega, L. (2003). Syntactic complexity measures and their relationship to L2 proficiency: A research synthesis of college-level L2 writing. Applied Linguistics, 24, 492–518. Rouhshad, A. (2014). The nature of negotiations in computer-mediated and face-to-face modes with/without writing modality (Doctoral dissertation). Retrieved from http://repository.unimelb.edu.au/10187/18441 (19 April, 2014). Sasaki, M. (2007). Effects of study-abroad experiences on EFL writers: A multiple-data analysis. The Modern Language Journal, 91, 602–620. Sasaki, M. (2009). Changes in English as foreign language students’ writing over 3.5 years: A sociocognitive account. In R. M. Manchon (Ed.), Writing in foreign language contexts: Learning, teaching and research (pp. 49–76). Clevedon, England: Multilingual Matters. Sasaki, M. (2011). Effects of varying lengths of study-abroad experiences on Japanese EFL students’ L2 writing ability and motivation: A longitudinal study. TESOL Quarterly, 45, 81–105http://dx.doi.org/10.5054/tq.2011.240861. Serrano, R., Tragant, E., & Llanes, A. (2012). A longitudinal analysis of the effects of one year abroad. Canadian Modern Language Review, 68, 138–163. Shaw, P., & Liu, E. T. (1998). What develops in the development of second-language writing? Applied Linguistics, 19, 225–254. Storch, N. (2007). Development in L2 writing after a semester of study in an Australian university. Indonesian Journal of English Language Teaching, 3, 173–190. Storch, N. (2009). The impact of studying in a second language (L2) medium university on the development of L2 writing. Journal of Second Language Writing, 18, 103–118http://dx.doi.org/10.1016/j.jslw.2009.02.003. Storch, N. (2013). Collaborative writing in L2 classrooms. Bristol, UK: Multilingual Matters. Storch, N., & Hill, K. (2008). What happens to international students’ English after one semester at university? Australian Review of Applied Linguistics, 31(1), 1–17. Storch, N., & Tapper, J. (2009). The impact of an EAP course on postgraduate writing. Journal of English for Academic Purposes, 8, 207–223http:// dx.doi.org/10.1016/j.jeap.2009.03.001. Tsang, W. K., & Wong, M. (2000). Giving grammar the place it deserves in process writing. Prospect, 15, 34–45. The University of Melbourne. (n.d.). Melbourne graduates. Retrieved from http://www.learningandteaching.unimelb.edu.au/curriculum/graduates. Van Beuningan, C. G., de Jong, N. H., & Kuiken, F. (2012). Evidence on the effectiveness of comprehensive error correction in second language writing. Language Learning, 62, 1–41. Verspoor, M., & Smiskova, H. (2012). Foreign language writing development from a dynamic usage based perspective. In R. Mancho´n (Ed.), L2 writing development: Multiple perspectives (pp. 17–46). Boston: deGruyter. Williams, J. (2012). The potential role(s) of writing in second language development. Journal of Second Language Writing, 21, 321–331. Ute Knoch is a senior research fellow and the Director of the Language Testing Research Centre at the University of Melbourne. Her research interests are in the areas of second language writing assessment, writing development, and assessing languages for academic and specific purposes. Amir Rouhshad received a PhD in Applied Linguistics at the University of Melbourne. His areas of interests include learner-learner interaction, second language writing development and second language attrition. Su Ping Oon is currently teaching General Paper in Nanyang Junior College, her alma mater. She graduated with a Master in Applied Linguistics from the University of Melbourne in 2012 and it was during this period that she developed an interest in second language writing and feedback. Neomy Storch is a senior lecturer in ESL and Applied Linguistics at the University of Melbourne, Australia. Her research has focused on issues related to second language (L2) pedagogy. These issues have included feedback on L2 writing, writing development, the nature of peer collaboration, and the role of L1 in L2 classes.