Improving Foreign Language Writing Competence

Carmen Schlig
Georgia State University
Abstract
Process writing has long been used in English-language composition and English as a Second Language courses, and in recent years it has been adopted in foreign language classes as well. Nevertheless, many teachers and learners still see foreign language writing as an exercise in perfecting grammar and vocabulary. Explicit instruction on the process of insightful writing is unusual in the foreign language classroom. This article presents a study of third-year Spanish courses and addresses (1) whether the different modalities of error correction promote improvement of writing skills , and (2) whether students’ awareness of the processes of reading and writing favors the development of the necessary abilities that will be required in more advanced courses. This study identifies the processes that allow students to acknowledge and address strengths and weaknesses that are not necessarily tied to grammatical competence. Findings show that regardless of the explicitness of the corrections, students performed better when electronic feedback was used and when they were aware of both reading and writing processes.
Background
The field of second language (L2) writing has raised theoretical concerns about how students improve writing skills. While some authors have looked into process writing as the optimum way to improve the skill (Roca de Larios, et al, 2002; Susser, 1994; Scott, 1996; Zamel, 1983), others have explored the effects of error feedback on writing proficiency (Ferris, 1995, 1999; Truscott, 1996, 1999; Ferris and Roberts, 2001; Robb, Ross, and Shortreed, 1986; Semke, 1984; Lalande, 1982), and the processes involved in reading-to-write (Carson, 1993; Flower, et al., 1990; Ruiz-Funes, 2001).
Since the 1970s, studies present writing as a “recursive, nonlinear cognitive process in which the writer moves back and forth between prewriting, writing, revising, and editing until he/she is satisfied with his/her creation” (Flower and Hayes, 1981). Research in the last 10 years has moved towards a more comprehensive theory of L2 writing, and an examination of the notion that first language (L1) and L2 writing are the same “has led L2 writing specialists to rely for direction almost exclusively on L1 composition theories…” (Silva, 1993). Reichelt (1999) has found an increase in studies on foreign language (FL) writing—other than English—since 1990. She has also noticed that many of the researchers consider themselves language teachers rather than writing teachers, thus supporting the notion that “FL writing is currently seen more as ‘foreign language’ than ‘writing’” (p. 182).
This study contributes to the discussion by incorporating several strategies into an intermediate course of the language sequence at a large state university that has become—based on students’ observations collected over the past five years—a place where many undergo a difficult transition between lower- and upper-division courses. The observations of students’ struggle to improve their foreign language skills raised the questions of whether certain conditions favor or hinder the development of necessary abilities that are required in advanced courses.
Kern and Schultz (1992) indicate that preparation of students is crucial to promotion because “in the upper-level courses students are expected to read articles and literary selections and to react and respond to them in an insightful and critical manner” (p. 2). Ruiz-Funes (1999) also emphasizes the contrast in tasks in advanced courses where “reading and writing requirements begin to demand more higher-order cognitive skills, instruction tends to focus on the mastery of the linguistic elements; and teachers tend to use writing-process and reading-process instruction independently from each other” (p. 45). Explicit instruction in the processes of critical reading and insightful writing is unusual in the FL classroom. As Kern and Schultz point out, some teachers and learners still see FL writing as an exercise in perfecting grammar and vocabulary:
Writing in lower-division language courses traditionally consists of fill-in-the-blank workbook exercises and occasional descriptive essays about personal topics such as friends, family, and vacations. In these writing tasks, the focus is usually on surface feature accuracy rather than on the development, organization, and effective expression of the students’ own thoughts or ideas (p. 2).
This study investigates the conditions that favor the development of writing skills and questions whether some areas are not meant to be improved at this point.
The term process writing, as used in this project, refers to pre-writing, drafting, feedback, and revising, as part of a non-linear model. It follows the two tenets of process writing as described by Susser (1994): awareness and intervention. The concept also departs from the personal introspective paradigm prevalent in L1 and L2 writing courses that obviously addresses more advanced writers than those who participated in this study.
According to Ruiz-Funes (1999), research on reading-to-write in FL belongs to two categories: (a) narrowing the intermediate/advanced-level gap, and (b) teaching literature in order to develop higher cognitive skills. Students in advanced levels are often required to write essays for which previous work has not prepared them intellectually. Since intermediate writing is usually circumscribed to narration and description, students lack the necessary tools to approach writing an argumentative essay and “engage in a complex process that includes exploration of a problem, evaluation of facts and evidences, generation and testing of hypotheses in relation to new ideas an evidence” (p. 521). In order to overcome these problems, students need to be aware that good writing is not just grammatically accurate and that other factors, such as organization, coherence, and use of cohesive devices, are essential elements of good writing.
This study also compares the use of handwritten comments to electronic corrections and their effects on students revisions of composition errors. For the handwritten comments, two items were implemented: an adaptation of the Essay Correction Code (ECCO) and the Error Awareness Sheet (EASE) (Lalande, 1982). For the electronic corrections Ferris’ (2001) model of “treatable” and “untreatable” errors was adopted. (Appendix A)
Process writing
The adaptation of L1 process writing approach into FL writing practices seems to have taken one of its least pedagogical modalities, breaking down the process into stages that “not only violate[s] what we know about the recursive nature of writing, it distorts a responsible pedagogy into a didactic one” (Susser, 1994: 35). Although English as a Second Language (ESL) writing theory is already in the “post-process” era, addressing genre and social issues, it is still necessary to point out the controversy about process writing. In his detailed discussion of process approaches in L2 writing, Susser mentions that disagreement was caused by the association of process pedagogies with numerous writing theories, the gap between educational theory and practice and the fact that process became a synonym for theories of writing. Process writing emerged as a response to pedagogies that emphasized the composed product rather than the writing process. Process writing is characterized by the awareness of the writer of the writing process and the intervention of a teacher, or peers, at any time during the process of writing in order to improve writing skills instead of exclusively fixing mistakes (pp. 34-35). In foreign language studies (i.e., other than ESL or EFL [English as a Foreign Language]) writing progress has been measured by accuracy. Similarly, Dvorak (1986) concludes that beyond the intermediate level “composition skill has been defined primarily in terms of language development” demonstrated by the main preoccupation of research that focuses on “how to reduce and repair error damage” (p. 162).
Reading-to-write
Reading-to-write is defined as tasks that require students to write an essay based on the reading of an assigned source text. Such tasks may require students to read texts that have a variety of topics and orientations (Ruiz-Funes, 1999; Kern and Schultz, 1992). Stein’s (1990) reading-to-write-in-L1-hypothesis proposes the following steps: (1) Monitoring: the writer uses the original text to supervise his/her progress; (2) elaborating: the writer combines the source text with prior knowledge of the topic creating new ideas; (3) structuring: the writer reorganizes information from the source to the new text; and (4) planning: the writer moves from reading to writing. If reading-to-write were similar in L2 and L1, it would be necessary to look at the cognitive processes involved in the act of reading in order to understand it and help students develop the skill. In addition to all of the cognitive processes, however the FL student has to decode the text in order to interact with it.
Effects of error correction on FL Writing
Most students expect and value the feedback they receive in writing, and research has shown that there seems to be a connection between active correction of errors and improvement in writing skills. Ferris (1995) emphasizes the importance that students give to writing accurately and their perceived need to obtain corrections from the instructor. An important factor mentioned by Truscott (1996) —who, incidentally, opposes grammar correction—is the necessity of not treating every linguistic category (lexicon, syntax, and morphology) as equivalent since these categories represent separate learning domains that are acquired during different stages and through different processes. Nevertheless, most researchers (Ferris and Roberts, 2001; Robb, Ross, and Shortreed, 1986; Semke, 1984; Lalande, 1982) agree that corrections are useful for the students as long as they are consistent and systematic. In his study about the effects of graded versus ungraded compositions, Chastain (1990) found that although there was no significant difference between the number and types of errors, “in some ways the expectation of a grade may influence student’s writing in some positive ways…. Students in this study wrote longer papers containing longer sentences and a higher number of complex sentences” (p. 14). Corrections place an importance on what is corrected, thus exclusive surface-level error correction places emphasis on form.
Over the years, correction of written production has provoked some controversy. Several studies (Ferris and Roberts, 2001; Terry, 1989; Zamel, 1985; Lalande, 1982) endorse differing approaches to written correction, which can be separated into two main categories: (a) explicit (direct): the instructor indicates the error and provides the correct form and (b) non-explicit (indirect): the instructor marks the error in some fashion (underlined, highlighted, coded, etc.) and the student has to decide the correction. Some researchers (Chastain, 1990; Scott, 1996; Ruiz-Funes 2001) have adapted Lalande’s (1982) ECCO to provide students with an indication of the type of errors without giving the correct answer. This procedure has become a common classroom practice that encourages learning through problem-solving. Ferris and Roberts (2001) established five main categories of errors—verbs, noun-endings, articles, lexicon, and sentences structure (p. 169), and she has found that by using these five categories students are more successful at correcting errors even in such areas as word choice and sentence structure. In a different study Ferris (1999) classifies errors as treatable (patterned and rule-governed) or untreatable for which there is no set of rules that students can consult to avoid making mistakes. For these errors she recommends a combination of direct correction and a set of strategies exclusive to this type of error.
For language students it is very important to understand that there is no ideal model for writing and that they are not being compared to native speakers, or more proficient classmates. Instructors need to remember that one cannot expect the second language learner to be able to write like an educated native speaker. As Yates and Kenkel (2002) point out, “To compare the learners’ knowledge to native speaker knowledge commits the comparative fallacy and provides incomplete insight into what principles the learner had” (p. 34). Therefore, the scoring instrument should reflect the kinds of tasks that are done in the classroom.
The Project: Description and Method
This research examines the following research questions:
1. What effect do different modalities of error correction have on students’ self-editing abilities?2. What is the effect of awareness of the processes of reading and writing on quality of writing?
Participants
The participants are 35 university students enrolled in Intermediate Spanish III (fifth-semester). In order to ensure the homogeneity of the sample all participants: (a) were native speakers of English, (b) had received two to four semesters of formal instruction in Spanish at the university level, (c) had scored 80% or less on the listening comprehension pretest, and (d) scored less than 80% on a diagnostic composition. Participants in group A (n = 16) were selected out of 25 students who took the course; group B participants (n = 19) were selected out of 32 students. Only those participants who met all criteria above listed were included in the study. The score in the diagnostic test (less than 80%) was determined as representative of the average student’s writing at this institution. Both sections were taught by the researcher in consecutive semesters.
Procedures
The fifth-semester course places an emphasis on the development of the four skills necessary to succeed in advanced courses. Grammar is never explicitly discussed. Students are encouraged to identify structures with which they have problems and work on them either by talking to the professor, consulting a tutor, or downloading handouts from the online course management system Web Course Tools (WebCT).
This study examines 140 samples of a corpus of N = 280 (first draft and final version of the two compositions). The first composition is used for diagnostic purposes and is part of other diagnostic instruments that measure motivation, grammatical judgment, listening comprehension, and oral proficiency. Students are also asked to rank the importance of language skills (Appendix A) using a survey adapted from Alalou and Chamberlain (1999).
Students wrote all their compositions in a word processor in the language lab. The use of word-processing for the writing component of the course presents many benefits for students and instructors. Grennia (1992) reports that students using the word processor wrote three to seven times more than those who used traditional methods and that the instructors “always have a clean, legible copy and unlimited space for responses” (p. 35). Smith (1990) found that students engaged in writing as a process using computers developed more fluency and their writing was more expressive. Scott (1996) recognizes more advantages in the use of word-processors for teaching foreign language writing: First, “the computer environment provides a good opportunity for implementing a process-oriented approach to teaching FL writing” (p. 94); second, there is a notable improvement of textual coherence attained by revising and correcting ideas and surface-level features using the computer.
Out-of-Class Writing
During the semester, students in Group A (n = 16) turned in 15 journal entries in four installments, after the discussions of every three or four topics. The instructor made comments to indicate that the entries had been read. Grammar mistakes were not marked, but the instructor made observations regarding the comprehensibility of each entry. The use of the journals was meant to build fluency. Casanave’s (2004) students reported that they spent less time writing the same number of words at the end of the semester. Some wrote more accurately or wrote with more details and expressiveness. In contrast, in this study, most students wrote extremely short entries, in many instances no more that four sentences per entry. In the last entry students were asked to comment on their strategies for journal writing and on the perceived benefits of the tasks. Ten participants indicated that they would have liked to have their grammar corrected in order to develop accuracy. Nine mentioned that they wrote all the entries in one session, thus defeating the purpose of journal writing all together.
In contrast, students in Group B (n = 19) submitted 15 reading comprehension exercises corresponding to class assignments. The entries were marked for content, organization, and accuracy. Students turned in their writing portfolio twice during the semester for a grade. Full credit was given only if all corrections had been made. In the introduction to the final portfolio, students were asked to comment on their approach to revising errors, their use of the source text, and the perceived benefits of writing a portfolio. Twelve students reported to have used their first-year Spanish book instead of a dictionary or a grammar book as a reference because the explanations were clear and concise. Eight students mentioned the difficulty of writing their own sentences using a source text . These students had a tendency to copy sentences literally from the source text, sometimes using fragments that did not convey complete ideas. These problems were not encountered in the compositions because tasks were designed to create a new text using the source instead of demonstrating reading comprehension. Regarding the benefits of writing a portfolio, nine participants admitted they had waited until the deadline to make corrections for all the entries, which resulted in a progressive decline of successful corrections.
In-Class Activities
Prewriting. Both groups engaged in prewriting activities in the classroom, the success or failure of which depended mainly on the completion of the homework assignments. In order to provide models of writing, participants read three to four short essays from the textbook or other sources. After each reading, the class discussed the content of the selection as well as the characteristics of the genre, the tone, the theme, and so forth. Participants worked in groups on one or more of the following tasks: surveying classmates, expressing agreement or disagreement, expressing preferences, comparing and contrasting, and narrating a personal experience to illustrate a point of view. During this stage students shared ideas and tested hypotheses orally, and each group presented a summary of its findings to the class.
Participants in Group A received a prewriting worksheet that required them to write an outline of their composition and a list of words (nouns, adjectives, verbs, adverbs, etc.) that they anticipated useful for the topic. Students were encouraged to compare their lists with their classmates and share information. The worksheets were evaluated for completion, and students were allowed to use them while writing their compositions as long as they had not written complete sentences on them.
Participants in Group B received a prewriting worksheet to be completed at home. They were instructed to organize ideas and vocabulary according to semantic categories. During the next meeting, students worked in groups and explained their choices to each other. Students who had not completed the tasks in advance did not sit with a group and finished the worksheet by themselves. By the end of class, each group presented its ideas and explained its choices, and the instructor answered questions about vocabulary and how to approach specific tasks (description, narration, comparisons, etc.). Participants were expected to note ideas during the discussion sessions. The last worksheet contained specific structures that were considered useful to the students according on the task (e.g., list advantages or disadvantages of tourism, list the positive and negative impacts of technology on every day life, list the characteristics of realist vs. surrealist art.)
Writing. The first sample (E 1) was written in the language lab the fourth week of classes for both groups. Participants had a 55-minute time limit to write and proofread, but there was not a required number of words. The last essay (E 4) was also written in the language lab the 13th week of classes. There was a 60-minute time limit to write, proofread, and print or send the essay via email. Both groups had a choice of three topics based on a source text. Group A printed copies and received handwritten feedback on their essays, while Group B dealt exclusively with electronic submissions and feedback.
Writing Assignments
Group A
E1: Source text: You have mail, Tienes un e-mail (Kiddle, et al., 2002: pp. 129-131). Argue if modern technology has more disadvantages or advantages, mention things you like and dislike about technology, and write a comparison of technology now and technology 50 years ago.
E 4: Source text: Thief of Minds, Ladrón de la mente (Muñoz, 2000a). Write an essay from the point of view of one of the characters (three choices).
Group B
E 1: Rain forests: The earth’s lungs should not die, Selvas tropicales: los pulmones del planeta no deben morir, (Kiddle, et al., 2002: pp. 23-26). Write an essay on a ecological problem, its causes, consequences, and possible solutions.
E 4: (1) Dangers of tongue piercing, El peligro de los piercings en la lengua (Saludmanía, 1998). Agree or disagree with a friend who wants to get piercings. Give advise on how to take care of it and how to deal with friends and family members who disapprove; (2) The Use of Animals in Scientific Experiments El uso de animales en los experimentos científicos, (Kiddle: p. 116). Argue in favor of or against the use of animals for food, medical research, defense, and so forth; or (3) TV Addiction, La teleadicción, (Kiddle: pp. 243-245). Argue that addiction to TV-watching is similar to or different from addiction to alcohol, tobacco, or other drugs.

Data collection and analysis
These essays were evaluated using an analytic scale that students received in advance. All students had an opportunity to submit two corrected versions, but only the last one was compared to the first draft in order to determine the number of successfully corrected errors. The essays were photocopied, and one copy was saved to be scored later in the semester. The essays were evaluated twice over a 4-month period in order to corroborate the validity of the instrument.
For group A, errors were identified according to an adaptation of Lalande’s (1982) ECCO (p. 148), and subsequently students had to record their mistakes on an error frequency chart also adapted from Lalande’s EASE (p.149). Errors for Group B were marked electronically using a code adapted from Ferris and Roberts (2001: p. 169) and students were instructed in the use of the editing tools in Microsoft Word in order to make corrections.

The results in Table 1 show errors made by group and by category in the first draft of Essay 4. Although participants successfully used new words from the source texts, the largest percentage of mistakes fall in the lexical category in both groups. Group A made a higher number of errors, but the average number of words for that group was 272 compared to Group B, which averaged 204 words. Although students in Group B made more mistakes in noun endings, many of them were agreement errors. Since grammatical gender and number are not salient features in Spanish, students tend to ignore them even when the gender assignment is semantic (e.g., mujer ‘woman,’ regardless of the ending, is a feminine noun) and not morphologically determined (e.g., artista refers to either a male or a female artist). Group A presented a higher mean of errors in sentence structure, and the majority of these mistakes were omitted prepositions and indirect object pronouns.
Table 1: Errors Essay 4, Draft 1—means/s.d.

V
N
A
L
SS
Total
Group A
(n =16)

mean
s.d.
%
8.8
2.9
(19.4)
5.9
3.9
(13)
4.0
2.4
(8.8)
16.8
7.5
(36.9)
9.9
7.2
(21.7)
46
17

Group B
(n = 19)

mean
s.d.
%
4.1
3.4
(17.42)
7.2
3.9
(26.49)
2.3
2.5
(10)
9.9
5.9
(39.14)
1.6
1.5
(6.9)
25
12

Note: V = all verb errors; N = errors in nouns, gender and number agreement, possessives; A = articles, demonstratives, adjectives, missing or used incorrectly; L = lexical errors, word choice, errors in pronouns and prepositions; SS = sentence structure, word order, omitted or unnecessary words.

Table 2 shows errors in the final version of Essay 4 after corrections. Students in Group A greatly reduce the mean of total errors after receiving feedback. Participants in this group, and those in Group B to a lesser extent, had difficulties correcting mistakes in the verb category. Participants in Group A corrected only 56% of them. The two areas that presented more problems were aspect (preterit vs. imperfect) and tense (present vs. past and future vs. conditional). Although students in Group A had a more detailed correction code, the mean of errors in the verb section is higher than that of Group B. The verb category, for example, has been condensed in these tables for the sake of comparison, but students in Group A received a more direct indication of the type of mistake they had made: VT = verb tense; VA = verb aspect; VM = verb mood; S-V = subject-verb agreement; VF = verb form. While students in Group B received the same marking for all verb errors (V = all errors in verbs), they were able to correct 71% of the mistakes.
Table 2: Errors Essay 4, Final Version—means/s.d.

V
N
A
L
SS
Total

Group A
(n =16)


mean
s.d.
%
5.0
2.9
(27.1)
1.5
1.5
(8.0)
0.93
1.0
(5.0)
7.0
4.2
(37.9)
4.0
2.9
(21.8)
18.6
2.9


Group B
(n = 19)


mean
s.d.
%
2.4
2.4
(21.2)
2.4
2.7
(20.7)
1.4
1.7
(11.6)
4.2
3.2
(36.3)
1.1
1.4
(10.1)
2.4
7.9


Table 3 shows the percentage of corrected errors. Students in Group A were more successful at correcting errors in the noun-ending and article categories, but these errors were also coded in a more specific manner: G = gender; N = number; F = word form for noun-endings; Art = missing article; AG = gender agreement; AN number agreement; AF = article form. We need also to consider that students have a 50% probability of successfully correcting gender and number agreement errors. It must be noted that students in Group B were more successful when correcting errors in what Ferris identifies as “untreatable” categories: sentence structure and word choice. For Group A the correction code indicated the following: L = wrong word; Pron. = error in pronoun; Prep. = error in preposition. It is recognized that students in Group B were somewhat more proficient than students in Group A from the onset.
Table 3: Correctly Identified Errors

V
N
A
L
SS
Total
Group A (n =16)
%

56.3

85.26

96.92

63.33

67.3

68.6

Group B (n = 19)
%

71.38

78.38

64.29

78.05

72.41

75.1

(percentage: number of errors corrected divided by number of errors in each category)
To further assess students’ perceptions about the importance of speaking and writing, they were given a survey (Appendix B) in which they were asked to rank their language skills. Students ranked speaking as the most important skill, followed by reading, listening, writing, grammar, and comprehension. Students saw writing as instrumental—a way to practice grammar but not to increase proficiency—rather than expressive. The participants in this study, who rated in the below-average range of overall scores, commented that they would have performed better had they known the grammatical emphasis of the assignments. They believed their lower scores were due to poor grammar knowledge.
Students also completed a survey on their reading and writing preferences and strategies (Appendix C). Most students who indicated that they like to read and write also avoided translation when writing and consulted the rules when correcting. These students reported that they spent more time working on assignments than the others. Regarding their primary concern in writing, 56% of students in Group A and 57% of students in Group B identified grammar as the most important, while organization and content were the least of their concerns (12.5% for Group A and 10.5% for Group B). These responses may be explained by the emphasis grammar instruction has in lower-division courses. In short, students expected to be corrected in the classroom; otherwise, they felt their language abilities would not improve.
Discussion
This study set out to answer the following research questions:
1. What effect do different modalities of error correction have on students’ self-editing abilities?2. What is the effect of awareness of the processes of reading and writing on quality of writing?
With regard to Research Question 1 there was substantial evidence that students in Group B, those who worked with a shorter code (five items), were more successful at correcting errors. Also the electronic modality allowed them to move within the text from error to error automatically, and it was easier to find corrections. Students in Group A used a 16-item correction code that was deemed rather cumbersome, and it seems the additional information carried on the code (i.e., what kind of error in the verb: tense, aspect, mood, etc.) was in some cases misinterpreted by students. Also, students who worked with hard copies had a harder time transferring the corrections to the computer. Several students in Group A (31%) indicated that they retyped the essays every time they had to make corrections an act that resulted in new mistakes, and in a few cases students opted to eliminate sentences rather than trying to fix them. The findings are consistent with Ferris and Roberts (2001) in regard to the treatment of lexical and sentence structure errors. Students in Group A were able to identify correctly 63.3% of lexical errors and 67% of sentence structure errors, while students in Group B correctly identified 78% of lexical and 72% of sentence structure errors. Students in both groups showed concerns about grammatical accuracy in writing and expected to have their errors marked. Students in Group A mentioned that the lack of grammar correction in their journal entries was as frustrating as the number of errors they had to correct in their essays.
While the debate regarding effectiveness of feedback continues, it is possible that students are motivated by the error –correction and that the effects will become evident over time (Ferris, 2004). The results also reveal that students show improvement from first draft to final version of the same essay and little improvement between first drafts of first and last essays; thus it is possible that feedback in this case has not had a long-lasting effect on writing. Nonetheless, students in both groups demonstrated an increased ability to successfully identify errors. As Casanave (2004) indicates, it is important that instructors define improvement for students and help them understand that the feedback is aimed at specific areas of improvement. The perceptions of students in Group A regarding the importance of grammar accuracy were reinforced by the correction code that emphasized what students did wrong. This outcome was not intended by the researcher and was partially rectified in the treatment of Group B.
Research Question 2 looks at the effects of awareness of reading and writing processes. Although both groups engaged in prewriting activities, the quality of their essays differed in content, organization, and completion. Students in Group A had brainstorming and planning sessions the objectives of which were to elicit useful vocabulary and to organize content. Nevertheless, they had trouble staying on task and often changed the function of the essay to one that seemed more comfortable for their proficiency level (i.e., description instead of narration, exposition instead of argumentation). Group B worked with brainstorming and planning as well, but the focus of the planning sessions was ideas and semantic grouping of vocabulary and concepts. This group also worked on discussing effective ways to describe, narrate, compare, and hypothesize, and the attention was drawn away from grammatical accuracy. Students who understood writing as a process and spent more time revising and editing were more successful at correctly identifying their errors. The use of electronic feedback coincides with Casanave’s observation that the instructor’s role can resemble the work of a professional editor more than that of a teacher, giving students a chance to reflect on their own meanings rather than adjusting to what the teacher thinks they mean ( p.95). The results of this study suggest that instructors need to find appropriate means to help students depart from translation to develop both proficiency and accuracy.
Conclusions
The findings of this study are similar to Chandler (2003) and Ferris and Roberts (2001) in the use of the corrections codes. The samples reveal that a longer, more detailed code lends itself to possible incorrect markings from the instructor. Also, the instructor spent more time marking the essays because the code had to be consulted regularly in order to maintain uniformity. Hand-written corrections also resulted in several mistakes that were not marked in the first drafts (these missed errors were not considered in the final tally). The simplified code was more efficient to use and clearer for the students. The researcher was aware of “the tenuous connections between their [the teachers’] labor-intensive work and real improvement in their students’ writing” (Casanave 2004, p. 95).
Although students’ preoccupations with grammatical accuracy in Group A possibly reflect a shortcoming of the correction code and the error chart, it is plausible that there is a need to address salient grammatical features in the classroom as a means of developing problem-solving techniques. Ferris (2004) recommends addressing grammatical problems on an as-needed basis and combining this grammar review with other aspects of error treatment like feedback. She also advocates the use of an error chart, which can help students to become aware of their weaknesses (p. 62). In this study the use of error charts only seemed to exacerbate the feelings of students who were already frustrated. A possible solution for the future may be the use of a simplified error chart that does not overwhelm students and helps them identify specific problem areas. Instead of focusing students’ attention on non-critical errors, such as gender agreement, instructors have to devise strategies that help students achieve better understanding of all the factors that are involved in FL writing. We must acknowledge that many instructors have little tolerance for mistakes in gender agreement or spelling since they are so common and thus interfere with fluency, but a better understanding of critical vs. non-critical mistakes among instructors may result in a better treatment of errors in writing.
Certainly the issue of error correction in FL writing is far from resolved. It is clear that more research needs to be undertaken in this area. With respect to foreign language teaching, researchers and instructors have to take into consideration the fact that students in intermediate courses may lack the motivation and focus of ESL students. Writing in the foreign language is often only done in the classroom, while ESL students may find numerous occasions for writing in English such as academic or professional uses. In addition, research in FL writing has yet to look into other issues for instance, the purposes of FL writing. Future research will have to follow Ferris’ suggestions of conducting longitudinal studies to document the long-term effects of error feedback (2004: p. 55). Keen observation of student behavior as well as the effects of treatment in the classroom must guide instruction and research in FL writing in order to be able to account for improvement.

References
American Council on the Teaching of Foreign Languages. (1986). ACTFL proficiency
guidelines. Yonkers, NY: Author.
Alalou, A.,& Chamberlain, E. (1999). Using student expectations and perceived needs to
rethink pedagogy and curriculum: A case study. Foreign Language Annals, 32, 27-44.
Carson, J. G. (1993). Reading for writing: Cognitive perspectives. In J.G. Carson & I, Leki
(Eds.), Reading in the composition classroom (pp. 85-104). Boston: Heinle
Casanave, C. P. (2004). Controversies in second language writing: Dilemmas and decisions in
research and instruction. Ann Arbor, MI: The University of Michigan Press.
Chandler, J. (2003). The efficacy of various kinds of error feedback for improvement in accuracy
and fluency of L2 student writing. Journal of Second Language Writing, 12, 267-296.
Chastain, K. (1990). Characteristics of graded and ungraded compositions. The Modern
Language Journal, 74, 10-14
Ferris, D.R. (1995). Can advanced ESL students be taught to correct their most serious and
frequent errors? CATESOL Journal, 8 , 41-62.
Ferris, D. R. (1999). The case for grammar correction in L2 writing classes: A response to
Truscott (1996). Journal of Second Language Writing, 8, 1-10.
Ferris, D. R. (2004). The “grammar correction” debate in L2 writing: Where are we, and where
do we go from here? (and what do we do in the meantime…?). Journal of Second Language Writing, 13, 49-62.
Ferris, D., & Roberts, B. (2001). Error feedback in L2 writing classes. How explicit does it need
to be? Journal of Second Language Writing,10, 161-184.
Flower, L. & Hayes, J. R. (1981). A cognitive process theory of writing. College Composition
and Communication, 32, 365-387.
Flower, L., et al. (1990). The role of task representation in reading-to-write. In L. Flower, V.
Stein, J. Ackerman, M.J. Kantz, K. McCormick, & W.C. Peck (Eds.), Reading-to-write:
Exploring a cognitive and social process. New York: Oxford University Press.
Greenia, G. (1992). Computers and teaching composition in a foreign language. Foreign
Language Annals, 25, 33-46.
Kern, R., & Schultz, J. M. (1992). The effects of composition instruction on intermediate level
French students’ writing performance: Some preliminary findings. The Modern Language Journal, 76, 1-13.
Kiddle, M. E., Wegmann, B., & Schreffler. S. (2002). Perspectivas. Boston: Heinle.
Lalande II, J. (1982). Reducing composition errors: An experiment. The Modern Language
Journal, 66, 140-149.
Muñoz, E. M. (2000). Ladrón de la mente. New York: McGraw-Hill
Reichelt, M. (1999). Toward a more comprehensive view of L2 writing: Foreign language
writing in the U.S. Journal of Second Language Writing, 8, 181-204.
Robb, T., Ross, S., & Shortreed, I. (1986). Salience of feedback on error and its effect on EFL
writing quality. TESOL Quarterly, 20, 83–91.
Roca de Larios, J., Murphy, L. & Marín, J. (2002). “A critical examination of L2 writing process
research.” In Sarah Randsell and Marie-Laurie Barbier (Eds.), New directions for research in L2 writing. Dordrecht: Kluwer Academic.
Ruiz-Funes, M. (1999). The process of reading-to-write used by a skilled Spanish-as-a-foreign-
language student: A case study. Foreign Language Annals, 32, 45-62.
Ruiz-Funes, M. (2001). Task representation in foreign language reading-to-write. Foreign
Language Annals 34, 226-234.
Saludmanía (1998). www.saludmania.com/noticia/8/98/1247
Semke, H. (1984). The effect of the red pen. Foreign Language Annals, 17, 195-202.
Scott, V. (1996). Rethinking foreign language writing. Boston: Heinle,.
Silva, T. (1993). Toward an understanding of the distinct nature of L2 writing: The ESL research
and its implications. TESOL Quarterly, 27, 657-677.
Smith, K. L. (1990). Collaborative and interactive writing for increasing communication skills.
Hispania, 73, 77-87.
Stein, V. (1990). Exploring the cognition of reading-to-write. In L. Flower, V. Stein, J.
Ackerman, M.J. Kantz, K. McCormick, & W.C. Peck (Eds.), Reading-to-write:
Exploring a cognitive and social process (pp. 119-143). New York: Oxford University Press.
Susser, B. (1994). Process approaches in ESL/EFL writing instruction. Journal of Second
Language Writing, 3 , 31-47.
Terry, R. (1989). Teaching and evaluating writing as a communicative skill. Foreign Language
Annals, 22, 43-54.
Truscott, J. (1996). The case against grammar correction in L2 writing classes. Language
Learning, 46, 327-369.
Truscott, J. (1999). The case for “The case against grammar correction in L2 writing classes”: A
response to Ferris. Journal of Second Language Writing, 8, 111-122.
Yates, R., & Kenkel, J. (2002). Responding to sentence-level errors in writing. Journal of Second
Language Writing, 11, 29-47.
Zamel, R. (1983). The composing processes of advanced ESL students: Six case studies. TESOL
Quarterly, 17, 165-187.
Appendix A: ECCO, EASE, and SC (Ferris)
EECO/EASE
Essay1/rewrite
Essay 2/rewrite
Verbs (S-V agreement)
Aspect (A)
Mood (M)
Auxiliary verbs (Aux)
Form (VF)
Gender (G)
Number (N)
Form (F)
Article (agreement/form)
Lexical error (L)
Pronoun (omitted/incorrect)
Preposition (Pre)
Word order (WO)
Idiomatic Expression (EX)
Spelling (Sp)
Total



Short Code (Ferris)
V all verb errors, tense, mood, aspect, s-v agreement
N noun-endings, gender and number
A articles or other determiners, incorrectly used or omitted
L lexical errors, nouns, pronouns and prepositions
ES sentence structure, word order, idiomatic expressions, unnecessary words

Appendix B: Language Skills Ranked by Students N = 35

Least important
(%)
Somewhat important
(%)
Most important
(%)
N/A
(%)
Reading
5.7
11.4
80
2.8
Writing
20
22.8
42.8
11.4
Listening
2.8
25.7
68.5
2.8
Comprehension
28.5
22.8
28.5
2.1
Speaking
2.8
11.4
82.8
2.8
Grammar
5.7
34.2
51.4
8.5

Appendix C: Writing and Reading perceptions (N= 35)

Group A
(n = 16)
Group B
(n = 19)
1. Do you like to write?
(%)
(%)

yes
68.8
68.4

no
31.3
26.3
2. Do you consider yourself a good writer?

yes
31.3
57.9

no
68.6
42.1
3. What is your primary concern when writing?

vocabulary and grammar
56.3
57.9

organization and content
12.5
10.5

meaning
31.3
31.6
4. How do you approach writing in a foreign language?

start in English and translate
75
47.4

start in Spanish and avoid translation
25
52.6
5. How do you approach composition corrections?



consult rules
25
42.1

intuition
75
57.9
6. Do you like to read?

yes
75
63.2

no
25
36.8
7. What do you do when you do not know the meaning
of a word?

consult a dictionary
56.3
47.4

ignore it
25
36.8

try to deduce meaning from context
18.8
15.8