profile 5.pmd ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○23 PROFILEPROFILEPROFILEPROFILEPROFILE Teacher Correction versus Peer-Marking* Corrección del maestro versus corrección por revisión de pares Mariana Correia Mourente MiguelMariana Correia Mourente MiguelMariana Correia Mourente MiguelMariana Correia Mourente MiguelMariana Correia Mourente Miguel** Universidade Federal do Rio de Janeiro, Brazil mariana_mourente@yahoo.com.br Written language is undoubtedly more often used than oral language in a variety of contexts, including both the professional and academic life. Consequently, developing strategies for correcting compositions and improving students’ written production is of vital importance. This article describes an experiment aimed at assessing the two most widely used methods of correction for compositions –traditional teacher correction and peer marking and their effect on the frequency of errors. Data was collected by asking students to write and revise a text. Statistical tests were performed to analyse it. At the end of the experiment, it was found that no significant difference in efficiency existed between the two methods, contradicting expectations (cf. Davies, 2002; Levine et al., 2002 and Ward, 2001). KKKKKey worey worey worey worey words: ds: ds: ds: ds: English-Teaching, Foreign Language-Teaching Writing, Evaluation, Assessment El lenguaje escrito es sin duda usado con más frecuencia que el lenguaje oral en una variedad de situaciones o contextos, incluyendo tanto la vida profesional como la académica. En consecuencia, el desarrollo de estrategias para corregir composiciones y mejorar la producción escrita de los estudiantes es de suma importancia. Este artículo describe un experimento cuyo objetivo es evaluar los dos métodos más usados para la corrección de composiciones, la corrección tradicional por el maestro y la corrección por revisión de pares, con respecto a su efecto en la frecuencia de errores. Se recogió información haciendo que estudiantes escribieran y revisaran un texto y sobre esos textos se aplicaron pruebas estadísticas para analizar los errores. Contrario a lo esperado, al final del experimento, no se encontró ninguna diferencia significativa entre los * This research project was carried out as part of a non-mandatory internship undertaken by the author during 2002-2003. This internship was offered by an institutional project called Projeto Cursos de Línguas Abertos à Comunidade and was funded by Universidade Federal do Rio de Janeiro. ** Mariana Correia Mourente MiguelMariana Correia Mourente MiguelMariana Correia Mourente MiguelMariana Correia Mourente MiguelMariana Correia Mourente Miguel is a graduate student and teacher whose main interest lies in written production in English as a foreign language as well as its evaluation and assessment. She has co-edited a collection of essays on foreign language teaching in 2003 and is currently co-editing another volume to be published this year. ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ 24 CORREIA PROFILEPROFILEPROFILEPROFILEPROFILE resultados encontrados por los dos métodos, (cfr. Davies, 2002; Levine et al., 2002 y Ward, 2001). Palabras claves: Palabras claves: Palabras claves: Palabras claves: Palabras claves: Inglés-Enseñanza, Idioma Extranjero-Enseñanza, Composición, Evaluación INTRODUCTIONINTRODUCTIONINTRODUCTIONINTRODUCTIONINTRODUCTION Nowadays, a great importance is attached to written language by many institutions since it is needed in order to succeed academically. Entrance exams for both undergraduate and graduate university courses require mastery of it. Written English is also more common than spoken English for many people both in their personal and professional lives. In Brazil, the federal government has helped to give prominence to written over spoken language by publishing the National Curriculum Programme for the teaching of foreign languages (Secretaria de Educação Fundamental, 1998). This document instructs teachers to focus on skills related to written language and states many reasons why they should be emphasised even when teaching students as young as ten years old. This state of affairs naturally leads teacher trainers and teacher trainees to consider the methods of correction for compositions. If teachers are supposed to devote a great part of their courses to the written language, they should equally make attempts to optimise the learning process of skills related to it. In this article, two of the most common methods of correction for written production –traditional teacher marking and peer marking– will be empirically assessed. THEORETICAL BACKGROUNDTHEORETICAL BACKGROUNDTHEORETICAL BACKGROUNDTHEORETICAL BACKGROUNDTHEORETICAL BACKGROUND Several studies such as Ward (2001), Levine et al. (2002), and Davies (2002), state that peer-marking is more efficient than traditional teacher marking. According to these researchers, peer marking presents students with more opportunities to assume an active role in the evaluation process. In order to be able to detect mistakes in colleagues’ compositions, students must be at least familiar with the vocabulary and structures that were taught in class. Besides that, these authors claim that students are more likely to pay attention in order not to make the same mistakes again. These authors also claim that this method provides students with opportunities to reflect on the learning process and critically evaluate it. Last but not least, students are also exposed more to the foreign language they are studying and may learn new vocabulary items or grammar points. HYPOTHESISHYPOTHESISHYPOTHESISHYPOTHESISHYPOTHESIS In the present research project, the objective is to verify which method is the most efficient: the traditional correction, when only the teacher has the authority to correct students’ mistakes, or peer-marking, when the students themselves evaluate their colleagues’ production. The aspect under analysis is the efficacy of these methods ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○25 PROFILEPROFILEPROFILEPROFILEPROFILE Teacher Correction versus Peer-Marking concerning the decrease in the frequency of errors after revision. The research question that was drawn from the cited bibliography is: Would students who practised peer marking commit fewer mistakes when revising their texts? METHODOLOGYMETHODOLOGYMETHODOLOGYMETHODOLOGYMETHODOLOGY In order to verify this hypothesis, two activities were conducted with groups of students belonging to the same level in order to gather the data required for our research. The students had had approximately 200 guided learning hours, being situated at the A2 level (or Key English Test level) according to the Common European Framework (see Cambridge ESOL). All students were adults, from 18 to 27 years old. They were attending a general English course aimed at developing the four abilities. This course was offered by a public institution with an especially low fee to attract disadvantaged students. The activities designed to gather data involved the writing of a composition by the students working individually. After the compositions were written, they were collected. Then each class was assigned a different procedure. In Class A, the traditional method was used while in Class B peer marking was chosen as the way of correcting the compositions. The corrected compositions were then given back to the students who were asked to check the mistakes that had been highlighted, as well as to try to assimilate the right answers. The next step for both classes was to give them an uncorrected copy of their compositions and ask them to rewrite it, eliminating the errors that had been corrected. This procedure is further clarified in the diagram below: Step 1: Students were instructed to write a composition Step 2—A: Teacher collected compositions and corrected them Step 2—B: Each student corrected a classmate’s composition Step 3: Students received their compositions and checked the markings Step 4: Students revised their compositions Diagram 1.Diagram 1.Diagram 1.Diagram 1.Diagram 1. Procedure for collecting data ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ 26 CORREIA PROFILEPROFILEPROFILEPROFILEPROFILE Both the original and the revised versions of the composition were carefully read and compared by the researcher in order to determine how many mistakes had been made in the original and in the revised versions. The data collected from this activity were then submitted to statistical analysis because this type of analysis allows the detection of tendencies and the comparison of groups, which proved useful given the aims of the experiment. The data were not normally distributed, as one can see in the histogram below (Graph 1), where “Erorig” stands for the number of mistakes found in the first version of the composition (“Erorig” is “erros no original”, or “mistakes in the first version”, shortened due to software limitations.). Considering the data distribution, the next step was to perform the chi-square test to see whether there was a significant difference between students’ performance before and after the correction (cf. Robson, 1975, for a detailed account of how to structure empirical research). As shown in the table below (Table 1), the difference for both groups was not significant, which means that neither of the methods tested in this experiment was considerably more efficient than the other in making students able to identify and correct their errors. Graph 1Graph 1Graph 1Graph 1Graph 1 ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○27 PROFILEPROFILEPROFILEPROFILEPROFILE Teacher Correction versus Peer-Marking TTTTTable 1able 1able 1able 1able 1 DISCUSSIONDISCUSSIONDISCUSSIONDISCUSSIONDISCUSSION From the data analysed in the previous section, one can infer that the assumptions made about peer marking are not entirely well founded. The experiment did not have the deep impact it was expected to have on students’ use of the foreign language. Mistake identification did not improve significantly when compared to students exposed to traditional evaluation methods, since there was roughly an equal improvement for both groups. Therefore, it may be more appropriate for teachers to include both styles of correction in their courses. This way, students who eventually feel inhibited with one of the methods will be given opportunities to profit from the other. A shortcoming in this research that has to be acknowledged is the fact that it dealt specifically with Brazilian students. As a result, this experiment does not account for the impact cultural background may have on students’ reaction to different evaluation methods. Further research, including other English as a foreign language teaching contexts such as other Latin American countries, is thus needed in order to reach a better understanding of how assessment functions during the foreign language acquisition process. It would also be pertinent to verify if these two methods vary in any other relevant aspect. ACKNOWLEDGEMENTSACKNOWLEDGEMENTSACKNOWLEDGEMENTSACKNOWLEDGEMENTSACKNOWLEDGEMENTS The author would like to thank fellow teacher Naíra Figueiredo Costa Monteiro de Almeida, for kindly discussing the design of the experiment in its early phase and helping to collect data as well as all the students who willingly participated in the experiment and consented on its publication under anonymity. The author would also like to thank in advance those who take an interest in this research and decide to conduct similar projects in their schools. REFERENCESREFERENCESREFERENCESREFERENCESREFERENCES Cambridge ESOL. How many hours of study are needed to reach the levels of the Main Suite exams? Retrieved from h t t p : / / w w w. c a m b r i d g e e s o l . o rg / f a q / ms_gen_faqs.cfm#1 Davies, P. Peer-assessment: No markings, just feedback? Retrieved from http:// www.comp.glam.ac.uk/pages/staff/pdavies/ caa.htm ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ 28 CORREIA PROFILEPROFILEPROFILEPROFILEPROFILE Levine, A., Oded, B., Connor, U., and Asons, I. (2002). Variation in EFL-ESL response. TESL-EJ, Berkeley 6: 3. Retrieved December 2002, from http://www-writing.berkeley.edu/TESL-EJ/ ej23/a1.html Robson, C. (1975). Experiment, design and statistics in psychology. Harmondsworth: Penguin Education. Secretaria de Educação Fundamental. (1998). Parâmetros curriculares nacionais: Terceiro e quarto ciclos do ensino fundamental: língua estrangeira. Brasília: MEC/SEF. Retrieved from http://www.mec.gov.br/sef/estrut2/pcn/pdf/ estrangeira.pdf Ward, A. (2001). Experimenting with web- enabled peer assessment. Retrieved from http:/ /www.dcs.warwick.ac.uk/~ashley/Research/ OASYS/forumarticle.html This article was received on May 11th, 2004 and accepted on August 23rd, 2004