Abstract
The present study assessed the effectiveness of oral recasts in an EFL classroom. Thirty-eight college students and a female teacher participated in the study in a western state in Venezuela. Students were expected to learn the right use of the auxiliary verb ‘to have’, and the use of past participles in the present perfect tense. They were divided into two groups. Each group was randomly assigned to one of two conditions: 1) recast or 2) explicit negative feedback. Positive feedback was provided for both groups. Pictures were used to elicit conversation and an oral test was performed to collect the data. Results supported the claim of the effectiveness of recast compared to explicit negative feedback. Based upon the results, the use of recast in college EFL classes is recommended.
Keywords: error correction, negative feedback, recasts, EFL
Introduction
The kind and amount of feedback the students should receive still remain an interesting research question because of the pedagogical implications of the issue on FL learning and teaching. However, there are not enough empirical data to support possible answers to that question, even though it is a relevant issue related to language teaching and learning practices. Most of the research on feedback has dealt with the role of negative feedback in foreign and second language acquisition. Among the studies that have addressed this issue, are those that have only described the kind of feedback that occurs between NS and NNS (e.g., Oliver, 1995). In second or foreign language contexts, authors have mostly described the kind of feedback that takes place in a second language classroom and the uptake of each one of them (e.g., Lyster & Ranta, 1997). Recently, some studies have addressed the issue within experimental designs in foreign or second language settings (e.g. DeKeyser, 1993; Long, Inagaky & Ortega, 1998; Perdomo & Rodríguez, 2002), and some other have started questioning about the cognitive processes related to feedback noticing (Nabei and Swain, 2002).
Negative feedback in classroom interaction can be classified into different types including recast (Oliver, 1995). Among these types of feedback, recast seems to be the most effective implicit negative feedback (Ayoun, 2001; Long, Inagaky& Ortega, 1998). Recast “involves the reformulation of all parts of student’s utterance, minus the error” (Lyster & Ranta 1997, p. 46). Long, Inagaki & Ortega (1998) refer to it as corrective recast and define it as responses which, although communicatively oriented and focused on meaning rather than form, incidentally reformulates all or part of the learner’s utterance, providing information that was missing or ill-produced. Other names for recast are repetition with change, repetition with change and emphasis; paraphrase (Spada & Fröhlich, 1995) and pedagogic corrective feedback (Llinares, 2005).
Studying and analyzing the kind, amount and effect of feedback that students receive during the process of language learning contribute to create an appropriate environment for language learning, especially in EFL contexts in which classrooms are the main sources for language learning and the teacher is (in some cases) are the only source learners expect to communicate with, and therefore, teachers are bound to correspond to the learners expectations in terms of questioning and feedback (Farooq, 2007). In this sense, as well as the language teaching methodology to be used, teachers must decide about the emphasis of the oral activities (i.e., choosing between form and content) and hence, the kind of feedback to be given.
In Venezuela, where it is common to observe that students still depend on language classes and the classroom is still the main environment in which they interact in English, very few studies have been conducted on the issue of feedback and oral production. For these reasons, this study was designed with the purpose of investigating the effectiveness of recast in an oral EFL context in a western college in Venezuela.
Literature Review
Nicholas, Lightbow and Spada (2001) made a documental review of the main studies in which recast was investigated. After that review, they concluded that recast is quite effective when students relate it to form and not to meaning. In relation to non-documental studies, Ayoun (2001) investigated the effectiveness of recast against metalinguistic information. His study was conducted with 145-second language students. These subjects were included in one of the conditions: Implicit, explicit and pre-emptive. He found that the hypothesis claiming that recast was superior to other feedback types was partially supported.
Other researchers, (Long, Inagaky & Ortega, 1998) conducted a two-experiment study to assess the utility of models and recasts in L2 Japanese and Spanish. Subjects in experiment one were 24 young adult learners of Japanese as a second language. They were 13 males and 11 females. Twenty of the subjects reported having studied Japanese in high school between one and four years. Half of the participants had visited Japan for short periods. Target Structures to be learned in this experiment were adjective ordering and locative construction. The researchers used a pretest, posttest, and a control group design. The results provided some evidence that adults are able to learn from explicit negative feedback, and that recasts can be more effective than preemptive positive feedback (models) at least for short-term improvement in an unknown L2 structure.
Experiment number two in the aforementioned study was conducted with 30 young adults learning Spanish. They were given a Latin American music cassette and extra credit to participate in the study. Direct object topicalization and adverb placement were the target structures. As in the previous one, this study used a pretest, a posttest, and a control group design. The target structures were presented in two conditions: model and recast. Treatments and structures were crossed and counterbalanced. Both, treatments and testing were performed individually. The findings for the adverb placement indicated that groups in model and recast conditions outperformed those in control group, and participants under the recast condition did better than those in model. No learning occurred in any group regarding object topicalization. The authors concluded from both experiments that implicit negative feedback facilitates L2 structures acquisition.
Also Rodríguez and Perdomo (2002) investigated the effect of negative feedback on oral production of college students. In the study, recast was the main implicit negative feedback used by the instructor. The study was conducted with intact classes of marketing majoring students enrolled in the second semester of English as a foreign language in a western college in Venezuela. Participants’ oral performance from the previous semester was used as a covariate to investigate a possible attribute by treatment interaction (ATI). The analysis revealed a statistically significant previous oral performance by treatment interaction. Those students with previous oral performance over 14 seemed to benefit more from implicit negative feedback. However, explicit negative feedback appeared to be more effective for those subjects with previous performance below that score. Inspection of the regression lines for the treatment groups indicated that incidental error correction was superior to explicit correction.
In a different context, Tsybina, Girolametto, Weitzman and Greenberg (2006) examined linguistic recasts provided by 16 early childhood educators to preschool children learning English as a second language. They divided children in two groups according to their expressive language skills: lower and higher. In this study recast seemed to be less effective for lower expressive language skills. However, the researchers suggested increasing the amount of simpler recast in classroom interactions in preschool contexts. Even when the subjects in this study were children, results seem consistent with those on behalf of recast for more proficient students like Rodríguez and Perdomo’s (2002).
Research on the role of recast in FL learning has gone beyond describing types of feedback during interaction and the effectiveness of recast compared to other kinds of feedback. It has also focused on the cognitive processes that occur during classroom interactions. In this sense, Nabeu and Swain (2002) presented a case study on a 19 years old Japanese college girl learning EFL. They examined how recasts were provided and the relationship between the student’s awareness of recast feedback and her L2 learning. They concluded, based upon the results, that recasting is a complex verbal behaviour influenced by the teaching environment, the interaction context, and the learner’s cognitive orientation. They also highlight that recasts are influenced not only by its linguistic elements, but also by paralinguistic elements, as well as by the learner’s autonomous use of the learning opportunities provided by the feedback. These results indicate that recast remain as a controversial issue in FL and SL teaching. Hence, the need for more research on recast and the elements involved in the interaction that can affect its effectiveness is evident.
Method
Design
The present study was conceived as an experimental research. Participants were randomly assigned to one of two conditions (recast – explicit negative feedback). A pretest and posttest with control group design was used. The pretest was used as a covariate in the analysis of the data.
Subjects
Participants were 38 college students in a western state in Venezuela. The students were enrolled in the second semester of English as a foreign language. They constituted an intact class attending 60-minute sessions twice a week with the same female teacher; hence no teacher effect was expected.
Materials and Procedures
The experiment was conducted during the study of Unit 2 of the official program and lasted three weeks. The grammatical structures they were expected to acquire were: 1) the right use of the auxiliary verb TO HAVE for present perfect tense, and 2) the use of past participles of regular and irregular verbs in present perfect tense.
The content was studied using pictures to elicit conversation. The procedure was as follows:
The teacher asked a question based on the picture. Then, the student was expected to answer the question using the right form of the auxiliary verb and the verbs in past participle. The teacher provided positive feedback under both conditions and the appropriate negative feedback for each one according to the condition.
The short dialogues had a similar structure for subjects in both conditions:
- Initial question (by the teacher)
- Students’ answer.
3. Positive or negative feedback.
The following example illustrates a dialogue in class in which positive feedback was provided:
- Teacher: (after showing a picture containing a man who had fallen down on the floor) “what has happened?”
-Student: “The man has fallen down.”
-Teacher: “That’s right.”
The next is an example illustrating a dialogue in which explicit negative feedback was provided:
- Teacher: (after showing a picture of a man who had blown a candle) “what has happened?”
-Student: “the man have blown the candle.”
-Teacher: “That’s incorrect. You should say: The man has blown the candle.”
The forthcoming dialogue exemplifies a case in which recast as negative feedback was supplied:
- Teacher: (after showing a picture containing a boy who had fallen down from a bike) “what has happened?”
-Student: “The boy has fell from the bicycle”
-Teacher: “Ah, the boy has fallen from the bicycle.”
-Student: “Yes, the boy has fallen from the bicycle.”
After the classes were carried out, the data were collected through an oral interview comprising some short dialogues similar to the ones practiced in class. To elicit conversation the teacher brought 15 pictures. The students were asked 10 questions about five of those pictures chosen at random. The answers were evaluated giving one mark (1) to each, the right use of the auxiliary form, and the right use of the past participle of the verbs to form the present perfect. Hence, the maximum possible score would be 20.
Results
Means, standard deviations, and groups’ numbers for previous and current oral performance are presented in Table 1. As it can be observed, control group outperformed the recast group in the pre-test, but these results were different for the post-test.
Table 1. Means and Standard deviations for both groups in Previous and Current Performance
Condition |
Proficiency |
Previous (pre-test) |
Current (post-test) |
M |
SD |
n . |
M |
SD |
N |
Control |
12,05 |
3,94 |
19 |
12.32 |
2.87 |
19 |
Recast |
10,65 |
3,60 |
20 |
15.45 |
3.89 |
20 |
Overall |
11,33 |
3,79 |
39 |
13.92 |
3.74 |
39 |
The data were submitted to an Analysis of Variance (ANOVA), using the General Lineal Model (GLM) (Table 2). A pretest was used as a covariate with Type I error rate set at p<.05. The error variance of the cued-recall data was equally distributed across the groups, according to the test of homogeneity of variance which did not reach statistical significance, F (2, 37) = .566, p <.723. In other words, both groups were homogeneous and could easily be statistically compared.
Table 2. Summary of the ANOVA
|
Sums of
Squares |
df |
Mean
Squared |
F |
P |
h2 |
Pretest |
33.461 |
1 |
33.461 |
3.067 |
.067 |
.092 |
Treatment |
122.434 |
1 |
122.434 |
13.03 |
.001 |
.271 |
Treat x Pretest |
46.148 |
1 |
46.148 |
4.913 |
.033 |
.123 |
Error |
328.727 |
35 |
9.392 |
|
|
|
A statistically significant difference was found for treatment F (2, 37) = 13.03, p <.001. These results seem to support the effectiveness recast as negative feedback in comparison to explicit negative feedback. However, a significant difference was also found for the interaction between treatment and pretest, p < .033. This interaction means that students who showed better performance on the pretest were more benefited by the recast condition. In other words, recast seems to benefit more those students with more previous knowledge of the language.
Figure 1. Distribution of scores by group.
Figure 1 shows the distribution of the scores under both conditions. As it can be observed, better scores for the post-test were distributed among students under recast conditions in general, while most of the students in the control group obtained similar scores on both tests. Almost all the students under recast condition increased their scores after treatment, indicating that recast benefits students’ students learning of some grammatical structures. Besides, students in the control group with high scores in the pretest, got lower scores in the post-test, indicating that explicit negative feedback can also affect students’ performance.
Discussion
This study was designed to investigate the effectiveness of recast in an oral EFL context in a western college in Venezuela. In this study, students were expected to acquire two grammatical structures: 1) the right use of the auxiliary verb TO HAVE for present perfect tense, and 2) the use of past participles of regular and irregular verbs in present perfect tense. Results in general show recast as an effective negative feedback when learning those structures.
According to the results, students in the recast condition did better than the control group, even thought they have been outperformed by controls in the pre-test. On the other hand, control group did not show differences between the pre-test and the post-test. The results indicate that students perform better when they receive recast as negative feedback. In other words, recast has an effect on students’ performance. However, that difference tends to be higher for more proficient students.
The results of the present study are consistent with those reported by Nicholas, Lightbow and Spada (2001) in which recast happened to be more effective than explicit negative feedback. They also seem to be consistent with the studies of Ayoun (2001) and Long, Inagaky, and Ortega (1998) who compared recast to other kinds of feedback. The findings of the present research also match with the results of Lightbow and Spada (1990) who although did not study recast separately suggested that recast might be beneficial in a foreign language setting. As it can be observed, although few studies have tested recast within experimental designs to provide empirical evidence of its effectiveness, in most of them, recast seems to be more effective than other kinds of negative feedback and preemptive feedback. However, its short and long-term effectiveness has scarcely been investigated; hence, more research in needed on this issue.
Results in the current study indicated a statistically significant interaction between previous knowledge and treatment, indicating that recast tends to benefit more proficient students. These results are consistent with Rodríguez and Perdomo’s (2002) findings who found implicit negative feedback as more effective for highly proficient students. In the present research, students with higher scores for previous knowledge (indicated here by the pretest) also did better under recast condition (a type of implicit negative feedback).
These findings together suggest the use of recast with more proficient students. Results in both studies let the author to believe that more advanced students are able to focus on form rather than just on meaning, condition in favor of the effectiveness of recast according to Long, Inagaki & Ortega (1998). In this sense, it is recommended to consider Marcosim’s (2003) claim that teachers must take into account their students’ level of L2 proficiency when making decisions about feedback.
It is possible to think that the results of this research could have been influenced by the kind of interaction in class. In this sense, this study seems to support Lyster’s (1998) claim that the type of interaction between teachers and students would determine if students notice recasts or not. As Lyster (1998) states, when the teacher does not give students the opportunity to reformulate the utterance, the attention is given to content rather than form. In the case of the current study, the kind of interaction let the students identify recast as negative feedback and hence take the advantage of it. Therefore, if teachers decide to include recast as feedback in oral activities, they must set a proper interaction environment to enhance its effectiveness. However, there are not empirical studies that directly assess this issue, for which more research is also needed on the kind of classroom interaction required to increase recasts effectiveness.
Conclusion
The issue of negative feedback is highly important in language teaching. Some studies have shown that implicit negative feedback can affect students’ attempts to communicate which is reflected in their oral production and recommended the use of implicit negative feedback (Rodríguez and Perdomo, 2002). Among all the choices of implicit negative feedback recast is seen in literature as one of the best, especially for those students with higher level of proficiency. However, several aspects of the effectiveness of recast still need to be studied. For example, short and long term effectiveness for grammatical structures learning, the kind of interaction in the classroom enhancing recast effectiveness, the relation between students’ learning styles and recast noticing, and the relation between students’ anxiety in oral activities and recast noticing, among other.
Pedagogical implications
As teachers are expected to seek the best methods and techniques to help students to learn, one important issue to consider is the kind of feedback students should receive in oral activities. Giving students the appropriate kind of feedback would improve conditions for FL learning. However, as it has been observed in this research and in previous ones (e.g., Rodríguez and Perdomo, 2002), students’ proficiency in the language affects recasts effectiveness. In this sense it is recommended for teachers to explore their students’ proficiency in the language before deciding the kind of feedback to provide. Hence, it would be useful to make some tests at the beginning of the courses, in order to determine students’ proficiency, because the use of recast is somehow determined by students’ proficiency.
References
Ayoun, D. (2001). The role of negative feedback in the second language acquisition of the passé compose and imparfait. Modern Language Journal, 85(ii), 226-243.
DeKeyser, R. (1993). The effect of error correction on L2 grammar knowledge and oral proficiency. The Modern Language Journal, 77(ii), 502-514.
Farooq, M. (2007). Exploring the Effectiveness of Spoken English Classes of Japanese EFL Learners. The Journal of Liberal Arts, 3,35-57.
Lightbow, P. & Spada, N. (1999) Focus on form and corrective feedback in communicative language teaching. Effects on second language learning. Studies in Second Language Acquisition, 12, 429-448.
Llinares, A. (2005). The effect of teacher feedback on EFL learners´ functional production in classroom discourse. ANGLOGERMANICA ONLINE. Available at: http://www.uv.es/anglogermanica/2005/llinares.pdf
Long, M., Inagaky, S. & Ortega, L. (1998). The role of implicit negative feedback in SLA: Models and recast in Japanese and Spanish. The Modern Language Journal, 82(iii), 357-371.
Lyster, R. (1998). Recasts, repetition, and ambiguity in L2 classroom discourse. Studies in Second Language Acquisition, 20(1), 51-81
Lyster, R. & Ranta, O. (1997). Corrective feedback and learner uptake. Negotiations of form in communicative classrooms. Studies in Second Language Acquisition, 19, 37-66.
Marcosim, M. (2003) Teacher feedback and learner’s uptake. Revista Linguagem en (Dis)curso 4 (1). Recuperado de http://www3.unisul.br/paginas/ensino/pos/ linguagem/0401/04.htm
Nabei, T. & Swain, M. (2002) Learner Awareness of Recasts in Classroom Interaction: A Case Study of an Adult EFL Student’s Second Language Learning. Language Awareness 11(1), 43.
Nicholas, H., Lightbow, P. & Spada, N. (2001). Recast as feedback to language learners. Language Learning, 5(4), 719-758.
Oliver, R. (1995). Negative feedback in child NS_NNS conversation. Studies in Second Language Acquisition, 17, 459-481.
Rodríguez, M. & Perdomo, B. (2002). Effects of Implicit and Explicit Negative Feedback on EFL Students’ Oral Performance. Paper presented at the annual TESOL convention, VenTESOL, May, 2002. Nueva Esparta, Venezuela.
Spada, N. & Fröhlich, M. (1995). The Communicative Orientation of Language Teaching Observation Scheme (COLT).
The National Centre for English Language Teaching and Research (NCELTR). Sydney: Australia.
Tsybina, I., Girolametto, L., Weitzman, E., Greenberg, J. (2006) Recasts Used with Preschoolers Learning English as their Second Language. Early Childhood Education Journal, 34(2), 177-185.
