Focused Written Corrective Feedback: What A Replication Stud

Focused Written Corrective Feedback: What a Replication Study Reveals About Linguistic Target Mastery

Focused Written Corrective Feedback: What a Replication Study Reveals About Linguistic Target Mastery Monika Ekiert, LaGuardia CC, City University of New York Kristen di Gennaro, Pace University

The debate over the effectiveness of written corrective feedback (WCF) in second language (L2) writing instruction has been longstanding, with prominent researchers such as Truscott (1996, 2007) arguing that corrective feedback is not only ineffective but potentially harmful to learners' language development. Conversely, scholars like Ferris (1999, 2004) have contested these claims, asserting that error correction can positively influence learner accuracy. This divergence highlights the necessity of empirical investigations to inform pedagogical practices.

WCF remains a highly contentious issue in SLA research, with over 300 studies addressing its efficacy. Researchers motivated by practical pedagogical concerns seek to determine whether WCF warrants the investment of instructional time. If WCF is ineffective, why do teachers devote hours to providing it? Conversely, if WCF is effective, what specific effects does it produce, and which types are most beneficial? The interface between theory and practice makes WCF a compelling area for research, as it allows for observable, measurable outcomes that can inform instructional strategies.

Among the specific areas studied, the effectiveness of WCF on article accuracy provides a nuanced understanding of how targeted feedback influences second language development. Articles—such as "a" and "the"—are particularly challenging due to their rule-governed usage and the complexity of their referential functions, which involve first and subsequent mentions, as well as nonreferential and idiomatic uses (Bitchener & Knoch, 2010; Ellis et al., 2008; Sheen, 2007). The body of research generally shows that treatment groups receiving WCF outperform control groups, indicating a positive effect on learners' accuracy with articles (Ellis et al., 2008; Sheen, 2007). However, this evidence appears to contradict Truscott’s (2007) meta-analysis suggesting minimal or negative effects of WCF, raising questions about the influence of target specific feedback versus broader grammatical features.

One unresolved issue concerns the linguistic target of feedback, especially regarding the contexts in which articles are used. Ellis et al. (2008) highlight that because there are occasions when the definite article "the" is necessary for first-time reference, WCF should not be applied indiscriminately. They also raise the possibility that correction could lead to overuse of certain structures, as learners might interpret correction as signaling frequent use, thereby overgeneralizing rules beyond appropriate contexts. Further complicating the issue are the exceptions to rules and idiomatic uses that challenge the clarity and consistency of targeted feedback.

Insights from studies focusing on linguistic targets reveal that the effects of WCF are not uniformly beneficial across all contexts. Bitchener and Knoch (2010) and Ellis et al. (2008) demonstrate that partial correction or the omission of correction in certain contexts can influence learning outcomes, sometimes leading to overcorrections or errors of overuse. These outcomes suggest that targeted feedback needs to be carefully designed and context-sensitive, emphasizing the importance of nuanced instructional approaches that account for real-world language use.

The current study aims to address gaps, particularly the often-overlooked impact of feedback on untreated or uncorrected features within the same linguistic system. Defined as how well second language learners can use articles appropriately, this study examines the effects of focusing feedback on specific functions—initial versus subsequent mentions—on learners' performance in other related contexts. Using a quasi-experimental design, the study involved three groups of ESL students at a low intermediate to intermediate level, engaging in pre-test, immediate post-test, and delayed post-test assessments across different treatments: direct correction, correction with metalinguistic explanation, and no correction.

The methodology involved direct error correction, where incorrect uses of "a" and "the" were corrected above errors, and a metalinguistic approach, where students received explicit rules explaining when to use each article. The control group did not receive corrective feedback but only overall comments on writing quality. The assessment instruments included narrative picture descriptions and correction of missing words embedded in texts, allowing for analysis of both free and controlled production. Specific attention was given to first and subsequent mentions, as well as nonreferential, idiomatic, and situational uses of articles, providing a comprehensive picture of learners' accuracy and overuse patterns.

Data analysis employed obligatory occasion analysis, calculating proportions of correctly supplied articles and overuse instances, followed by statistical analysis with mixed ANOVAs to examine treatment effects over time and across groups. Results indicated that all groups showed some improvement over time, but the experimental groups did not outperform the control. Notably, the control group either matched or exceeded the treated groups in accuracy, challenging assumptions that targeted feedback necessarily enhances linguistic target mastery. Treatment effects appeared to be specific to the features directly corrected, with the experimental groups improving in treated areas but sometimes degrading in untreated areas—a finding consistent with the negative transfer or overgeneralization effects highlighted in prior research.

These findings raise critical questions about the broader implications of WCF. The observed pattern suggests that focusing on specific features—such as initial and subsequent article use—may inadvertently lead to neglect or deterioration of uncorrected features. Such trade-offs imply that learners’ grammatical systems are interconnected, and targeted correction can produce unintended consequences. Furthermore, the provision of partial metalinguistic information, while potentially aiding specific features, may also foster overgeneralization, leading to overuse errors in contexts not directly targeted by the feedback.

Discussion in the current study emphasizes the importance of selecting target features thoughtfully and considering their complexity and contextual variability. While engaging in writing tasks with frequent and natural occurrences of targeted structures may facilitate acquisition, pedagogical strategies must balance focus and comprehensiveness. The research supports a cautious view of WCF, acknowledging that corrections focusing narrowly on specific features can produce localized gains but may also induce negative effects on the broader system— an instance of "scaffolding" that does not support full system integration.

Moreover, the findings underscore the need for instructors to modify their expectations of language development, recognizing that correction alone is insufficient for holistic mastery. Instead, a more balanced approach combining explicit correction, contextualized practice, and awareness of interrelated linguistic features could foster more sustainable learning outcomes. The debate remains open regarding the overall efficacy of WCF, with some arguing that its benefits are limited or context-dependent, while others advocate for its strategic use within a comprehensive instructional framework (Lyster & Saito, 2010; Norris & Ortega, 2000).

In conclusion, the present replication study confirms that WCF can improve accuracy concerning specific targeted features, yet it also highlights the potential for unintended negative effects on untreated features. The findings advocate for a nuanced, system-aware approach to error correction, emphasizing careful target feature selection, context-sensitive correction, and ongoing research to elucidate the complex dynamics of second language development. Such insights should inform both theoretical models of SLA and practical pedagogical strategies to optimize language learners' outcomes.

References

• Bitchener, J., & Knoch, U. (2010). The value of written Corrective feedback for migrant L2 learners. Language Learning & Technology, 14(2), 198-218.
• Ellis, R., Sheen, Y., Murakami, M., & Takashima, J. (2008). The effects of focused written feedback and re-study on L2 pronunciation. Language Learning, 58(4), 683-722.
• Ferris, D. (1999). The case for grammar correction in L2 writing. College Composition and Communication, 50(4), 515-538.
• Ferris, D. (2004). The relationship between student and teacher modifications in college ESL writing classes. Language Learning & Technology, 8(2), 83–105.
• Lyster, R., & Saito, K. (2010). Oral feedback in classroom SLA: Research, theory, and practice. Language Teaching, 43(01), 1-45.
• Norris, J. M., & Ortega, L. (2000). Effectiveness of L2 instruction: A research synthesis and quantitative meta-analysis. Language Learning, 50(3), 417-528.
• Sheen, Y. (2007). The effects of focused written feedback and oral feedback in improving the accuracy of ESL learners’ writing. Studies in Second Language Acquisition, 29(3), 317–351.
• Truscott, J. (1996). The case against grammar correction in L2 writing classes. Language Learning, 46(2), 327-369.
• Truscott, J. (2007). The effect of error correction on learners’ language development: A meta-analysis. Language Learning, 57(3), 589-627.
• Ellis, R. (2010). Principles of instructed second language acquisition. European Second Language Association.