Este projeto de tese analisa como o método do ciclo de counterbalance, articulado por Lyster (2007), pode influenciar no aumento da precisão da língua dos alunos nos programas de imersão. O objetivo do projeto foi descobrir como tal modelo de lições e atividades pode vir a afetar a precisão dos alunos no uso da língua portuguesa, focando principalmente na concordância nominal e verbal. O projeto foi implementado para duas turmas de imersão de português da 4a série em um programa de imersão no estado de Utah. As descobertas revelam diferenças significativamente relevantes entre a avaliação previa e final no desempenho dos alunos e trazem implicações pedagógicas em relação ao método aplicado, embora, na ausência de um grupo de controle, os resultados devem ser interpretados com cautela.

Grammatical accuracy in second language (L2) writing is one of the key issues that English as a Second Language (ESL) learners struggle with, both in intensive English language programs and continuing after their university matriculation. Numerous instructional methodologies exist that center around the concept of error correction—how can or should ESL instructors correct grammatical errors in L2 students' writing to best facilitate improvements in written linguistic accuracy? Error correction in L2 writing has been a controversial issue for over a decade (e.g., Ferris, 1999; Truscott, 1996), and in an effort to contribute to an understanding of this controversial topic, this study investigated an innovative method of error correction known as dynamic written corrective feedback (WCF). For 15 weeks, 24 students at the Brigham Young University (BYU) English Language Center (ELC) received a form of dynamic WCF dramatically modified from Hartshorn's (2008) original method with the objective of increased practicality. These students produced a 30-minute pretest and posttest essay, and researchers calculated the complexity, accuracy, and fluency of each pretest and posttest. Data from the current study is compared against data from Hartshorn (2008), which found dynamic WCF to be successful in improving accuracy after carrying out similar research. The results validate previous findings and confirm that dynamic WCF is an effective approach to error correction, even when dramatically modified.