Since the release of advanced Large Language Models (LLMs) like ChatGPT in the winter of 2022, educators have been challenged to integrate AI in educational contexts without adequate preparation. This multimethod study aims to provide valuable insights into how students and teachers can make effective, responsible use of AI tools for writing. Specifically, it focuses on the impact of ChatGPT-powered English writing instruction on writing development among emerging multilinguals in Korea. The study examines students' usage patterns of ChatGPT's five key functionalities--"vocabulary support," "grammar checking," "feedback and revision," "sentence generation," and "paraphrasing/summary"--and explores how these patterns correlate with writing development. 50 undergraduate students from four different universities voluntarily participated in a four-week ChatGPT-powered English writing course designed based on the AI literacy framework (Warschauer et al., 2023). Data were collected from pre- and post-TOEFL writing tests, ChatGPT chat histories, demographic surveys, perception surveys, and interviews. Quantitative data were analyzed using linear multiple regressions, correlation analyses, and paired-samples t-test, while qualitative data underwent content analysis. The findings revealed that students showed significant improvement in their writing scores after receiving ChatGPT-powered instruction. Students expressed high satisfaction with ChatGPT, particularly appreciating its immediate feedback and its contribution to writing fluency. Self-perceived English proficiency levels and prior English writing achievement emerged as significant factors influencing writing development, while qualitative analysis further highlighted the importance of ChatGPT familiarity as a contributing factor. Among the five key functionalities of ChatGPT, "feedback and revision" emerged as the most frequently used and the only statistically significant contributor to writing development. However, qualitative data indicated that students also found value in each of the other functionalities. Additionally, when students were divided into higher and lower prior English writing achievement groups, the lower achievement group demonstrated a significant correlation between the "paraphrasing/summary" functionality and writing development. The higher achievement group showed more balanced usage across functionalities, but there were no strong correlations with writing development. This study makes significant contributions to the field of Intelligent ComputerAssisted Language Learning (ICALL) by providing empirical evidence on the effectiveness of ChatGPT in fostering writing skills. By examining how students with different writing proficiency levels interact with specific ChatGPT functionalities and identifying their relationships to writing development, this research can help educators develop more effective AI-driven instructional strategies, better tailored to the diverse needs of their students. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com.bibliotheek.ehb.be/en-US/products/dissertations/individuals.shtml.]