The Impact of Large Language Models on EFL Learners Higher-Order Thinking Skills in Iraqi Context
Generative AI can be utilized for EFL teaching to develop higher-order thinking skills, but it may also present challenges in academic integrity. Little is known about the combined influence of DeepSeek over three crucial factors of critical thinking, creativity, and self-reflection in the Iraqi EFL context. Therefore, this study seeks to bring to light how these skills are affected by the assistance of DeepSeek qualify students' impressions of the learning process. Qualitative methods and analyses: The researcher used a sequential explanatory design. The subjects of the study were 200 undergraduate students who were randomly split into two groups (experimental group [N = 100]; control group [N = 100]). Data were collected by means of pre-validated scales: Critical Thinking, Reflective thinking Creative thinking. Additionally, the researcher obtained qualitative data utilizing a semi-structured interview guide. As for the quantitative data they were analyzed with ANCOVA. As for the analyses of qualitative data thematic coding was utilized. Significant differences were found in measures of critical, creative, and reflective thinking scores between the DeepSeek group and the control group. Qualitative results reflected a paradox in students' perceptions of AI; despite appreciating how AI facilitated skills through the generation of ideas and opportunities for safe practice, students simultaneously expressed concern about cognitive offloading and skill atrophy.