1 Introduction

The importance of self-regulation ability in human-computer collaboration, especially in generative artificial intelligence (GAI) feedback, is a key factor affecting learning outcomes. The dynamic and complex nature of GAI feedback requires learners to use different strategies for cognitive, behavioral, and emotional regulation in order to maintain positive feedback engagement and achieve modification goals. Effective feedback intake behavior can promote the internalization of self-regulation and enhance the proactivity of human-computer interaction. Although previous studies have shown a correlation between feedback investment and self-regulation strategies, further empirical research is needed to investigate how self-regulation ability affects the specific mechanism of GAI feedback investment. This study aims to explore the mechanism of self-regulation ability on the three-dimensional feedback input of behavior, cognition, and emotion, in order to reveal the internal mechanism of GAI empowering writing feedback and provide inspiration for smart writing teaching.

2 Literature Review

This article summarizes the research on AI assisted writing and feedback investment, and points out that GAI assisted writing feedback can improve writing quality and ability development. The degree of feedback investment is a key factor affecting the learning effect of AI. Feedback investment has diverse and dynamic characteristics, covering three core dimensions: cognition, behavior, and emotion. The study explored the characteristics and influencing factors of learners' feedback engagement under different feedback modes, and found that there is a close relationship between the dimensions of feedback engagement, but there are differences in the level of engagement. Feedback investment is also influenced by a combination of individual language proficiency, learning beliefs, motivation, emotions, and other factors. Self regulatory ability is closely related to feedback input. Self regulatory ability is a multidimensional and dynamic ability system, involving individuals' regulatory behaviors in cognition, motivation, emotion, and social environment. It is closely related to academic achievement and self-directed learning ability. In different learning scenarios, self-regulation ability is closely related to engagement. GAI assisted writing feedback has complex and dynamic characteristics, requiring learners to have stronger self-regulation abilities to integrate information, flexibly adjust strategies, and efficiently process machine feedback. This study adopts a case study method to deeply track the cognitive, behavioral, and emotional engagement characteristics, changes, and causes of foreign language learners with different self-regulation abilities in GAI assisted writing feedback.

3 Research Design

Introduced the research design, including the research subjects, data collection, and data analysis. The research subjects are six first-year non English major students, divided into high and low self-regulation ability groups, and their writing proficiency is evaluated by completing the CET4 writing task. The data collection adopts multi-source methods, including scales, human-computer interaction dialogues, writing and editing texts, and retrospective interviews, to explore the impact of self-regulation ability on feedback input. Data analysis involves two rounds of data collection, obtaining human-computer interaction data, writing modified texts, and retrospective interviews to construct a framework for analyzing writing feedback investment, covering three dimensions: cognitive investment, behavioral investment, and emotional investment. Using content analysis method to cyclically encode and classify data, calculate the frequency of each dimension input and its secondary indicators, and ensure reliability and validity through text rereading, result verification, and expert discussion, forming a triangulation to ensure the reliability and effectiveness of research results.

4 research findings

Research has found that students with high self-regulation abilities have higher levels of cognitive, behavioral, and emotional engagement than those with low self-regulation abilities, and there are individual differences within the group. In terms of cognitive engagement, students with high self-regulation abilities are better at using higher-order cognitive strategies such as comparing, analyzing, and finely processing feedback information, while low-level students selectively process feedback. In terms of cognitive monitoring, students with high self-regulation ability can adjust their interaction instructions with AI based on modification goals, while low-level students mechanically receive feedback. In terms of feedback understanding, students with high self-regulation ability can recognize the limitations of AI feedback and pay attention to issues such as discourse coherence and sentence structure, while low-level students mainly focus on feedback at the lexical and grammatical levels. In terms of behavioral engagement, students with high self-regulation abilities exhibit more successful uptake behaviors and diverse modification behaviors, while low-level students rely on shallow strategies. In terms of emotional engagement, students with high self-regulation abilities hold a positive attitude towards feedback content and quality, and exhibit deep cognitive processing behavior, while low-level students acknowledge the efficiency of feedback, but their modification behavior is mostly limited to the language level. The difference in negative attitudes towards feedback between the two groups of students is relatively small, mainly focusing on issues such as insufficient feedback intelligence, single content, and inconsistent grading standards. In terms of changes in feedback input, students with high self-regulation ability showed a significant decrease in the use of information organization strategies and feedback intake frequency, but remained stable in terms of modification behavior and feedback understanding, while students with low self-regulation ability showed weaker feedback information understanding ability and fewer modification behaviors in the second round of feedback. In terms of emotional investment, both groups of students showed a positive change in their attitudes towards GAI feedback. Students with high self-regulation abilities recognized the role of GAI feedback in solving language problems and alleviating psychological pressure during modification, while students with low self-regulation abilities gradually found the modification process interesting as their proficiency increased.

5 Discussions

This study found through multi-source data analysis that self-regulation ability is a key factor affecting multidimensional feedback input in the context of GAI assisted writing. Students with high self-regulation abilities perform better in cognitive and behavioral engagement dimensions, and are able to organize and understand feedback information more effectively, achieving synergy between personal modification goals and GAI feedback. Learners with higher language proficiency demonstrate stronger feedback assessment and selective processing abilities. Students with high self-regulation abilities exhibit more diverse and proactive modification behaviors, while those with low self-regulation abilities rely on shallow modification strategies. The limitations of GAI feedback can inhibit the reflective space of students with low self-regulation ability. The successful feedback intake behavior of two groups of students promoted the internalization of individual self-regulation ability and enhanced positive emotional experiences. There is an interactive relationship between learners' cognitive, behavioral, and emotional engagement dimensions, and self-regulation ability moderates learners' feedback engagement patterns by influencing cognitive processing depth, behavioral flexibility, and emotional experience. Students with high self-regulation ability maintained stable feedback understanding and behavior modification through active use of moderation strategies, while students with low self-regulation ability showed a significant decline in these two aspects. Feedback investment has the characteristics of internal interaction and dynamic changes, and self-regulation ability is a key factor affecting cognitive, behavioral, and emotional feedback investment changes. This study has important implications for GAI empowered writing teaching practice. It is recommended to implement layered teaching for students with different self-regulation abilities, enhance personalized feedback, cultivate human-computer interaction literacy, implement metacognitive training, carry out diverse interactive feedback models, promote learners' transition from passive acceptance to active negotiation, and improve human-computer collaborative construction ability.

6 Conclusion

This chapter summarizes the impact of six students' self-regulation ability on multidimensional feedback input in GAI assisted writing. Research has found that students with high self-regulation abilities exhibit more positive feedback engagement in cognitive, behavioral, and emotional dimensions, while students with low self-regulation abilities exhibit cognitive fatigue and superficial behavior. The limitations of the study lie in the small sample size and single analysis method. Future research can expand the sample range and combine multimodal data analysis to explore the neurocognitive mechanisms of feedback acceptance and modification behavior in depth.