From Confucian Values to AI-Supported Learning in Vietnam: Motivational Profiles and Academic Outcomes

Description The dataset “From Confucian Values to AI-Supported Learning in Vietnam: Motivational Profiles and Academic Outcomes” contains three-wave longitudinal data collected from Grade 12 students in Vietnam during the 2025 academic year. The dataset is designed to examine how motivational profiles, Confucian belief systems, and AI-supported learning behaviors jointly influence student engagement and academic performance in a high-pressure educational context. Research Objectives The primary objective of this dataset is to support the investigation of dynamic motivational configurations among students and their implications for learning outcomes. Specifically, the dataset enables the examination of how Confucian belief systems relate to motivational profiles, how AI-supported learning behaviors vary across different profiles, and how these factors together explain variation in engagement and academic achievement. The dataset is also structured to support person-centered longitudinal analysis, allowing researchers to identify motivational profiles, track transitions across time, and evaluate the stability and change of motivational patterns under sustained academic pressure. Study Design and Context The dataset is based on a three-wave longitudinal design conducted during the second semester of the 2025 academic year. Data were collected at three key time points, including the beginning, middle, and end of the semester. This design enables the analysis of temporal dynamics in motivation, learning behavior, and academic outcomes. Participants consist of Grade 12 students in Vietnam, a context characterized by high-stakes national examinations and strong cultural expectations related to academic success. The dataset reflects a learning environment in which effort, persistence, and performance are shaped by both internal beliefs and external pressures. Variable Structure Control Variables The control block includes baseline characteristics of students. The variable gender is coded as an observed binary variable with 0 representing male and 1 representing female students. The variable PAA (Prior Academic Achievement) captures students’ prior GPA on a 0–10 scale. The variable PPAP (Perceived Parental Achievement Pressure) is measured as a scale mean on a 1–7 Likert scale and reflects perceived parental expectations regarding academic success. Participation Variables Participation across waves is tracked using three observed variables: t1_participated, t2_participated, and t3_participated, each coded as 0 (no participation) or 1 (participation). These variables allow for the assessment of attrition and longitudinal data completeness. Motivational Profile Variables Motivational profile membership is represented through latent categorical variables derived from clustering procedures. The variables profile_t1, profile_t2, and profile_t3 indicate profile membership at each wave and are coded from 1 to 5. These variables enable person-centered and longitudinal transition analyses. Motivation Variables The motivation block captures core dimensions of academic motivation. The variable MGO (Mastery Goal Orientation) reflects a focus on learning and competence development. The variable PAP (Performance-Approach Goal Orientation) represents the tendency to demonstrate competence relative to others. The variable PAV (Performance-Avoidance Goal Orientation) captures the tendency to avoid failure and negative evaluation. The variable ATV (Academic Task Value) measures perceived importance and usefulness of learning tasks. The variable AAM (Academic Amotivation) reflects lack of motivation or disengagement. All variables are measured as scale means using a 1–7 Likert scale. Confucian Belief Variables Culturally grounded learning beliefs are represented by two variables. The variable SCB (Self-Cultivation Belief) reflects belief in continuous self-improvement through learning. The variable EBI (Effort Belief for Improvement) captures belief in effort as a pathway to success. Both variables are measured on a 1–7 Likert scale. AI Behavior Variables AI-supported learning behaviors are captured through two variables. The variable AGAI (Agentic AI Learning Use) reflects active and intentional use of AI to support learning. The variable AICO (AI Cognitive Offloading) represents the delegation of cognitive processes to AI systems. Both variables are measured as scale means on a 1–7 Likert scale. Engagement Variable The variable EPE (Exam Preparation Engagement) represents effort, persistence, and focus during exam preparation. This variable is measured as a scale mean on a 1–7 Likert scale and serves as a behavioral indicator of learning engagement. Academic Outcome Variables Academic performance is measured using two observed variables. The variable GPA (Semester Grade Point Average) reflects overall academic performance during the semester on a 0–10 scale. The variable EXAM (Exam Composite Score) represents average performance across subjects, also measured on a 0–10 scale. Missing Data Missing values are consistently coded using the value 999, allowing for standardized handling across statistical analyses, including clustering, mediation modeling, and machine learning procedures. Data Characteristics and Usage All psychological constructs are measured using Likert-scale instruments and reported as scale means. The dataset satisfies standard requirements for reliability and validity, supporting advanced analytical approaches. The dataset can be used for clustering analysis, longitudinal transition modeling, mediation analysis, and machine learning prediction. Its structured design allows integration into both statistical modeling frameworks and data-driven approaches for educational research. Ethical Considerations Participation was voluntary, and all respondents were informed about the purpose of the study prior to data collection. The dataset has been anonymized to ensure that no personally identifiable information is included. All procedures followed institutional ethical standards.