Structured Questionnaire and Dataset: Feedforward, Learner Agency and GenAI, A Cybernetic Perspective on AI-mediated Learning
Description
This dataset contains anonymised responses collected through a structured mixed-method questionnaire administered to undergraduate engineering students enrolled in a Lean Manufacturing module at a private university. The survey was designed to investigate how learners enact feedforward, self-regulation, learner agency, and interaction with generative artificial intelligence (GenAI) within higher education. The instrument combines closed-ended Likert-scale items with open-ended questions. The quantitative component operationalises seven dimensions: (1) Intended Learning Outcome (ILO) awareness, (2) Feedforward practices, (3) Self-regulation, (4) GenAI interaction and co-regulation, (5) Learner agency, (6) Opportunities and challenges associated with GenAI, and (7) Recursive learning and adaptation. The qualitative component consists of six open-ended questions exploring learners' experiences of using GenAI during learning and assessment. The dataset underpins the article Feedforward, Learner Agency, and GenAI: A Cybernetic Perspective on AI-mediated Learning and was used to develop and illustrate a cybernetics-informed conceptual framework of feedforward, learner agency, and human–AI interaction.
Files
Steps to reproduce
1. Open the Excel dataset and identify the worksheet containing the anonymised questionnaire responses. 2. Separate quantitative Likert-scale responses from qualitative open-ended responses. 3. Group quantitative items according to the seven constructs defined in the questionnaire: - Intended Learning Outcome (ILO) Awareness - Feedforward Practices - Self-Regulation - GenAI Interaction - Learner Agency - Opportunities and Challenges - Recursive Learning and Adaptation 4. Compute descriptive statistics (means, modes, and standard deviations) for individual items and construct-level summaries. 5. Examine frequency distributions to identify patterns and variability across constructs. 6. Analyse open-ended responses using an inductive thematic analysis involving familiarisation, initial coding, category development, theme refinement, and review. 7. Integrate quantitative patterns and qualitative themes to identify areas of convergence, complementarity, and divergence across the conceptual framework. 8. Compare the resulting findings with the cybernetics-informed framework presented in the associated publication.