Understanding AI-enabled Teaching Innovation: An Ecosystem-based Perspective (Vietnam)

Understanding AI-enabled Teaching Innovation: An Ecosystem-based Perspective

Study overview

This repository contains the anonymized dataset associated with the study “Understanding AI-enabled Teaching Innovation: An Ecosystem-based Perspective.”

The study examines how institutional capacity, collaborative professional learning, and faculty artificial intelligence competence are associated with artificial intelligence-enabled teaching innovation in higher education. It adopts an ecosystem-based perspective in which meaningful artificial intelligence integration depends not only on access to technology, but also on institutional leadership, infrastructure, resources, professional support, collaboration, and faculty capability.

Survey data were collected from 729 university faculty members in Vietnam representing different academic disciplines, academic ranks, institutional types, and levels of teaching experience.

Main constructs

The dataset includes four main constructs.

Artificial Intelligence-Enabled Education Ecosystem

This construct represents the institutional conditions that support sustainable artificial intelligence integration in higher education. It is modelled as a reflective–reflective higher-order construct with five dimensions:

• Artificial Intelligence Leadership – strategic direction and institutional leadership for artificial intelligence integration.
• Artificial Intelligence Infrastructure – technological infrastructure and systems supporting artificial intelligence use.
• Artificial Intelligence Resources – institutional resources available for artificial intelligence-supported teaching and learning.
• Artificial Intelligence Support – professional and organizational support provided to faculty members.
• Artificial Intelligence Partnership – collaboration with external organizations, experts, and partners supporting artificial intelligence development.

Together, the five dimensions contain 27 survey items.

Professional Learning Community

This construct captures the collaborative professional environment in which university faculty members exchange knowledge, engage in collective inquiry, reflect on teaching practices, and share responsibility for professional improvement.

It is measured using 12 survey items and reflects collaborative learning, reflective dialogue, professional knowledge exchange, and continuous improvement among faculty members.

Artificial Intelligence Competence

This construct represents faculty members’ capability to use artificial intelligence effectively, critically, responsibly, and pedagogically in teaching and learning.

The construct contains 12 survey items and captures the professional knowledge and judgement required to evaluate artificial intelligence tools, select appropriate applications, consider ethical implications, and integrate artificial intelligence into instructional practice.

Artificial Intelligence-Enabled Teaching Innovation

This construct represents the extent to which faculty members use artificial intelligence to redesign and improve teaching practices. It is modelled as a reflective–reflective higher-order construct consisting of five dimensions:

• Content Innovation – using artificial intelligence to develop, adapt, and improve learning content.
• Instructional Innovation – integrating artificial intelligence into instructional strategies and teaching activities.
• Assessment Innovation – using artificial intelligence to redesign assessment, feedback, and evaluation practices.
• Engagement Innovation – using artificial intelligence to support student participation and engagement.
• Personalization Innovation – using artificial intelligence to provide adaptive and differentiated learning experiences.

Together, the five dimensions contain 23 survey items.

Dataset structure

The dataset contains 74 substantive survey items measuring the four main constructs.

It also includes background variables describing participants’:

• gender;
• academic rank;
• academic discipline;
• institution type; and
• years of teaching experience.

An anonymous identification code is provided for each respondent. No direct personal identifiers are included in the shared dataset.

Survey items were measured using a seven-point Likert scale, ranging from 1 = Strongly disagree to 7 = Strongly agree.

The accompanying codebook provides detailed information on variable names, item wording, construct membership, dimensions, and response coding.

Research model

The study examines a structural model in which the Artificial Intelligence-Enabled Education Ecosystem provides the broader institutional foundation for professional learning, competence development, and teaching innovation.

The model evaluates relationships between:

• Artificial Intelligence-Enabled Education Ecosystem and Professional Learning Community;
• Artificial Intelligence-Enabled Education Ecosystem and Artificial Intelligence Competence;
• Professional Learning Community and Artificial Intelligence Competence;
• Artificial Intelligence-Enabled Education Ecosystem and Artificial Intelligence-Enabled Teaching Innovation;
• Professional Learning Community and Artificial Intelligence-Enabled Teaching Innovation; and
• Artificial Intelligence Competence and Artificial Intelligence-Enabled Teaching Innovation.

The study additionally evaluates indirect pathways through Professional Learning Community and Artificial Intelligence Competence, including a sequential pathway connecting institutional ecosystem capacity with teaching innovation through collaborative professional learning and faculty competence.

Data analysis

Data were analysed using partial least squares structural equation modelling with SmartPLS 4.

A two-stage approach was applied to the reflective–reflective higher-order constructs. The analysis included:

• measurement model assessment;
• reliability and validity assessment;
• higher-order construct assessment;
• structural model assessment;
• bootstrapping;
• mediation analysis;
• predictive assessment;
• measurement invariance assessment; and
• gender-based multi-group analysis.

Because the research uses a cross-sectional survey design, the estimated structural relationships should be interpreted as associations rather than evidence of causal or temporal effects.

Ethics and confidentiality

The study received ethical approval from Vinmec International General Hospital JSC under Decision No. 65/2026/GCN/HĐĐĐ VMEC, dated March 24, 2026, for research involving human subject participation.

Participation was voluntary, and the dataset shared in this repository contains anonymized research data without direct personal identifiers.

Funding

This research was supported by the National Foundation for Science and Technology Development under Grant No. 503.01-2025.22 and by VinUniversity under Grant No. VUNI.2526.AREP.031.

Intended use

The dataset may support replication and secondary analyses related to:

• artificial intelligence ecosystems in higher education;
• faculty artificial intelligence competence;
• professional learning communities;
• artificial intelligence-enabled teaching innovation;
• higher-order construct modelling;
• mediation analysis;
• predictive modelling; and
• multi-group analysis.

Researchers using the dataset should consult the accompanying codebook before constructing composite scores or reproducing the measurement and structural models.