Expert Validation Instrument for AI-Enhanced Project-Based Learning Model

An expert validation instrument assesses the validity and feasibility of an AI-Based Project-Based Learning model across five main aspects using a 5-point Likert scale. The instrument, created by Ahmad Swandi, contains structured evaluations for conceptual, pedagogical, technological, and practical dimensions. Validation results analyzed with Aiken's V indicate high validity scores (V ≥ 0.80) for most aspects, particularly in conceptual and pedagogical domains.

Use Cases

• Analyze expert ratings on the 'conceptual feasibility' aspect to validate the model's alignment with constructivist and PjBL theories.
• Compare scores across 'pedagogical validity', 'technological validity', and 'practical feasibility' aspects to identify areas needing revision in the AI-PjBL model.
• Use the 'overall assessment' Likert scale ratings and Aiken's V validity scores to quantify the instrument's reliability for evaluating educational interventions.
• Examine expert feedback on 'technological adaptability' indicators to inform model adjustments for implementation under limited infrastructure conditions.

Strengths

• Structured instrument covering five distinct validation aspects: conceptual, pedagogical, technological, practical, and overall.
• Validation results quantified using Aiken's V, with most aspects achieving high validity scores of V ≥ 0.80.

Limitations

• The dataset is very small at 22.2 KB, representing a single validation instrument rather than a large-scale dataset.
• No raw expert response data or sample size (number of experts, rows of data) is provided, limiting statistical analysis.
• The instrument is specific to one proposed AI-PjBL model, limiting generalizability to other educational frameworks.

Provenance

Source

Ahmad Swandi via figshare.

Collection Method

Designed as a structured validation instrument for expert judgment, with results analyzed using Aiken's V coefficient.

Freshness

Last updated on April 8, 2026.