DOI
10.5703/1288284317859
Description
Effective assessment is essential in construction management (CM) education to evaluate student understanding of industry-demanded skills. Integrating the Bloom’s Taxonomy, American Council for Construction Education (ACCE) provides student learning outcomes (SLOs) for targeted learning objectives, as well as guiding assessment development. However, crafting high-quality assessment questions remains challenging for CM instructors. This study explores the potential of ChatGPT to generate assessment questions aligned with Bloom’s Taxonomy and using two ACCE student learning outcomes (SLOs): 1) creating construction project cost estimates and 2) analyzing methods, materials, and equipment. Generated questions were evaluated for alignment with Bloom’s levels, clarity, sufficiency, feasibility, and relevance. Results indicate ChatGPT can produce cognitively appropriate questions, though issues with clarity and completeness remain. The study concludes that ChatGPT can assist CM educators in assessment development when supported by effective prompt design and added context. Future research should explore expanded data collection and domain-specific training for improved outcomes.
A Preliminary Examination of ChatGPT’s Capability in Developing Construction Education Instruments Through Bloom’s Taxonomy
Effective assessment is essential in construction management (CM) education to evaluate student understanding of industry-demanded skills. Integrating the Bloom’s Taxonomy, American Council for Construction Education (ACCE) provides student learning outcomes (SLOs) for targeted learning objectives, as well as guiding assessment development. However, crafting high-quality assessment questions remains challenging for CM instructors. This study explores the potential of ChatGPT to generate assessment questions aligned with Bloom’s Taxonomy and using two ACCE student learning outcomes (SLOs): 1) creating construction project cost estimates and 2) analyzing methods, materials, and equipment. Generated questions were evaluated for alignment with Bloom’s levels, clarity, sufficiency, feasibility, and relevance. Results indicate ChatGPT can produce cognitively appropriate questions, though issues with clarity and completeness remain. The study concludes that ChatGPT can assist CM educators in assessment development when supported by effective prompt design and added context. Future research should explore expanded data collection and domain-specific training for improved outcomes.