•  
  •  
 

CIB Conferences

Abstract

The rapid uptake of generative AI in higher education has made student AI use increasingly common, but frequent use does not necessarily imply sound academic judgement. This study examines the relationship between self-reported AI literacy and scenario-based judgement among 379 undergraduate construction economics students at Hanoi University of Civil Engineering (HUCE). A cross-sectional survey combined 32 self-reported AI literacy items across four dimensions with six scenario-based judgement tasks covering academic writing, problem solving, use of project data, and source verification. Descriptive statistics, reliability checks, and correlational analyses were used. Students reported moderately positive AI literacy across all four dimensions, yet judgement was uneven across scenarios. Performance was strongest on source verification and bounded assistance in individual assignments, but weaker on privacy-sensitive use of project data and boundaries of authorship in AI-assisted writing. Scenario scores were positively associated with critical thinking, safe use, and lecturer guidance, whereas AI-use frequency alone was not a clear positive predictor. The findings suggest that using AI often is not the same as using it well in academically responsible ways. For construction economics education, the results point to the value of explicit, scenario-based guidance on redaction, verification, disclosure, and authorship.

Keywords

generative AI; AI literacy; scenario-based judgement; higher education; construction economics; academic integrity; responsible AI use

Download

Share

COinS