The scale of artificial intelligence literacy for all (SAIL4ALL): assessing knowledge of artificial intelligence in all adult populations

Abstract

Abstract This study presents a new artificial intelligence (AI) literacy scale for comprehensive assessment of the concept across adult populations, regardless of the setting in which it is applied: the SAIL4ALL. The scale contains 56 items distributed across four different themes [(1) What is AI? (a: Recognizing AI, Understanding Intelligence and Interdisciplinarity; b: General vs. Narrow AI); (2) What can AI do?; (3) How does AI work?; and (4) How should AI be used?] and has two different response formats (true/false and 5-point Likert scale), each of which can be applied depending on the context. The study provides quantitative evidence of psychometric quality in three different UK samples. It also presents evidence of internal structure validity through confirmatory factor analysis and adequate internal consistency for most of the scales and formats. Moreover, it shows measurement invariance tested for gender and education level. Finally, the study also assesses the relationship of AI literacy with external measures, examining the nomological network. SAIL4ALL demonstrates positive evidence of psychometric quality, and serves as a valuable tool for determining both actual and perceived knowledge of AI, thus guiding educational, organizational, and institutional AI literacy initiatives.

Ähnliche Arbeiten