Content Validation of a Scale for Assessing Artificial Intelligence Use in Learning

Abstract

The purpose of this study was to determine the content validity of a scale designed to evaluate the use of artificial intelligence in university learning (EPIA-U). The instrument consists of 28 items distributed across four dimensions—Academic support, Study personalization, Satisfaction with use, and Perceived dependence—and was subjected to the judgment of nine experts with backgrounds in education, technology, and research. The experts assessed each item according to the criteria of clarity, relevance, and congruence, using a four-point Likert scale. Data were analyzed through Aiken’s V coefficient (for items and dimensions) and the Content Validity Coefficient (CVC) proposed by Hernández-Nieto (2011) for the global index. The results showed Aiken’s V values equal to or above the acceptance threshold (≥ .75) for almost all items and across the evaluated dimensions. However, some lower confidence interval bounds fell below the .70 criterion. Complementarily, a global CVC of .90 was obtained, reflecting a high level of agreement among the judges. A minor wording adjustment was made to one item with lower indices, improving its semantic precision. In conclusion, the results provide favorable evidence of content validity for the EPIA-U and support its progression to subsequent phases of psychometric evaluation among university students