Design of the Attitudinal Assessment Scale Towards Artificial Intelligence (EVAIA-1)

Abstract

In recent years, the exponential growth of artificial intelligence as a technological tool at the service of human beings has led to an ethical debate about its future implication. The existing instruments to evaluate attitudes towards artificial intelligence have non-specific dimensions and are designed for populations different from the Spanish-speaking. In this sense, it is necessary to have valid, reliable, and contextualized tools to evaluate people’s attitudes toward the use of artificial intelligence. Therefore, the present study aimed to develop an attitudinal rating scale for artificial intelligence. There were 604 volunteer participants between 18 and 55 years of age, 311 men and 293 women. Bartlett’s test of sphericity showed a significant result (approximate chi-square = 1502. 7862387833S;p <.001), and the Kaiser-Meyer-Olkin test of sample adequacy showed an index of.825. With this, it was considered feasible to factorize the data matrix, and thanks to the factor analysis, three components explain 52.76% of the total rotated variance. In addition, a high internal consistency index was obtained for the 12 items of the inventory (0.768). These findings indicate that the EVAIA-I is a valid and reliable tool to evaluate the attitude towards artificial intelligence in Ecuador and other Latin American countries.