Development and Validation of the Scale for Artificial Intelligence Anxiety

Abstract

The aim of this study was to develop a self-administering scale to assess Artificial Intelligence Anxiety (AI Anxiety). Inter-rater reliability, Cronbach’s Alpha reliability, Construct validity, and Concurrent Validity of the Scale for Artificial Intelligence Anxiety have been demonstrated. Inter-rater reliability (Fleiss’ Multirater Kappa) was calculated to be 0.534. The final scale was administered to a total of 881 adults from India. The sample comprised 432 males and 449 females, with ages ranging from 20-59. Cronbach’s alpha was found to be 0.922. Construct Validity was established as 0.543 through Exploratory Factor Analysis. Concurrent validity established by correlating the scores to those obtained on the Artificial Intelligence Anxiety Scale (Wang & Wang, 2019) was 0.772. Mean score for the total sample was 69.41 and S.D. was 15.43. For the male sample, the mean score and S.D. were found to be 69.815 and 15.677 respectively. For the female sample, the values were 69.015 and 15.195 respectively. Higher mean values indicate greater Artificial Intelligence Anxiety.