The scale measures the degree to which a person believes an AI assistant (e.g., Alexa, Siri) has a mind of its own — intentions, free will, consciousness, and emotions. Uysal, Alavi, and Bezçenon (2022) created the scale by drawing key words and phrases from an established measure of perceived agency by Epley et al. (2008), and it was later reused by Castelo et al. (2023).
7-point Likert (varies by study: Not at all (1) to Very much (7), or strongly disagree (1) to strongly agree (7))
Full scale items shown. Always cite the original source when using this scale in research.
Reliability was α = .86 (Study 1) and α = .90 (Study 2) for Uysal, Alavi, and Bezçenon (2022, p. 1162), and α = .91 when reused by Castelo et al. (2023, p. 857) with Prolific participants. CFA supported convergent and discriminant validity, with AVEs of .57 (Study 1) and .67 (Study 2) — the Study 1 AVE is on the lower side of acceptable. The item stem asks respondents to rate the focal AI assistant on each characteristic.
Uysal, Ertugrul, Sascha Alavi, and Valéry Bezçenon (2022), "Trojan Horse or Useful Helper? A Relationship Perspective on Artificial Intelligence Assistants with Humanlike Features," Journal of the Academy of Marketing Science, 50 (6), 1153–1175.
Free access to 2500+ psychometric scales at PsychScales