Abstract
We conducted questionnaire investigations how liberal arts and informatics students estimated the agents' abilities on three different artificial agents that have different appearances; a PaPeRo human-like robot, an AIBO dog-like robot and a MindStorms machine-like robot. The results showed that 1) the PaPeRo robot received the better estimations from all students, 2) the AIBO and the MindStorms received the significantly better estimations from the informatics students whereas did the worse ones from the liberal arts students.
