Effect of Interactional condition on total speech of participants
Interactional condition was found to have a significant effect on the total number of participant utterances (ANOVA, F (1.9, 33.4) = 8.13, p < 0.001). T-tests showed participants produced more utterances with robot (M = 43.0, SD = 19.4) than with adult interactional conditions (M = 36.8, SD = 19.2, t(23) = 1.97, p < 0.05) and that robot and adult interactional conditions produced more utterances than the touchscreen computer game condition (M = 25.2, SD = 13.4, t(23) = 3.61, p < 0.001).
Effect of interactional condition on participant speech towards confederate
Interactional condition was found to have a significant effect on the number of participant utterances directed toward the confederate (ANOVA, F (1.8, 33.0) = 3.46, p < 0.05). T-tests show participants with ASD directed a higher number of utterances towards their confederate in the robot (M = 29.5, SD = 16.6) than in the adult interactional condition (M = 25.5, SD = 15.5), t (23) = 1.87, p < 0.05) and more in both the robot (t(23) = 3.05, p < 0.01) and adult (t(23) = 2.15, p < 0.01) than the game condition (M = 20.5, SD = 10.1).
Effect of interactional condition on participant speech: comparison between robot, adult
and game interaction partners The number of utterances directed towards the interaction partner was found to vary depending on interactional condition (ANOVA, F(1.5, 26.9) = 15.20, p < 0.001). T-tests show there were significantly more utterances directed towards the robot (t(23) = 5.40, p < 0.001) and the adult interaction partners (t(23) = 8.22, p < 0.001) than towards the touchscreen computer game. However, there was no difference between the number of utterances directed towards the interaction partner in the robot (M = 13.5, SD = 12.0) and adult conditions (M = 13.5, SD = 7.8); (t(23) = 0.02).