The effects of virtual human’s verbal persuasion strategies on user intention and behavior

•Relative effectiveness of Cialdini’s persuasion strategies varies in the virtual human context.•Implementing Cialdini’s persuasion strategies using virtual humans can help overcome the limitations of prior work in human-computer persuasion.•Latent Profile Analysis was used to identify personality types in the study population.•User personality influences the effectiveness of Cialdini’s persuasion strategies in the context of virtual humans. Prior work has found that computers can effectively use six persuasion strategies characterized by Cialdini to influence people’s intentions and behaviors. However, researchers are yet to examine the effectiveness of Cialdini’s persuasion strategies with virtual humans. Virtual humans provide a human representation to computers, which influences how people respond to persuasion attempts from computers. To evaluate Cialdini’s persuasion strategies with virtual humans, we conducted an online study (N=183), where a virtual human promoted a coping skill for good mental health using Cialdini’s six persuasion strategies. Our results reveal that strategies that persuaded users by increasing the feeling of liking and reciprocity towards the virtual human were most successful in changing user intentions to perform the recommended behavior. Furthermore, the relative effectiveness of strategies that involved persuasion using expertise and normative beliefs varied for different user personality types. We draw conclusions and design implications for using Cialdini’s persuasion strategies with virtual humans.