This research project looked at the impact of a robot math tutor’s social interaction skills (i.e., scaffolding and personalization) on children’s math performance and their social perception of the robot. It investigated the impact of an autonomous robot math tutor among the same children after 4 interactions over an 8-9-month interval. In a 2 (scaffolding: without vs. with) x 2 (personalization: without vs. with) between-subjects design, 113 children (9-12 years) were randomly assigned to one of the four conditions.