This systematic review and meta-analysis aims to quantify the overall effect of robot presence on human cognitive task performance and to examine the moderating roles of two key characteristics: interactive capability and anthropomorphic morphology. A total of 10 studies were included for synthesis. We used the standardized mean difference (SMD) to assess task time efficiency (9 studies, n=540) and task accuracy (4 studies, n=129). The meta-analysis results showed no significant overall net effect of robot presence on either task time efficiency or accuracy. However, subgroup analyses revealed a significant differential pattern for task time efficiency: interactive robots were associated with improved efficiency (SMD = -0.31, P=0.03) while non-interactive robots showed the opposite (SMD = 0.22, P=0.009); additionally, non-anthropomorphic robots were associated with decreased efficiency (SMD = 0.47, P=0.0003).These findings indicate that robot presence does not have a uniform impact on human performance; rather, the direction of its effect is bidirectionally moderated by robot characteristics. In particular, interactive capability plays a crucial role in determining whether robots facilitate or inhibit task efficiency, which underscores the importance of designing robot social roles and behaviors in future human-robot collaboration.
Ren et al. (Fri,) studied this question.