Psychological safety is a group-level concept defined as the shared belief that a team is safe for interpersonal risk-taking; however, contrary to its definition, the belief is not always shared. Previous studies have demonstrated that, even in a same group, low-status members perceive lower levels of psychological safety than high-status members. To understand this gap between theoretical assumption and empirical findings, we distinguished between perceived safety (i.e., how much each member believes other members are tolerant of interpersonal risk-taking) and actual safety (i.e., how much each member is actually tolerant of other members' interpersonal risk-taking) and examined the difference between them. This approach can help us understand whether lower-status members underestimate psychological safety and whether higher-status members overestimate it. We surveyed 739 students in a school to measure each student's perceived level of psychological safety (e.g., "My classmates would not blame me, even when I point out an issue in the class") and their status (e.g., "My friendship group has a central position in my class"). We also calculated each class's actual level of psychological safety by averaging students' actual attitudes (e.g., "I would not blame my classmates, even when they point out an issue in class"). The results demonstrated that perceived psychological safety was lower than actual safety, and this gap was greater among lower-status students. Perceived psychological safety among lower-status students was lower than actual safety, while no difference was found among higher-status students. The results imply that lower-status members underestimate psychological safety, while higher-status members accurately estimate it. Since psychological safety can increase a member's performance, job satisfaction, and work-engagement, informing lower-status members of their actual psychological safety and modifying their underestimated psychological safety can increase their performance.