Based on the grounded theory, 46 interviewees were interviewed through personal in-depth interviews and focus group interviews, and the interview data was analyzed in three stages to construct an integrated analysis model of social media users' anxiety. User anxiety is affected by emotional cognition and experience, platform environment and operation quality. Under the effect of user personality traits, it is characterized by state anxiety and general anxiety. Anxiety further leads to the user's avoidance behavior. The intensity and behavior of the avoidance behavior are regulated by the characteristics of the individual's ability, and positive avoidance behavior or passive avoidance behavior occurs. Active avoidance behavior will have a feedback effect on the Weibo platform environment and the quality of Weibo operations when users use it personally.