For a long time, the need for belongingness has been considered as one of the most important basic human needs. Peer interactions make important and unique contributions to children's social, emotional, cognitive, and moral development (Rubin, Bukowski, & Bowker, 2015), and Lacking it may cause mental illness and maladjustment (Rubin, Coplan, & Bowker, 2009). However, socially withdrawn children habitually remove themselves from opportunities for peer interaction. In order to better understand this group of children, Asendorpf (1990) proposed a conceptual model postulating different child psychosocial outcomes as a function of social approach and social avoidance motivations. According to this model, avoidant children were portrayed as both desiring being solitude and actively seeking to avoid social interaction, whom were consequently at the most risk for social and emotional maladjustment. Although very little is known about social avoidance in childhood (Coplan et al. 2015), three postulations hypothesize the processes that may underlie the development of avoidant children (Asendorpf, 1990; Bowker & Raja, 2011; Chen, 2010; Coplan & Armer 2007; Coplan et al., 2015; Rubin et al., 2009; Schmidt & Fox, 1999), which are social anxiety, negative peer experiences and depression. However, there has been very limited empirical study of social avoidance to compare those three conceptual models directly. To our knowledge, only one study using short-term longitudinal design examined those hypotheses, and provided evidence supporting depressive symptoms as a salient predictor of social avoidance (Ding et al., 2018). Two approaches are available to measure social avoidance. One indirect approach is to select avoidant children based on the model of social motivations. Researchers speculated that children characterized by low approach motivation and high avoidance motivation would be defined as socially avoidant children according to Asendorpf (1990) before. The other direct approach for measurement on social avoidance, researcher used self-reported or parental-reported Social Avoidance Scale to test the level of avoidance directly.