Dinghui Ma, Hourong Yu, Weijia Li, Yi Tang
Appearance anxiety and appearance-oriented social media behaviors show prospective bidirectional between-persons associations in a Chinese platform ecology, robust to sensitivity checks. FNE may function as one candidate link between habitual appearance anxiety and downstream appearance-management behavior, though a reverse-order robustness check could not rule out FNE preceding anxiety. The present study did not find statistically reliable evidence that cognitive-level media literacy buffers this pathway, tentatively suggesting that effective intervention may need to target appearance-specific anxiety directly rather than relying solely on cognitive media literacy education. An exploratory finding that media literacy may strengthen, rather than weaken, the FNE-behavior link for beauty-filter use is counter-theoretical and warrants confirmatory replication before informing intervention design.
BACKGROUND: Research on appearance anxiety and social media use has predominantly treated social media exposure as the cause and appearance anxiety as the outcome. This study examined the reverse and bidirectional pathways: whether appearance anxiety is prospectively associated with three functionally distinct appearance-oriented social media behaviors, and whether these behaviors are associated with subsequent anxiety, in young Chinese women.
METHOD: A two-wave panel design with a 6-month interval was employed with 339 young Chinese women (aged 18-35; 87.1% Wave-2 retention). Cross-lagged panel models (CLPM) estimated bidirectional prospective between-persons associations, with an assumption-dependent phantom-variable sensitivity analysis used only to assess robustness under fixed stable-variance assumptions. Fear of negative evaluation (FNE) was examined as a candidate dispositional link using a partially longitudinal indirect-association model (T1 anxiety -> T1 FNE -> T2 behavior, controlling for baseline behavior), since FNE was assessed only at baseline. Whether cognitive-level media literacy attenuated this indirect association was addressed as a research question.
RESULTS: Appearance anxiety was prospectively associated with all three T2 behaviors at the between-persons level (βs = 0.28-0.39, ps < 0.001), and reverse-direction paths were of comparable or larger magnitude (βs = 0.38-0.50); both directions held in sensitivity analyses. Findings were consistent with a modest indirect association through FNE (indirect effects = 0.057-0.095), smaller than the direct effects of anxiety (0.235-0.339). An exploratory extension testing media literacy as a moderator of the FNE-to-behavior link found a significant, counter-theoretical effect for beauty-filter use only.
CONCLUSIONS: Appearance anxiety and appearance-oriented social media behaviors show prospective bidirectional between-persons associations in a Chinese platform ecology, robust to sensitivity checks. FNE may function as one candidate link between habitual appearance anxiety and downstream appearance-management behavior, though a reverse-order robustness check could not rule out FNE preceding anxiety. The present study did not find statistically reliable evidence that cognitive-level media literacy buffers this pathway, tentatively suggesting that effective intervention may need to target appearance-specific anxiety directly rather than relying solely on cognitive media literacy education. An exploratory finding that media literacy may strengthen, rather than weaken, the FNE-behavior link for beauty-filter use is counter-theoretical and warrants confirmatory replication before informing intervention design.