“…The sample size needed for serial mediation was calculated using the Monte Carlo Power Analysis webpage (Schoemann et al., 2017). Referencing the correlation coefficients reported in previous studies, the correlation coefficients between self‐stigma and social support, between self‐stigma and perceived hope, between self‐stigma and perceived recovery, between social support and perceived hope, between social support and perceived recovery, and between perceived hope and perceived recovery were set to −.353 (Li et al., 2021), −0.582 (Lim et al., 2020), 0.560 (Lim et al., 2020), 0.219 (Mutschler et al., 2022), and 0.641 (Sari et al., 2021), respectively; referencing previous relevant studies that used the tools expected to be employed in this study, the standard deviations for self‐stigma, social support, perceived hope and perceived recovery were set to 10.83 (Lin et al., 2022), 15.48 (Prabhakaran et al., 2021), 5.02 (Lim et al., 2020), and 9.04 (Yang & Lo, 2017), respectively. The calculation by the webpage indicated that when the sample size was 110, the power was 0.97.…”