Author(s): ZHANG, S., PAN, Z., MICHALCZYK, M., Institution: HUNAN UNIVERSITY OF TECHNOLOGY, Country: CHINA, Abstract-ID: 650

Introduction
Enriched research evidence supports the benefits of self-compassion in sport. However, athletes from diverse backgrounds have consistently expressed resistance to, or fear of self-compassion; they view embracing self-compassion as indicating an inferior position, signalling weakness, and as potentially compromising their performance. In such a context, one key challenge in delivering self-compassion practices to athletes is the lack of concrete evidence supporting the performance (rather than health) benefits of self-compassion. The present research, therefore, aimed to examine the causal relationships among self-compassion, fear of self-compassion, and their interaction with athlete mental toughness, which is key to maintaining performance levels and achieving performance goals under high pressure.

Method
We recruited 303 competitive club players (Mage = 29.02, SD = 11.41; 52.8% male). We followed these participants through three waves of online surveys assessing state self-compassion (SSC), fear of self-compassion (FSC), and mental toughness (MTI) with a four-week interval between successive pairs of data collection time points (each data collection window lasted one week). At the end of three months of data collection, 52.5% of participants remained throughout all data collections; 40.6% dropped out at Time 2, and an additional 6.9% at Time 3. No systematic pattern was observed among those who dropped out; thus, the dropouts were considered missing at random. All missing or dropouts were thereby retained for panel analysis and hypothesis testing using the robust Full Information Maximum Likelihood (FIML) approach for longitudinal path analysis.

Results
Overall, no missing values were found except for dropouts. Skewness and kurtosis for all study variables were within ±1.04 and ±0.75, respectively, supporting the assumption of normality. McDonald’s ω ranged .80-.83, .73-.86, and .89-.91 for the SCSS, FSCS, and MTI, respectively. The robust chi-square difference test indicated that a fixed-lagged, fixed-autoregressive effects model fit the data best. According to this best-fit model, the autoregressive effects of all study variables were significant and positive. Importantly, the lagged effect of T1/T2 SSC × FSC on T2/T3 MTI was significant (β = -.04, p = .04; 95% CI [-.08, -.01]). To unfold this interaction, the lagged effect of SSC on MTI was greater when FSC was low (β = .14, p = .00; 95% CI [.07, .21]) compared to high (β = .06, p = .04; 95% CI [.01, .11]).

Discussion
The present study provided novel, robust evidence supporting the performance benefits of self-compassion. Specifically, state self-compassion manifested a lagged effect on mental toughness. Nevertheless, it is noteworthy that the mental toughness benefits of self-compassion can be suppressed as the fear of self-compassion increases. Self-compassion is an effective yet overlooked component that should be considered in mental toughness training.