Validation of the Korean Version of the Self-Compassion Scale-Short Form Among Young Adults: A Comparison of Factor Structures

Danbi Oh; Hyunjoo Na; Young-Eun Jung

doi:10.30773/pi.2025.0073

Psychiatry Investig > Volume 22(10); 2025 > Article

Oh, Na, and Jung: Validation of the Korean Version of the Self-Compassion Scale-Short Form Among Young Adults: A Comparison of Factor Structures

Original Article

Psychiatry Investigation 2025;22(10):1139-1146.

Published online: September 16, 2025

DOI: https://doi.org/10.30773/pi.2025.0073

Validation of the Korean Version of the Self-Compassion Scale-Short Form Among Young Adults: A Comparison of Factor Structures

Danbi Oh¹

, Hyunjoo Na¹

, Young-Eun Jung²

¹College of Nursing, The Catholic University of Korea, Seoul, Republic of Korea

²Department of Psychiatry, Jeju National University Hospital, Jeju, Republic of Korea

Correspondence: Hyunjoo Na, RN, PhD College of Nursing, The Catholic University of Korea, 222 Banpo-daero, Seocho-gu, Seoul 06591, Republic of Korea Tel: +82-2-3147-8442, Fax: +82-2-3147-8660, E-mail: hjna@catholic.ac.kr

Correspondence: Young-Eun Jung, MD, PhD Department of Psychiatry, Jeju National University Hospital, 15 Aran 13-gil, Jeju 63241, Republic of Korea Tel: +82-64-717-1234, Fax: +82-64-717-1849, E-mail: jyejye77@daum.net

Received February 24, 2025 Revised June 18, 2025 Accepted July 28, 2025

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Objective

Self-compassion plays a crucial role in emotional well-being and positive mental health. Self-Compassion Scale (SCS) can facilitate measurement of psychological distress and positive affect. SCS-Short Form (SCS-SF) offers an economical and efficient means of reducing burden of both respondents and researchers.

Methods

This study conducted secondary data analysis to construct and validate the Korean version of the SCS-SF (K-SCS-SF) using a sample of 650 young adults residing in Korea. To evaluate the factorial structure of the scale, confirmatory factor analyses were performed on two-, three-, and six-factor models, and model comparisons were conducted using multiple fit indices. Concurrent validity was assessed by analyzing correlations of the SCS-SF with measures of resilience, depression, and anxiety. Reliability was examined based on item-total correlations and internal consistency analysis.

Results

Findings revealed that the two-factor model comprising negative and positive indicators was the best-fitting model. The two-factor model demonstrated strong validity and reliability in its application, highlighting its robustness in assessing self-compassion and its associations with resilience, depression, and anxiety.

Conclusion

The validated K-SCS-SF serves as a reliable, efficient, and practical assessment tool, advancing self-compassion research and its applications in mental health assessment, intervention studies, and psychological well-being research.

Keywords: Self-compassion, Short form, Validation, Young adults, Korean

INTRODUCTION

Compassion refers to the feeling that arises when witnessing someone else’s suffering, which then prompts a desire to alleviate that suffering by taking action to help that person [1]. Compassion not only can be felt for the suffering of others, but also can be directed toward oneself during hardship (termed self-compassion), whether the suffering stems from external factors or from personal mistakes, failures, and inadequacies [2]. Self-compassion involves acknowledging suffering as part of the human experience, treating oneself with kindness, and maintaining a balanced awareness by distancing oneself from suffering [3]. A number of studies have reported that self-compassion plays a crucial role in emotional well-being and positive mental health [4,5]. Adults with higher levels of self-compassion generally experience fewer symptoms of depression, anxiety, and stress, while maintaining equanimity, happiness, optimism, and positive affect [6,7]. This may be particularly relevant in young adults, a population known to be psychologically vulnerable due to major life transitions including graduation, career initiation, and family formation [8]. According to recent research, the level of self-compassion in young adults is significantly associated with various indicators of mental health, including psychological distress and well-being [9,10].

Self-compassion is conceptualized as a multifaceted construct that has been hypothesized to encompass two, three, or six factors. First, it has been framed as a bipolar continuum of negative and positive indicators, ranging from uncompassionate self-responding to compassionate self-responding when experiencing distress [11,12]. Second, it has been suggested to have three broad domains: attention to suffering, cognitive understanding of one’s situation, and emotional response [3]. According to Neff [13], self-compassion encourages a balanced approach to suffering through mindfulness rather than overidentification, viewing struggles as part of the shared human experience rather than as isolating events, and responding with kindness instead of self-judgment. Third, these three broad domains are further divided into six subscales, with each component consisting of two opposing factors: over-identification versus mindfulness, isolation versus common humanity, and self-judgment versus self-kindness [14]. Over-identification refers to being overly absorbed in one’s painful thoughts and feelings, whereas mindfulness involves maintaining balanced awareness of them. Isolation reflects perceiving one’s experiences as separate and isolated while common humanity emphasizes viewing them as part of the broader human experience. Self-judgment is characterized by harsh self-criticism, in contrast to self-kindness, which involves understanding oneself [3].

A recent meta-analysis has highlighted that compassion-focused therapy can increase self-compassion while reducing self-criticism, depression, and eating disorders [15]. Similarly, compassionate mind training has been found to be able to reduce self-criticism [16]. Furthermore, another meta-analysis has demonstrated that mindfulness-based stress reduction can enhance empathy and spiritual values, while alleviating stress, anxiety, and rumination [17]. In contrast, individuals with lower levels of self-compassion tend to exhibit avoidance of problems and rumination on negative thoughts and emotions, which may contribute to psychopathologies such as depression, anxiety, stress, eating disorders, and bipolar disorder [12,18]. To demonstrate the effectiveness of such interventions and to accurately assess self-compassion, highly valid and efficient measurement scale needs to be implemented beforehand.

In recent years, research on self-compassion has garnered significant attention in fields of psychology and mental health. Self-Compassion Scale (SCS) with 26 items developed by Neff [3] is widely used as a tool to measure self-compassion levels. In Korea, Kim et al. [19] have adapted the SCS and validated it, developing a Korean version of the SCS (K-SCS). However, the original SCS might be burdensome under time-constrained situations and result in less thoughtful and more uniform responses to later-positioned questions [20]. Therefore, reducing the number of items is critical for enhancing the efficiency of psychological assessments while allowing for the inclusion of multiple psychological variables within a limited timeframe. To address this issue, Raes et al. [21] have developed the SCS-Short Form (SCS-SF), a 12-item version that maintains core elements of the original scale while reducing respondent burden. Although the SCS-SF was originally developed based on the sixfactor structure of the long form, its development was also motivated by concerns regarding the structural instability of the original scale. In light of these concerns, subsequent studies have proposed a two-factor model, which has demonstrated superior model fit across diverse populations [22,23]. These findings informed the decision to examine multiple factor structure models in the present study. Beyond differences in the number of items, the original and short forms share the same scale types and scoring methods. The validity of the short form has already been established across various populations, including students in the USA [24] and the UK [25], as well as nurses in Spain [26]. Given the cultural variations in the concept of self-compassion, it is essential to ensure its reliability and validity across diverse linguistic and cultural contexts.

Therefore, this study aimed to validate the Korean version of the SCS-SF (K-SCS-SF) based on the items of the Korean version developed by Kim et al. [19]. Specifically, this study seeks to validate a more efficient assessment of self-compassion by preserving the reliability and validity of the original scale while reducing response burden and enhancing its applicability for assessing self-compassion in Korea.

METHODS

Study design and sample

This study conducted a secondary analysis to evaluate the K-SCS-SF using data from Oh and Na [27]. Participants were recruited via online advertisements placed at mental health and suicide prevention centers in Seoul, Gyeonggi-do, and Jeju, Korea. Inclusion criteria were as follows: 1) adults aged 19 to 29 years at the time of the survey, 2) those who could comprehend the study’s purpose, and 3) those who provided informed consent. Exclusion criteria were: 1) those with cognitive or thought-related impairments that could hinder their understanding of the study’s objectives and 2) those with psychiatric symptoms that could impair assessment of suicidal intentions. A total of 133 participants who provided duplicate or identical responses across all items were excluded, resulting in a final sample size of 650. Ethical approval for this study was obtained from the Institutional Review Board (IRB approval number: MC22QISI0122).

Measurements

Self-compassion

Self-compassion was assessed using the K-SCS-SF, which was adapted from the original SCS-SF developed by Raes et al. [21]. The K-SCS-SF utilized the existing Korean translation of the original scale established by Kim et al. [19], in accordance with the item structure of the original SCS-SF [21]. No additional translation was performed to ensure consistency and comparability. The K-SCS-SF comprises 12 items divided into three main components, each paired with two subscales: over-identification versus mindfulness, isolation versus common humanity, and self-judgment versus self-kindness. Each subscale consists of two items and responses are rated on a 5-point Likert scale, with a higher mean score indicating greater self-compassion. Subscale scores were calculated as the mean of individual item scores. The overall self-compassion score was derived by reversing the scores for over-identification, isolation, and self-judgment, and then averaging all subscale scores.

Resilience

Resilience was measured using the Korean version of the Brief Resilience Scale (BRS) [28,29] consisting of six items rated on a 5-point scale. The mean score was calculated, with a higher value indicating a greater resilience [30]. In this study, the internal consistency of the scale had a Cronbach’s alpha value of 0.90.

Depression

Depression was assessed using the Korean version of the Patient Health Questionnaire-9 (PHQ-9) [31,32] comprising nine items for evaluating depressive symptoms. Each item was rated on a scale from 0 (not at all) to 3 (nearly every day) based on symptom frequency over the preceding 2 weeks. The total score ranged from 0 to 27, with a higher score indicating a more severe depression. In this study, the Cronbach’s alpha value for the scale was 0.93.

Anxiety

Anxiety was evaluated using the Generalized Anxiety Disorder Scale (GAD-7) [32,33] consisting of seven items for measuring anxiety symptoms, with each rated on a scale from 0 (not at all) to 3 (nearly every day) based on the frequency of symptoms over the last 2 weeks. The total score ranged from 0 to 21, with a higher score indicating a more severe anxiety. The internal consistency in this study had a Cronbach’s alpha value of 0.93.

Statistical analysis

Data were analyzed using IBM SPSS version 27 and AMOS 27. Descriptive statistical methods were employed to summarize sociodemographic characteristics of the sample. To assess construct validity, confirmatory factor analysis (CFA) was conducted without a prior exploratory factor analysis, based on the premise that the factor structure of the K-SCS-SF has been theoretically and empirically validated in previous studies [3,21,24]. This approach aligns with prior validation studies that have consistently supported the robustness of the factors across diverse populations.

The analysis employed a range of fit indices, including incremental fit indices such as the Comparative Fit Index (CFI) and the Non-Normed Fit Index (NNFI), and absolute fit indices such as the Root Mean Square Error of Approximation (RMSEA), the Standardized Root Mean Square Residual (SRMR), and the normed chi-square (χ²/df). Parsimony fit indices were also considered, including the Akaike Information Criterion (AIC), the Bayesian Information Criterion (BIC), the Parsimony Normed Fit Index (PNFI), and the change in the Comparative Fit Index (ΔCFI).

Concurrent validity was assessed by analyzing correlations of the K-SCS-SF with measures of resilience, depression, and anxiety. Reliability was examined based on item-total correlations and internal consistency analysis, including Cronbach’s alpha and McDonald’s omega coefficients. McDonald’s omega provides a more robust estimate of internal consistency, particularly for scales with multidimensional.

RESULTS

Socio-demographics of the study participants

Socio-demographic characteristics of participants are presented in Table 1. A total of 650 young adults with a mean age of 23.99 years (SD=2.42) participated in this study, including 379 (58.31%) females. Participants having a college degree or higher education accounted for 46% (n=299). Most participants were unemployed (54.92%, n=357) and living with their family (66.15%, n=430). Regarding average household income, 206 (31.69%) participants reported an average household income of 1 million to 3 million won monthly. Most participants had no history of psychiatric treatment (70.15%, n=456) and had never taken psychiatric medication (77.08%, n=501).

CFA of the K-SCS-SF

The structure of the K-SCS-SF was evaluated through CFA using multiple competing models. Three CFAs were tested: a two-factor model (Model 1), a three-factor model (Model 2), and a six-factor model (Model 3). Model 1 supported by Costa et al. [34] and Hayes et al. [24] identified two distinct factors: over-identification, isolation, and self-judgement as negative components, and mindfulness, common humanity, and self-kindness as positive components. Model 2 grounded in Neff ’s [3] framework proposed three factors: mindfulness, common humanity, and self-kindness. Model 3 based on Garcia-Campayo et al. [14] and Raes et al. [21] suggested six factors: over-identification, mindfulness, isolation, common humanity, self-judgement, and self-kindness (Figure 1).

Model fit indices for the three CFAs are presented in Table 2. Model fit was evaluated using χ², the CFI, the NNFI, the RMSEA, and the SRMR. According to established guidelines for evaluating model fit in structural equation modeling, models with sample sizes exceeding 250 with 12 to 30 observed variables, an acceptable model fit is indicated by p<0.05, CFI and NNFI ≥0.92, an RMSEA ≤0.07, and SRMR ≤0.08.35 Additionally, a χ²/df ratio of 3 or less is generally considered indicative of good model fit [35].

Model 1, which comprised two factors, demonstrated an acceptable fit (χ²=284.648, χ²/df=5.37, CFI=0.947, NNFI=0.934, RMSEA=0.082, SRMR=0.076), all of which met or exceeded conventional cutoff values. Conversely, Model 2 with three factors failed to adequately fit the data (χ²=2140.221, χ²/df=41.96, CFI=0.519, NNFI=0.378, RMSEA=0.251, SRMR=0.251). Model 3 containing six factors also exhibited an acceptable fit (χ²=230.119, χ²/df=5.90, CFI=0.956, NNFI=0.926, RMSEA=0.087, SRMR=0.067). Therefore, both Model 1 and Model 3 showed acceptable overall fit, even though their RMSEA and χ²/df values slightly exceeded conventional thresholds (0.08 and 3.0, respectively). According to Browne and Cudeck [36], RMSEA values between 0.08 and 0.10 may still indicate mediocre but acceptable fit. Moreover, as MacCallum et al. [37] suggest, RMSEA should be interpreted flexibly by taking model complexity and degrees of freedom into account. Although χ²/df ratios below 3.0 are generally preferred, the large sample size and favorable values on other fit indices support the overall acceptability of the models [35].

For comparing Model 1 and Model 3, the principle of model parsimony was employed to select the simplest model with an optimal fit [38]. To this end, additional comparison indices such as ΔCFI, AIC, BIC, and the PNFI were also considered. Model 1 showed slightly lower AIC and BIC values (358.648 and 332.119, respectively) than Model 3 (AIC=360.160, BIC=334.204), suggesting that Model 1 may be preferable [39]. Furthermore, Model 1 yielded a PNFI of 0.751, higher than 0.560 for Model 3, indicating superior parsimony-adjusted fit [35]. Although Model 3 showed a ΔCFI of only 0.009 compared to Model 1, this minimal improvement suggests that the added complexity of Model 3 may not be justified, and that Model 1 exhibited a superior overall fit.

Descriptive statistics and correlations of the K-SCS-SF

To assess the normality of each item within negative (F1) and positive (F2) indicators, means, standard deviations, skewness, and kurtosis were examined. Item means ranged from 2.57 to 3.22, with standard deviations ranging from 1.11 to 1.26. Mean scores for F1 and F2 were 2.69 (SD=0.99) and 3.09 (SD=0.90), respectively. Skewness values ranged from -0.12 to 0.33, all within the acceptable absolute threshold of 2. Kurtosis values ranged from -0.95 to -0.68, well within the acceptable absolute threshold of 7. These results indicate that univariate distributions of all items met the criteria for normality [40]. Additionally, a correlation analysis between indicators F1 and F2 revealed a significant negative correlation of -0.78 (p=0.045).

Construct validity analysis of the K-SCS-SF

Convergent validity of items used in the reliability analysis for each factor is outlined in Table 3. Modification indices were used to detect potential cross-loadings. Explicit evidence of model misspecification was not observed. Standardized estimates (β) ranged from 0.60 to 0.87. Average variance extracted (AVE) values for factors F1 and F2 were 0.60 and 0.57, respectively, while composite reliability (CR) was 0.99 for both factors. Based on the criteria (β exceeding 0.5, AVE surpassing 0.5, and CR greater than 0.7) proposed by Hair et al. [35], the convergent validity of items used in this study was confirmed.

To assess discriminant validity, AVE values, correlation coefficient, and squared correlation coefficient (R2) were evaluated. AVE values for F1 and F2 were 0.60 and 0.57, respectively, both exceeding R2 value of 0.02. As suggested by Fornell and Larcker [41], AVE values should surpass the R2 between factors to substantiate discriminant validity.

The results of the concurrent validity analysis, displaying the correlations between the two factors of the K-SCS-SF and resilience (BRS), depression (PHQ-9), and anxiety (GAD), are presented in Table 4. The negative indicator (F1) showed a negative correlation with resilience (r=-0.52, p<0.001) but positive correlations with both depression and anxiety (r=0.56, p<0.001). In contrast, the positive indicator (F2) demonstrated a positive correlation with resilience (r=0.57, p<0.001) and negative correlations with depression and anxiety (r=-0.23 and r=-0.20, p<0.001 respectively).

Reliability analysis of the K-SCS-SF

Reliability of the measure was assessed using Cronbach’s alpha and McDonald’s omega. The Cronbach’s alpha was 0.83 for the total K-SCS-SF, 0.90 for F1, and 0.89 for F2. Similarly, McDonald’s omega was 0.76 for the total scale, 0.90 for F1, and 0.89 for F2. The internal consistency of all items demonstrated acceptable reliability [42].

DISCUSSION

This study aimed to assess factor structure and validate the K-SCS-SF using a sample of 650 young adults. The original SCS and SCS-SF were primarily validated in populations composed mostly of individuals in their twenties [3,21]; accordingly, the present study was conducted with Korean young adults in this age group. Based on comparisons among alternative models derived from prior literature, the two-factor model (Model 1) was identified as the most appropriate structure. This model, comprising negative and positive indicators, demonstrated strong construct validity and reliability.

The two-factor model offered greater parsimony and conceptual clarity in this study. Although the original developers of the SCS-SF proposed a six-factor structure reflecting that of the full version [21], subsequent validation studies in the USA [24], the UK [25] and Spain [26] have consistently reported better model fit for the two-factor structure when applied to the short form. Additionally, recent studies comparing multiple factors, including both two-factor and six-factor models, have found that the two-factor model has the best fit for assessing SCS-SF [25,26]. It is noteworthy that while the K-SCS adopted a six-factor model, this study identified a two-factor model as the optimal structure for the K-SCS-SF. This difference may be attributed to the fact that the original K-SCS validation did not include a two-factor model in its analysis, thereby overlooking the potential suitability of this structure. Moreover, the short-form nature of the K-SCS-SF, with fewer items, may limit the differentiation of nuanced subcomponents, making a more parsimonious two-factor model theoretically and practically more appropriate. The reliability analysis for two-factor model revealed that both the negative and positive indicators of self-compassion exhibited high reliability, with values close to 0.90. These values were higher than those reported in the UK [25] and Spain versions [26], where the two-factor model demonstrated reliability estimates of approximately 0.80, thereby supporting the robustness of the current model.

The findings of this study support a two-factor model of self-compassion, distinguishing between negative and positive indicators. This distinction is particularly valuable in understanding the relationship between self-compassion and psychopathology. For example, the negative composite score consistently differs among individuals with no depressive symptoms, those with depressive syndromes, and those diagnosed with major depressive disorder [43]. However, studies using total self-compassion scores without distinguishing negative and positive indicators have shown inconsistent results or an inflated relationship with psychopathological symptoms [12]. Given these findings, adopting a two-factor model may enhance the accuracy of evaluating self-compassion interventions by distinguishing their differential effects on positive and negative components. Research has shown that self-compassion training increases positive indicators while decreasing negative indicators [13], reinforcing the idea that self-compassion involves both cultivating compassionate responses and reducing uncompassionate tendencies. These findings emphasize the need for a two-factor model in assessing self-compassion and its psychological implications.

The concurrent validity was analyzed to determine the extent to which the two-factors (negative and positive indicators) correlate with resilience, depression, and anxiety. Negative indicators— comprising over-identification, isolation, and self-judgment— showed a negative correlation with resilience and positive correlations with depression and anxiety. In contrast, positive indicators-including mindfulness, common humanity, and self-kindness—exhibited a positive correlation with resilience and negative correlations with depression and anxiety. Previous studies have also supported the concurrent validity of these factors, showing that self-compassion is associated with depression, either through both factors [11] or solely through positive indicators negatively correlating with depression [44]. Additionally, another study found that the two factors were related to various psychopathological symptoms, including depression, generalized anxiety, social anxiety, eating concerns, substance use, hostility, academic distress, and family concerns [24]. While these findings align with previous research, this study further demonstrates that positive indicators are significantly associated with resilience, highlighting an additional protective role of self-compassion. By incorporating both negative and positive aspects of mental health, the two-factor model provided a more comprehensive understanding of the relationship between self-compassion and psychological well-being, supporting its relevance as an effective assessment framework.

This study validated the K-SCS-SF to provide a more efficient assessment of self-compassion by ensuring that its reliability and validity meet basic standards while enhancing relevance to Korean culture. Korean young adults often experience psychological distress, including stress from intense academic and employment-related competition, anxiety about the future, and increased symptoms of depression [45]. In light of these challenges, self-compassion may function as a positive psychological resource for coping with such pressures. The validated K-SCS-SF offers a practical means of evaluating self-compassion within the cultural context.

Nevertheless, certain limitations should be acknowledged. Convenience sampling from limited geographic regions (e.g., Seoul, Gyeonggi-do, and Jeju) may limit the generalizability of the findings. To enhance external validity and broader applicability, future studies should employ more diverse and representative samples across regions, age groups, clinical populations, and cultural contexts. Furthermore, as the scale’s validity was based on self-reported responses, there might be potential biases related to social desirability or memory recall. Future research should consider adopting mixed-method approaches or incorporating observer ratings to minimize these limitations. In addition, although the scale demonstrated high internal consistency, test-retest reliability was not assessed, which is essential for evaluating temporal stability of the measure. To further establish the robustness of the K-SCS-SF, future studies should incorporate test-retest reliability analysis.

This study validated the K-SCS-SF among young adults in their 20s residing in the community, confirming its strong reliability and validity. The findings suggest that this 12-item scale provides an economical and efficient means of assessing self-compassion while maintaining psychometric robustness. By reducing response burden, the K-SCS-SF allows for the inclusion of multiple variables in research, making it a valuable tool for evaluating various aspects of mental health. Moreover, confirming the two-factor structure enhances the understanding of both positive and negative aspects of self-compassion, supporting research on its dual nature. This distinction also enables more precise evaluations of self-compassion interventions, contributing to evidence-based practices aimed at improving well-being. In conclusion, the validated K-SCS-SF serves as a reliable, efficient, and practical assessment tool, advancing self-compassion research and its applications in mental health assessment, intervention studies, and psychological well-being research in Korea.

Notes

Availability of Data and Material

This study used data from Oh and Na (2024), which is available under ethical approval and cannot be publicly shared due to the sensitive nature of the research and the lack of participant consent for data sharing.

Conflicts of Interest

The authors have no potential conflicts of interest to disclose.

Author Contributions

Conceptualization: all authors. Data curation: Danbi Oh. Funding acquisition: Hyunjoo Na. Methodology: all authors. Project administration: Hyunjoo Na. Validation: all authors. Visualization: all authors. Writing—original draft: all authors. Writing—review & editing: all authors.

Funding Statement

This research was supported by the National Research Foundation of Korea (NRF-2022R1A2C1011568).

Acknowledgments

The authors thank Dr. Kim for granting us permission to use and validate the translated version of the Self-Compassion Scale into Korean, which served as the foundation for validating the Self-Compassion Scale-Short Form in this study.

Figure 1.

Model comparison for confirmatory factor analysis.

Table 1.

Socio-demographics of study participants (N=650)

Variables	Value
Age (years)	23.99±2.42
Sex
Female	379 (58.31)
Male	271 (41.69)
Education
≤High school	98 (15.08)
In college	253 (38.92)
≥College	299 (46.00)
Occupational status
Employed	293 (45.08)
Unemployed	357 (54.92)
Living arrangement
Alone	197 (30.31)
With family	430 (66.15)
With others (friends etc.)	23 (3.54)
Monthly household income (10 thousand won)
<100	51 (7.85)
100-299	206 (31.69)
300-499	202 (31.08)
≥500	191 (29.38)
Psychiatric treatment experience
Yes	194 (29.85)
No	456 (70.15)
Psychiatric medication experience
Yes	149 (22.92)
No	501 (77.08)

Data are presented as mean±standard deviation or number (%).

Table 2.

Model fit of confirmatory factor analysis (N=650)

Model	χ²	df	CFI	ΔCFI	NNFI	RMSEA (90% CI)	SRMR	AIC	BIC	PNFI
Model 1	284.648 (p<0.001)	53	0.947	-	0.934	0.082 (0.073, 0.092)	0.076	358.648	332.119	0.751
Model 2	2,140.221 (p<0.001)	51	0.519	0.428	0.378	0.251 (0.242, 0.260)	0.251	2,242.221	2,470.497	0.398
Model 3	230.119 (p<0.001)	39	0.956	0.009	0.926	0.087 (0.076, 0.098)	0.067	360.160	334.204	0.560

df, degree of freedom; CFI, Comparative Fit Index; ΔCFI, change in Comparative Fit Index between model and comparison model in the row above; NNFI, Non-Normed Fit Index; RMSEA, Root Mean Square Error of Approximation; SRMR, Standardized Root Mean Square Residual; AIC, Akaike Information Criterion; BIC, Bayesian Information Criterion; PNFI, Parsimony Normed Fit Index.

Table 3.

Construct validity evaluation (N=650)

Factor	Items	Β	β	SE	C.R.	AVE	CR
Negative indicators (F1)	11	1.00	0.87			0.60	0.99
	12	0.94	0.84	0.04	27.06
	9	0.97	0.83	0.04	26.44
	8	0.87	0.76	0.04	22.95
	1	0.78	0.70	0.04	20.35
	4	0.70	0.63	0.04	17.72
Positive indicators (F2)	5	1.00	0.82			0.57	0.99
	6	0.98	0.79	0.04	22.43
	3	0.96	0.79	0.04	22.31
	2	0.91	0.75	0.04	20.97
	7	0.94	0.77	0.04	21.75
	10	0.77	0.60	0.05	15.68

SE, standard error; C.R., critical ratio; AVE, average variance extracted; CR, composite reliability.

Table 4.

Correlations of K-SCS-SF with BRS, MSPSS, PHQ-9, GAD-7 (N=650)

Measures	K-SCS-SF			BRS	PHQ-9	GAD
Measures	F1	F2	Total	BRS	PHQ-9	GAD
K-SCS-SF
F1	1
F2	-0.78^*	1
BRS	-0.52^**	0.57^**	0.74^**	1
PHQ-9	0.56^**	-0.23^**	-0.55^**	0.53^**	1
GAD	0.56^**	-0.20^**	-0.53^**	-0.47^**	0.87^**	1

Negative factor scores of the SCS-SF were not reversed.

^* p<0.05;

^** p<0.01.

K-SCS-SF, Korean version of Self-Compassion Scale-Short Form; BRS, Brief Resilience Scale; PHQ-9, Patient Health Questionnaire-9; GAD, Generalized Anxiety Disorder Scale.

REFERENCES

1. Goetz JL, Keltner D, Simon-Thomas E. Compassion: an evolutionary analysis and empirical review. Psychol Bull 2010;136:351-374.

2. Seppälä EM, Simon-Thomas E, Brown SL, Worline MC, Cameron CD, Doty JR (1st ed). The Oxford handbook of compassion science. New York: Oxford University Press, 2017, p.478-493.

3. Neff KD. The development and validation of a scale to measure self-compassion. Self and Identity 2003;2:223-250.

4. Ewert C, Vater A, Schröder-Abé M. Self-compassion and coping: a meta-analysis. Mindfulness 2021;12:1063-1077.

5. Zessin U, Dickhäuser O, Garbade S. The relationship between self-compassion and well-being: a meta-analysis. Appl Psychol Health Well Being 2015;7:340-364.

6. Barnard LK, Curry JF. Self-compassion: conceptualizations, correlates, & interventions. Rev Gen Psychol 2011;15:289-303.

7. MacBeth A, Gumley A. Exploring compassion: a meta-analysis of the association between self-compassion and psychopathology. Clin Psychol Rev 2012;32:545-552.

8. Rod NH, Davies M, de Vries TR, Kreshpaj B, Drews H, Nguyen TL, Elsenburg LK. Young adulthood: a transitional period with lifelong implications for health and wellbeing. BMC Glob Public Health 2025;3:25

9. Egan SJ, Rees CS, Delalande J, Greene D, Fitzallen G, Brown S, et al. A review of self-compassion as an active ingredient in the prevention and treatment of anxiety and depression in young people. Adm Policy Ment Health 2022;49:385-403.

10. Shin NY, Lim YJ. Contribution of self-compassion to positive mental health among Korean university students. Int J Psychol 2019;54:800-806.

11. Brenner RE, Heath PJ, Vogel DL, Credé M. Two is more valid than one: examining the factor structure of the Self-Compassion Scale (SCS). J Couns Psychol 2017;64:696-707.

12. Muris P, Petrocchi N. Protection or vulnerability? A meta-analysis of the relations between the positive and negative components of self-compassion and psychopathology. Clin Psychol Psychother 2017;24:373-383.

13. Neff KD. The self-compassion scale is a valid and theoretically coherent measure of self-compassion. Mindfulness 2016;7:264-274.

14. Garcia-Campayo J, Navarro-Gil M, Andrés E, Montero-Marin J, López-Artal L, Demarzo MM. Validation of the Spanish versions of the long (26 items) and short (12 items) forms of the Self-Compassion Scale (SCS). Health Qual Life Outcomes 2014;12:4

15. Millard LA, Wan MW, Smith DM, Wittkowski A. The effectiveness of compassion focused therapy with clinical populations: a systematic review and meta-analysis. J Affect Disord 2023;326:168-192.

16. Wakelin KE, Perman G, Simonds LM. Effectiveness of self-compassion-related interventions for reducing self-criticism: a systematic review and meta-analysis. Clin Psychol Psychother 2022;29:1-25.

17. Chiesa A, Serretti A. Mindfulness-based stress reduction for stress management in healthy people: a review and meta-analysis. J Altern Complement Med 2009;15:593-600.

18. Krieger T, Altenstein D, Baettig I, Doerig N, Holtforth MG. Self-compassion in depression: associations with depressive symptoms, rumination, and avoidance in depressed outpatients. Behav Ther 2013;44:501-513.

19. Kim K, Yi G, Cho Y, Chai S, Lee W. [The validation study of the Korean version of the Self-Compassion Scale]. Korean J Health Psychol 2008;13:1023-1044. Korean.

20. Galesic M, Bosnjak M. Effects of questionnaire length on participation and indicators of response quality in a web survey. Public Opin Q 2009;73:349-360.

21. Raes F, Pommier E, Neff KD, Van Gucht D. Construction and factorial validation of a short form of the Self-Compassion Scale. Clin Psychol Psychother 2011;18:250-255.

22. Babenko O, Guo Q. Measuring self-compassion in medical students: factorial validation of the Self-Compassion Scale-Short Form (SCSSF). Acad Psychiatry 2019;43:590-594.

23. Bratt A, Fagerström C. Self-compassion in old age: confirmatory factor analysis of the 6-factor model and the internal consistency of the self-compassion scale-short form. Aging Ment Health 2020;24:642-648.

24. Hayes JA, Lockard AJ, Janis RA, Locke BD. Construct validity of the Self-Compassion Scale-Short Form among psychotherapy clients. Couns Psychol Q 2016;29:405-422.

25. Kotera Y, Sheffield D. Revisiting the self-compassion scale-short form: stronger associations with self-inadequacy and resilience. SN Comprehensive Clinical Medicine 2020;2:761-769.

26. Lluch-Sanz C, Galiana L, Vidal-Blanco G, Sansó N. Psychometric properties of the Self-Compassion Scale-Short Form: study of its role as a protector of Spanish nurses professional quality of life and well-being during the COVID-19 pandemic. Nurs Rep 2022;12:65-76.

27. Oh D, Na H. Identifying protective and risk factors for non-suicidal self-injury among young adults in Korea: insights from problem behaviour theory. Int J Ment Health Nurs 2024;33:2215-2226.

28. Choi N, Leach SM, Hart JM, Woo H. Further validation of the Brief Resilience Scale from a Korean college sample. Journal of Asia Pacific Counseling 2019;9:39-56.

29. Smith BW, Dalen J, Wiggins K, Tooley E, Christopher P, Bernard J. The brief resilience scale: assessing the ability to bounce back. Int J Behav Med 2008;15:194-200.

30. Smith BW, Epstein EM, Ortiz JA, Christopher PJ, Tooley EM. The foundations of resilience: What are the critical resources for bouncing back from stress? In: Prince-Embury S, Saklofske DH, editors. Resilience in children, adolescents, and adults: Translating research into practice. New York: Springer Science+Business Media, 2013, p.167-187.

31. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med 2001;16:606-613.

32. Spitzer RL, Williams JBW, Kroenke. Patient Health Questionnaire (PHQ) screeners [Internet] Pfizer Inc. Available at: https://www.phqscreeners.com. Accessed February 3 2025.

33. Spitzer RL, Kroenke K, Williams JB, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med 2006;166:1092-1097.

34. Costa J, Marôco J, Pinto-Gouveia J, Ferreira C, Castilho P. Validation of the psychometric properties of the self-compassion scale. Testing the factorial validity and factorial invariance of the measure among borderline personality disorder, anxiety disorder, eating disorder and general populations. Clin Psychol Psychother 2016;23:460-468.

35. Hair JF, Babin BJ, Anderson RE, Black WC. Multivariate data analysis. 8th ed. Hampshire: Cengage; 2018. p.639-679.

36. Browne MW, Cudeck R. Alternative ways of assessing model fit. Sociological methods & research 1992;21:230-258.

37. MacCallum RC, Browne MW, Sugawara HM. Power analysis and determination of sample size for covariance structure modeling. Psychol Methods 1996;1:130-149.

38. Hong S-H. The criteria for selecting appropriate fit indices in structural equation modeling and their rationales. Korean J Clin Psychol 2000;19:161-177.

39. West SG, Taylor AB, Wu W. Model fit and model selection in structural equation modeling. In: Hoyle RH, editor. Handbook of structural equation modeling. 2nd ed. New York: Guilford Press; 2023. p.184-205.

40. Curran PJ, West SG, Finch JF. The robustness of test statistics to non-normality and specification error in confirmatory factor analysis. Psychological Methods 1996;1:16-29.

41. Fornell C, Larcker DF. Evaluating structural equation models with unobservable variables and measurement error. J Mark Res 1981;18:39-50.

42. Kalkbrenner MT. Alpha, omega, and H internal consistency reliability estimates: reviewing these options and when to use them. Couns Outcome Res Eval 2023;14:77-88.

43. Körner A, Coroiu A, Copeland L, Gomez-Garibello C, Albani C, Zenger M, et al. The role of self-compassion in buffering symptoms of depression in the general population. PLoS One 2015;10:e0136598.

44. Kumlander S, Lahtinen O, Turunen T, Salmivalli C. Two is more valid than one, but is six even better? The factor structure of the Self-Compassion Scale (SCS). PLoS One 2018;13:e0207706.

45. Cheong S, Ryu J, Kang Y, Kim S, Ham S, Kim D, et al. 2022 survey on the living conditions of youth in Korea. Sejong: Office for Government Policy Coordination, Korea Institute for Health and Social Affairs; 2022 [Internet]. Available at: https://repository.kihasa.re.kr/handle/201002/42426. Accessed June 6, 2025.