Effects of Self-Compassion and Self-Disclosure on Mental Health through Spirituality: A Study on Adolescents in Orphanages

Some adolescents who live in orphanages mayfeel a lack of affection from their parents during their development; So, some orphanage adolescents experience a lack of guidance in dealing with problems in life, including mental health problems. To prevent this, it is necessary to have an attitude of self-compassion and self-disclosure of adolescents to their foster parents. In addition, it is necessary to strengthen the spirituality of orphanage adolescents to prevent adolescents from mental health disorders. This study aims to find a match and habit of the model of the influence of self-compassion and self-disclosure through spirituality on the mental health of orphanage adolescenets. This research uses a survey method with a quantitative descriptive approach to correlational models. The study subjects were orphanage as many as 165 samples were taken randomly at 3 orphans in Yogyakarta. Research data were obtained from observations, interviews, and research questionnaires. Analyze the research data using SmartPLS version 3 software with a two-stage embedded second-order approach. The results of this study show that a match of the research model was found between structural results and empirical data. This study implies that strengthening mental health for adolescents living in foster care can be done by consistently instilling the value of spirituality and improving self-compassion attitudes.


INTRODUCTION
The issue of mental health has become a hot topic to discuss lately, one of them is mental health problems experienced in adolescence (Garcia-Carrion et al., 2019).According to Schreuders et al. (2019), adolescence is the age of entering a person's critical cycle to find one's identity and reach maturity.So, adolescents often experience conflicts between fellow teenagers and the surrounding environment.Therefore, adolescents are very likely to be exposed to mental health disorders when faced with various problems, especially when adolescents still have unstable emotions (Vharensie, 2021).There is a need for parental assistance during the adolescent development phase because it will help adolescents get through the consequences of their lives and avoid mental health disorders.
However, some teenagers are do not get parental assistance during the development phase.This is due to parents are left to work outside the city or parents of teenagers who have passed away (Babedi & Pillay, 2019).Coupled with the arrival of the Covid-19 pandemic, it has increased the number of orphans in Indonesia so that adolescent care is entrusted to orphanages (Guzder, 2021).The Ministry of Social Affairs of the Republic of Indonesia Neff (2003) defines self-compassion as a person's ability to act reasonably with oneself and not judge oneself for the problems, shortcomings, or suffering experienced.Murfield et al. (2020) believe self-compassion is generally a recognition of oneself and others of the suffering that is being experienced, coupled with a commitment to solving and preventing the problem.Therefore, self-compassion can be expressed as compassion or compassion for oneself in every misfortune/problem experienced with a positive attitude and not excessively judging the problem.

Self-Compassion
Self-compassion in orphanage adolescents means that when faced with various problems in life, adolescents can accept them and not blame themselves when they come.Some of the roles of self-compassion in overcoming mental health disorders of adolescents living in orphanages are: First, reducing feelings of isolation.Self-compassion can help them to feel more connected to themselves and develop compassion towards themselves, which can be a source of support when they feel lonely (Kawitri et al., 2020).Second, reduce feelings of guilt and shame.Self-compassion can help them to accept themselves better and understand that everyone makes mistakes and experiences difficulties (Kawitri et al., 2019).Third, manage trauma and negative experiences.Self-compassion can help them cope with this trauma healthier by acknowledging their feelings without judging themselves (Beaton et al., 2022).

Self-Disclosure
Melumad & Meyer (2020) defines self-disclosure as voluntarily telling stories about feelings, thoughts, and private information to others they trust.Another opinion from DeVito (2019), self-disclosure is to communicate one's secret message regarding the problem at hand to others.Some of the definitions above can be concluded that self-disclosure is telling information about secrets or problems experienced by people closest to them, such as friends, parents, or loved ones.
Self-disclosure is important in maintaining the mental health of adolescents living in orphanages.Here are some of the roles of self-disclosure in maintaining adolescents' mental health in orphanages: First, reducing isolation and loneliness.Through self-disclosure, they can find a friend or caregiver they can trust to talk about their feelings.It helps reduce feelings of loneliness and increases the sense of connectedness (Aziz et al., 2023).Second, overcoming trauma and stored emotions.Self-disclosure can help them confront and cope with these feelings by discussing their experiences.This is the first step to recovery (Fitranti, 2021).Third, get support and guidance.Self-disclosure to orphanage caregivers can help teens get emotional support and guidance in coping with their mental health issues.This is important for developing effective coping strategies (Khan et al., 2023).

Spirituality
According to Makkar & Singh (2021), spirituality is an interpersonal and metaphysical relationship with a higher or transcendent power that can provide motivation, purpose, and a sense of connectedness with others.Hodapp & Zwingmann (2019) said that the concept of transcendent in spirituality is described by relationships vertically and horizontally.This follows the statement by Abdollahyar et al. (2019) that the vertical and horizontal relationship both have interrelated and inseparable dimensions.Based on several expert opinions regarding the description of spirituality above, spirituality is an awareness of the unity of the universe and the interrelationship of individuals within the universe.
Life Control [LC]; 6) Commitment [CM].Each mental health dimension in this study has 3 objective indicators, with 18 mental health assessment indicators.

Self-Compassion
Furthermore, the second measuring instrument for the self-compassion variable uses a measurement scale from the study by Neff et al. (2021), the Self-Compassion Scale for Youth, with a Cronbach's Alpha coefficient value of 0.7 (reliable) in the reliability test.Research from Jiao & Segrin (2022) found that there are six indicators of self-compassion, namely: 1) Self-Kindness [SK]; 2) Common Humanity [CH]; 3) Mindfulness [MF]; 4) Self-Judgment [SJ]; 5) Isolation [IS]; and 6) Over Identification [OI].Each dimension of self-compassion in this study has 3 indicators, with 18 self-compassion assessment indicators.

Self-Disclosure
The Third measuring instrument that researchers use for self-disclosure variables is the measurement scale from Gamayanti et al. (2018)'s research, namely the Resived Self-Disclosure Scale by Whelees adapted from Albes (2013); the instrument has been tested for the reliability of its research instrument items with a Cronbach's Alpha coefficient value of 0.89 (reliable).According to the results of research from Wheeles, it is stated that there are five indicators in self-disclosure, namely: 1) Amount [A]; 2) Valence [V]; 3) Honesty [H]; 4) Intent [I]; 5) Depth [D].Each dimension in the self-disclosure of this study has 3 or 4 indicators, so there are 18 self-disclosure assessment indicators.

Spirituality
Finally, instruments on the spirituality variable use measuring instruments from research by Westbrook et al. (2018), the Trait Sources of Spirituality Scale (TSSS), which has a Cronbach's Alpha coefficient value of 0. 96 (reliable) on its reliability test.According to Westbrook et al. (2018) there are 4 indicators in adolescent spirituality, including: 1) Theistic Spirituality [TS]; 2) Nature Spirituality [NS]; 3) Human Spirituality [HS]; 4) Transcendence Spirituality [TRS].Each dimension of spirituality in this study has 6 indicators of assessment, so there are 24 indicators of spirituality assessment.

Construct Validity and Reliability
Evaluation of the measurement model at the construct stage of the validity and reliability test is called the outer model test.Hair et al. mentions that the validity construct test explains how suitable the measuring instrument can be used to measure the construct to be measured.The validity construct test is divided into convergent validity and discriminant validity.Convergent validity can be seen from the loading factor value used in determining the validity of a construct, and a construct is declared valid if it has a loading factor value of > .7 (Sarstedt 2017).In addition, convergent validity can also be seen from the Average Variance Extracted (AVE) value which is > .5 (Hair et al., 2019).
Furthermore, the discriminant validity is seen from the value of cross-loading value on the dimension statement items to the dimension itself must be greater than the value of the correlation in the statement item to other dimensions.If these conditions are met, the statement item only passes the discriminant validity test.The data processing results in the discriminant validity test are first described in the table of Fornell and Larcker Criterion value calculation results.Whereas the construct reliability test was conducted to show the internal consistency of the measuring instrument.Construct reliability testing includes composite reliability with expected criteria for composite reliability value≥ .7 (Ghozali & Latan, 2012).
Based on the results of testing on the inner model, the results can be used as data to see whether the hypotheses formulated in this study can be declared accepted or rejected.The results of the inner model test obtained the T-statistical value, and the hypothesis can be accepted if the T-statistical value is greater than the T-table, which is 1,96 (α 5%).In addition, the hypothesis test can be seen from the value of the P-value, provided that the value must be < .05,meaning that the hypothesis is accepted.

Data Analysis
The research data was analyzed with the SmartPLS version 3.0 application with a secondorder embedded two stages approach.Hair et al. (2019) mention that the second-order embedded two stages approach is carried out because each variable has a dimension, and each dimension has an indicator.Therefore, data analysis is carried out in two stages, namely dimensional-level and variable-level data analysis with inner model tests (validity and reliability tests), outer models (R-Square, F-Square, Q-Square, VIF, and model fit/goodness tests), and hypothesis tests.

Outer Model Test
The SmartPLS version 3.0 software was employed to test the outer and inner models, and the result is shown in Figure 1.The outer model analysis includes convergent validity, discriminant validity, and composite reliability.

Stage 1: Dimensional-level model evaluation (Convergent Validity Test)
The results of data processing in the convergent validity test at the dimension level are explained as follows in table 1.Based on Table 1, from all indicators of each dimension, there are 8 invalid indicators due to the loading factor value < .7,so they will be removed from data processing to not change the AVE value.Furthermore, the AVE values at the variable level are described in the table below.Based on the values presented in Table 2 above, the AVE value in each dimension > .5 is obtained, so all dimensions have met the convergent validity test requirements.

Reliability
The reliability test of statement items on the research dimension is carried out with a composite reliability test.The results of the composite reliability value are described in the table 4. Based on data from Table 4 above, each variable has a composite reliability value of ≥ .7.These results show that each variable has met the requirements of the composite reliability test, so it can be concluded that these four research variables can be said to be reliable.

Stage 2: Evaluation of the Variable-level Model (Convergent Validity Test)
The results of data processing in the convergent validity test at the variable level are explained as follows, in table 5. Based on Table 5 above, from all the dimensions of each variable, there are 5 invalid dimensions due to the loading factor value < .7.So, it will be removed from data processing not to change the AVE value.Furthermore, the AVE value at the variable level is described in the table below, in table 6.Based on the values in Table 6 above, the AVE value in each dimension > .5 is obtained, so all dimensions have met the convergent validity test requirements.

Discriminant Validity Test
The results of the data processing in the discriminant validity test are first described in the table of Fornell and Larcker Criterion value calculation results below, in table 7. The data results in Table 7 above show that the correlation value of each dimension to the dimension itself is more significant when compared to the value of the dimension correlation to other dimensions.So, the variables in this study have met the discriminant validity test based on the values of the Fornell and Larcker Criterion.In addition, the discriminant validity test can be seen from the cross-loading value to see the correlation value between variables.

Reliability
The results of the composite reliability value are described in the table below, table 8. Based on data from Table 8 above, each variable has a composite reliability value of ≥ .7.These results show that each variable has met the requirements of the composite reliability test, so it can be concluded that these four research variables can be said to be reliable.

Inner Model Test
After the test on the outer model has been fulfilled, the next step is testing the inner model.The structural model analysis test (inner model) in this study consists of 5 tests, including the R Square, F Square, Predictive Relevance (Q 2 ), Collinearity Statistics (VIF), and the model fit and goodness test.The results of the inner analysis of the research model it is described in Figure 2.

R-Square
The results of the research data processing obtained the R-Square value as follows, table 9. Based on the results of the calculation of the R-Square value in Table 9 above, it was found that the spirituality variable has an R-Square value of .5378.This shows that the selfcompassion and self-disclosure variables affect the spirituality variable by 53,78% (moderation), and other variables outside this research model influence the other 46,22%.Meanwhile, the R-Square value for the mental health variable was .5697.This shows that selfcompassion, self-disclosure, and spirituality influenced the mental health variables by 56,97% (moderation), and other variables outside this research model influenced another 43,3%.

Q-Square
The Stone Geisser Value (Q 2 ) is used to describe predictive relevance, whether the structural relevance of the model is good or not.The results of data processing for Q 2 values are explained in the following table 10.
Based on the results of the data presented in Table 10 above, it is known that the value of Q 2 in the mental health variable is .3184> 0, so the mental health variable has an excellent predictive relevance value.While the value of Q 2 in the spirituality variable is .3559> 0, or in other words, the spirituality variable also has an excellent predictive value of relevance.

Goodness of Fit
The match and goodness test of the model can be seen from the data processing results on the Standardized Root Mean Square Residual (SMSR) value and the Normal Fit Index (NFI) value.As for the results of processing data on the match and goodness test, it is explained in the fit summary table below, in table 11.
Based on the results of the data presented in Table 11 above, it was found that: SMSR values of .0714< .10(meet the criteria), so the model can be said to be fit.The model test is seen from the NFI test value; in this study, it shows an NFI value of .7855; the research model can be said to be fit and has a match or goodness of the model with research data of 78,55%.The results of the hypothesis test in this study obtained hypothesis test data of direct and indirect influence as described in the table 12 and 13.

H1: The Influence of Self-Compassion on Spirituality
Based on the results as shown in Table 12 above, the first hypothesis (H1) has a Tstatistical value of 3,9731 > 1,96 (α 5%).In addition, the P-value is .0001< .05.So that the first hypothesis (H1) is declared accepted, it means that self-compassion has proven to have a positive and significant effect on the spirituality of orphanage adolescents.The higher the adolescent's self-compassion, the higher the level of spirituality.The basis of the hypothesis test of this study proves this.The findings of the first hypothesis (H1) corroborate the previous theory: self-compassion has a positive and significant effect on spirituality (Hidayati et al., 2019).In addition, the theory of (Dewi et al., 2020) says that self-compassion attitudes influence the spirituality of former drug abusers.

H2: The Effect of Self-Disclosure on Spirituality
Based on the results as shown in Table 12 above, the second hypothesis (H2) has a Tstatistical value of 5,8738 > 1,96 (α 5%).In addition, the P-value is .000< .05.So that the second hypothesis (H2) is declared accepted, it means that self-disclosure has proven to positively and significantly affect the spirituality of orphanage adolescents.The higher the selfdisclosure of adolescents, the higher the level of spirituality.The basis of the hypothesis test of this study proves this.The findings of the second hypothesis (H2) above corroborate a preexisting theory, among others, according to the theory (Hapsari, 2021) that self-disclosure can affect spirituality with script therapy.Another theory from (Wiyono & Muhid, 2020) says that the self-disclosure of women with disabilities is to bring worship closer to God.

H3: The Effect of Spirituality on Mental Health
Based on the results as shown in Table 12 above, the third hypothesis (H3) has a Tstatistical value of 4,278 > 1,96 (α 5%).In addition, the P-value is .000< .05.So that the third hypothesis (H3) is declared accepted, it means that spirituality has proven to have a positive and significant effect on the mental health of orphanage adolescents.The higher the spirituality value of adolescents, the higher the level of mental health.The basis of the hypothesis test of this study proves this.The third hypothesis (H3) findings corroborate the previous theory that spirituality can affect students' mental health levels (Wahyuni & Bariyyah, 2019).In addition, according to (Kao et al., 2020), spirituality and mental health have a positive influence.Another opinion (Hodapp & Zwingmann, 2019) is that spirituality and mental health have a positive and significant relationship and influence.

H4: The Effect of Self-Compassion on Mental Health
Based on the results as shown in Table 12 above, the fourth hypothesis (H4) has a Tstatistical value of 4,5294 > 1,96 (α 5%).In addition, the P-value is .000< .05.So that the fourth hypothesis (H4) is declared accepted, it means that self-compassion has been shown to positively and significantly affect the mental health of orphanage adolescents.The higher the self-compassion of orphanage teenagers, the higher the level of mental health.The basis of the hypothesis test of this study proves this.The findings of the fourth hypothesis (H4) corroborate the previous theory, namely the role of self-compassion in improving health and quality of life in orphanage adolescents (Kawitri et al., 2020).In addition, self-compassion positively and significantly influences the resilience of orphanage adolescents (Kawitri et al., 2019).Another theory says low self-compassion attitudes can worsen mental health (Beaton et al., 2022).
Furthermore, look at the hypothesis test results based on Table 13.In that case, it shows that the influence of the self-compassion variable through spirituality on mental health, has a T-statistical value of 2,9036 > 1,96 (α 5%).In addition, the P-value is .0037< .05.So that the fourth hypothesis (H4) is declared accepted, it means that self-compassion through spirituality has positively and significantly affected the mental health of orphanage adolescents.The higher the self-compassion of orphanage adolescents through the moderation of spirituality, the higher the level of mental health.The basis of the hypothesis test of this study proves this.The research findings on the fourth hypothesis (H4) corroborate previous theories, including the role of selfcompassion and spirituality in reducing depression in pregnant women (Chairunnisa & Fourianalistyawati, 2019).In addition, the role of self-compassion through spirituality can predict a person's depressive symptoms (Fenzel & Richardson, 2022).

H5: The Effect of Self-Disclosure on Mental Health
Based on the results as shown in Table 12 above, the fifth hypothesis (H5) has a Tstatistical value of 1,8389 < 1,96 (α 5%).In addition, the P-value is .0661> .05.So that the fifth hypothesis (H5) is stated to be rejected, it means that self-disclosure has not been shown to affect the mental health of orphanage adolescents significantly.The basis of the hypothesis test of this study proves this.
The absence of a significant influence of the self-disclosure variable on mental health is directly caused because people who are too open (oversharing) about the problems faced by others can result in an increase in depression, where depression is one of the symptoms in a person experiencing mental health disorders (Laurentius et al., 2020).Over-sharing behavior is defined as conveying too much private information to others, which can be through direct communication or social media (Mawarniningsih et al., 2022). in addition, according to (Akhtar, 2020), a person with an over-sharing attitude has a negative psychological impact.Therefore, the attitude of orphanage adolescents who are too open (self-disclosure) can result in the emergence of over-sharing behaviors that trigger the onset of mental health disorders.The findings of the fifth hypothesis (H5) reject the theory from (Yani & Nilawati, 2021) that there is a positive and significant influence between self-disclosure variables and mental health.
However, the results of the hypothesis test data based on Table 13 show that the influence of self-disclosure variables through spirituality on mental health, has a T-statistical value of 3,4428 > 1,96 (α 5%).In addition, the P-value is .0006< .05.So that the fifth hypothesis (H5) is declared accepted, it means that self-disclosure through spirituality has proven to positively and significantly affect the mental health of orphanage adolescents.The higher the selfdisclosure of orphanage adolescents through the moderation of spirituality, the higher the level of mental health.The basis of the hypothesis test of this study proves this.The findings that there is a positive and significant influence of the self-disclosure variable through spirituality on the mental health of adolescents in orphanages, are new findings because no previous studies or theories mention it.

CONCLUSION
Based on the findings and discussion above, this study can be concluded that there is a positive and significant influence of self-compassion on the mental health of orphanage adolescents, both direct influence and influence through the medium of spirituality.There was no significant influence between the self-disclosure variable and the mental health of orphanage adolescents directly.However, there is a positive and significant influence of the self-disclosure variable and the mental health for orphanage adolescents through the spirituality mediator variable.Spirituality and adolescent mental health variables have a positive and significant influence.A matching research model was found between the structural analysis results and empirical data.Therefore, this study implies that strengthening mental health for adolescents living in orphanages is carried out by consistently instilling spirituality and increasing selfcompassion.

Figure 1 .
Figure 1.Outer Model Test Result

Figure 2 .
Figure 2. Inner Model Test Result

Table 1 .
The value of the Loading Factor of each Variable

Table 3 .
Variable Rate Fornell and Larcker Criterion Values

Table 4 .
Dimension-Level Composite Reliability Value

Table 5 .
Variable Level Loading Factor Value

Table 6 .
Variable Level Average Variance Expected (AVE) Value

Table 7 .
Variable Rate Fornell and Larcker Values

Table 8 .
Dimension-Level Composite Reliability Value

Table 12 .
Hypothesis Test Value -Direct

Table 13 .
Hypothesis Test Value -Moderation