ADAPTATION AND VALIDATION OF SELF-COMPASSION SCALE-SHORT FORM FOR ADOLESCENT DRUG ABUSERS AT SOCIAL REHABILITATION FOUNDATIONS IN INDONESIA

: Self-compassion is one of the adaptive self-attitudes that can be used by adolescents who experience personal dissonance or facing difficult situations and implemented as a normal experience of adolescents who drugs abuse. However, studies on standardized instruments and self-compassion profiles of drug abusers in Indonesia are still limited. Therefore, an intensive study of standardized instruments is needed to measure the compassion of adolescent drug abusers. This study uses an adaptation of a instrument procedures. As many of 38 as teenage drug users of six Social Rehabilitation Foundations in Indonesia. Data is measured using Self-Compassion Scale-Short Form. Operationally, the data analysis used Rash Model 3.75. The results showed that SCS-SF has satisfactory psychometric properties. Further studies are required to examine the psychometric properties of the SCS-SF of various participant socio-demographic factors


INTRODUCTION
The illicit circulation, abuse, and dependence of drugs on adolescents in Indonesia shows an increasing trend, already very concerning and endangering the lives of the community, nation and state. Indonesia is not only a transit point in illicit trade and circulation, but has become a place of marketing, use and has even become a place for illicit drug production. The National Narcotics Agency of the Republic of Indonesia (BNN RI, 2019) reported that there was an increase of 24-28% of adolescents using drugs. P4GN in 2018 also reported that the number of drug abusers among students in 13 provinces reached 2.29 million with an age range of 15-35 years.
The dramatic increase in drug abuse in adolescents in Indonesia occurs along with life in the 21st century which has brought psychological and sociological changes, in addition to having a positive impact on the progress of life also causing side effects in the form of paradigm changes and views on adolescent values and morals that are very concerning. Changes in a teenager's outlook accompanied by uncertainty about the future can make him confused or make him dissolve in the situation. If the teenager gets psychological difficulties, then the person concerned looks for shortcuts in solving the problem, even though it is actually pseudo. It is not surprising that there is a widespread "drug for every problem" attitude in adolescents. Teenagers assume that the problem will disappear when using the drug. In fact, the reality raises new problems in the form of abuse and dependence on drugs.
The increase in drug abusers from year to year involving adolescents can be seen as a threat to the development of the nation and a national danger. In addition, drug addiction can cause various physical and psychological symptoms and impacts such as: experiencing disturbances in perception, learning and memory, self-harm, depression, psychomotor function, thinking orientation, sensory processes, anxiety, depression, tendency to self-harm, and other psychological impacts.
In order for adolescents to avoid drug abuse, they need to develop and increase selfcompassion.
Self-compassion is a multidimensional construct that includes three conflicting components, namely: self-kindess vs. self-judgment, common humanity vs. isolation, and mindfulness vs. over-identified (Neff K. D., 2003a). Recent studies suggest that self-compassion is one of the healthiest attitudes (Neff K. D., 2003a) and self-adaptive that needs to be studied, considered, improved, and used by adolescents who experience personal disappropriateness or face difficult situations and are understood as crucial experiences of adolescents who are normal and who abuse drugs (Cunha, Xavier, &;Castilho, 2015). Self-compassion can influence adolescents in overcoming negative self-views and prevent them from behaving and behaving maladaptively Neff & Knox, 2017), such as abusing drugs.
Empirical studies show that self-compassion in adolescents consistently has a significant positive effect on mindfulness, physical and psychological well-being, psychological wellbeing and adjustment, self-concept, selfesteem, and compassion towards others, empathic responses, and other prosocial goals (Sutton, Schonert-Reichl, &;Wu, 2018). Research produces two dimensions of SC and is proven to have a positive effect on recovery from illness (Geller, Iyar, Kelly, &;Srikameswaran, 2019), predicting psychological well-being and emotional regulation (Phillips, 2019). Other studies have shown that self-compassion has a positive effect on attachment and negatively affects fatigue (Babenko &;Guo, 2019), positively correlates with self-awareness and negatively affects anxiety, stress, and depression (Garcia-Campanyo, et al., 2014). A comprehensive understanding of the selfcompassion of adolescent drug abusers is needed so that the guidance and counseling approach can be ongoing and useful. Early identification of drug abusers needs to be done inconspicuously in the community. The more successful the effort to describe the sociopsychological condition of adolescent drug abusers, the greater the likelihood of successful counseling in the recovery process or prevention of relapse.
On the other hand, studies on standardized instruments and self-compassion profiles of adolescent drug abusers in Indonesia are still limited. Therefore, intensive studies are needed on standardized instruments to measure self-compassion of adolescent drug abusers.
One of the self-compassion instruments that has good psychometric properties and is widely used in various studies is the Self-Compassion Scale (Neff K. D., 2003b;2016;Neff & Knox, 2017), and Self-Compassion Scale-Short Form (Babenko & Guo, 2019;Garcia-Campanyo, et al., 2014;Sutton, Schonert-Reichl, &;Wu, 2018). In this study, SCS-SF was selected and used to obtain a valid, reliable, fast, practical, and efficient instrument in measuring self-compassion of adolescent drug abusers.
Based on this rationale, this study focused on the adaptation and validation of Self-Compassion Scale-Short Form for adolescent drug abusers in Indonesia. Furthermore, the data from this assessment can be used as a pilot project for developing self-compassion for assisted citizens at the Social Rehabilitation Center for Drug Abuse. The long-term goal is to create Indonesian adolescents who are able to feel sorry for themselves at a high level and free from drug abuse

RESEARCH METHODS
This study used standard instrument adaptation and validation procedures from the International Test Commission (Bartram, Hambleton, Gregoire, &;Muniz, 2018). This method was chosen because this study aims to adapt and test the psychometric properties of the standard Self-Compassion Scale-Short Form (SCS-SF) instrument for adolescent drug abusers in Indonesia.
The study participants were 38 adolescent drug abusers from six Social Rehabilitation Foundations in Indonesia. This is based on expert opinion that a minimum number of participants of 30-50 people is feasible / sufficient for pilot studies to produce stable estimation results on a scale of ± 1 logit (Linacre, 2019; Sumintono &; Widhiarso, 2014).
Research instruments using Self-

SCS-SF Item Eligibility Overview
The first question about the eligibility picture of Self-Compassion Scale-Short Form (SCS-SF) items was analyzed from item measure criteria, item fit orders, and DIF using the Rasch Model. Item Measure is used to measure the difficulty of an item. The test results showed that S5 items with +.26 logit showed the most difficult items to approve, while S12 items with a value of -.26 logit were the items most easily approved by participants.
Item Fit Order is used to test items that fit and misfit. The results found six of the 12 items that were fit, namely the numbers S2, S5, S6, S7, S11, S12. This decision is based on the criterion that an item is said to be fit if the Infit and Outfit Mnsq values are within the acceptable value range (.5 MNSQ < 1.5), Outfit Zstd (-2 < Zstd > +2), Pt-MCorr (.4 < pt-MCorr .85) (Bond &;Fox, 2015;Linacre, 2019;Widhiarso, 2014;2015;Boone, Staver, & Yale, 2014;Taufiq, 2019; DIF is used to detect item bias in certain categories of participants. Detected items can be known based on probability values that are less than 5% (.05) (Bond & Fox, 2015;Dimitrov, 2012;Linacre, 2019;Widhiarso, 2014;2015;Taufiq, 2019;. The test results show that all (12) items are not biased towards a particular gender. Based on the three item eligibility criteria, six SCS-SF items were obtained for a fit version of adolescent drug abusers . Reliability. Table 2 of the statistical summary shows person measure = +.13 logit. A higher average value than a .0 logit indicates a tendency for participants to answer more highly on each item. Cronbach's Alpha score to measure the reliability of the test, namely the interaction between person and item as a whole, showed .87 which is high. Likewise, the value of person reliability is .83 and item reliability is .82 so that it can be concluded that the consistency of participants' answers and a set of tests is good Widhiarso, 2014;2015;Taufiq, 2019; Finalizing and Drafting Usage Manuals. Taking into account the results of feasibility testing, six items of adolescent drug abuse versions of SCS-SF were determined in this study. SCS-SF is expected to be an assessment tool for the level of self-love for adolescent drug abusers in Indonesia. This data can be used as a basis for developing guidance and counseling programs to rehabilitate adolescent drug abusers. SCS-SF is a self-report measurement tool in the form of a summated rating with a Likert scale of 5 levels from 1 = almost never to 5 = almost always.

SCS-SF Instrument Quality Overview
Self-love profiles were analyzed using averages and percentages. The classification of self-love, both as a whole and per dimension using criteria, namely: X ̅ < 3 = low, ≥ 3 ≤ 4 = medium, and > 4 = high A high score reflects a more positive/kind level of self-love, and vice versa

Feasibility Item SCS-SF
The feasibility analysis of these SCS-SF items includes: item measure, item fit order, and DIF (Bond &;Fox, 2015;Linacre, 2019;Widhiarso, 2014;2015;Boone, Staver, & Yale, 2014;Taufiq, 2019;. Item measure is used to measure the difficulty level of SCS-SF items. Meanwhile, fit order items are used to identify items that are fit or unfit ). The results found that six of the 12 items were fit, namely S2, S5, S6, S7, S11, S12 numbers. This decision is taken based on the criteria that an item is said to be fit if the Infit and Outfit Mnsq values are in the acceptable value range (.5 MNSQ < 1.5), Outfit Zstd (-2 < Zstd > +2), Pt-MCorr (.4 < pt-MCorr .85) (Bond & Fox, 2015;Linacre, 2019;Widhiarso, 2014;2015;Boone, Staver, & Yale, 2014;Taufiq, 2019;. On the other hand, DIF is used to identify SCS-SF items that are biased or not. The results showed that all items did not experience bias towards a particular gender. This decision is made based on the probability value of each item which is more than 5% (.05). As is known that items that can be detected from the probability value of items that are less than 5% (.05) (Bond &;Fox, 2015;Dimitrov, 2012;Linacre, 2019;Widhiarso, 2014;2015).

SCS-SF Instrument Quality
The quality of SCS-SF instruments is known by conducting tests such as: unidimensionality, rating scale, and reliability.
First, unidimensionality is analyzed by principal component analysis (PCA) of residuals, which measures the level of uniformity of the instrument in measuring what should be measured (Linacre, 2019;Sumintono & Widhiarso, 2014;2015;Herdi, Kartadinata, & Taufiq, 2019;. In the development and adaptation of instruments, unidimensionality is a prerequisite that must be met. The measurement results show a raw variance of 41.9% and an unexplainable variance by the instrument of 15.4%. This means that the unidimensionality requirement of at least 20% is met, while the variance that cannot be explained by the instrument ≤ 15% is not met (Bond & Fox, 2015;Linacre, 2019;Sumintono & Widhiarso, 2014;2015;Herdi, Kartadinata, & Taufiq, 2019;. The second prerequisite of the quality of the SCS-SF instrument tested is the rating scale. The rating scale is a category of statements that express attitudes, feelings, or perceptions about something (Dimitrov, 2012). This rating scale is important to analyze and use to verify the rating choices used in SCS-SF whether it can confuse participants or not (Bond & Fox, 2015;Linacre, 2019;Sumintono & Widhiarso, 2014;2015;Herdi, Kartadinata, & Taufiq, 2019;. The rating scale in an instrument is important to make and analyze because it has practical value and aims to place the subject / participant on an assessment continuum so that the relative position of the participant according to a measured attribute (group counseling competence) can be obtained (Azwar, 2012). The results of the analysis using the Rasch Model show that the average observation value and Andrich Threshold value move from NONE to negative and continue to lead to positive sequentially. This means that the choice of rating scale used is valid and appropriate for and does not confuse participants.
The third SCS-SF quality prerequisite tested is reliability. The results showed that the Alpha Cronbach test kit value (interaction between person and item), person reliability, and item reliability was in the very low / less good category. In other words, it can be concluded that the consistency of participants' answers and a set of tests is very low / not good Widhiarso, 2014;2015). This result is well below the consensus of psychometric experts that the acceptable test reliability index is at least .70. Reliability is one of the key concepts in measurement and plays an important role in determining the quality of instruments and the validity of data collected (Dimitrov, 2012). Reliability indicates the degree of accuracy, consistency, and replicability of a set of instruments, when: (1) different people take measurements; (2) different instruments are used to measure the same attributes; and (3) there are incidental variations in measurement conditions (Dimitrov, 2012;Herdi, Kartadinata, & Taufiq, 2019;.

CONCLUSION
This pioneering research has produced an adapted version of SCS-SF to measure the level of self-love of adolescent drug abusers in Indonesia. However, this adapted version of SCS-SF has satisfactory psychometric properties. This is evident from the criteria of item eligibility, unidimensionality, as well as item reliability, person reliability, and test reliability. Studies are needed to develop and test the CAT version of SCS-SF by conducting confirmatory factor analysis in participants for drug abusing adolescents in Indonesia that are broader and representative based on ethno-socio-demographic factors.