Meta-analysis of dimension of on the Meta-Analysis of Dimension of Autonomy on the Psychological Well-Being Measurement in Indonesia

. Since it was raised more than two decades ago, research on Psychological Well-Being (PWB) in Indonesia has invited many interested people to follow up. This situation is in line with the development of the overall orientation of Positive Psychology as a form of renewal from conventional psychological orientation. A large number of PWB studies carried out in Indonesia provides quite broad theoretical and practical implications. In general, the data obtained through these studies use group norms on respondents who are relatively small in size, so that their findings are not accurate in providing a description of PWB in the group of respondents studied. Among the six dimensions psychological well-being (self-acceptance, personal growth, positive relations with others, purpose in life, environmental mental mastery, and autonomy), dimensions autonomy often show findings with lower scores than other dimensions. The purpose of this study was to find a profile of the dimensions of autonomy. The research method used is a meta-analysis and confirmation analysis of approximately 100 results of existing research, with a respondent size of around 3,000 people. After getting the score standard autonomy from the meta-analysis study, it will be continued by distributing the PWB questionnaire to around 844 respondents emerging adulthood. This research moved from the basic assumption that Indonesian culture was different from a western culture where the initial research of this PWB was developed. These cultural differences will then form different behaviors in displaying dimensions autonomy in everyday life


INTRODUCTION
The focus of mental health research and practice puts a spotlight on overcoming pathological cases, such as depression and anxiety, which is often accompanied by prevention measures, has been ongoing for a long time (Huppert & So, 2013). Implicitly, it is assumed that well-being will be perceived if pathology is not present (Huppert & So, 2013). Over time, however, more and more evidence indicates that a high welfare level is excellent for the lives of individuals and society (Huppert & So, 2013). This perspective-shifting has inspired many people to conduct research with the new paradigm. Positive psychology, which has been the trigger of the paradigm shift about mental health, advocates well-being as a substitute for a perspective in psychology from which initially focused on disease Bhullar et al. (2014) to a comprehensively perspective-oriented. Likewise, the World Health Organization (WHO, 2005) defines mental health as "a state of wellbeing in which an individual realizes his or her own abilities, can cope with the normal stresses e-ISSN: 2541-450X p-ISSN: 0854-2880 Indigenous: Jurnal Ilmiah Psikologi 2021, 6(2), [25][26][27][28][29][30][31][32][33][34][35][36] of life, can work productively and is able to make a contribution to his or her community". If elaborated, there are three components in the definition, namely (1) well-being, (2) individuals with effective functioning, and (3) effective functioning in the community (WHO, 2005) Over the past two decades, there has been a strong urge to study various psychological traits and processes that are perceived as positive, that is, those beneficial to well-being (McNulty & Fincham, 2012). It is the reason behind psychological well-being (PWB,) which emerged more than 25 years ago by Ryff, is pointed at overcoming negligence in a positive human functioning formulation which entails implications for psychotherapy (Ryff, 2014). Other influential approaches, regarding wellbeing or positive mental health, require both hedonic and eudaimonic components, comprising a combination of feeling good and functioning well (Huppert & So, 2013).
As an Aristotelianism, Seligman at first suggested that real well-being or happiness consists of three essential elements; comfort, engagement, and meaning. In his following journey, Seligman appended two more elements, encompassing relationships and achievements (Forgeard et al., 2011;Seligman, 2011). According to Seligman, five important elements (positive emotions, engagement, relationship, meaning, accomplishment), or PERMA for short, are the best achievements that humans will pursue themselves (Seligman, 2011) and contribute to the definition of well-being. The concept of well-being expressed by Keyes et al. (2002) also combines elements of feeling and functioning. Specifically, Keyes' measures the combination of hedonic or emotional well-being (happiness or life satisfaction) with Ryff's six eudaimonic dimensions of psychological well-being (Huppert & So, 2013).
Most researchers believe that every individual requires both hedonia and eudaimonia to grow (Huta, 2015). Hedonia and eudaimonia are neither contradictory nor independent, but both are complementary psychological functions (Huta, 2015).
Even though the positive psychology approach, which is the umbrella for psychological well-being, is still relatively new, it has called forth the interest of many young researchers to conduct various researches for the purposes of theses and dissertations. This fact is evidenced by 170 studies conducted by students from four universities in Java (Jatinangor, Bandung, Depok, and Yogyakarta) covering the topic of positive psychology, in the last 10 years (Ginting et al., 2018). This fact also supports the great contribution of positive psychology to the development of psychology in Indonesia.
Psychological well-being (hereinafter referred to as PWB), is one of the concepts in positive psychology that has been widely studied. The steps for measuring PWB are following these dimensions: (1) the extent to which a person feels that his or her life has meaning, purpose, and direction (purpose in life); (2) whether a person views himself or herself as able to live according to his personal beliefs (autonomy); (3) the extent to which a person is able to take advantage of his or her talents and potential (personal growth); (4) how well one's ability to manage the situation in his or her life (environmental mastery); (5) how deeply connected he or she is in the form of bonds with meaningful people (positive relationships), and (6) knowledge and acceptance of oneself, including awareness of personal limitations (self-acceptance) (Ryff, 2014). Together these six dimensions hone the contrast with previous views on well-being that focused on indicators of feeling good, happy, positive, or life satisfaction (Ryff, 2014). As suggested by Ryan and Deci (2018) that a psychologically good person does not mean free from psychopathology, nor those who are solely happy and satisfied, but those who are able to mobilize and utilize psychological and psychological energy to carry out meaningful activities, especially activities that make the person being part of the activity and are therefore motivated to carry out such action.
In Ryff's PWB model, to determine high or low PWB, one can use absolute norms or group norms. The use of group norms is more commonly used in psychology research in Indonesia. In fact, the use of group norms is not suitable for research with small respondent numbers (Ginting et al., 2018). Group norms in a test are based on the distribution of scores obtained by several predetermined individual samples so that the use of group norms will describe the position of the subject in the population. Surprisingly, of the six dimensions of PWB that have been extensively studied, the autonomy dimension has always shown its distinguish characteristic findings. For instance, research by Wardani (2015) on the effect of academic hardiness on PWB indicated that the autonomy dimension was not influenced by academic hardiness. Likewise, Wardani's dissertation (2014) probing the strength of the influence of each personality trait (Neuroticism, Openness, Conscientiousness, Agreeableness, and Extraversion) on the PWB dimensions showed that the influence of the trait strength on the autonomy dimension obtained the lowest number, which was in the range of 0.661 -0.665, compared to the other five dimensions of 0.756 -0.878 range. The research by Sagone and Elvira De Caroli (2014) discovered that the dimensions of psychological well-being (except autonomy) were positively related to commitment, control, and challenge. If PWB is associated with coping strategies, it was found that autonomy, environmental mastery, personal growth, positive relations with others, and self-acceptance were negatively correlated with the avoidance strategy; on the other hand, autonomy, environmental mastery, personal growth, and self-acceptance were positively correlated with problem-solving coping (Sagone & Elvira De Caroli, 2014).
Research results on PWB in Indonesia are in line with the aforementioned explanation, for example, Saraswati and Teja (2018) who examined employee PWB in Java revealed that the average PWB of employees was high but the dimensions that the lowest was Autonomy. Research by Kurniasari et al. (2019) regarding the description of students' psychological well-being indicates that the aspects of autonomy and personal development had the lowest level of accomplishment. In addition, the PWB study on post-stroke patient caretakers conducted by Lestari (2018) obtained that the participants' autonomy dimension was low. Likewise, a study by Rahmadhani (2015) on PWB on Female Prisoners in Prison Class IIA Sukamiskin, Bandung, found that autonomy was the dimension with the lowest degree compared to other dimensions. Marsha and Nurpatria (2018) in their research on PWB among women who married at an early age due to poverty proved similar results that all participants showed limitations in terms of autonomy due to the dominant role of a husband in the family life.
Autonomy refers to a person's ability to self-determination and independence; able to withstand social pressures and think and act according to their own desires and ways; able to regulate behavior according to the internal assessment framework; and evaluate self on personal standard-based (Ryff, 2014). In contrast, individuals with low autonomy are very concerned with the expectations and judgments by others; relying on others to make important decisions; adapt to social pressures to think and act in certain ways.
The concept of autonomy in Ryff's theory, as aforementioned, may provide an understanding that is unfitting if it is adapted to Indonesian culture which highlights deliberation, togetherness, and harmony. Autonomy in a conceptual sense based on the framework of personal internal judgment is almost impossible to find. In line with this, if PWB studies often use group norms, meaning that most of the respondents' scores fall into the high category. For example, if the PWB has a 6 (six) scale, the minimum score is 1 (one) and the maximum score is 6 (six), so the median is 3.5 (three point five). Respondent of a score less than or below 3.5 (three point five) will be categorized as a person who has a low PWB, and those who obtain a score more than or above 3.5 (three point five) will be categorized as high PWB. However, in the autonomy score, it is possible to find a floor effect in the research results, which shows a low score so that respondents are categorized as less autonomous.
On the other hand, the use of absolute norms frequently cannot represent the picture of PWB in Indonesia. In her model, Ryff took foreigners (western culture) as respondents who are not necessarily in accordance with Indonesian culture. It can be seen from many studies on PWB that describe low PWB symptoms in the phenomena under the study but the research results show the opposite fact. Such circumstances may be emerged due to differences in the dimensions of people's lives in different countries. Therefore, this research specifically focuses on identifying the profile of autonomy dimension in Indonesia using the Ryff model, as well as creating group norms from the measuring instrument for autonomy dimension.

METHOD
This study will examine the features of the autonomy dimension in Indonesia. The research designs used are meta-analysis and confirmatory factor analysis. Meta-analysis is a common research method used to re-analyze existing research data. Prevailing research results (student theses) were combined and reprocessed in a more comprehensive manner. Furthermore, the lists of questions from the questionnaire used in these theses were retested using confirmatory factor analysis (CFA). Based on those processes, broadly speaking, this research is divided into two stages; the metaanalysis stage and the CFA stage. In the meta-analysis stage, the data obtained from the results of previous studies were used. In this case, approximately 100 previous studies on PWB with a sample size of around 3,000 respondents. These data were obtained from the students' theses of various universities. Following that, the CFA stage was carried out. At this stage, the research instrument was a questionnaire containing autonomy dimension which was compiled based on Ryff's theory of PWB. The following chart describes the procedure of this research: The PWB questionnaires were distributed to roughly 844 respondents who were in the emerging adulthood category or those in their late teens to 25 years old (Santrock, 2011). Participants who are in the emerging adulthood age range were chosen since it is following Arnett (2006) that the main criteria for emerging adults are individuals who must accept and assume responsibility for themselves, individuals who are able to make decisions independently, and individuals who are building independence financially. It signifies that emerging adulthood is full of developmental tasks that require one to rely on independence as a preparation to become a full-fledged adult.
The reliability test of the research measurement instrument employed Cronbach's Alpha while the validity of the tool relied on construct and content validity. The data was processed by means of statistical analysis of two stages, comprising the metaanalysis stage and the confirmatory analysis stage. At the meta-analysis stage, data processing administered descriptive statistics. In this stage, mean and standard deviation of each dimension and overall become the basis for constructing autonomy norms. In the CFA stage, data processing applied factor and correlation analysis. During this stage, testing was conducted to determine whether the autonomy dimension had been represented accurately or not.

RESULTS AND DISCUSSION
The PWB studies collected were 112 quantitative studies. Of the 112 studies, 57.14% could not be used for the purposes of meta-analysis as they did not attach research data and/or they had inaccurate research premises. Therefore, the studies whose data could be analyzed were 48 studies, conducted in the course of 2009-2018. Given these 48 studies, it was obtained 2,574 individual data (mean age = 34.55, SD = 17.71; female = 69.8%). Most of the respondents had a high school education background (31.9%) and 56% of respondents were unmarried.
Following data collection, in general, each study had a small sample size with an average of 54 samples/study. With this small sample size per a study, factor analysis could not be performed. Therefore, data processing of meta-analysis utilized frequency distribution and non-parametric statistics to obtain a profile of autonomy dimension in Indonesia.
This data processing began with generalizing the data range. For the most part, the research range applies a scale of 1-6. However, some studies take advantage of different Likert scale ranges, such as 1-4 or 1-5. Thus, for all prevailing data to have the same weight, weighting was carried out so that all data would be in the range 1-6. It was completed by converting the existing scale ranges. In principle, the range 1-6 is the same as the range 1-4, in the sense that the score 4 and the score 6 have the same concept (maximum score) and 2.5 and 3.5 also have the same score (mean). Therefore, arithmetic calculations for scale adjustments from 1-4 to 1-6 were conducted using the following formula: [(score-1 x 5/3) + 1]; as well as for a scale of 1-5 [(score-1 x 5/4)+1]. With this calculation, a weighted score of 3 on a scale of 1-4 will have a weighted score of 4.3 on a scale of 1-6. After that, the data normality test was done. Based on the normality test, it was found that the autonomy dimension had an abnormal distribution so that all data processing would use nonparametric statistics for the PWB dimension (Table 1). Overall, in Table 1, when compared with the scores of other PWB dimensions, the autonomy dimension showed a lower score (mean=3.873, SD=.773) from a score range of 1-6. This finding is in line with previous studies by Saraswati and Teja (2018)  The analysis in Table 1 was proceeded by comparing the scores between dimensions to see e-ISSN: 2541-450X p-ISSN: 0854-2880 Indigenous: Jurnal Ilmiah Psikologi 2021, 6(2), 25-36 the distribution of scores. Friedman-test showed that each PWB dimension had a significantly different score (sig. < .001). It indicates that the PWB autonomy dimension was significantly different from the other dimensions. Subsequently, a correlational analysis was conducted between demographic data (gender) and the autonomy dimension ( Table 2). The results showed gender differences, that male and female had different PWB scores on autonomy (sig. < .000). In this case, male respondents were more likely to have higher autonomy scores. This finding corresponds to Perez's (2012) study of gender differences in PWB among students at the Filipino College. Perez (2012) found that all participants showed low scores on autonomy. But specifically on the autonomy dimension, male students exhibited significantly higher scores than female students. In addition, the results of this study correlate with the research of Li et al. (2015) which examined gender differences in PWB, discovering that females had significantly lower independence scores than males. To obtain a further picture of the research, it was found that in general, PWB research was carried out on several sample categories, which are the general population, the population experiencing social rejection, the population experiencing crisis, the population experiencing conflict (support vs. demands), and population more likely to be positive. Therefore, each research sample was included in one of these categories. Then, a differential analysis was carried out on the dimensions of PWB with sample categories. The Kruskal-Wallis test proved that, in general, five sample categories had different degrees of autonomy in the PWB dimension. An illustration of the difference in scores for each sample category on each dimension is presented in Figure 1. Figure 1 displays that autonomy scores in populations experiencing crisis and populations experiencing conflict inclined to be higher than scores in other PWB dimensions. The opposite situation was discovered in the general population, revealing that autonomy scores appeared to be lower than scores of other dimensions. The group experiencing social rejection had e-ISSN: 2541-450X p-ISSN: 0854-2880 Indigenous: Jurnal Ilmiah Psikologi 2021, 6(2), 25-36 Kiswantomo & Wardani | 31 a higher autonomy score than the scores of other dimensions. These findings confirm the results of Maryam's (2013) qualitative study regarding PWB in transgender women that two participants under the research demonstrated autonomy in terms of initiative, decision making, and finance. Furthermore, the research of Sanchez et al. (2016) found that the preliminary findings of frequent discriminatory treatment, stigmatization, and minorities could help individuals to regain autonomy in the process so as to improve psychological well-being. From Figure 1, it can be determined that the autonomy dimension in the 5 (five) sample categories shows that samples with general categories (such as students, employees) with no peculiarities description of individual conditions appeared to have lower autonomy than other psychological well-being dimensions. This fact is understood since individuals without peculiarities very rarely experience confrontation in the environment, making their autonomy is less developed. A different situation is displayed by groups experiencing crisis and experiencing conflict. Crisis and conflict can expose individuals to frequent confrontations with those around them. A high frequency of confrontation may increase the autonomy of the individual in this group.
A critical review of the autonomy scale findings on one of the PWB dimensions based on exploratory factor analysis (EFA) of all samples more than 59 found that the vast majority of autonomy scale coped two or three sub-dimensions, implying that autonomy is not unidimensional as what has been assumed theoretically. To validate these findings, re-testing was carried out through a questionnaire constructed after completing the meta-analysis phase, and then the data were recollected using a new instrument. Moreover, regardless of autonomy is a dimension of PWB that plays an important role in human life in general, considering that Indonesian society does not adhere to an individualistic cultural orientation, it would be better if autonomy dimension is categorized as a dimension whose existence is needed but not the principal one. To ensure, interdependence factor (counter-A) needs to be performed. Table 3 indicates that there was a positive relationship between autonomy and age (P = .255, sig < .001). These findings correspond with the research results of Springer et al. (2011) that the autonomy dimension might change along with increasing age. Subsequently, a psychological well-being measurement instrument was examined (autonomy as one of the dimensions being tested) and proposed a standard measurement instrument according to circumstances in Indonesia. Researchers re-collected data on students with the number of PWB items equivalent to the original text, including 84 statement items. The data collection was done on the ground that the sample in data collection from previous studies was considered to have a small sample per a study, hence factor analysis could not be carried out.
The next data collection was applied in two stages as a method to determine the Test-Retest Validity and Reliability of measurement instruments. In the first stage, the test was conducted on 151 respondents, and 84 students in the second stage of the same respondents as the first stage. So, respondents that could be used in the validity and reliability test-retest stage were 84 respondents. With n = 84 respondents, each statement item was correlated with the same statement in the second data collection (two weeks after the first data collection). Initially, each PWB dimension had a 14item statement. After processing the data, the number of reliable statement items in each dimension ranged from 8-14 statement items, which is shown in Table 4.
Based on the aforementioned data, 8 (eight) statements with the highest correlation index were used for the second data collection (n = 844). The purpose of this data collection is to perform factor analysis (EFA and CFA) on psychological wellbeing variable. The number of statement items with good discriminating power in each dimension ranged from 6-8 statements. Therefore, 6 (six) statements with the highest factor loading were selected.  Table 4 lists the items from the autonomy dimensions which are "Used" and "Not Used." It is worth noting that the used or not used items from the autonomy dimension could not be directly related to the sample category (as shown in Figure 1). The "Not Used" items represent general content, namely content that people generally regard autonomy, such as "decisions do not depend on what other people do", or "it is difficult to express a controversial opinion." After carrying out EFA on the autonomy dimension, the next step was conducting CFA on the autonomy dimension. The EFA of each dimension is described in Table 5. Table 5 shows the results of the EFA test from the autonomy dimension. Based on the results of the exploratory factor analysis (EFA) the statement items (items) of the autonomy dimension have provided preliminary evidence that this instrument is valid and has positive (+) and negative (-) factor loading simultaneously. This preliminary evidence supports that the statement items of the autonomy dimension can be differentiated into positive factors as well as negative factors. Results of data processing on 844 respondents found that the score of "the ability not to depend on others, and thoughts and actions that are not affected by social pressure" (autonomy) shows the lowest means score. This result corresponds with the findings in stage 1, which has been presented in Table 1, with the result that the mean of autonomy was the lowest compared to other dimensions.  Table 8 describes the norms for the autonomy dimension in the emerging adulthood respondents according to the results of descriptive data processing:

CONCLUSION
This research has resulted in the standardization of measurement instrument for autonomy dimension that can be used for emerging adulthood (19-25 years) in Indonesia. Autonomy norms have been successfully compiled based on the mean and standard deviation of each dimension. The results of the EFA on the autonomy dimension recommended 6 (six) statements with the highest factor loading, comprising 3 (three) positive and negative statements, respectively. From the profile of PWB dimensions, autonomy is the dimension with the lowest mean score. The finding of the lowest score on the autonomy dimension confirms the findings of previous studies.
As customary, this research entails weaknesses. The main weakness comes from the data on the results of theses that have been collected as initial data for meta-analysis. Regardless of the PWB measurement instrument used is a standard measurement instrument constructed by Ryff (2014), consisting of 84 statement items, the non-uniform interpretation techniques, heterogeneous age, and education of research respondents allow the initial research data for meta-analysis for further study.
Autonomy in the context of Indonesian participants does not yet fully reflect individuals who are completely independent of the environment, but individuals who at a certain level wish to express different opinions but still take into account the harmonious relationship with the people around them.
Likewise, even though there is a multidimensional indication of the autonomy dimension, it still requires further and comprehensive studies to confirm it.