Muslim millennials fashion self-congruity:How the religiosity and spiritual well-being affect?

Abstract Individual attitudes and behavior can be influenced by religiosity and spiritual well-being. Religiosity has four dimensions: belief, emotional involvement, and guidance, while spiritual well-being can influence behavior and attitudes by promoting transcendence and a harmonious relationship with the environment. In this context, the study examined how religiosity and spiritual well-being influence how millennials relate to fashion. The researchers collected data from an online survey completed by 287 participants and used least-squares structural equation modeling (PLS-SEM) to analyze the data. The results showed that religious relatedness and guidance were the most significant drivers of fashion self-congruence, specifically IIFSC. In addition, spiritual well-being is a mediator variable between dimensions of religiosity, including busyness, faith, and emotional involvement, and fashion self-congruity measures such as AFSC and IIFSC. This study gives fashion industry managers insights into how religiosity and spiritual well-being influence millennials’ fashion purchase decisions. By understanding these key factors, fashion industry managers can improve their marketing strategies to fulfill the needs of millennial customers better.


PUBLIC INTEREST STATEMENT
Online harassment as an undesirable phenomenon has become a dilemma for humanity. It causes threatening, bothering, fearing and other challenges for the victims. Nowadays, university students in Afghanistan utilize Internet and social media tools for various purposes. Many individuals in Afghanistan think that Afghan female university students who have access to Internet and social media are harassed by males online considerably. Hence, this study aimed to explore the experiences of Afghan female university students towards online harassment, its effects and coping strategies. The finding of the study indicated that the majority of the participants have not experienced online harassment. Also the finding revealed that the effects of online harassment on the victims are low. Besides, the result showed that forty five percent of the respondents used some coping strategies like talking face to face or online with the harasser, reporting the case, cleaning the received messages, keeping silence, etc.

Introduction
The millennial generation is very articulate when expressing themselves in various aspects of life (Ali, 2020;Rahman, 2018). The reason is, referring to Okumus (2021) view, that the millennial generation physically and psychologically has a high activity level. This makes the millennial group a very articulate age, even adaptive compared to other groups (Okumus, 2021). In this context, millennials are described as individuals with characteristics such as being confident, goal-oriented, involved, enthusiastic, sheltered, and active (Gursoy et al., 2013;Okumus, 2021;Stewart et al., 2017). This unique character is decisive in influencing the millennial generation's behavior and articulation pattern. In addition, the value system adopted by the millennial generation is also believed to determine self-articulation, even coloring their behavior in various aspects of life, including consumption and other cultural behaviors, such as dressing.
In choices and decisions, in terms of consumption and cultural behavior, for example, displaying identity in social spaces, millennials have specific rationality (Heyes & Aluri, 2018). This rationality is inherent in the motives and motivations behind every millennial generation's attitude, preference, and behavior towards the object of choice that lies before them. The inspiration that often grows among millennials is more complex than in other ages-for example, the motivation of subsistence and mainstreaming identity. Likewise, not only considering adherence to religion but the choice and behavior of consumption is also aimed at modifying religion (Hasan et al., 2022). That is, Muslim millennials have more complex rationality. It can be identified from a high degree of instrumental rationality. In addition, Muslim millennials also place religiosity as a shaper and the foundation of consumption choices and behaviors. This can be seen in several studies: for example, the study of Sholihin et al. (2022); Osman and Solehah Bt Aziz (2019); Widyanto and Sitohang (2021), who succeeded in identifying that religiosity is the determining factor and can increase the awareness of Muslim millennials towards halal products. In this context, the question is worth asking: "is religiosity capable of and significantly influencing the choice of Muslim millennial dress styles in Indonesia? So, what is the position of spiritual well-being concerning the Muslim millennial dress style in Indonesia?" Two questions arise from several fundamental issues, namely: First, it is assumed that religiosity under any conditions is the most potent factor determining the behavior and pattern of individual choices in various aspects of life (Hill & Hood, 1999;Sholihin et al., 2022;Zwingmann et al., 2011), both economic, political, and even cultural behavior.
Secondly, it is assumed that religiosity is one factor that determines the pattern and style of dressing among millennials. Whether the choice of type of dressing is following self-concept and the identity to be conveyed to others, e.g., actual fashion self-congruity, or thus the selection is aimed at representing the superior value of Islam, e.g., the ideal Islamic fashion self-congruity (Anand & Kaur, 2018;Salam et al., 2019). These two issues will become unique and relevant when connected with the Indonesian context. This is because the Millennial Muslim generation is an age group that is not only large in quantity but also has the potential to influence market trends in Indonesia. At least 61.45% of Millennials in Indonesia were born between 1980 and 2000 (Widyanto & Sitohang, 2021). In this regard, the Millennial generation is the dominant group in Indonesia, which is not only believed to be able to influence Indonesia's social, economic, and political structure. This group is also assumed to be "trendsetters", especially in the context of the dress style of the younger generation in Indonesia. Several studies have found an essential point on this issue, which becomes the foundation to develop on this study. Muposhi and Chuchu's (2022) study that seeks to examine the factors that influence avoidance behavior related to sustainable fashion among Millennial buyers in South Africa. The findings that are considered interesting in this study are materialism and symbolic discrepancies emerging as the main predictors that influence millennial consumers to avoid sustainable fashion (Muposhi & Chuchu, 2022). In addition, the study of Anand and Kaur (2018) emphasizes the development of scales for measuring fashion self-congruity, and this study is the foundation for other studies that seek to map and understand how individual dress style choices, e.g., follow self-conception or thus consider other aspects and values (Anand & Kaur, 2018). In line with the research of Anand and Kaur (2018), Willems study et al. (Willems et al., 2011) first sought to develop a scale termed fashion store personality. This scale measures the degree of immersion of a fashion model in the minds of consumers (Willems et al., 2011). Several studies have attempted and are concerned with examining the factors influencing the choice of millennial dress style, related explicitly to selfcongruity fashion. It is difficult to find specific literature identifying the influence of religiosity and spiritual well-being on fashion self-congruity among Millennials, especially in Indonesia.
This study addresses the gaps in previous research by using several methods. Firstly, this study modeled the relationship between religiosity and fashion self-congruity variables such as actual fashion self-congruity (AFSC) and ideal Islamic fashion self-congruity (IIFSC). Secondly, it investigates the role of spiritual well-being as a mediator variable in linking the exogenous variable (religiosity) with the endogenous variable (fashion self-congruity). From these two strategies, this study has novelty in several ways, namely: Firstly, by involving religiosity variables, with four dimensions, i.e., religious preoccupation; religious conviction; religious, and emotional involvement; and religious guidance, it is expected to be able to model the relationship between religiosity and fashion self-congruity. The uniqueness of this study lies in the dimensions chosen to explain religiosity, which is the main variable under study. Generally, religiosity is explained through its two dimensions, namely "religious beliefs" and "participation" (Bergan & McConatha, 2001). In this regard, this study uses four dimensions of religiosity, i.e., religious pre-occupation, religious conviction, religious, emotional involvement, and religious guidance, to clarify how religiosity affects "fashion self-congruity".
Secondly, this study used spiritual well-being to mediate between religiosity and fashion selfcongruity. This was done to determine the specific aspects of religiosity that spiritual well-being can influence concerning fashion self-congruity. It is considered relatively new because spiritual well-being in contemporary studies is generally explained through transcendence to God and efforts to build harmonious relationships with the environment (Ekşi & Kardaş, 2017;L. Ellison, 2006). However, this study also adds one dimension of SWB, namely the absence of anomalous feelings towards life as a product of the process of transcendence to God. Thirdly, related to the dimensions of fashion selfcongruity, this study offers a relatively new and exploratory dimension, namely the Islamic fashion self-congruity dimension: a dimension that refers to the conformity of the style of dressing with Islamic values and rules. As a response, this paper is written in several parts: the first is an introduction that parses the problem statements, gaps, and novelty. The second part parses the literature review, e.g., fashion self-congruity, religiosity, and spiritual well-being. The third part describes the methodology. The fourth part describes the findings using the PLS-SEM (partial least square-structural equation model) approach. The fifth part of the analysis compares empirical results and existing studies. Finally, the conclusions confirm the study's contribution to the "body of knowledge" and the limitations and recommendations for subsequent research.

Literature review
This section will break down the literature related to three variables used in this study, namely: fashion self-congruity (Anand & Kaur, 2018;Willems et al., 2011); dimensions of religiosity, i.e., religious pre-occupation; religious conviction; religious, emotional involvement; and spiritual guidance (A. B. Cohen et al., 2005;Hill & Hood, 1999;Ul-Haq et al., 2020); and spiritual well-being (Ekşi & Kardaş, 2017;Moberg & Brusek, 1978). The primary purpose of the review of this literature-based variable is to formulate a hypothesis, which will be verified through the PLS-SEM approach.

Fashion self-congruity: An overview
The concept of fashion self-congruity (FSC) is a development of the theory of "self-congruity" in the context of fashion. In general, this concept explains the compatibility between fashion style and one's self-concept (Anand & Kaur, 2018;Casidy et al., 2015). It rests on the assumption that fashion choices communicate one's taste, values, and even identity (Au et al., 2007;J. Lee, 2009). This means that needs determine not only one's clothing choices but are more broadly influenced by the identity and self-concept one wants to externalize to others. Based on this assumption, the FSC scale was introduced by Anand and Kaur (2018) to measure the congruence between one's fashion choices and self-concept. Anand and Kaur (2018) present various dimensions in their fashion self-congruity scale, including (Firstly) actual fashion self-congruity (AFSC). This dimension relates to the extent to which a person views their fashion style as a reflection of their actual attributes. Secondly, ideal fashion self-congruity (IFSC), where this dimension refers to the extent to which a person considers his or her fashion style to represent the attributes he or she has. Thirdly, social fashion self-congruity (SFSC), where this dimension refers to the extent to which a person views his or her fashion style as consistent with what others think about him or her. Fourthly, ideal social fashion self-congruity (ISFSC), where this dimension refers to the extent to which a person perceives her fashion style as a reflection of the image she wants others to have of her (Anand & Kaur, 2018). The "fashion self-congruity" dimension proposed by Anand and Kaur (2018) is still very general, and is not aimed at one particular community, such as Muslims, so the dimension used does not specifically represent the cultural characteristics of the Muslim community. In this context, it is not superfluous to propose one dimension to complement it. It is called "Islamic ideal fashion self-congruity" (IIFSC): a dimension that refers to an individual's perception of clothing choices as an attempt to reflect ideal Islamic values, i.e., morality, ethics, and Islamic principles.
There are several arguments behind the novelty of the new dimension (IIFSC) of fashion selfcongruity: First, Islam for Muslim communities is not just a religion that only provides ritual guidance and transcendental dimensions, but Islam itself is a world-view that is the foundation behind choices and behaviors in daily life (Abdullah & Nadvi, 2011;Johan et al., 2020). Second, Islam is believed to be not only lived as an effort to enhance spirituality alone but at the most authentic level, passion for Islam thus colors, even shapes the unique and specific rationality of Muslims (Furqani & Masudul Alam Choudhury, 2015;Hardivizon et al., 2016;Ngasifudin, 2018). Third, Islam in Muslim societies has become a force capable of driving cultural and social transformation (Erden, 2019;Lai & Thornton, 2015). Islam is lived by individuals, not just by rites, but has colored practical, behavioral, and cultural matters. In this case, Islam has shaped people's cultural behavior, including the tradition of dress in Muslim communities. These three arguments are then used to propose a new dimension to fashion self-congruity theory to identify the clothing choices of Muslim communities.

Dress standards: within an Islamic framework
The framework and rules of good dress according to Islam refer normatively to Q.S. 7: 26, where through this text, Islam teaches that all Muslims choose clothes capable of carrying aurat (intimate parts), which has been set limits by Islamic law (fiqh). Islam has established moral and ethical guidelines for Muslims regarding their dress, including how to acquire clothing and the type of clothing chosen. The Quran and the Sunnah of Prophet Muhammad emphasized the importance of dressing modestly and appropriately to avoid drawing unnecessary attention from others ( (Kamarulzaman & Shaari, 2020). In simpler terms, dressing in Islamic traditions must be preventive and prevent the lust of others when paying attention to someone. In this context, Islam only defines the boundaries of "aurat" (intimate parts), which should not be shown to others, both men and women. Meanwhile, the models and patterns of clothing are entirely culturally appropriate (Rosida, 2019)so each Muslim community has a different way of dress but meets the criteria for good dress according to Islamic rules. Sabiq (1991), for example, sets limits on women's aurat other than the face and two palms, which must be covered according to the text of the Quran (24): 31 ( (Sabiq, 1991;Sesse, 2016). The restrictions put forward by Sabiq (1991) and the majority of Islamic scholars agree that in addition to the face and two palms are mandatory, it is even a legal requirement for the prayer of every Muslim (Hafidz & Ibn, 1994). But when outside of worship, women and men face two conditions. When a woman is in the presence of someone else, who can marry her (mahram), the woman's aurat is all the limbs, other than the face and palms. While for men, their aurat is between the navel and the knee. This framework and scholars' explanations regarding the rules and ethics of dress in Islam are considered for every Muslim, both women and men, in choosing clothes and dressing. Thus, the choice of clothing in Islam is not only limited to needs but is an indicator of adherence to Islamic values, ethics, and morality.

Religiosity: Definition, dimensions, and hypotheses
Scholars put forward many definitions regarding religiosity. This difference depends on many factors, one of which is the perspective used by scholars in formulating the meaning of religiosity (Sholihin et al., 2022). In addition, differences in the sense and definition of religiosity are also driven by the difficulty in measuring and identifying dimensions of religiosity (Meguella et al., 2015). However, some explanations seem to be easy and holistic to interpret religiosity. For example, Mookherjee (1994) reveals that religiosity is a term that is transcendental and, at the same time, participatory. Transcendental because religiosity refers to an individual passion for doctrine and the faith expressed through the rituals of worship and chanting prayers. In addition, religiosity is defined as a participatory concept since this term refers to individual participation in religious activities that are colossal and collective (Mookherjee, 1994). In a different editorial, Koenig (1998) defined religiosity as a collection of beliefs and practices related to transcendental matters (Koenig, 1998). In addition, Zullig et al. (2006) interpret religiosity as a concept consisting of a belief system with rituals that are binding, well-established, and consist of various religious practices that require a specific place, for example, mosques, churches, and monasteries (Zullig et al., 2006). These multiple definitions and meanings of religiosity affirm that religiosity has relatively dynamic dimensions. Therefore, various scholars propose variations in the dimensions of religiosity. Joseph and DiDuca (2007) deeply introduce four dimensions of religiosity, namely: (i) pre-occupation: this dimension refers to individual psychological feelings, which include a sense of peace and comfort when believing in God; (ii) conviction: dimension refers to faith and individual recognition of God's existence; (iii) religious emotional involvement: the dimension attached to religiosity as the emergence of things that are emotionally appropriate when living and practicing religious doctrines, such as feelings of pleasure, happiness, peace, and feeling overwhelmed; and (iv) guidance: this dimension refers to individual expressions and beliefs that religion is a guide for various aspects of life (Joseph & DiDuca, 2007). The four aspects introduced by Joseph and DiDuca (2007) affirm that religiosity is complex and dynamic. Its means that this dimension of religiosity is emotional in shaping and influencing individual behavior. Sometimes one dimension simultaneously affects individual aspects of life, but sometimes the existing dimensions partially affect aspects of human life. Although this study refers to the dimensions of religiosity (i.e., pre-occupation, conviction, religious, emotional involvement, and guidance) introduced by Joseph and DiDuca (2007), it modifies these four dimensions. It adapts them to the character of the Muslim community. In this context, the hypotheses that can be formulated related to this study are as follows: H1. The religiosity dimension of "pre-occupation" directly and significantly affects the "actual fashion self-congruity" aspect.
H2. The religiosity dimension of "pre-occupation" directly and significantly affects the aspect of "Islamic ideal fashion self-congruity." H3. The dimension of religiosity of "conviction" directly and significantly affects the "actual fashion self-congruity" aspect.
H4. The religiosity dimension of "conviction" directly and significantly influences the aspect of "Islamic ideal fashion self-congruity."

H5.
The religiosity dimension of "religious emotional involvement" directly and significantly affects the "actual fashion self-congruity" aspect.
H6. The religiosity dimension of "religious emotional involvement" directly and significantly affects the aspect of "Islamic ideal fashion self-congruity." H7. The religiosity dimension of "guidance" directly and significantly affects the "actual fashion selfcongruity" aspect.
H8. The religiosity dimension of "guidance" directly and significantly affects the aspect of "Islamic ideal fashion self-congruity." Mathad et al. (2017) view and identify "spiritual well-being" as a multidimensional concept. Therefore, Fisher (2010) defines "spiritual well-being" as an affirmation of a strong and holistic relationship between life and God, self, society, and the environment (Fisher, 2010). Spiritual well-being (SWB) affirms one inseparable relationship; it can even be considered unity between God, man, and the environment. Furthermore, Monod et al. (2010) assert that SWB is interpreted and known as one consequence and as the operationalization of spirituality and the spiritual experience (Monod et al., 2010). More specifically, Chowdhury and Fernando (2013) interpret SWB as a construct that can measure personal development in the spiritual dimension, primarily related to individual well-being (Chowdhury & Fernando, 2013). The meaning is understandable because Chowdhury and Fernando (2013) interpret SWB as a unity of two constructs, namely "spirituality" and "well-being." Shortly, it can be interpreted as an index of life of spirituality that indicates the quality of life within the framework of the spiritual dimension (Chowdhury & Fernando, 2013;Pellengahr, 2018); it is also in this context that Moberg and Brusek (1978) believe and affirm that SWB is a subjective experience derived from the process of personally searching for the meaning and purpose of life concerning the transcendental dimension (Moberg & Brusek, 1978). The various substances and concepts of SWB proposed by these scholars presumably have in common that SWB is a subjective experience and is the product of the reflection of spirituality in the broader dimension of life. SWB can be interpreted as a willingness to seek the meaning and purpose of human existence. The tendency to scrutinize and try to derive knowledge from intangible or incomprehensible concepts. (Opatz, 1986). The definition of SWB, although it has a reasonably diverse variety, thus delivers on the dimensions of SWB vertically, i.e., the relationship between man and God, and horizontally, i.e., the relationship between humans and their environment (L. Ellison, 2006;Moberg & Brusek, 1978). Moberg and Brusek (1978;L. Ellison (2006); Opatz (1986) divide the spiritual wellbeing dimension into two forms, namely: "horizontal" and "vertical"-horizontal refers to the extent to which a person perceives the relationship between self and environment to reflect the values of spiritual goodness. The vertical dimension refers to how a person interprets the metaphysical relationship between his existence as a religious believer and his God. Although these dimensions are most commonly used in social research, this study developed one dimension that reflects the extent to which a person is grateful for life and feels satisfied with the life they have lived. This dimension is called "internal well-being." In addition, these gradually connected relationships, horizontal and vertical, thus assert that the concept of spiritual well-being has broad implications for individual dimensions of life, behavior, and choice. It is just that through this study, the SWB variable is not directly positioned as an exogenous or independent variable that directly affects other variables, such as the dimensions of the fashion self-congruity variable. However, SWB is included in the model as a mediator variable against religiosity and fashion self-congruity. With regard to the hypothesis to be tested through empirical means, it is stated as follows:

Spiritual well-being: Meaning and hypothesis
H9. The relationship of the religiosity dimension of "pre-occupation" to the "actual fashion selfcongruity" aspect is mediated by the SWB variable.
H10. The relationship of the religiosity dimension of "pre-occupation" to the aspect of "Islamic ideal fashion self-congruity" is mediated by the SWB variable.
H11. The relationship of the religiosity dimension of "conviction" to the "actual fashion selfcongruity" aspect is mediated by the SWB variable.
H12. The relationship of the religiosity dimension of "conviction" to the aspect of "Islamic ideal fashion self-congruity" is mediated by the SWB variable H13. The relationship of the religiosity dimension of "religious emotional involvement" to the "actual fashion self-congruity" aspect is mediated by the SWB variable.
H14. The relationship of the religiosity dimension of "religious emotional involvement" to the aspect of "Islamic ideal fashion self-congruity" is mediated by the SWB variable.
H15. The relationship of the religiosity dimension of "guidance" to the "actual fashion selfcongruity" aspect is mediated by the SWB variable.
H16. The relationship of the religiosity dimension of "guidance" to the aspect of "Islamic ideal fashion self-congruity" is mediated by the SWB variable.

Method
There are three groups of variables used in this study, namely: First, exogenous variables consisting of four dimensions of religiosity, i.e., "pre-occupation," "conviction," "religious, emotional involvement," and "religious guidance" (Holdcroft, 2006;Schnitker et al., 2021). Both endogenous variables are "fashion self-congruity," with two dimensions: "actual fashion self-congruity" and "Islamic ideal fashion self-congruity" (Anand & Kaur, 2018). The three mediator variables are "spiritual well-being, " which reflects the spiritual health of individuals, especially millennials (C. W. Ellison, 1983;Ekşi & Kardaş, 2017). This model involving three groups of variables requires the right approach and can identify dynamic relationships between three groups of variables. Therefore, this study uses a partial least square structural equation model (PLS-SEM) approach. This choice is based on Wong's (2013) view that PLS-SEM is the second generation of multivariate data analysis methods, often used widely in the social sciences, especially marketing research. This is because this approach is reliable in testing theories with "additives" and "linear causal modalities" (Wong, 2013). In addition, PLS-SEM is thus believed to be a capable method for estimating complex models with many constructs, variable indicators, and structural paths without imposing distributive assumptions on the data (J. F. Hair et al., 2019). That is, this approach is very suitable for conducting relatively complex model testing (Figure 1) related to the influence of religiosity dimensions on fashion self-congruity, as well as how the mediator variable "spiritual well-being" can mediate exogenous variables, i.e., religiosity and endogenous variables, i.e., fashion self-congruity.

Measures
Figure 1 is a research model or framework and, at the same time, informs that there are three groups of variables with specific items. The first is the exogenous group of variables that refer to the dimension of religiosity: (i) pre-occupation. This dimension refers to personal psychological feelings who feel peace when believing in God. In this case, there are four items used as measurement scales, namely: (a) my thoughts are often fixed on God, (b) I think of God all the time; (c) my mind is fixed on God every day; (d) everything that happens to me is reminiscent of the Lord (Joseph & DiDuca, 2007); (ii) the conviction dimension: the dimension that refers to the faith and individual recognition of the existence of God. There are four measurement items in this dimension, namely: (a) I have always believed in God; (b) I believe that God exists; (c) I am sure God jacked up the prayer; and (d) I know that God is always there for His servant (Joseph & DiDuca, 2007;Reitsma et al., 2006); (iii) the dimension of "religious emotional involvement": this refers to the appearance of psychological emotions when carrying out religious activities, such as a sense of happiness, a sense of blessedness; and a sense of pleasure. Items used in this dimension include: (a) I feel happy when I think of God; (b) being a Muslim is fun; (c) praying increases morale; (d) when feeling miserable, then thinking of God becomes a solace; (e) God has always loved His servant (Galen & Kloet, 2011;Joseph & DiDuca, 2007); and (iv) the dimension of "religion as a guide": the dimension refers to the function of religion as a guide to life; hence the items used include: (a) consistently invoking God's instructions; (b) I consistently follow the laws established by the Quran; and (c) I am incapable of making decisions without the Lord's guidance (Joseph & DiDuca, 2007;Torgler, 2006).
The second is an endogenous variable that refers to two dimensions of "fashion self-congruity". . The third is the mediator variable, namely SWB. This variable refers to the dimensions of spiritual health ranging from the discovery and achievement of satisfaction when praying to God to feeling that life is more positive when believing in God (Ekşi & Kardaş, 2017). In addition, another group of variables is the control variable which includes two socioeconomic dimensions of the respondents, namely the cost of living each month.

Data collection and pre-test survey
This study collected data online from early July to late August 2022. The responders are volunteers (Vehovar et al., 2016), and there is no coercion in filling out questionnaires distributed online at various universities in Indonesia, especially Public and Private Islamic Universities. This kind of data collection refers to the models that Rowley (2014) offered and introduced. Online surveys were chosen because they are more time-efficient and effective for homogeneously-compliant target populations (Ball, 2019;Rowley, 2014). Before distributing the questionnaire, two stages are carried out: instrument tests, which include validity and reliability tests (Ball, 2019). At this stage, 60 data were used for validity, and reliability tests, where the results showed that the items used were valid because it indicates the value of r count > r table 0.25. At the same time, the alpha value reaches 0.94. This means that it has qualified as required by Nunnally (1990). In addition, this study uses Harman-single factor analysis to identify the normality of the data and ensure that there is no bias in the data (Podsakoff et al., 2003). In this context, Rahi et al. (2020) assert that the variance value in Harman's single factor test must be less than 40%, so it can be concluded that the data is unbiased and there are no outliers (Rahi et al., 2020). The results show the variance of each item used <.40. This means that the data used in this study does not contain "common bias." Second, the minimum sample amount used in this study was determined after the instrument test. The G-Power application calcifies the sample size on the "linear multiple regression: random model" statistical test by entering a standard error value of 0.05 and a probabilistic power level at 0.95 points (L. Cohen et al., 2007), then the minimum total sample was 164. This means that by using the number of data of 287 responders, this number of pieces is already qualified. This is per the criteria proposed by L. Cohen et al. (2007).

Data Analysis
This study uses the PLS-SEM approach, believing that this approach can evaluate multiple factors with complex variables and items (Rahi & Mazuri, 2019). In the analysis process, based on PLS-SEM analysis techniques, this study uses two stages of analysis, namely: "the measurement" and "structural models" (J. Hair, 2017;Wong, 2013). For a valid and accurate model, the study used SmartPLS 3.5 software for instrument tests using Stata 16.0.

Results
This study has a complex model character because it involves four variables: exogenous, endogenous, mediator, and control. Therefore, "partial least square-structural modeling" (PLS-SEM) is used to perform estimation and is assisted by SmartPLS version 3.3.0 in analyzing the data. In context, PLS-SEM is recommended by many social scholars, one of which is J. Hair (2017). Unlike other approaches, PLS-SEM does not require data normality (J. Hair, 2017). Because basically, PLS-SEM is recommended to test models with complex constructs (Isa et al., 2020) and exploratory studies. This argument then uses the complete PL-SEM approach and seems relevant for analyzing demographic features, as found in Table 1. 67% of the participants were female, and the rest were male. As for the most dominant age, participants were 20, as much as 72%. The participant's cost of living is 72% IDR 500,000-1,000,000.

Measurement model assessment
Theoretically, the measurement model assessment consists of "internal consistency reliability," "convergent validity," and "discriminant validity." The "internal consistency reliability" is checked using Cronbach's alpha (CA) and composite reliability (CR). Table A1 shows the values of Cronbach's alpha and the value of composite reliability, where all item values for CA and CR values above 0.70, as required by J. Hair (2017), the existing values show a vigorous intensity of consistency reliability (J. Hair, 2017). In addition, "convergent validity" can be achieved as shown in Table A1, where the value of the loading factor is above 0.40 as required by J. Hair (2017), the CR value of all items is above 0.70, while the AVE value is above 0.50, only the constructional spiritual well-being is below 0.50, but this is acceptable as asserted by Fornell and Larcker (1981), with a note cr value above 0.60 (Fornell & Larcker, 1981;Lam, 2012). The Heterotrait-Mononitrate ratio (HTMT) in the context of this study was used to measure "discriminate validity" as recommended by Henseler et al. (2015), where HTMT is the best criterion for determining "discriminate validity" compared to the Fornell-larcker criterion (Henseler et al., 2015). Table A1 indicates that the requirements for discriminant validity (Table A2) are met because they are below 0.90 (Henseler et al., 2015;J. F. Hair et al., 2018;Roemer et al., 2021). Meanwhile, the issue of multicollinearity was not found in this data because the VIF value was below 5 (J. Hair, 2017), where the VIF value is between 1-2.59.

Structural model assessment
The issue of multicollinearity between exogenous variables was tested using inner VIF values, where it was found to be 1.35 for the religiosity dimension of "pre-occupation," 1.79 for the religiosity dimension of "conviction," 2.71 for the religiosity dimension of "emotional involvement," and 2.22 for the religiosity dimension of "guidance." In addition, to evaluate "the structural model" for path coefficients, the study used a "bootstrapping" technique with a "confidence interval" rate of 95%, with 5,000 resamples (Table 2). Overall, the exogenous variables meet the criteria required by J. Hair (2017), below 5 (J. Hair, 2017).
From the results of "the structural model" carried out through PLS-SEM, for the "total direct effect," it was found that the religiosity dimension of "guidance" positively affects the fashion self-congruity dimension "ideal Islamic fashion self-congruity" (Table 2), where the value of t-value = 2.16, p < 0.05. Meanwhile, for mediator effects (specific indirect effects), there are four exogenous relationships to endogenous variables mediated by spiritual well-being variables, namely: (i) pre-occupation ->SWB->IIFSC; (ii) Religious emotional involvement->SWB->AFSC; (iii) Religious emotional involvement->SWB->IIFSC; (iv) Religious guidance->SWB->AFSC; and (v) Religious guidance->SWB->IIFSC. The level of mediation between the religiosity and fashion selfcongruity dimensions through the spiritual well-being variable can be calculated through the Variance Account for (VAF). This refers to the criteria of J. Hair (2017); there are three forms of mediation, namely: (1) the VAF value<20% then the no-mediation effect; (2) VAF value>20% and VAF<80% then mediation is partial; and (3) VAF value>80% then the media is perfectly fast or full-mediation (J. Hair, 2017). From the calculation using the VAF method, full mediation only occurs between the religiosity dimensions of "guidance"->SWB->IIFSC, with a VAF value of 99%. Meanwhile, partial mediation occurs between (1) the religiosity dimension of "pre-occupation ->SWB->IIFSC (VAF 76.94%); (2) Dimensions of religiosity "emotional involvement" ->SWB->AFSC (VAF 72.20%); (3) the religiosity dimension of "emotional involvement"->SWB->IIFSC (VAF 70.40%); and (3) the religiosity dimension of "guidance"->SWB->IIFSC (VAF 65.63%). Meanwhile, the religiosity dimension of "guidance" has a detrimental effect on the dimension of fashion self-congruity, as it is indicated by an f value of 0.02. This refers to Cohen's criteria (J. Cohen, 1988), the values of f 2 0.02 (small), 0.15 (moderate), and 0.35 (significant effects). In addition, the coefficient of determination R 2 is obtained at 0.13 for IIFSC and 0.24 for AFSC. That is, the variation of exogenous variables (dimensions of religiosity) can explain endogenous variables (dimensions of fashion self-congruity) at a moderate level (J. Cohen, 1988). This refers to Cohen's criteria (J. Cohen, 1988), where 0.26 (good), 0.13 (moderate), and 0.02 (weak) are.   Table 3 shows that, statistically, the models offered and tested in this study have met the criteria for the goodness of fit. The standardized root mean square residual (SRMR) value is 0.071, below 0.08 (Hu & Bentler, 1999). Finally, to identify the relevance and predictive strength of the model, a blindfolding method was used and successfully estimated that the Q 2 for AFSC and IIFSC was 0.06 and 0.15, respectively. This means the resulting model has a predictive value and is relevant to the data since the value is greater than 0 (Hu & Bentler, 1999;J. F. Hair, 2014;Shaffer, 2015). From the estimation of "the structural model" carried out through the PLS-SEM approach, the question that is then worth asking is: "what meaning can be built from the results and statistical estimates of the fashion model of the millennial generation in Indonesia?" Next, this matter will be critically discussed through empirical and theoretical comparisons.

Robustness: Moderator's role and IPMA Analysis
The principle of robustness in modern statistics is necessary to test the consistency of one model when a new variable is inserted or another analysis of the same variable is used (Aldrich, 2006;Parks et al., 2010). This study uses two strategies to identify the robustness of the resulting model: First, by including the moderator variable, namely "monthly spending" ( Figure A1). In this case, the variable "monthly spending" can moderate the relationship of the religiosity dimension of "emotional involvement" to the fashion self-congruity dimension of "AFSC." It is characterized by a p-value < 0.05, which is 0.02. The second is by using importance-performance map analysis (IPMA) on two dimensions of fashion self-congruity, namely "AFSC" and "IIFSC." Regarding this analysis, Figures A2 and A3 show that the religiosity dimension of "conviction" has a high performance in explaining the measurement of fashion self-congruity, which reaches 94.37, and this also applies to the dimension of fashion self-congruity "IIFSC." The meaning that can be formulated from these two strategies is that the dimension of religiosity consistently shapes fashion choices (fashion self-congruity) among Millennials in Indonesia.

Discussion
Empirically, this model shows dynamic relationships. Table 2 shows two forms of influence, namely direct and indirect. In direct relation, only the religiosity dimension of "guidance" significantly influences the fashion self-congruity measurement of "IIFSC." Interestingly, the change occurs when the mediator variable "spiritual well-being" is entered. Dimensions of religiosity such as "pre-occupation," "emotional involvement," and "guidance" show a significant influence on the fashion self-congruity dimensions of "AFSC" and "IIFSC" through the mediation of the Spiritual Well-Being variable ( Table 2). The study findings confirm that religiosity and its dimensions can influence individual choices, including foaming. This finding is understandable, because "guidance" is holistically contained in fiqh or Islamic law, and then becomes an institution in the actual life of every Muslim, including in determining the style of dress (Bayoumi et al., 2020). In this context than the meaning of Wilson et al. (2013) that religion in the modern urban world is a form of "neo-spiritualism." The term refers to the expansion of "consumer culture theory," where Islam has become a reference in shaping the culture of consumerism in the 20th century (Wilson et al., 2013). That is, religion in the modern era has become one force shaping the expression of religiosity and spirituality in various aspects of contemporary society's life, including the choice of dress in the millennial group.
The study of Krisjanous et al. (2020) previously confirmed that religiosity is a determinant of forming normative perceptions based on religious values among Muslim female consumers in choosing a dress style (Krisjanous et al., 2022). In line with that study, Bachleda et al. (2013) identified that religiosity dynamically affects how women choose clothes. The dynamic influence of religiosity is formed when other variables, such as education level, marital status, age, and income, are incorporated into the model (Bachleda et al., 2014). In addition, the study of Ashraf et al. (2022) shows that the younger generation of Muslims tends and can even be said to be consistent in aligning the choice of dress with the religious identity embraced; in this case, it is Islam (Ashraf et al., 2022). This condition then encourages the growth of the fashion industry that pays attention to and considers Islamic identity and then offers Islamic-clothing products in various countries and major cities, especially in Indonesia. However, the existing literature and studies do not emphasize which aspects of religiosity strongly influence dress choices among millennials. In response, the study presumably offers an alternative model to explain dress choices among millennialswhether the choice of dress is an essential aspect of fashion self-congruity or is it precisely the ideal Islamic fashion self-congruity. Thus, the choice is influenced by which dimension of religiosity -pre-occupation, conviction, emotional involvement, or religious guidance. This study succeeded in finding and convincingly producing a theory that the choice of dress that relies on ideal Islamic values among millennials in Indonesia is directly and significantly influenced by the religiosity dimension of "guidance." That is faith and passion for Islam as a guide to behavior by actively encouraging the choice of millennials to dress per Islamic norms and rules.
Meanwhile, other dimensions of religiosity, such as pre-occupation, conviction, and religious and emotional involvement, are insignificant and do not even affect the choice of dress without the growth of spiritual well-being in the Millennial generation in Indonesia. This presumably confirms the view of functionalism that religion will experience a degradation of meaning without being followed by the strengthening and passion of spirituality. Spirituality can strengthen one's religious vision so that religiosity and existing dimensions can color and shape individual behavior patterns and choices, including dress choices (Moberg & Brusek, 1978;Y. Lee, 2014). Meanwhile, Chowdhury and Fernando (2013) more deeply managed to identify the reasons why the relationship between well-being spirituality and religiosity is coexistent, i.e., religiosity does not form too much of an individual pattern of choice without being supported by well-being spirituality because spiritual well-being is actively able to activate consumer awareness and vision of "ethical beliefs" such as choice does not violate religious values, does not harm the soul, and able to increase personal problems (Chowdhury & Fernando, 2013). The explanation proposed by Chowdhury and Fernando (2013) seems to be an argument and theoretical justification for why the SWB (spiritual well-being) variable strengthens, even able to activate the dimensions of religiosity, i.e., pre-occupation, conviction, and emotional involvement, to then influence the choice of dress among millennials in Indonesia.

Conclusion
The study of dress choices among Millennials in Indonesia-whether the selection is based more on the desire to conform to self-concept (self-congruity) or follow and adhere to Islamic idealsconfirms that the choice of dress is complex. The choice of dressing at the individual level, especially among Millennials, is influenced by various factors. Still, the dimension of religiosity is believed to be one of the critical factors determining the pattern and choice of dress in the Millennial generation. This study partially confirmed this, where only the religiosity dimension of "guidance" can encourage millennials to dress consistently per ideal Islamic values. While other dimensions of religiosity, i.e., pre-occupation, conviction, and emotional involvement, only affect the choice of dress, i.e., AFSC or IIFSC, among Millennials when mediated by spiritual well-being variables, The meaning that can be inferred from this empirical fact is that awareness of Islam as a guide will linearly form the consistency of choices and behaviours of the millennial generation, which is per the values and norms Islam. In addition, spiritual and well-institutionalised qualities in the Millennial era will activate broader dimensions of religiosity to influence choices and behaviours, including how to dress well and in accordance with the values of religiosity. To ensure that fashion choices are in line with Islamic values, there are several approaches that can be used based on the results of this study: Firstly through a structural approach, where educational institutions ideally play a role in raising awareness of Islamic values and morality. Secondly through a cultural approach, especially with the role of ulama and families to consistently foster awareness of Islamic values and morality.

Theoretical implications
The study is speculated to be the first study to develop the dimension of "ideal Islamic fashion selfcongruity." In addition, using spiritual well-being as a mediator variable for the dimensions of religiosity-pre-occupation, belief, emotional involvement, and religious guidance-and the dimensions of fashion self-congruence-AFSC and IIFSC could potentially enhance the theoretical understanding of individual clothing preferences. This model could add to the "body of knowledge" on this topic, particularly regarding millennials, who may find it difficult to make clothing choices.

Practical implications
The results of this study show that coexistence and spirituality are significant factors influencing millennials' clothing choices. Therefore, the clothing industry catering to this market segment can use these findings as a reference and endeavour to offer accessories and designs aligned with Islamic values, tastes and principles. This can potentially increase interest among millennials and foster brand loyalty towards specific Muslim clothing brands.

Limitations and recommendations for future research
The data collected is limited to millennials aged 17 to 27; most are recorded as students at Islamic universities in Indonesia. This is undoubtedly one limitation, even a weakness, that can be corrected in the future. Therefore, the next researcher can expand the population, not only in the millennial generation of Islamic universities but also in public universities and the wider community, so that the relationship between religiosity, spiritual well-being, and fashion selfcongruity can be appropriately mapped and more accurately.

Item
Loading Like CR AVE  Source: data analysis Shalihin et al., Cogent Social Sciences (2023)