Intention as a determinant of Islamic work culture in Indonesia based on the modified theory of planned behavior

work culture values in the Ministry of Religious Affairs of the Republic of Indonesia. Motivation for the study: The TPB has been considerably modified and expanded to predict various types of organizational behaviour in different fields. However, the modifications did not reduce or change the substance and have not yet been applied to religious organizations. In contrast to previous studies, this article modifies the TPB from its original form and applies it to religious organizations in Indonesia. Research approach/design and method: This article employs a quantitative method. Primary data include samples collected from all Regional Offices of the Ministry of Religious Affairs in 34 provinces in Indonesia, while data analysis uses structural equation modelling and structural models. Main findings: This article finds that the most decisive work culture value in the organisation of the Ministry of Religious Affairs of the Republic of Indonesia is spirituality (intention). Implications: Results of this study have extended the implementation of the TPB by modifying perceived behaviour control (PBC) with organisational culture, subjective norm with peer reference and attitude with spirituality to predict organisational values (Nilai Budaya Kerja – NBK)-based work behaviour at the Regional Offices of the Ministry of Religious Affairs with compliance intention as a mediating/intervening variable, which acquired positive and significant results. Contribution: Theoretically speaking, the findings in this study provide empirical evidence that modifying TPB reinforces TPB’s applicability to predict compliance intention and work behaviour. Practically speaking, this research provides recommendations for improving the implementation of the five organisational values to the Ministry of Religious Affairs in three aspects. First, enhance Civil Servants’ (Pegawai Negeri Sipil – PNS) spirituality by providing substantial understanding of religious values. Second, intensify the role and support of structural leaders, particularly echelon 1, 2 and 3 officials so they provide good examples in the implementation of NBK. Third, strengthen organisational culture at the Ministry of Religious Affairs by creating result/impact-oriented programme planning (result/impact oriented government).


Introduction
The theory of planned behaviour (TPB) had initially been applied in various studies to provide better understanding on various behaviours in the field of social psychology. The TPB had subsequently been used in the field of marketing to predict consumer behaviours relating to product purchase. In its current development, TPB continues to be developed to predict various types of behaviour within organisations (Armitage & Conner, 2001), behaviours in the use of information technology (Ozer & Yilmaz, 2011), health-related behaviours (Conner & Sparks, 1996), behaviours in transportation (Bamberg et al., 2003;Kerr et al., 2009) and others. The advantage of TBP is that it maintains a greater prediction power compared with most other behavioural change models (Taylor et al., 2006).
Previous researchers have developed the three main variables of TPB, namely attitude towards behaviour, subjective norm, perceived behavioural control (PBC), and intention by way of extending or modifying them with other variables (Guerin et al., 2018;Mustapha et al., 2016;Sardžoska & Tang, 2012;Urumsah et al., 2016). Meng et al. (2014) modified the attitude variable with ethical ideology. Trisna (2019) modified the subjective norm variable with peer reference. Peredaryenko (2019) modified the PBC variable with subjective product knowledge. Modifications are acceptable as TPB has an extensive definition allowing it to be continuously developed further so that it could provide even more accurate elaborations concerning variables that influence behavioural intention in accordance with the observed phenomenon.
Several prior studies have also extended the variable of intentions into compliance intentions to predict certain behaviour. For instance, Ajzen (1991) used TPB to predict compliance intentions of taxpayers in Turkey, which indicates that the normative expectation of compliance and the intensity of penalty have a significant effect on compliance intentions. Appah and Wosowei (2016) found a significant relation between compliance intention and equity attitude of taxpayers in Nigeria. In research by Sarrah et al. (2016), it was found that higher perceptions of control and ability taxpayers will encourage their compliance intentions. Afdalia et al. (2014) have proven that compliance intentions can predict the implementation of Governmental Regulation No. 71 of 2010 on the Accrual Based Government Accounting Standards. Hartoni and Riana (2015) used compliance intentions to predict work behaviour of employees in implementing Occupational Safety, Health, and Environment (OSHE) policy. Wong and Lee (2016) used the TPB model with the variable of compliance intention as a predictor of work protocol and occupational safety behaviour. Compliance intentions can also predict the work behaviour of hospital employees in carrying out Standard Operating Procedure (SOP) of medical services (Yami, 2015).
Compliance intentions are at times affected by subjective norms (Yasa & Prayudi, 2019), perceived behaviour control (PBC) (Sarrah et al., 2016), peer reference and family (Nuryana, 2016). However, in previous studies, compliance intentions/intention to comply has never been used as an intervening/mediating variable affected by organisational culture, peer reference and spirituality to predict work behaviour in a modified TPB model.
In this study, the variables of TPB were modified to predict the work behaviour of civil servants (pegawai negeri sipil -PNS) at Indonesia's Ministry of Religious Affairs based on organisational values (Nilai Budaya Kerja -NBK) with the mediation of compliance intention. In 2015, the Ministry of Religious Affairs issued a policy concerning the implementation of five organisational values (5 NBK), which are integrity, professionality, innovation, responsibility and exemplar. As of 2018, efforts to implement 5 NBK at the Ministry of Religious Affairs had not been entirely successful as reported in a study carried out by the Center for Research and Development of Religious Literature and Heritage, and Organizational Management, Agency for Research and Development and Training of the Ministry of Religious Affairs of the Republic of Indonesia (2018). Accordingly, the Ministry of Religious Affairs has continuously made efforts to improve and find the right formula for implementing 5 NBK.
The 5 NBK implementation policy is a part of Indonesia's government bureaucratic reform programme stipulated in the Decree of the Minister of Religious Affairs No. 447/2015 on the Road Map for Bureaucratic Reform of the Ministry of Religious Affairs in 2015-2019. As the implementation of NBK is an institutional policy, it undoubtedly correlates with the ministry's encompassing organisational culture and social influence. In addition, because the policy is implemented at the Ministry of Religious Affairs, the factor of spirituality should be considered as having an effect on the work behaviour of the Ministry's civil servants. This is on account of spirituality being closely associated with religiosity (Amir & Lesmawati, 2016).
Based on the given background, this study attempts to modify TPB by replacing the PBC variable with organisational culture, subjective norm with peer reference and attitude with spirituality in order to predict the NBK-based work behaviour at the Ministry of Religious Affairs with compliance intention as an intervening/mediating variable. Studies conducted by Ampofo et al. (2011) and Razi et al. (2014) corroborate that PBC can be replaced by organisational culture. Hu et al. (2012) also indicate that organisational culture is able to predict compliance intention.
Subsequently, subjective norm can be replaced by peer reference (Trisna, 2019). Foltz et al. (2016) show that peer reference can serve as a predictor of behavioural intention. Meanwhile, Mustikasari (2007), Yasa and Prayudi (2019) indicate that peer reference has an effect on compliance intention. Kuo (2009), Fang (2010 and Guerin et al. (2018) show the influence of peer reference on work behaviour. Furthermore, the attitude variable can also be replaced by spirituality. Widuri et al. (2017), Yasa and Prayudi (2019) show that the spirituality variable in modified TPB can predict tax compliance intention. Afif and Astuti (2009) demonstrate that the spiritual dimension is one of the factors that influence consumer's behavioural intention. Ahmed et al. (2019) found that Islamic work ethic-based spiritual values have positive and significant effect on work behaviour. Meanwhile, Coetzer et al. (2008) developed a non-theistic spirituality model that has positive influence on workplace behaviour.

Research purpose, objectives and hypotheses
The significance of this study is to corroborate the results of modified TPB by replacing a number of variables, that is PBC with organisational culture, subjective norm with peer reference and attitude with spirituality in order to predict organisational values (NBK)-based work behaviour at the Indonesian Ministry of Religious Affairs with compliance intentions as the intervening/mediating variable.
Organisational culture contains common values, norms and beliefs that serve as the core characteristics of how things are performed in the organisation. Those values, norms and beliefs function as a guide for all personnel in the organisation (Mukarom & Laksana, 2016;Wibowo, 2018). Accordingly, a strong organisational culture makes it possible to increasingly drive intention (Mahal, 2009;Schein, 1990), including intention to comply with all organisational policies and rules. Several studies have verified the correlation between organisational culture and individual's compliance intention in an organisation. As an example, Hu et al. (2012) have proven the positive and significant influence that organisational culture has as a key predictor of employee compliance with company's information security policy. Based on such study, the first hypothesis relating to the correlation between organisational culture and compliance intention to implement the NBK policy at the Ministry of Religious Affairs is developed as follows: Hypothesis 1: Organisational culture positively affects compliance intention to implement NBK. This implies that the stronger the organisational culture, the greater the compliance intention to implement NBK.
Hypothesis 2: Peer reference positively affects compliance intention to implement NBK. This implies that the greater the influence of peer reference, the greater the compliance intention to implement NBK.
Hypothesis 3: Spirituality positively affects compliance intention to implement NBK. This implies that the greater the influence of spirituality, the greater the compliance intention to implement NBK.
Hypothesis 4: Organisational culture positively affects NBK-based work behaviour. This implies that the stronger the organisational culture, the better the NBK-based work behaviour.
Hypothesis 5: Peer reference positively affects NBK-based work behaviour. This implies that the greater the influence of peer reference, the better theNBK-based work behaviour.
Hypothesis 6: Spirituality positively affects NBK-based work behaviour. This implies that the higher the spirituality level, the better the NBK-based work behaviour.
Hypothesis 7: Compliance intention to implement NBK positively affects NBK-based work behaviour. This implies that the greater the compliance intention to implement NBK, the better theNBKbased work behaviour.
Based on the research hypotheses elaborated in the previous passages, the relationships between variables can be illustrated in the following model, as shown in Figure 1.

Theory of planned behaviour
There are three key concepts in the TPB: attitude towards behaviour, subjective norm and PBC (Ajzen & Fishbein, 1970). Attitude is an internal state that affects individual action choices to accept or reject a particular object, person or event. Attitude is learned behavioural, cognitive, affective inclinations for responding either positively or negatively towards an object, situation, intuition, concept or person. Attitude is acquired from one's belief of consequences brought about by behaviour, which is defined as behavioural beliefs (Ajzen, 2002).
Subjective norm is an individual's perception or point of view of others' beliefs that will influence one's intention to act on the considered behaviour or otherwise. Subjective norm describes the extent that a person is motivated to follow the perspective of others in relation to the behaviour he or she will perform (normative belief). Ajzen and Fishbein (1970) use the term 'motivation to comply' to describe this phenomenon, which l refer to whether an individual complies with the perspective of other people who are influential in their life or not. Normative belief refers to expectations coming from referents or persons/ groups considered influential to the individuals (significant others) such as parents, spouse/partner, close friends, work colleagues or others depending on the behaviour involved (Ajzen, 2005).
Perceived behaviour control is a new construct incorporated in TPB for controlling behaviour. Perceived behavioural control is defined as the ease or difficulty of performing a behaviour (Ajzen, 1991). According to TPB, intention is a cognitive representation of one's preparedness to engage in a particular behaviour or action (Ajzen, 2005). This means that one's behavior or action can be realized if she or he has intention to engage in a behaviour/action. There are numerous factors that may disrupt the correlation between intention and behaviour. The manifestation of a behaviour depends on one's ability to control factors that may influence the said behaviour, either to facilitate or impede its execution.
As a theory that has been applied in various extensive fields, the three exogenous variables of TPB (attitude towards behaviour, subjective norm and PBC) have generally been proven to be capable of predicting intention and behaviour (Ajzen & Fishbein, 1970;Ozer & Yilmaz, 2011;Sarrah et  Notes: Hypotheses 8, 9 and 10 are indirect effects that cannot be shown in the given figure but were directly tested using data analysis.
http://www.sajhrm.co.za Open Access 2016). As a result of the broad application of TPB in predicting various behaviour phenomenon, there have been many researchers who made TPB modifications. This is also the case in the present article, TPB's three main constructs have been modified and replaced by other variables. The PBC variable has been replaced by organisational culture; subjective norm has been replaced by peer reference and attitude by spirituality. As for intention, it is focused on compliance intention. These modifications neither undermine nor alter the substance of TPB.

Modifying perceived behavioural control by using organisational culture
Perceived behavioural control refers to external influences that can affect intention and behaviour based on one's perception of ease or difficulty in performing the behaviour (Ajzen, 1991). In this case, the position of PBC can be replaced by organisational culture as it contains common values, norms and beliefs that function as the core characteristics as to how things should be carried out in the organisation. The five organisational values (Nilai Budaya Kerja -NBK) are a part of the organisational culture developed by Indonesia's Ministry of Religious Affairs. Hence, incorporating organisational culture as a replacement for PBC is considered correctly in line with the objective of this study.
Organisational culture refers to values, principles, traditions and attitudes that influence the behaviour of organisational members (Robbins & Coulter, 2012). To understand whether the Ministry of Religious Affairs has a well-functioning organisational culture, the current research used indicators referring to the seven dimensions of organisational culture proposed by Robbins and Coulter (2012), namely innovation and risk taking, outcome orientation, people orientation, team orientation, aggressiveness and stability. Innovation is not used as an organisational culture indicator in this case because it is included as an NBK-based work behaviour indicator.
Organisational culture has an influence on compliance intentions, be it compliance with rules, state laws and policies within an organisation or company. Studies on the influence of organisational culture on compliance intention in the TPB model are generally used to predict the intention to stay or leave an organisation, such as turnover intention (Omeluzor, 2018), intention to leave (Leisanyane & Khaola, 2013;MacIntosh & Doherty, 2010), withdrawal intention (Carmeli, 2005) and intention to stay (Abualrub & Nasrallah, 2017).
Other studies put more efforts in associating organisational culture with a particular compliance, such as compliance with code of medical ethics (Brønstad, 2011), tax compliance (Mollanazari et al., 2016), compliance with financial administration rules (Yanto et al., 2017) and compliance with companies' information security policies (Hu et al., 2012).

Modifying subjective norm by using peer reference
Subjective norm refers to the extent of one's perception to follow the perspectives of others concerning the behaviour they will perform (normative belief). Such normative belief deals with expectations originating from referents or persons/groups (significant others) that are influential to the individual such as parents, spouse/partner, close friends, work colleagues, superiors or others depending on the relevant behaviour (Ajzen, 2005). Accordingly, subjective norm can be replaced by peer reference. In line with the objective of this research, peer reference is defined as work colleagues and superiors/leaders who serve as a good example or role model for the respondent.
Peer reference denotes individuals or group of individuals who can be considered role models or patrons when engaging in a particular behaviour (Vitoria et al., 2009 as cited in Trisna, 2019). Peer reference has direct (face to face) or indirect effects on one's behaviour and attitude (Kotler & Keller, 2009 Previous studies have confirmed the correlation between peer reference and compliance intention. Generally speaking, those studies analysed the correlation between peer reference and compliance intention in the context of tax and zakat (Mustikasari, 2007;Nuryana, 2016;Yasa & Prayudi, 2019). Another study shows the influence of peer reference on work behaviour (Fang, 2010). Peers influence employee behaviour in using new technology (Morris et al., 2005). Kuo (2009) found that deviant behaviour at the work place and breach of contract being triggered by influence of peers at work.

Modifying attitude by using spirituality
Attitude is defined as an internal state that affects an individual's choice of action concerning a particular object, person or occurrence. An individual's attitude towards a certain behaviour is obtained from a belief in the consequences brought about by said behaviour, which is termed behavioural belief (Ajzen, 2002). A person who believes that a certain behaviour can have positive outcome will have a positive attitude and vice versa. This is why attitude can be replaced by spirituality. The reason is that spirituality also refers to one's internal state, which contains a series of motivational traits, common emotional powers and non-specific affective power that encourage, direct and select various behaviours (Piedmont, 2001).
This article is inclined towards the perspective that views spirituality as part of religion or religiosity (Love, 2002;Selvam, 2013). To measure the indicator of spirituality, we used the Daily Spiritual Experience Scale (DSES) developed by Underwood and Teresi (2002). Daily Spiritual Experience Scale has been used as a measuring tool for spirituality in over 200 research publications, and it has been translated into 40 languages (Underwood, 2011).
Previous studies frequently used the spirituality variable to modify TPB for predicting tax compliance intention (Elkins et al., 1988;Widuri et al., 2017;Yasa & Prayudi, 2019). Spirituality has been known to affect work behaviour.  found that spirituality has a significant effect in increasing employees' proactive work behaviour. Mahmud and Yusof (2018) used spirituality as a predictor of work safety behaviour. Spirituality also has an influence on ethical behaviour (Pio et al., 2015), pro-environment work behaviour (Afsar et al., 2016) and customer-oriented work behaviour (Moghaddampour & Karimian, 2013

Research methodology Population and sampling
The research population consisted of Civil Servants (Pegawai Negeri Sipil -PNS) with echelon III rank working in 34

Data collection procedures
Data collection was performed using a questionnaire. As a result of the wide distribution of research subjects (respondents), the questionnaire was given in google form format so that it could be uploaded online. The questions in the questionnaire were close ended. The data collection process was conducted as follows. First, the questionnaire was made in google form format then uploaded on the internet. Second, the pre-determined respondents were asked to fill in the questionnaire, and after completing the questionnaire correctly, it will be automatically sent to the researcher's data base. Third, a filter was put in the questionnaire as control, so the respondents were filtered according to the researcher's requirement. Fourth, the researcher closed access to the questionnaire once the submitted questionnaires have met the expected amount.

Method of analysis
As the research model is considered as a rather complex multivariate analysis, structural equation modelling (SEM) was used to analyse the data. The main step required in this process is conducting testing of the measurement model in order to ensure that the instruments have satisfied validity and reliability requirements. The confirmatory factor analysis (CFA) approach was used to test instrument validity and reliability. The subsequent step was to test the structural model. Once the model was deemed fit, the next step was the hypothesis test. The hypotheses that claimed influence were tested using 't-test'. In accordance with most hypothesis testing, a standard significance level of 5% (=0.05) was used in the process.

Direct influence of organisational culture, peer reference and spirituality in compliance intentions to implement organisational culture
As mentioned here, the research samples totaled 216 respondents consisting of Heads of Administrative Division (Kabag TU), Heads of Division (Kabid) and Heads of Section (Kasi) at 34 Regional Offices (Kanwil) of the Ministry of Religious Affairs. After the selection and verification of submitted questionnaires, as many as 212 samples had met the required criteria as shown in Table 1.
Once data were collected, SEM was used for data analysis. The steps carried out in applying SEM analysis included two stages. First, testing of measurement model by using CFA. Second, testing of the structural model. By reason of simplification, confirmatory testing in this study was performed simultaneously for all the concepts.
To verify whether the empirical (statistical) model or statistics covariance matrix developed by the researchers is similar to or fitting with the theoretical (parameter) model or parameter covariance matrix, five elements or goodness-of-fit index were used, which are chi-square, significance, the goodnessof-fit index (GFI), adjusted goodness of fit index (AGFI) and the root mean square error of approximation (RMSEA). Results of confirmatory testing after modifying some of the indicators had indicated a result of fit or meeting the required criteria. This means that the empirical model developed in the study is deemed compatible with the theoretical model. Confirmatory factor analysis results are provided in Table 2.
Subsequently, instrument validity and reliability testing were carried out with a CFA approach using factor loading (lambda = λ) value with significance value of 0.05. If the significance value produced from the testing is < 0.05, then it can be concluded that the tested questions are valid (Hair et al., 2006;Sekaran, 2006;Sekaran & Bougie, 2017 The next analysis refers to the structural model test, which is preceded by a goodness-of-fit test. In the first stage, the fit test result between the empirical model and the theoretical model indicates that some parts of the model did not fit. Accordingly, the model had to be modified. Once modifications were applied on the structural model by eliminating or correlating indicators or items that have high correlations in one construct, the acquired results are shown in Figure 2: Given that SEM had deemed the model to be of good fit, the following analysis could be conducted, that is hypothesis testing for direct effect, indirect effect and total effect coefficients. Considering that the analysed data are an interval scale, analyses of testing results used standardised values, namely values of standardised direct effect, standardised indirect effect and standardised total effect. The results of analyses are summarised in Table 3. Table 3 shows that each independent variable (organisational culture, peer reference and spirituality) partially has significant direct effect on compliance intention to implement NBK. This is verified with the significance values produced through testing; each variable has a significance level of < 0.05. Nonetheless, the three variables partially have indirect effect on NBK-based work behaviour. This is corroborated by the significance level produced in the testing being > 0.05. Meanwhile, the compliance intention to implement NBK variable has direct effect on NBK-based work behaviour with a significance level < 0.05.

Intention as a work culture determinant at the ministry of religious affairs in Indonesia
According to the test results, the effect of X1 on Y1 has been proven significant as (p) = 0.050 (equal to 0.05). Therefore, hypothesis 1 is accepted. That is, organisational culture positively affects compliance intention to implement NBK. The better the organisational culture, the greater the compliance intention to implement NBK. This also applies vice versa. The result indicates that organisational culture has been proven effective as a predictor of intention. Organisational culture is considered an external influence/ situational influence, as a modification of the PBC variable, capable of predicting compliance intention to implement NBK at the Regional Offices of the Ministry of Religious Affairs. The result of this study is relevant with previous studies that consider organisational culture as a predictor of compliance intention. For instance, Hu et al. (2012) corroborated the positive and significant effect of organisational culture as a key predictor of employee's compliance intention with company information security policy. Yanto et al. (2017) employed the organisational culture variable as a predictor of small medium enterprise (SME) manager's compliance intention in making reports according to the Accounting Standards without Public Accountability (Standar Akuntansi Tanpa Akuntabilitas Publik).
Subsequently, based on test results, the effect of X2 on Y1 has been proven significant as (p) = 0.036 (less than 0.05). Therefore, hypothesis 2 is accepted. That is, peer reference positively affects compliance intention to implement NBK. The greater the influence of peer reference, the greater the compliance intention to implement NBK and vice versa. This finding indicates that peer reference has been proven effective as a predictor of intention. Peer reference is considered a social influence, as a modification of the subjective norm variable, capable of predicting compliance intention to implement NBK at Regional Offices of the Ministry of Religious Affairs.  This finding is relevant with several past studies associating peer reference with compliance intention (Haryadi & Wiyono, 2014;Suwarto, 2016;Yusuf, 2006). Not many studies used peer reference as a variable to modify TPB for predicting compliance intention and work behaviour. Peer reference is commonly included as one of the indicators of subjective norm instead of replacing subjective norm (Mustikasari, 2007;Yasa & Prayudi, 2019).
The role that leaders play as an exemplary model also has a substantial influence on employee's compliance intention to work particularly in government institutions (Riyaniningsih & Triyaningsih, 2010). This is in line with the finding of this research, which shows that respondents perceive that peer reference (superiors: echelon I and II officials; co-workers: fellow echelon III officials) supports NBK implementation at the Ministry of Religious Affairs. Peer reference even suggests, urges and instructs employees to implement NBK. This, consequently, instigates respondent's belief, which triggers compliance intention to implement NBK as it is supported by echelon I, II and III officials serving as their points of reference.
Subsequently, based on test results, the effect of X3 on Y1 has been proven significant as the significance level is less than 0.05. Therefore, hypothesis 3 is accepted, that is spirituality positively affects compliance intention to implement NBK. The higher a person's level of spirituality, the greater the compliance intent to implement NBK and vice versa. This finding demonstrates that spirituality has been proven effective as a predictor of intention. Spirituality is considered an internal influence (personal influence), as a modification of the attitude variable, capable of predicting compliance intention to implement NBK at Regional Offices of the Ministry of Religious Affairs. This implies that positive attitude towards spirituality will bring about compliance intention to implement NBK as it is aligned with values of spirituality and teachings of Islam adhered by the majority of the respondents.
The finding of this study shows that respondents have a high level of compliance intention to implement NBK because they are driven by the belief that God constantly watches over human behaviour and the belief that religion is a source of power in terms of work. Based on the hypotheses testing results, it is known that spirituality partially has the most   significant direct effect on compliance intention to implement NBK compared with the other two independent variables (organisational culture and peer reference). This indicates that spirituality is the most influential factor in compliance intention to implement NBK at the Ministry of Religious Affairs.
The respondents' spirituality as a behavioural belief led to actual work behaviour based on NBK. As stated by Ajzen and Fishbein (1970), Ajzen (2002), behavioural beliefs connect behaviour with a certain outcome that is achieved when a certain behaviour is performed. In other words, a person who believes that a behaviour can generate positive outcome will have a positive attitude and vice versa. In the context of spirituality, positive outcome does not refer to material or financial gains, it is more of a psychological benefit such as inner peace or religious benefit such as belief in receiving God's affection and gift. This finding is relevant with a study by Passos et al. (2013), which highlights the influence factors that organisational culture has as a subjective norm to predict the behaviour of developing software technology using behaviour intention as an intervening variable in the TPB model. However, Passos et al. (2013) placed organisational culture as a subjective norm, while this study used organisational culture to replace PBC. Other studies show that intention has a significant effect as an intervening variable between attitude and behaviour (Tan et al., 2018), attitude and decision to purchase products (Haikal, 2018), attitude and performance (Başar & Varoğlu, 2016 The sixth hypothesis is also not accepted because the significance level = 0.598 (greater than 0.05). Meanwhile, the indirect effect of X3 on Y2 as mediated by Y1 has been proven significant, it even has the highest score when compared with organisational culture and peer reference. This shows that spirituality does not have direct effect on NBK-based work behaviour, but it does have indirect effect when mediated by compliance intention to implement NBK as an intervening (mediating) variable. The stronger the respondent's spiritual belief, the greater the compliance intention to implement NBK, which further improves NBK-based work behaviour. This finding differs with several prior studies, which found that spirituality affects work behaviour without being mediated by intention as an intervening variable (Ahmed et al., 2019;Choerudin, 2015;Coetzer et al., 2008;Mahmud & Yusof, 2018;Yilmaz & Bahçekapili, 2015).
As for the seventh hypothesis, it is accepted because the significance level is less than 0.05. The greater or higher the compliance intention to implement NBK, the better the NBKbased work behaviour and vice versa. This is in accordance with Ajzen's (2005) theory, which states that intention is a cognitive representation of one's readiness to perform certain behaviour/act. Meanwhile, compliance intention in the organisational context may be defined as a cognitive representation of one's readiness in the form of impetuses that consciously emerge to believe, accept and act on all organisational policies, provisions and rules voluntarily (Darley et al., 2001). Several prior studies have also found compliance intention as a predictor of work behaviour (Afdalia et al., 2014;Hartoni & Riana, 2015;Yami, 2015).
These results indicate that compliance intention to implement NBK plays a role as a mediating variable of influence between the three independent variables (organisational culture, peer reference and spirituality) and NBK-based work behaviour. The magnitude of indirect effect and total effect that the three independent variables have on NBK-based work behaviour is shown in Table 4.
According to the elaboration of the hypotheses given here, it is known that organisational culture, peer reference and spirituality, respectively, affect NBK-based work behaviour through the mediation of compliance intention to implement NBK. This indicates that the three variables have an effect on NBK-based work behaviour when mediated by compliance intention to implement NBK. It is, thus, known that compliance intention in this modified TPB model has a significant role as an intervening variable or mediator. Among the three modified TPB variables, spirituality has the strongest influence on NBKbased work behaviour as mediated by compliance intention.

Conclusion
Results of this study have extended the implementation of the TPB by modifying PBC with organisational culture, subjective norm with peer reference and attitude with spirituality to predict NBK-based work behaviour at the Regional Offices of the Ministry of Religious Affairs with compliance intention as a mediating/intervening variable, which acquired positive and significant results. Theoretically speaking, the findings in this study provide empirical evidence that modifying TPB reinforces TPB's applicability to predict compliance intention and work behaviour. Practically speaking, this research provides recommendations for improving the implementation of the five NBK values to the Ministry of Religious Affairs in three aspects. First, enhance civil servants' (PNS) spirituality by providing substantial understanding of religious values. Second, intensify the role and support of structural leaders, particularly echelon 1, 2 and 3 officials so they provide good examples in the implementation of NBK. Third, strengthen organisational culture at the Ministry of Religious Affairs by creating result/impact oriented programme planning (result/impact oriented government).

Limitations and future research
The present study did not distinguish the respondents in terms of work duration, career path and education level. Future research should consider these three aspects to complement this research model. In addition, subsequent research may be carried out in non-religious institutions to expand the general applicability of the theoretical model proposed in this study.