The positive role of religiosity in dealing with academic dishonesty

Abstract This study examines the role of religious knowledge and activity in dealing with academic dishonesty. Furthermore, we investigate whether the locus of control and personality strengthen the relationship between religious knowledge and activity and academic dishonesty. The study investigates 235 respondents from accounting and medical undergraduate students of Syiah Kuala University, the largest university in Aceh Province, Indonesia. They are prospective accountants, auditors, and doctors. The sample is determined by using a random sampling technique, where all students have the opportunity to fill out the online questionnaire provided through Google form. The questionnaire can be accessed in 2 months, June-August 2019. Respondents are directed to answer 97 questions divided into five specific sections of the variables. Respondents must complete each section before going to the next section and unable to change the previous section’s answers. To test the hypothesis, we employ path analysis. The results show that religious knowledge is the main predictor in dealing with academic dishonesty and the basis for building the character of the students. Locus of control strengthens the effect of religious knowledge on academic dishonesty, but personality does not. Religious knowledge and activity should be synergized in the higher education system and must be well arranged and structured to function well in preparing honest accountants, auditors, and doctors.


PUBLIC INTEREST STATEMENT
Academic dishonesty has become a serious unsolved problem at all levels of education. A person who commits academic fraud during the studies has a higher tendency to behave in an unethical and fraudulent manner at the workplace.
People who have better religious knowledge must become an honest person. However, some people with better religious education are not honorable. The causes are personality and locus of control. The different personalities and locus of control cause different views and perceptions in dealing with something. Some of them are shy and have low self-esteem, while others are aggressive, etc. Things that are not common in someone's view become a customary view for others. Things that make someone embarrass will not necessarily embarrass others.
The results show that religious knowledge would become the main predictor of honest behavior and the basis for building the character of the students depending on their locus of control.

Introduction
Academic dishonesty has become a serious unsolved problem in all levels of education, including higher education, throughout the world (Hadjar, 2017;Ives et al., 2017). The problem of dishonesty in higher education institutions has been widely studied and succeeded in attracting the researcher's attention from various countries (Khan et al., 2019). Several studies prove that more than 50% of students in various universities have cheated at least once in examinations or assignments during their study (Ives et al., 2017;Winardi et al., 2017). Furthermore, various studies have been carried out to find out the cause and effect of a student committing academic fraud and the factors influencing it (Ampuni et al., 2019;Ballantine et al., 2014;Ives et al., 2017). Ampuni et al. (2019) find that lower moral integrity and higher level of moral disengagement affect academic dishonesty among students. Ballantine et al. (2014) show that academic dishonesty is caused by ethical ideology and idealism. Ives et al. (2017) demonstrate that a student cheating because he saw another friend cheating frequently.
The other issues related to academic cheating is religiosity. A high level of religiosity fortifies someone to commit cheating. Al-Ebel et al. (2020) examine the impact of religiosity and accounting expertise on audit report lag and the results show that a religious top leader with accounting expertise avoids audit report lag. However, the other studies show dissimilar results. Williamson and Assadi (2005), Huelsman et al. (2006), and Hadjar (2017) find that religiosity does not have a significant influence on academic cheating behavior. Meanwhile, Sofyani and Rahma (2015) prove that Islamic religious knowledge has a significant influence on the honest behavior of the students. Based on the inconclusive results, this paper intends to examine whether Islamic religious knowledge and Islamic religious activity affect students' academic dishonesty.
Academic cheating is related also to attitude and behavior (Khan et al., 2019). The willingness to cheat depends on someone's personality (Giluk & Postlethwaite, 2015;Hendy & Montargot, 2019) and locus of control (Sofyani & Pramita, 2015). Different from previous studies, we examine the role of locus of control and personality on the effect of Islamic religious knowledge and Islamic religious activity on students' academic dishonesty. Previous studies investigate the effect of religiosity, locus of control, and personality as independent variables on academic dishonesty as dependent variable. This study demonstrates that locus of control and personality are moderating variables in the relationship between religiosity and academic dishonesty. Someone's locus of control and personality can strengthen or weaken the correlation between religiosity and academic dishonesty.
Given the empirical evidence of the impact of religious knowledge and religious activity on academic dishonesty, our study contributes to the development of the literature on cheating behavior in terms of the research model. In contrast to previous studies that test directly the impact of religiosity on dishonesty (Hadjar, 2017;Sofyani & Rahma, 2015); Khan et al., 2019), locus of control on dishonesty (Sofyani & Pramita, 2015;Winardi et al., 2017), and personality on dishonesty (Giluk & Postlethwaite, 2015;Hendy & Biderman, 2019;Hendy & Montargot, 2019), this study employs locus of control and personality as moderating variables which strengthen or weaken the impact of religious knowledge on students' dishonesty. This is the novelty of this study.
Strong logical reasoning why these two factors can moderate the relationship between religiosity and academic dishonesty is a tendency to commit academic dishonesty even though they have good religious knowledge depending on his/her locus of control and personality.
In fact, the attributes inherent in a person are not the only locus of control and personality, but also self-control (Flores et al., 2020;Zhang et al., 2019). Self-control is a person's action of controlling the desire to commit a crime (Cochran et al., 1998). Religious knowledge is a form of thought that is obtained systematically, philosophically, and scientifically from religious education and religious science (Burns, 1914). Higher religious knowledge combined with lower self-control encourages a person to commit crimes (Kerley et al., 2011). Therefore, the relationship between religious knowledge and academic dishonesty is not only influenced by the locus of control and personality but is also influenced by self-control (Kerley et al., 2011). This means that self-control also influences people to commit crimes or cheating even though their religious knowledge is high. This study does not test self-control to avoid a wider relationship among the moderating variables. Personality and locus of control are closely related to self-control (Flores et al., 2020;Zhang et al., 2019). Besides, this study employs the theory of planned behavior. Using the locus of control and personality as moderating variables support the theory of planned behavior (Ajzen, 1991). We suggest to include self-control in a more comprehensive model in future studies.
The second novelty is juxtaposing locus of control and personality concurrently in one model as moderating variables that strengthen or weakens the relationship between religiosity and academic dishonesty. Previous studies have not used these two factors as moderating variables either individually or simultaneously. Employing locus of control and personality as moderating variables simultaneously to get a comprehensive picture of why someone who has a good religious knowledge commits academic dishonesty.

Background
The state of dishonest behavior in academics has become the spotlight all over the world, where it occurs due to the engagement of students in unethical academics behavior (Ballantine et al., 2014). From the 1960s, academic dishonesty had been seen as a trivial problem related to one's academic achievement (Ampuni et al., 2019;Bonjean & McGee, 1965). However, this problem becomes a major issue when researchers found a significant relationship between academic dishonesty and someone's unethical behavior at work (Ives et al., 2017). A person who commits academic fraud during the studies has a higher tendency to behave in an unethical and fraudulent manner at the workplace (Khan et al., 2019;Williams & Williams, 2012). The forms of fraud resulted from the cheating habit during someone's study are corrupting, giving false documents, duplicating documents or reports from others for a similar activity, not being professional at work, etc.
In the world of health, the common fraud in Indonesia according to Minister of Health Regulation No. 36 of 2015(Kementerian Kesehatan RI, 2015, that often occurs is divided into two classifications. First, the act of medical fraud committed by a doctor, such as manipulating diagnoses to increase the size of claims by falsifying medical diagnoses and/or actions; plagiarism of claims from other patients (cloning) by copying and pasting all or part of existing medical records and/or patient data; unnecessary treatment; no medical value; increase the length of time on the ventilator usage. Second, administrative fraud such as false claims (phantom billing) are claims for services that have never been performed/provided to patients; Inflated bills for drugs and/or medical devices (inflated bills); solving episodes of service in accordance with medical indications but not in accordance with statutory provisions; services unbundling or fragmentation; prolonged length of stay. The act of cheating is generally an act that someone feels is reasonable (Rettinger & Jordan, 2005).
An accountant is a profession promoting honesty and moral responsibility, due to the obligation for producing financial information used by other parties in making business decisions. When an accountant does not uphold the value of honesty and moral responsibility, the financial information produced can be misleading and detrimental to the user. Therefore, starting their study as an accountant, the students are taught to raise their honesty value in all aspects including truthfulness in academics. Similarly, a medical doctor as a profession related to human life has also prioritized honesty and high moral responsibility to prevent malpractice that can harm human life. When a doctor does not follow the honesty value and high moral responsibility, they may not report the actual diagnosis, but tend to report the results that benefit himself. Furthermore, the forms of honesty and moral responsibility between accountants and medical doctors are different. However, they promote a similar concept of honesty which is regulated in their professional code of ethics. Besides, accountants and doctors must also have high moral responsibility and professionalism. Therefore, they have to be equipped with religious education and high professional knowledge that can foster feelings, empathy, honesty, and moral responsibility.
People who have better religious knowledge must become an honest person. However, some people with better religious knowledge are not honorable. The causes are personality and locus of control in which the personality and locus of control among people are different. The different personalities and locus of control cause different views and perceptions in dealing with something (Ajzen, 1991). Some of them are shy and have low self-esteem, while others are aggressive, etc. Things that are not common in someone's view become a customary view for others. Things that make someone embarrass will not necessarily humiliate others. Personality differences also encourage someone to have different views about the concept of honesty (Giluk & Postlethwaite, 2015;Hendy & Biderman, 2019). As an example, giving assignments to be copied by other students will be considered as a natural reason for helping friends, but others will view this activity as unnatural and violate the concept of honesty. Thus, this study examines the influence of personality on the relationship between religiosity and academic dishonesty. Also, another factor moderates the effect of religious education and religious activity on academic dishonesty is a locus of control. Investigating these factors, Sofyani and Pramita (2015) find a positive significant correlation between locus of control and students' academic dishonesty. Winardi et al. (2017) agree that personality and locus of control influence ethical behavior. Hendy and Biderman (2019) and Giluk and Postlethwaite (2015) also support the finding that certain types of personality tend to commit fraud or dishonest behavior.
In section 3, we explain the theoretical literature review, Theory of Planned Behavior. In section 4, we discuss empirical literature review and hypothesis development. In this section, we explain the relationship between variables. Section 5 will describe the research design. The empirical results will be presented in section 6, followed by the discussion in section 7. Finally, the conclusions are presented in section 8.

Theory of planned behavior
The phenomenon described in this paper is explained by the Theory of Planned Behavior which is developed by Icek Ajzen (Ajzen, 1985(Ajzen, , 1991. The theory describes how to predict and explain human behavior in specific contexts. Social, personality traits (Ajzen, 1987), and cognitive selfregulation (Ajzen, 1991) help to predict and explain human behavior. The general attitude is influenced by the institutions or organizations that are frequently visited, certain groups and people who frequently interact (Ajzen, 1991). Someone is more likely to adopt a behavior if he has a positive attitude towards the behavior, gets approval from the other one who is close to and related to the behavior, and believes that the behavior can be done well (Ajzen, 1985).
There is an intention-action relationship (Ajzen, 1985). Intention to accomplish a specific behavior is affected by attitudes toward the behavior, subjective norms, and perceived behavioral control (Ajzen, 1991). The intention to do something is not necessarily carried out into real action. The actual behavior is determined by the interaction between the intention and perceived behavioral control (Ajzen, 1991). In the context of cheating, the intention to do cheating is not an action. It is influenced by religious knowledge and religiosity activity as perceived behavioral control. The study examines the effect of perceived behavioral control directly on dishonest behavior (action).
The intention-action relationship can be moderated by psychological characteristics, namely locus of control and personality traits (Ajzen, 1991). The theory is empirically supported by Bernardus et al. (2020). Further, Bernardus et al. (2020) examine which characteristic of psychology strengthen the intention-action relationship and conclude that internal locus of control reinforces the intention-action relationship. Even thought they examine the theory with the other context, the principle of the relationship can be adopted. Giluk and Postlethwaite (2015) and Hendy and Montargot (2019) examine the impact of personality traits on dishonesty by implementing the Theory of Planned Behavior. The results show that individuals personality affect students' cheating behavior.

Institutional setting
As an institution organizing higher education and generating prospective accountants and medical doctors, a university is expecting to formulate strategies that can foster the honesty and professionalism of the students (Van Dyk et al., 2020). It is also the place where the formation of character, morals, and ethics are continuously being taught at an in-depth level (Asif et al., 2020;Juanda & Sofyani, 2016). This circumstance is reflected in the curriculum containing religious content, which are religious education, religious activities, and ethics education (Asif et al., 2020;Juanda & Sofyani, 2016). The goal is to shape the character and morals of students to become honest and noble students with the curriculum designed in such a way to contain courses that can shape and develop the nature of honesty (Asif et al., 2020;Juanda & Sofyani, 2016). In most state universities in Indonesia, Religious Education is included in their curriculum as a general course taught in 2 nd semester for both theoretical and practical methods. While, universities with a Moslem majority, the Islamic religious education approach is used, in which the learning materials are divided into 3 major groups: Tawhid, Aqedah, and Akhlak (Moral) (Akko, 2018).
This study examines the effect of religious knowledge and religious activity on academic dishonesty with the locus of control and personality as the moderating variables. This study is expected to contribute to higher education institutions in Indonesia regarding the importance of religious education and activities to be included in the curriculum and provide a material consideration in developing learning curricula in higher education that can reduce dishonest behavior among students.

Religious knowledge, religious activity and academic honesty
Religious knowledge is a form of thought that is obtained systematically, philosophically, and scientifically from religious education and religious science (Burns, 1914). Religious knowledge is divine teaching containing values that are universal for human life, which becomes a perfect guide in solving all human life problems (Akko, 2018). The results of the divine teaching shape one's attitude, mindset, and outlook. From the Islamic perspective, there are three important points in the religious teachings referred to as religious characters, which are: divinity reflected in one's faith, obedience in carrying out what is ordered and leaving out what is prohibited, and doing good deeds and far from the nature of treason (Akko, 2018). These points shape a person's behavior and knowledge as a whole which requires serious attention and a long and well-structured process. Therefore, the formation of these religious characters needs to be well planned and programmed in the long-term to form an obedient, honest, and responsible person (Sofyani & Rahma, 2015).
Related to students' attitudes toward academic honesty, religious values formed from the participation in religious education and activities can be factors influencing these individuals in performing academic honesty. This process refers to the planned behavior theory developed by Ajzen (1991), in which the intention is the basis of a person's behavior and it is influenced by attitudes and behavior. The attitude in behaving itself is a function of belief in behaves. This belief may be shaped by religious values embedded in individuals' religiosity. Thus, religious values are factors that influence individuals in taking actions.
An experimental study by Williamson and Assadi (2005) using 65 samples of undergraduate students conclude that religious orientation is not related to the tendency of fraudulent behavior in participants. Similar research by Rettinger and Jordan (2005) with the students from business and liberal art that has religious courses shows that religiosity influences the academic cheating behavior of students, where business students are known to commit more fraud than liberal arts students. The result is consistent with Khan et al. (2019) who show that religiosity but not spirituality is a factor of the mindset of university students toward cheating and cheating behavior.
Empirical research related to academic fraud is conducted by Hadjar (2017) using 239 students from the Islamic religious education study programs and study programs of da'wah from one of the leading universities in Indonesia. The results present that 98% of the respondents claim to involve in one or several forms of academic cheating, which conclude that religion (belief and behavior) does not influence the fraudulent behavior of students. Other quantitative studies are carried out at several universities in Malaysia by Mustapha et al. (2016) examining the relationship between Islamic religion and intention of academic fraud among Malaysian Muslim students. The results from 221 respondents found no significant relationship between the Islamic religion and student academic cheating behavior.
From prior literature and mixed research findings, it argues that the influence of religious education on academic dishonesty must be extended to find a robust pattern. This study also expects that religious activity held to support and accelerate the objective of formal religious education influences the academic dishonest behavior of students. Hence, the proposed hypotheses in this study are: H1: Religious knowledge has a positive role in dealing with students' dishonest behavior.

H2:
Religious activities negatively affect the academically dishonest behavior of prospective accountants/auditors and doctors.

The effect of locus of control on the relationship between religious knowledge, religious activity and academic dishonesty of prospective accountants/auditors and doctors
The concept of locus of control is coined by Rotter, a social learning theorist in 1966. Locus of control is a psychological concept that refers to how strongly people believe that they have control over the situations and affect their life (Rotter, 1966). Someone with an internal locus of control believes that the success he achieves is due to his hard work and the failures he suffers are due to his own mistakes. Conversely, someone with an external locus of control believes that the success he achieves is due to his abilities, but his failure is caused by other external factors, like fate, luck, destiny (Rotter, 1966). Someone who believes that events and destiny are caused by control over himself is called an internal locus of control, while things are caused by outside factors called an external locus of control (Robbins & Judge, 2012).
In education, locus of control typically refers to how students perceive the causes of their academic success or failure in school. Students with an internal locus of control generally believe that their success or failure is a consequence of their effort and hard work. Adversely, students with an external locus of control generally believe that their successes or failures outcome of external factors beyond their control, such as luck, fate, circumstance, injustice, bias, or teachers who are unfair, prejudiced, or unskilled. Students with an internal locus of control might blame poor grades on their failure to study, whereas students with an external locus of control may blame an unfair teacher or test for their poor performance (Rotter, 1966).
The influence of religious knowledge and religious activities on academic dishonesty depends on which category of locus of control that students have. Students who tend to an internal locus of control consider that religious education and activities encourage students to be honest when they are completed with confidence and self-motivation (Sofyani & Pramita, 2015). They argue that religious forces control the honesty of prospective accountants and doctors. Thus, it concludes that external locus of control is relying more on someone's hopes to depend on others and more seeking and choosing favorable situations. While the internal locus of control is relying more on someone's expectations of themselves and prefer skills rather than just a favorable situation (Winardi et al., 2017).

H3:
Locus of control moderates the influence of religious knowledge on the academically dishonest behavior of prospective accountants/auditors and doctors.
H4: Locus of control moderates the influence of religious activities on the academically dishonest behavior of prospective accountants/auditors and doctors.

The impact of personality on relationship between religious education, religious activity and academic dishonesty of prospective accountants/auditors and doctors
One issue that has received much attention regarding the rise of academic cheating at various levels of education is personality. Various studies have shown that certain personality traits have a relationship with the tendency to cheat (Giluk & Postlethwaite, 2015;Hendy & Biderman, 2019;Hendy & Montargot, 2019). The personality itself is divided into five types based on the Big Five Trait model, which are: Extraversion, Agreeableness, Conscientiousness, Emotional Stability, and Openness to Experience.
Extraversion is a type of personality where individuals tend to enjoy being in social situations. They are characterized by friendliness, warmth, full of positive vigor, and have high vitality (Giluk & Postlethwaite, 2015). De Bruin and Rudnick (2007) define someone who has a high score in extraversion has a higher tendency to cheat in the test because he tends to be excitementseeking leading him to become a risk-taker. Agreeableness is a type of personality where individuals have higher social concerns and tend to avoid problems with others (Graziano et al., 1996). These individuals tend to have less involvement in academic cheating practices because they avoid conflicts with teachers (Giluk & Postlethwaite, 2015). Besides, conscientiousness is a personality type that drives someone to be full of planning and following the existing rules and norms. This type has a negative relationship with academic cheating (De Bruin & Rudnick, 2007;Muntada, 2013). Openness to Experience is a type of personality where individuals are full of thoughts and like new experiences. These individuals tend to prefer being less involved in academic cheating. Furthermore, neuroticism is a type of personality where individuals have emotional instability; easy to get angry, stressed, depressed, mood-swing, and low self-confident (Costa & McCrae, 1992). Previous researches have associated this type with academic dishonesty because neurotic individuals will prefer to cheat to get success (Barrick et al., 2013).
H5: Personality moderates the influence of religious knowledge on the academically dishonest behavior of prospective accountants/auditors and doctors.
H6: Personality moderates the influence of religious activity on the academically dishonest behavior of prospective accountants/auditors and doctors.

Sample and procedure
The Samples are active students between the second and the fourth year at both Faculty of Economics and Business and Faculty of Medical, Universitas Syiah Kuala, Indonesia. The secondyear students are chosen because they have completed the religious activity (UP3AI activities) which become the indicators to measure respondent's participation in religious activities. The sample is determined by using a random sampling technique, where all students have the opportunity to fill out the online questionnaire provided through Google form. This application is user-friendliness and can be shared with a wider community to obtain a large number of samples in a short time. It also provides a report describing the responses in a simple form and the collected data can be exported to Excel. The questionnaires can be accessed by all students in 2 months, from Juni until August 10 th , 2019.
The respondents voluntarily fill questionnaires including personal information such as initial or name, major and faculty, grade, CGPA, and gender (see Table 2). This information is necessary because they are the predictors of academic dishonesty (Ives et al., 2017). Moreover, respondents are directed to answer 97 questions divided into five specific sections of the variables; 16 questions about academic dishonesty, 50 questions about personality, 12 questions about religiosity, 6 questions about religious activity, and 13 questions about the locus of control. The questions represent the indicators of each variables as shown in Table 1. Respondents must complete each section before going to the next section and unable to change the previous section's answers.

Data collection
Data are collected using a questionnaire designed based on previous research and adapted to the conditions of students in a state university. Questionnaires for academic dishonesty, religious education, religious activity, and personality variables are built using a 5-Likert scale with different descriptions, while the locus of control variable used closed questions where respondents have to This study involved 235 respondents, which are 169 female and 66 male respondents (see Table  2). Responses from respondents are then processed by using Eviews and the validity of the model is tested by using a diagnostic test (normality test, heteroscedasticity test, autocorrelation test, and multicollinearity test) before the regression test is performed.

Academic dishonesty
Academic dishonesty is measured by indicators developed by (; Lin & Wen, 2007;McCabe & Trevino, 1993;Sajid Nazir & Shakeel Aslam, 2010), 16 questions ask how often the respondent cheats during exams, on assignments, plagiarism, and other academic cheats. The 5-Likert scale is used to present the answers of respondents starting from the lowest value which is never (1), to the highest value of always (5). (never, once, rarely, several times, always).

Religious knowledge
The instrument is developed from the Religious Orientation Scale revised (Darvyri et al., 2014), 12 items of statements that have been adapted to Indonesian students are used to determine the level of student religiosity. The following are examples of adjusted statements: I like reading newspapers & its translations, as well as other religious books; I am sure there will be another life in the hereafter; and so on. The 5-Likert scale is used to interpret the responses starting from the lowest 1 strongly disagree to 5 strongly agree.

Religious activity
There are 6 questions to determine students' participation in a mandatory religious activity held by UP3AI (Assistance Development Program Unit for Islamic Religious Subject) and other religious activities outside the campus, such as reciting Holy Quran, etc as Juanda and Sofyani (2016).

Locus of control
The indicator is developed from Sofyani and Pramita (2015) using the categories of external and internal locus of control. According to Murphy and Mayhew (2012) external locus of control is someone who tends to commit ethical violations in the profession compared to an internal locus of control. This is because they depend on the fate of the boss's attitude, while the internal locus of control has more control over self-attitude and is responsible for all good and bad actions taken. To determine the locus of control category, the number of participants' answers is used to see their position from the median. When the value obtained is higher than the median, it is classified as an external locus of control and vice versa consequently.

Personality
The indicator used in this study similar to those of Hendy & Biderman (2019). The indicator is developed based on Goldberg (1992). Personality is seen as one of the important factors that influence one's behavior. 5 types of personality reflected in someone's behavior are Extraversion, Agreeableness, Conscientiousness, emotional stability, and Openness to experience. Each type is measured by 10 questions. For instance, to reflect the type of extraversion, the behavior that appears is someone feels comfortable in acrowd, feels comfortable despite being the center of attention, someone who is friendly, has the idea to start aconversation, not rigid, flexible, etc. Respondents were asked to answer each statement that reflects their true nature, not how ideal behavior they want. The 5-Likert scale is used to interpret the responses starting from the lowest 1 strongly disagree to 5 strongly agree.

Data analysis technique
The relationship among variables is described in the research model as shown in Figure 1. Hypothesis testing is conducted by using path analysis with the following formula.
Where AD i is academic dishonesty of student i; RK i is religious knowledge of student i; RA i is the religious activity of student i; PS i is the personality of student i; LoC i is the locus of control of student i; RK i *PS i is the interaction between religious knowledge and personality; RA i *PS i is the interaction between religious activity and personality; RK i *LoC i is the interaction between religious knowledge and locus of control; RA i *LoC i is the interaction between religious activity and locus of control, and ε i is the error term.

Diagnostic test
Classic Assumption Tests perform to examine the validity of the model before running the regression test. The model meets the rule of thumb criteria with a sample size of 235 and is free from heteroscedasticity problems as indicated by the Chi-Square of Obs*R-squared that is equal to 0.2677, where this value is greater than the significance level, 0.05. From multicollinearity test, the results show that there are no intercorrelations or multicollinearity problems indicated by the Centered VIF value that is smaller than 10. Finally, for the autocorrelation test, the regression model shows no positive autocorrelation problems with the value of d (2.223) are greater than dU (1.819) and no negative autocorrelation problem with the value of d (1.978) is greater than dU (1.819).

Respondent characteristics
After almost two months collecting responses from 235 students (206 accounting students and 29 medical students), the results show that the respondents are dominated by female students (71% accounting students and 76% medical students) and mostly studying in the third semester (43% accounting students and 72% medical students). We also obtain information about the CGPA of each respondent, where accounting students have CGPA around 3.26-3.50 and medical doctor students around 3.00-3.25 (see Table 2).
Based on the data as illustrated in Table 3, the top three academic dishonesty are: 1) sharing answers to other students during the examination (92% accounting students and 86% medical students committed it at least once in one semester); 2) duplicating answers from other student's homework (82% accounting students and 83% medical students committed it at least once in one semester); 3) sharing finished individual homework to be duplicated by other students (87% accounting students and 75% medical students committed it at least once in one semester).
Also, other interesting findings, 62% accounting students and 20% medical students use electronics/gadgets during examination with different frequency categories. Besides, 66% accounting students and 93% medical students ask the questions' clues to other students who have passed the examination, and 53% accounting students and 86% medical students share the questions' clues about examination to other students who will take it.
Other findings show that 52% accounting students and 34% medical students have copied other students' answer during examination with and without permission at least once during one semester, and 50% accounting students and 17% medical students have used cheat sheet during the examination at least once during one semester.
In terms of religious education, accounting and medical students have a good understanding of what they have learned. 99% accounting students and 97% medical students believe that God always observes everything, 78% accounting students and 69% medical students like to recite Holy Quran, 96% accounting students and 93% medical students believe that everything happens to them is their destiny, 83% accounting students and 90% medical students like to hear a religious lecture, 52% accounting students and 59% medical students will not litter when they do not find any trash can (see Table 4).
In terms of religious activity, less than 35% of accounting and medical students attended all Iqra class meetings, mentoring class, and halaqah during one semester (see Table 5).
Another finding shows 97% accounting students and 100% medical students recited Holy Quran at least once a day during the past year, 21% accounting students and 31% medical students never recited Holy Quran in a day during the past year, the rest recited Holy Quran at least once in Table 3

Table 4. Religious education
No. a day, and 97% accounting students and 100% medical students agree that understanding interpretation of Holy Quran is important (see Table 5).

Frequency
For the locus of control, 60% of accounting students and 52% of medical students have an external locus of control, while the rest of them have an internal locus of control (see Table 6).
Special case for the personality, it is difficult to identify the specific personality traits of each respondent due to the questionnaire issue, in which respondents have to fill all the questions that represent all traits of personality at once. This type of questionnaire is built according to the theory that explains, at the very basic, each individual has all traits of personality, but there will be only one dominant trait in each individual. Consequently, the responses given by respondents vary and thus disable for us to exactly identify each personality. Therefore, this study examines the influence of personality in general without specifically examining the influence of each type of personality. Table 7 illustrates the regression analysis examining the influence of religious knowledge, religious activity, locus of control, and personality on academic dishonesty. Religious knowledge (c = −0.1920, p = 0.0062) has a negatively significant impact on academic dishonesty. Higher religious knowledge, lower the academic dishonesty. The results indicate that religious knowledge has a positive role to reduce the intention to cheat. The result supports hypothesis 1. Other independent variables: religious activity (c = 0.0459, p = 0.1773), locus of control (c = 0.0787, p = 0.6196), and personality (c = −0.0678, p = 0.4821) show no influence on academic dishonesty. Besides, the results show both locus of control and personality as the moderating variables are not the independent variables influencing the academic dishonesty.

Hypothesis testing
Further examination on the role of locus of control and personality as the moderating variables in the influence of religious knowledge and religious activity on academic dishonesty illustrates in   Furthermore, Table 8 shows that personality (C = 1.5150, p = 0.2657) fails to moderate the impact of religious knowledge (C = −0.3725, p = 0.2331) and religious activity on academic dishonesty (C = −0.0008, p = 0.9942). While the relationship between religious activity and academic dishonesty remains the same in the second model. This finding triggers us to do further analysis to employ religious knowledge as the moderating variable between religious activity and academic dishonesty (see Tables 7 and Tables 8). Table 9 shows that without the intervention of other variables, the relationship between religious activity (C = 0.0393, p = 0.2367) and academic dishonesty remains unsupported, meanwhile religious knowledge (C = −0.2018, p = 0.0034) is found as the predictor of academic dishonesty of students. Further examination, we employ religious knowledge as the moderating variable, as shown in Table 10, where the influence of religious activity on academic dishonesty is identified (C = −1.264 p = 0.0004) significantly after the involvement of religious knowledge as the moderating variable. These results indicate that religious activities organized by universities discourage students from committing academic fraud if students have good enough religious knowledge. These results are very important for universities in developing religious education curricula in universities, especially universities that provide religious studies for all students. The first step that must be considered by universities is to increase students' understanding of Tawhid, Aqedah, and Akhlak (Moral), the next step to increase their religious activities to shape good behavior, attitudes, ethics, and honesty.

Discussion
This study examines the effect of religious knowledge and activity on student academic dishonesty using locus of control and personality as the moderating variables. It is expected that the study contributes as a material consideration to the development of curriculum at higher education institutions, in which it improves honest behavior for students. The overall results show that religious knowledge contributes to reducing the desire to commit academic dishonest behavior as Akko (2018), Sofyani and Rahma (2015), and Rettinger and Jordan (2005). The negatively significant relationship explains that the higher religiosity through better religious knowledge, the lower intention of students to commit academic dishonesty. The results show the positive role of religious knowledge in dealing with academic dishonesty. The possible reason for this finding is because materials delivered in the scope of religious education leads to the development of moral and behavior (Azis, 2008). Moral and behavior developed under the religious guidance lessen the intentions to commit unethical and dishonest behavior because students obtained better understanding and responsibility in doing good and bad deeds. The results support the Theory of Planned Behaviour that "actions, then, are controlled by intentions, but not all intentions are carried out; some are abandoned altogether while others are revised to fit changing circumstances" (Ajzen, 1991).
Further results show that the influence of religiosity on academic dishonesty becomes more concentrated by a locus of control, internal and external locus of control as Rustiarini (2014). Students who tend to an internal locus of control consider that religious knowledge and activities encourage students to be honest when they are completed with confidence and self-motivation (Sofyani & Pramita, 2015). They argue that religious forces control the honesty of prospective accountants, auditors, and doctors. Students with an external locus of control are relying more on someone's hopes to depend on others and more seeking and choosing favorable situations. While students with an internal locus of control are relying more on someone's expectations of themselves and prefers skills rather than just a favorable situation (Winardi et al., 2017). Students with an internal locus of control generally believe that their success or failure is a consequence of their effort and hard work. Adversely, students with an external locus of control generally believe that their successes or failures outcome of external factors beyond their control, such as luck, fate, circumstance, injustice, bias, or teachers who are unfair, prejudiced, or unskilled. Students with an internal locus of control might blame poor grades on their failure to study, whereas students with an external locus of control may blame an unfair teacher or test for their poor performance (Rotter, 1966).
When the importance of religious knowledge and locus of control are neglected at an early stage of school, in a longer period it will drive unethical behavior at the workplace (Asif et al., 2020; Juanda & Sofyani, 2016). The results support Rustiarini (2014) who studies the influence of locus of control on the auditor to demonstrate that internal and external locus of control is the predictor of the audit's dysfunctional behavior. As Rustiarini (2014), Al-Ebel et al. (2020) show that a religious top leader with accounting expertise avoids audit report lag.
Furthermore, although we cannot find the direct influence of religious activity on academic dishonesty, the further analysis concludes that religious activity will function better in proper behavior development when the students have understood the core values delivered in religious education which is shaped in religious knowledge. Besides, religiosity plays a major role as the basis that fortifies individuals to guide students' behavior and prevent them from being dishonest and unethical. In other words, intense religious activity without strong religiosity is not enough to prevent students from committing cheating. Therefore, the results suggest that both factors must be synergized and performed structurally.
Extending the study by Hendy and Montargot (2019), Hendy and Biderman (2019), and Giluk and Postlethwaite (2015) regarding personality and academic dishonesty, this study proposes a distinctive result. Personality, as measured by Big Five Trait, does not influence academic dishonesty and does not moderate religiosity and religious activity. This probably occurs because the behavior of the individuals is not necessarily represented by their personality but sometimes by external factors. In certain circumstances, individuals will decide and act spontaneously as a quick response to what they are facing, which sometimes are contradic their true personality or beliefs (Ellison & Taylor, 1996;Kohlberg, 1973). Pressure, trends, chance, and encouragement from other parties are some of the conditions that might drive individuals in behaving inappropriately.

Summary and conclusions
The novelty of this study is examining the influence of religious knowledge and religious activity on academic dishonesty and the role of locus of control and personality as the moderating variables. The study suggests that formal religion is documented as the predictor of honest behavior committed by students and is the basis in building the students' characters. Although the study is unable to prove the direct relationship between academic activity and personality with academic dishonesty, it will motivate other researchers to develop a further study to generate a robust pattern that will enrich the knowledge in explaining the relationship among the variables. It is suggested that religiosity and religious activities should be synergized in the higher education system and must be arranged and structured well to generate honest accountants, auditors, and doctors.
The result shows that religious knowledge contributes to reducing the desire to commit academic dishonesty. The negatively significant relationship explains that the higher religiosity through better religious knowledge, the lower intention of students to commit academic dishonesty. These findings are consistent with Al-Ebel et al. (2020), Akko (2018), Rustiarini (2014), Sofyani and Rahma (2015), and Rettinger and Jordan (2005). The possible reason for this finding is because materials delivered in the scope of religious education leads to the development of the students moral and behavior (Azis, 2008). Moral and behavior developed under the religious guidance reduce the intentions to commit unethical and dishonest behavior because students obtained better understanding and responsibility in doing good and bad deeds.

Limitation and directions for future research
The main objective of this study is to examine the role of religious knowledge and activities in dealing with the academic dishonesty of prospective accountants, auditors, and doctors. Implicitly, this study emphasizes the honesty that is examined by using an online questionnaire. By employing this method, in general, points of academic dishonesty can be well delineated and can be obtained more objectively. Students who fill out the questionnaire have plenty of time to remember how often and in what form they committed academic dishonesty. In addition, they had plenty of time to recall the religious activities they had gone through and their impact on them because the online questionnaire is not time-limited. However, in-depth analysis and analytical components of their honesty that are private and confidential are not acquired as well as if we use indepth interviews as described by Allmark et al. (2009).
The limitations of this study are not using in-depth interviews with selected respondents to obtain specific information to achieve an in-depth analysis of student honesty. For future studies, it is suggested to use in-depth interviews as an implication of this study. Besides, this study uses respondent prospective accountants, auditors, and doctors, not continuing to investigate how common or not dishonesty is committed by doctors, accountants, auditors to confirm the statement in the background section above that "a person who commits academic fraud during the studies has a higher tendency to behave in an unethical and fraudulent manner at the workplace (Khan et al., 2019;Williams & Williams, 2012)".
This study focuses on the locus of control and personality as moderating variables. In fact, the attributes inherent in a person are not the only locus of control and personality, but also selfcontrol (Flores et al., 2020;Zhang et al., 2019) as mentioned in the introduction above. This study does not test self-control to avoid a wider relationship among the moderating variables. We suggest to include self-control in a more comprehensive model in the next studies.