Contents Introduction
Since the COVID-19 global pandemic, there has been more reliance on social media platforms such as Facebook, TikTok, and Instagram [1]. Almost sixty percent of the global population, or nearly five billion people, now use social media. Primarily, social media platforms work as a medium for social networking and media sharing –inspiring users to share information about themselves and encouraging them to self-disclose [2]. Its usage has expanded to more than just for socializing, where users build online communities with similar interests and values [3]. Today, social media helps individuals build their online personas [4], businesses narrow the gap with their audience, ensuring increased loyalty [3], [4], and professionals communicate with their specific audience in an active participatory manner [5]. As a result, individuals, businesses, professionals, and government bodies now broadly integrate personal and professional communications with social media.
Although social media usage has many benefits, its use has risks. For example, studies have found that users’ behavior on social media contributes to cyber victimization [6]. In particular, risky practices on social media, such as openness and extroversion in revealing personal information [7], [8] or self-disclosures [6], [9], [10], [11].
Compared to face-to-face communication, communications on social media platforms lead to higher self-disclosure as social media provides perceived anonymity and lacks social cues [12], [13]. Self-disclosure is any information one shares, such as personal information, experience, opinion, or belief, making one known to others [14], [15], [16]. Typically done to build or maintain relationships, self-disclosure in communication helps create depth and build trust in relationships. Previously, self-disclosure attracted researchers from psychology and communications [17]. However, discussions on self-disclosure now include topics such as online interactions, privacy, e-commerce, and cybersecurity [18], showing that as people widely communicate through online mediums, specifically social media platforms, the narratives for self-disclosure are also changing.
Studies also found that social media features and affordances contribute to self-disclosure, such as the ability to conduct synchronous and asynchronous communication, where users can self-disclose to audiences without simultaneously interacting with one another [19], [20]. Furthermore, on social media, users have visibility preference, editing freedom, and content continuity linked to their network, consequently affecting disclosure behavior [21], [22]. With features like online communities [3], [14], data storage [23], and interactivity [12], users have more reasons to self-disclose on social media, indicating that with the evolving communication technology, new affordances of social media platforms can contribute to users’ self-disclosures, resulting in higher privacy risks.
Self-disclosure can be positive or negative [24] and can be categorized into evaluative or descriptive disclosures [25]. Evaluative disclosures include voicing feelings, experiences, opinions, judgment, or critique, whereas descriptive disclosures can be in the form of descriptions or explanations [6], [11], [14], [15], [19], [26], [27], [28]. Users could accidentally self-disclose on social media when expressing descriptive or evaluative information, such as activities carried out throughout the day and reviewing their feelings during an event that occurred in their lives [18], [29]. For example, employees disclose themselves on Facebook confession pages by expressing anger towards the employer and their coworkers, frustration towards organizational policies, and more [30]. Accordingly, self-disclosure on social media could lead to severe repercussions, such as dismissal from employment, legal actions, or a tarnished reputation [31], [32]. For instance, nurses found disclosing improper or unprofessional comments, opinions, or behavior shared on social media were terminated by their employers [33]. Furthermore, self-disclosing on social media could also lead to regret after posting [34], victim blaming by bystanders [35], and privacy violations [3]. Thus, although self-disclosure has many benefits, it must be done discreetly to avoid grave repercussions.
Another concern arising from social media self-disclosure is the ethical aspect: whether online communication substituting face-to-face communication would provide a positive or negative outcome since online communication happens on platforms equipped with various tools and affordances that mask users’ identities, thus providing perceived anonymity [36].
Legal and technological solutions alone are not enough to control online behavior; therefore, like offline behavior, internalized norms of behavior can be a solution [37]. For example, Peluchette et al. [11] found that indiscreet self-disclosure is the main predictor of cybervictimization on social media. As such, [38] stated that users’ ethical ability on social media could potentially prevent cybervictimization. Furthermore, Müller et al. [38] describe ethical ability as the information, drive, and capability of online interpersonal communication that follows the law and social norms. Thus, social media users should be able to recognize the acceptable, proper, desirable, and legal aspects of their actions before being aware of their consequences [37], [39].
However, when searching the literature, ethical behavior has yet to be examined as a factor for self-disclosure protection. But, many existing discussions regarding online communications revolve around cyber safety, cyber security, and cyber ethics. These include acting safely and responsibly online to help protect users’ personal information and reputation and reduce dangers from behavioral-based threats, consequently ensuring cyber safety [40]. Thus, in addition to adhering to social media etiquette, studies suggest users should take responsibility for their online safety [41] by protecting their privacy [42] and having control over their self-disclosures online [42], [43]; these are the main research questions of this study.
Therefore, to address this gap, this study examines users’ protection behavior and social media etiquette as self-disclosure protection factors while examining whether offline ethical behavior affects social media etiquette.
Literature Review and Hypotheses Development
Revealing and concealing personal information are opposing processes [44], whereby revealing personal information involves self-disclosure, and concealing personal information involves keeping privacy. Thus, to help users avoid privacy threats due to self-disclosure, users should be aware of its potential risks [45] and countermeasures by educating them on privacy protection to prevent threats such as cybervictimization due to the disclosure of information [46], [47], [48].
Self-disclosure is multidimensional and measured using five dimensions [15]: intentionally conscious disclosure, the positive or negative nature of the disclosure, honesty and accuracy, and the depth or intimacy of the disclosures. In contrast, Omarzu [49] discusses three dimensions of self-disclosure, which are breadth, duration, and depth; where breadth is measured using the number of disclosed topics, and duration is measured using the amount of disclosure, either by the time spent disclosing or the length of the disclosure. Depth is measured with the emotional intimacy of the disclosure. Other studies measured self-disclosure using awareness of the discloser, the disclosure amount, valence, honesty, and the intimacy of the disclosures [11], [22], [50], [51]. Users’ self-awareness or the intention to disclose information about themselves is closely related to the perceived benefits and risks when disclosing the information [52]. While valence is the form of the content, including a positive, neutral, or negative emotional tone [20], [53], [54], and honesty is the accuracy or credibility of the information disclosed [51]. Disclosure frequency refers to the disclosure amount, duration, or breadth, while intimacy refers to the depth or sensitivity of the disclosed information [50], [51].
Scholars have deliberated on staying safe while benefiting from social media use, although balancing maintaining privacy and social media participation seems contradictory. For example, social media platforms such as Facebook encourage sharing personal information, contact information, and work or education information in the profile section and uploading photos in the album section containing the users’ identifiable information [55], [56]. Thus, balancing self-disclosure and protecting privacy on social media is challenging [57], [58], [59], [60]. The situation creates discrepancies between privacy concerns and behavior, also known as the privacy paradox whereby although users are concerned about their online privacy, they keep disclosing information about themselves on social media [1] to achieve certain gratification [59], such as distress relief [61] and gaining perceived social support [62].
Several studies have examined the factors of self-disclosure protection. The factors mainly include privacy-related issues, such as perceived privacy risks [12], [63], [64] [65], [66], [67], privacy concerns [12], [60], [68], [69], the perceived severity of privacy breach [70], [71], the perceived vulnerability of privacy [70], [71], self-efficacy of privacy protection [71], response efficacy of privacy protection [70], privacy attitude and behavior [68], [72], and privacy invasion experience [12]. Other self-disclosure protection factors found in the literature include perceived control over information [43], self-presentation of a positive online persona [73], and self-control [74].
Discussions on privacy state that privacy protection has three main aspects: control over personal information, defense against threats, and users’ perception of risk regarding social media communication [75]. Additionally, studies describe that social media users must have relevant security skills and knowledge to adopt protective behavior [37], [38]. Overall, previous studies suggest that victim-centric prevention for cyber safety can be achieved by raising awareness of risky behavior on social media and adopting protective measures to defend against threats and ensure privacy [9], [75], [76], [77], [78], [79], [80].
A. Protection Behavior
Some theories have been used to explain self-disclosure protection, such as the Protection Motivation Theory [67], [68], Communication Privacy Management Theory [64], [72], Privacy Calculus [61], [62], Social Exchange Theory [61], [64], Theory of Reasoned Action [43], [63] and Theory of Planned Behavior [43].
The Protection Motivation Theory (PMT) has been widely applied and adapted to understand the motivations behind the adoption of online protective behavior for information security [31], [81], [82], [83], [84], [85], online safety [86], cybervictimization prevention [87], information privacy protection [78], [88], [89], risk-reducing behavior [90], and information sharing intention [91].
PMT posits that users’ fear appeals (perceived severity and perceived vulnerability) and the effectiveness of coping appraisal (response efficacy and self-efficacy) influence users’ intention to adopt protective behavior [92]. For example, [93] investigated the relationship between perceived severity, perceived vulnerability, response efficacy, self-efficacy, users’ risky online behavior, and intentions for safe communication. They argue that the Protection Motivation Theory (PMT) can help understand users’ online behavior that leads to cyber victimization. Furthermore, their results show that perceived severity, response efficacy, and self-efficacy predicted increased safe online communication behavioral intention. Also, [41] conducted a study to investigate users’ motivations toward safe online behavior. Their findings reveal that personal responsibility, intervention approach, and prior knowledge are essential for effective safety interventions.
However, studies have shown that measuring actual behavior instead of intention provides more reliable results [94], [95], [96]. Thus, different from existing studies, instead of investigating intention for protective behavior, we adapted and applied the protection motivation theory to examine the relationship between users’ protection behavior and social media self-disclosures, with the following hypothesis:
H1a: Perceived vulnerability influences self-disclosure on social media.
H1b: Perceived severity influences self-disclosure on social media.
H2a: Self-efficacy influences self-disclosure on social media.
H2b: Response efficacy influences self-disclosure on social media.
B. Social Media Etiquette
Cyber ethics, which is composed of computer ethics and internet ethics [97], is a discipline that discusses appropriate ethical behavior, moral duties, and obligations in online environments and digital media [40], discussing issues such as copyright, hacking, online addiction, and etiquette [98].
Online etiquette is defined as using the internet responsibly and respecting others during online communication [99]. As in offline communication, etiquette in online communication is essential as the lack of it would influence the audiences’ positive or negative perception [100]. Today, learning and exploring online etiquette is inevitable with the change in socializing and communicating using online platforms [101].
Additionally, online communications conducted with etiquette reduce online risks. For example, learning online etiquette and applying it minimizes the chance of bullying [102]. Spinks et al. [100] mentioned the importance of etiquette in online communication, as in offline communication, a lack of it would influence the audiences’ positive or negative perception. Several studies have also discussed the ethics and etiquette in online communication [103], such as writing emails [100], [104]. These studies highlighted the appropriate rules of behavior users should adhere to when using emails to avoid misunderstandings, harmful reactions, and reaching the wrong audience [104]. For example, respecting others’ time, privacy, and feelings by paying attention to the different rules of behavior with various mediums, focusing on sharing knowledge, and being careful with self-presentation [103]. Additionally, Spinks et al. [100] described that online communication can lead to miscommunication in tone and content, failing to convey a successful communication, thus suggesting applying positive manners or online etiquette when sending emails.
Thus, focusing on the social media context, we investigate whether online etiquette affects communication involving self-disclosures with the following hypothesis:
H3: Users’ social media etiquette influences their self-disclosure on social media.
C. Ethical Behavior
Although people face ethical dilemmas daily, now that their daily lives have been integrated widely into social media, whether a person’s online behavior reflects their offline ethical behavior arises. Research argues that people behave differently online than in face-to-face communications [105]. Studies found that offline activities affect users’ online social behavior [99], [106]. However, it has been recently identified that studies on online etiquette are still in their initial state, and discussions are scarce [36]—specifically, investigations of social media communications involving self-disclosure and its linkage to online etiquette. Thus, we investigate whether offline ethical behavior affects users’ social media etiquette.
OxfordLanguages defines ethics as the “moral principles that govern a person’s behavior or the conducting of an activity” and “the branch of knowledge that deals with moral principles.” Philosophers and theologians have long discussed ethical behavior and right and wrong actions [107]. However, due to the complexity of ethics, no one particular definition is widely accepted by philosophers and psychologists [108].
In our investigations of users’ offline behavior, we look at ethics through the lens of Islamic perspectives. We do so for several reasons; first, using religion as a source of ethics is justifiably acceptable, as many of the world’s population now embrace religious beliefs [109]. Additionally, with Islam having enormous followers globally, we explored offline ethical behavior from within its perspective. Moreover, Islam links religion and behavior [110], where good ethical conduct is one of the essential facets of the fundamental pillars of Islam [111], thus calling for the investigation of whether offline behavior is reflected online. Mohammed [112] describes the following about Islamic ethics:
“Islamic ethics prescribe its followers to zealously guard their behavior, words, thoughts, and intentions and observe certain norms and moral codes in their family affairs; in dealings with relatives, neighbors, and friends; in their business transactions; in their social affairs; and in private and public life” (pp. 873).
Agreeing with the Divine Command Theory, which posits that right and wrong actions are grounded on religious belief, the basic tenets of Islamic ethical conduct or Akhlaq Islamiyah are based on the Al-Qur’an and the Sunnah—the two primary knowledge sources in Islam. The Al-Qur’an contains the word of Allah (God), which commands its followers to good behavior and forbids evil. The Quran constitutes Islamic law (shari’a) or the Islamic legal and social system that states that a morally good person leads to the welfare of individuals and society, and whatever is injurious is morally wrong [112], [113]. Meanwhile, the Sunnah or conducts are based on the normative example set by the Prophet Muhammad (peace be upon him). The best of Akhlaq, according to Islam, is as portrayed by the Prophet, as mentioned in the Al-Qur’an, “Indeed, in the Messenger of Allah you have an excellent example for whoever has hope in Allah and the Last Day, and remembers Allah often” (Quran 33:21).
The word that best describes ethics in Islamic teachings is Akhlaq. Al-Khuluq or Al-Alkhlaq in plural translates to conducts, morality, ethics, and behaviors [114]. Akhlaq is the practice of virtue, morality, and manners, which in English translates to a person’s character, disposition, personality, or ethics [115]. Beekun [116] described, in the Islamic context, ethical behavior or the concept of goodness can be interpreted into many words as mentioned in the Qur’an, such as ma’ruf (approved), khayr (goodness), haqq (truth and right), birr(righteousness), qist (equity), ’adl (equilibrium and justice), and taqwa (piety).
Many prominent scholars have discussed ethics in their work, such as al-Farabi (d. 950), Miskawayh (d. 1030), Ibn Sina (d. 1037), Ibn Rushd (d. 1198), and Al-Ghazali (d. 1111). The work of Miskawayh, The Refinement of Character (Tahdhib al-Akhlaq), offers an analytical system of Islamic ethics that separates ethics from other disciplines [117]. Miskawayh defined akhlaq as “a state of the soul which causes it to perform its actions without thought or deliberation” [118].
Miskawayh’s work, Tahdhib al-Akhlaq, also inspired other scholars like Al-Ghazali, Al-Tusi (d. 1274), and Al-Dawani (d. 1501) [117]. Al-Ghazali describes man in two forms: the physical form (khalaq) and the spiritual form (khuluq or akhlaq) [119]. Akhlaq is engrained in the soul and manifested through one’s actions, which, if the soul is virtuous, will manifest the right action and vice versa; therefore, Islamic ethics helps acquire the soul’s well-being and guard against vices [119].
Online activities should not be considered a separate section from offline life but rather an extension; most online behavior and manners reflect what we cultured offline, such as ways to address others and expectations in conversation and engagement [120]. Thus, we pursue our investigations with the following hypothesis:
H4: Users’ offline ethical behavior (Akhlaq) influences social media etiquette
Furthermore, Al-Ghazali, in his work, the Revival of Religious Learning (Ihya Ulum id Din), discusses the four natures of the soul, which are the power of discerning knowledge (rational), the power of administration or anger (irascible), the power of greed (concupiscent), and the power of adjustment of the three previously mentioned natures. Al-Ghazali explained that to form good conduct, one must first bring the four qualities of the soul (fi al-batin arba‘ arkan) to perfection [121]. In perfecting the soul, one must govern the nature of the soul with the four cardinal virtues, namely, wisdom, temperance, courage, and justice, to achieve good conduct [122].
As mentioned in the Al-Quran, “Whoever is given wisdom has been given abundant good” (Quran, 2:269). Al-Ghazali [122] described wisdom as the cardinal virtue of the rational, where wisdom is the root of good conduct. A person poses wisdom with the strength of knowledge to differentiate truth from false and good from evil. In the work of both Miskawayh and Al-Ghazali, wisdom is further divided into intelligence, retention, prudence, lucidity, sound judgment, and easy comprehension [123], [124].
Temperance is managing one’s desire by applying moderation in the struggle with possessions, which calls for a balance between indulgence and insensibility [123], [125]. With wisdom and shame, one can manage concupiscent power [122]. Miskawayh and Al-Ghazali, in their work, further divided temperance into modesty, meekness, self-control, liberality, honesty, contentedness, equableness, self-discipline, good disposition, docility, composure, and piety [123], [124].
Courage is when one can control anger and conduct it with knowledge and shari’a; he is said to possess courage [122]. Courage corresponds to managing irascible power, where a courageous person has traits such as magnanimity, fearlessness, fortitude, steadfastness, patience, self-assurance, manliness, and endurance [123], [124]. A person of courage would behave constructively while sustaining virtue and preventing falsehood in their conduct [126].
Justice is the virtue that controls the three virtues (wisdom, temperance, and courage) by ensuring the right balance in managing the irascible and concupiscent, guided by wisdom [122], [123], [125]. Among the subdivisions of justice are friendship, concord, kindness towards relatives, reciprocity, honest dealings, gracious dispensation, affability, and piety [123], [124].
Thus, first, we investigate the relationship between wisdom, temperance, courage, and justice towards Akhlaq. Therefore, focussing on adult social media users, we conducted a quantitative study with the following hypotheses:
H5a: Wisdom positively influences Akhlaq.
H5b: Temperance influences Akhlaq.
H5c: Courage influences Akhlaq.
H5d: Justice influences Akhlaq.
Methods
This quantitative research used a cross-sectional electronic survey to collect data and draw conclusions on the effect of social media etiquette and protection behavior. It also investigates the impact of offline ethical behavior on social media etiquette. The data collected from the survey was analyzed using Structural Equation Modeling with the Partial Least Square (PLS) technique. The hypotheses for the research are tested through the structural assessment model, which assesses the nature of the relationship or association between latent variables [127].
A. Research Design
We administered an online survey with convenient sampling from November 2021 to Dec 2022. 221 surveys were completed, with 219 usable surveys for analysis. We aimed the questionnaire for Muslim social media users above 18 to capture adults’ ethical behavior and self-disclosure on social media. Accordingly, we included filter questions addressing the age limit for only 18-year-old and above social media users to answer the survey.
We depict the demographic composition of the sample in Table 1. Half of the respondents in the sample consist of young adults, and females are slightly overrepresented. The respondent sample is mostly highly educated. More than fifty percent of respondents spend more than two hours on social media daily. However, eighty percent of the respondents indicated that they applied privacy settings to limit the audience for their social media posts, and almost sixty percent showed that they include risky connections in their network. These risky connections include adding the social media profile of someone famous they did not know personally, someone they knew only online, an online shop, or someone they do not know.
Memon et al. [128] suggest a sample between 160 and 300 valid observations for multivariate statistical analysis techniques such as PLS-SEM. Power analysis can assist in deciding the suitable sample size by considering the magnitude of the measurement error, complexity of the model, and number of indicators [62]. Thus, we also conducted Power Analysis using G* Power version 3.1.9.4 to estimate the needed sample size. As our model includes formative measurements, we used the number of indicators (8) forming the formative construct for the number of predictors in estimating the needed minimum sample size (n =109).
B. Measures
We adapted past survey questions from existing studies and developed new instruments validated by three field experts to measure users’ offline behavior from the Islamic perspective.
We adapted the questions for self-disclosure on social media using [15] and [129]. We asked the respondents, “How much do you agree or disagree with the following statements (in the context of social media)?”. Like Hanus and Wu [81], we measure the relationships between PMT and behavior instead of intentions due to the unclear connection between the intents and cognitive variables in PMT. Thus, respondents were asked about their self-disclosure behavior on social media instead of their intention.
To measure the respondents’ protection behavior, we used questionnaires from previous studies [87], [130] and adapted them to the context of self-disclosures on social media.
For threat awareness, the constructs consist of perceived severity and perceived vulnerability. To measure the respondents’ perceived severity (PS), they were asked, “How serious are these issues to you? (in the context of social media),” with responses made using a 5-point Likert scale ranging from “Not at all serious” to “Extremely serious.” We measured perceived vulnerability (PV) by asking users, “How likely will the following experiences happen to you?” with responses made using a 5-point Likert scale ranging from “not at all likely” to “extremely likely.”
Meanwhile, the countermeasure awareness constructs consist of self-efficacy and response-efficacy for countermeasure awareness. Regarding self-efficacy, we asked the respondents to indicate agreement or disagreement on their knowledge and to conduct protective actions. Respondents rated their ability using a 5-point Likert scale ranging from “strongly disagree” to “strongly agree.” We then measured the respondents’ response efficacy by asking them, “To what extent do you agree or disagree that doing the following would prevent you from being bullied electronically.” Again, respondents can choose from “strongly agree” to “strongly disagree” on a 5-point Likert scale.
We measure users’ offline ethics based on the four virtues (wisdom, temperance, courage, and justice) that make up a good Akhlaq before measuring Akhlaq. Operationalizing the variable requires defining the construct, identifying the instruments measuring the construct, deciding on the response format, and assessing the validity and reliability of the measurement scales [131]. Therefore, we the measurements based on the discussions by Al-Ghazali in his book Ihya’-Ulumuddin and Miskawayh in his book Tahdhib al-Akhlaq, which explained the four cardinal virtues and validated by three field experts for the study’s four constructs and items. We then made corrections according to the expert’s comments. Specific definitions are ensured to avoid abstract and conceptual definitions [132].
We asked the respondents whether they agreed with the statements about their behavior regarding wisdom, temperance, courage, and justice. Users can select a response from a 5-point Likert scale ranging from “strongly disagree” to “strongly agree.” Next, we asked the users about the frequency of the items regarding Akhlaq’s conduct using a 5-point Likert scale ranging from “never” to “always.” Finally, we list the construct and items used in the questionnaire in Appendix 1.
We developed the questionnaire for social media etiquette by adapting to cyber-safety guidelines by Cyber Security Malaysia [133] and previous studies discussing social media etiquette [11], [104], [134], [135], [136], [137], [138], [139], [140]. We asked the respondents how frequently they conducted the behavior on social media using a 5-point Likert scale ranging from never to always.
C. Analysis Technique
We test the research model with partial least squares - structural equation modeling (PLS-SEM) using SmartPLS version 3.3.3. The collected data must be analyzed to assess the relationship between the constructs. Before conducting the analysis, the research framework is first translated into the relationship model in SmartPLS, as depicted in Figure 1, where all the latent variables are connected accordingly. Thus, using PLS-SEM, this research can assess the relationship between the independent and dependent variables and evaluate if the data supports the hypotheses.
Results
A. The Measurement Model
For reflective measurement, we removed the items with low factor loadings of less than 0.4 and retained the indicators with more than 0.7. However, we retained the indicators with loadings between 0.4 and 0.7 that contribute to an adequate Average Variance Extracted (AVE) of greater than 0.5 and Composite Reliability (CR) of greater than 0.7 [141]. We checked for the constructs’ reliability using Cronbach’s Alpha and CR. All the constructs exceeded CR and Cronbach’s Alpha threshold, thus satisfactory. We present the result of the measurement model in the Appendix 2
For discriminant validity assessment, Henseler et al. [142] suggests using the Heterotrait-Monotrait (HTMT) ratio of correlations for its superior performance compared to Fornell and Larcker’s assessment. Thus, we assessed the HTMT criterion and found values lower than the required thresholds of HTMT.85 and HTMT.90, indicating that discriminant validity is established for the constructs of this study. Additionally, the results for HTMTinference of the constructs show that the confidence interval has no value of 1, thus confirming discriminant validity [142]. The HTMT, Fornell and Larcker results, and cross-loadings are presented in Appendix 4, 5, and 6.
When assessing the formative measured construct, we applied Mode A estimations, which offer better performance for a moderate sample size [143]. We assessed the convergent validity of the formatively measured constructs by carrying out the redundancy analysis. The statement for redundancy analysis is displayed in Table 2. The path coefficient result (0.784) is larger than 0.700 [141], displaying sufficient degrees of convergent validity; hence, the formative indicators contribute to its intended content at an adequate level.
Next, we assessed the VIF values for collinearity among the indicators. VIF values are below the threshold values of 5 [144] and 3.3 [145], indicating collinearity is not an issue in estimating the PLS path model. Next, we analyze the outer weights for their significance and relevance. The bootstrapping result suggests that the constructs’ outer weights are significant and thus retained. The loading result is above 0.5, and the t-value is more than 1.645.
B. The Structural Model
It is vital to evaluate lateral collinearity issues in the initial stage of structural model assessment [127]. Thus, we measured and found that the VIF values of all the independent variables are less than 3.3 [145] and 5 [144], indicating that collinearity is not an issue [141]. Therefore, we tabulate the results in Appendix 3.
We then proceeded with bootstrapping analysis (5000 subsamples, one-tailed tests) to assess the relevance and significance of the structural model relationships. Wisdom is the most important predictor of the dependent construct Akhlaq, followed by justice, temperance, and courage. For another endogenous construct, social media etiquette (SME), Akhlaq strongly affects SMEs. At the same time, SME and perceived vulnerability negatively affect self-disclosure (SD), followed by perceived severity, response efficacy, and self-efficacy. Then, we checked the confidence interval bias corrected of the upper and lower bounds and found that value zero does not straddle between the results, showing a significant result.
Next, we assessed the model’s predictive accuracy via the
We then assessed the predictor construct’s effect size (
C. Overview of the Hypothesis
We formulated 11 hypotheses in this study, as observed in Figure 1. All path coefficients related to the hypotheses were significant except for perceived severity (H1b), self-efficacy (H2a), and response efficacy (H2b) on social media self-disclosure. Wisdom, temperance, justice, and courage have a significant positive relationship with Akhlaq, thus confirming hypotheses H5a, H5b, H5c, and H5d. Akhlaq significantly and positively affects social media etiquette, which confirms hypothesis H4. We found that social media etiquette and perceived vulnerability significantly negatively affected self-disclosure on social media, confirming hypotheses H3 and H1a. Finally, the hypotheses that perceived severity (H1b), self-efficacy (H2a), and response-efficacy (H2b) significantly affected self-disclosure were rejected. We present the hypotheses testing results in Table 4.
Discussion
This study applied the Protection Motivation Theory and Islamic Ethics to investigate the self-disclosure protection factors. The relationship between the variables was evaluated using the Structural Equation Modeling technique with the Partial Least Squares-Structured Equation Modeling (PLS-SEM) method.
The results revealed that wisdom, temperance, courage, and justice significantly affect offline ethical behavior (Akhlaq Islamiyah), which is in line with what was outlined in Thadhib Al-Akhlaq by Miskawayh and Revival of Religious Learning (Ihya Ulum id Din) by Al-Ghazali [121], [122], [124]. The result also showed that users’ offline ethical behavior (Akhlaq Islamiyah) influences users’ social media etiquette, as suggested by several studies that offline behavior affects how users behave online [37], [99], [100], [106], [147]. Additionally, users’ social media etiquette influences their self-disclosure behavior, similar to findings from previous studies [120], [139].
The relationship between protection behavior and self-disclosure shows an interesting result: three components of protection behavior (perceived severity, self-efficacy, and response efficacy) do not significantly affect self-disclosure. Although users perceive risks regarding self-disclosure as highly serious, possess high self-efficacy, and find the recommended response effective in safeguarding them from self-disclosure-based victimization, these conditions do not influence their self-disclosure behavior. In contrast, perceived vulnerability significantly influences self-disclosure.
The study finds three possible explanations for this phenomenon. First, users’ belief in their ability to protect information boosts their confidence and creates a sense of false security [70], [71] for them to conduct self-disclosure. Secondly, it could be due to users’ perceived benefits in self-disclosing, which outweigh their perception of the severity of the risks [78]. The findings align with previous studies that discuss privacy paradox phenomena, whereby users self-disclose themselves even though risks await. For example, if users perceive that the benefits outweigh the risks they face upon self-disclosing, they will self-disclose themselves [1], [148]. Third, users’ trust in the social media platform and other users regarding the information being disclosed will be handled with integrity, and benevolence influences their self-disclosure behavior, resulting in higher self-disclosure, as discussed in previous studies [1].
Furthermore, studies show that users disclose themselves when they highly trust the platforms’ features, such as privacy settings that enable disclosure visibility to only a particular audience segment [149], [150]. So, in line with previous studies [66], our analysis implies that although the users know the severity of self-disclosing on social media and have the technical skills to protect themselves, their self-disclosure behavior is not affected by them. Also, Rauf [151] stresses the need to evaluate online technology’s role in aiding varying hostile acts, for example, the freedom of speech on social media, which consequently opens doors for freedom to widely and freely express hate. Hence, mitigations for risky self-disclosure efforts should look beyond technical knowledge and risk awareness integration. As such, these findings contribute to theoretical and practical implications.
A. Theoretical Implication
The results present that offline ethical behavior based on the Islamic perspective (Akhlaq Islamiyah) affects social media etiquette. It analyses the relationship between offline and online ethical behavior, which the previous research has mentioned is currently lacking in the literature. Thus, this research provides theoretical implications whereby users’ social media etiquette is linked to users’ adherence to Islamic ethical conduct. This finding opens an interesting avenue for future investigations regarding social media etiquette and its link to offline ethical behavior in other religious or cultural contexts.
Additionally, findings suggest that although beneficial in many circumstances, self-disclosure reaps severe repercussions in other instances. Therefore, this research contributes to the body of knowledge regarding the determinants of self-disclosure protection factors, which can benefit future research on a similar mitigation approach for other types of risky social media behavior. For example, practitioners designing cyber safety prevention approaches can embed offline ethical behavior and social media etiquette elements for safe self-disclosure.
Also, previous studies that have explored varying antecedents to self-disclosure decisions, our findings provide several vital contributions regarding social media self-disclosure. For example, [148] recently classified four main factors contributing to self-disclosure on social media: discloser-centric, audience-centric, platform and affordances-centric, perceived risk, and cost-centric, whereby the discloser-centric factor discusses the users’ character (extraversion, openness, agreeableness, conscientiousness, and neuroticism), experience (negative online experience and life events), and intra-personal goals. Also, as discussed in [152], cultural differences affect privacy-related decisions, as privacy behavior is deeply rooted in users’ distinct cultural beliefs and values. Therefore, future assessments associated with the discloser-centric factors contributing to self-disclosure can incorporate our findings regarding ethical and cultural factors affecting self-disclosure decisions.
B. Practical Implication
As social media platforms progress, new features will always exist, thus leading to indirect self-disclosures, such as reacting using an emoticon [153], online stickers, using the “like” button, commenting [154], location-based check-ins [24], and being tagged on others’ posts that is visible to other people outside the users’ network. Thus, technical solutions (e.g., privacy settings, anonymity, and two-factor authentication) alone are not enough to guarantee safety on social media from behavioral-based risks [155]; they also warrant ethical solutions [40]. Therefore, as our findings demonstrate that social media etiquette can help curb risky self-disclosure on social media, social media platforms and policymakers intending to prevent risky self-disclosure can incorporate social media etiquette as a preventive measure. For example, nudges are prompted before publishing posts or comments, reminding users to adhere to community, security, and safety guidelines. Of course, social media platforms and policymakers must first outline the proper guidelines and policies for ethical behavior on social media, and users must be made aware of the risks and guidelines for cyber safety linked to self-disclosures on social media.
Additionally, educators may embed the findings regarding self-disclosure protection factors, such as social media etiquette and protection behavior, into discussing cyber safety, particularly in social media usage. Also, as offline ethical behavior has been found to affect online behavior, applying what has been cultivated and taught offline must be highlighted when discussing cyber safety and ethical behavior in online settings.
C. Limitations and Recommendation
Our study has several limitations. First, we only conducted a step-wise analysis of the model and did not conceptualize any mediating effect. However, future research may extend the model to assess the mediating effect of online etiquette on the relationship between offline ethics and self-disclosure behavior.
Next, the relatively low
This study only investigated protection behavior (threat and countermeasure awareness) using the adapted version of Protection Motivation without including the cost and reward component of Protection Motivation and social media etiquette as factors affecting social media protection on social media. However, as Schyff and Flowerday [158] discussed, other factors, such as perceived risks and benefits, contribute to self-disclosure protection behavior on social media and thus can be further investigated in future studies. Future studies may consider integrating the cost and reward components of the Protection Motivation Theory to explain self-disclosure behavior on social media, particularly regarding higher self-disclosure when users consider self-disclosing to be more rewarding than protecting privacy with lower self-disclosure. Past negative experiences or knowledge may also affect social media self-disclosure protection, as explained in previous studies [4], [12], thus worth investigating in future research.
Finally, the following limitation of the study is the sample population investigated, which is heavily focused on Muslim social media users; thus, findings may vary in other religious or cultural settings. However, this limitation also provides a broader avenue for future research investigating online social media etiquette across different religious and cultural settings.
Conclusion
In conclusion, the unceasing progression of communications technology widely influences our society, both positively and negatively. However, the inattention towards ethical values and morals in using technology as a medium for communicating and socializing could lead to a more destructive impact.
Thus, with technological progress, further research should also consider the ethical elements within their usage. As self-disclosure on social media becomes more prevalent, varying risk arises. Thus, future studies must investigate its protection factors, including potential factors other than security tools and privacy protection aspects, such as ethical communication behavior.
Appendix 2Measurement Properties of Reflective Construct
Measurement Properties of Reflective Construct
See Table 6.
