Digital media literacy and perceived fake news handling skills among Arab social media users: evidence from a cross-national survey

1 Introduction

In recent years, social media networks have become the primary source of news consumption among young people across the Arab world (Hassan et al., 2024). These platforms provide rapid and immediate access to information while also offering broad opportunities for interaction and participation in public discussions (Fayez et al., 2026a). However, this digital transformation of the news environment has been accompanied by an unprecedented rise in the spread of fake news, misinformation, and media disinformation, posing a significant challenge to both individuals and societies.

The Arab world has not been immune to this global phenomenon. Arab social media users’ are increasingly exposed to misleading news content, fabricated media narratives, and unreliable information circulated through platforms such as Facebook, X (formerly Twitter), Instagram, TikTok, and WhatsApp. Digital algorithms, rapid sharing mechanisms, dependence on sensational and emotionally charged headlines, and weak verification practices all contribute to the widespread dissemination of such content, negatively affecting public opinion formation, trust in media institutions, and social stability (Al Adwan et al., 2024).

Within this context, digital media literacy has emerged as one of the most significant contemporary educational and communicative approaches to confronting the risks posed by fake news and media disinformation. Digital media literacy refers to individuals’ ability to access digital media messages, critically analyze and evaluate them, and understand the contexts in which they are produced and circulated. This enables individuals to acquire essential skills such as critical thinking, source verification, detecting media manipulation, and engaging responsibly in digital environments (Dame Adjin, 2022; Khalifa et al., 2026).

The importance of developing these skills among young people is particularly pronounced, given that they constitute the demographic group most engaged with social media networks and, at the same time, among the most vulnerable to misleading content. Strengthening their ability to critically and consciously engage with digital news has therefore become an urgent necessity for promoting media citizenship and enhancing societal awareness (Al-Makhzoomy et al., 2026; Vosoughi et al., 2018).

Despite growing international interest in examining the relationship between digital media literacy and resistance to misinformation, the Arab context continues to suffer from several notable research gaps. The research gap addressed by the present study is based on three main dimensions. First, the majority of previous studies have focused on Western societies, where institutionalized media literacy programs are more firmly established (Dumitru et al., 2022; Washington, 2023). In contrast, Arab studies examining levels of digital media literacy among Arab social media users’ and their ability to identify fake news remain relatively limited.

Secondly, a considerable portion of the Arab literature has addressed fake news primarily from political or technological perspectives, with insufficient attention given to the educational dimension of digital media literacy as a preventive and developmental framework for combating disinformation.

Thirdly, there is a shortage of large-scale survey studies targeting Arab social media users’ across a broad regional scope, which limits a deeper understanding of the relationship between digital media literacy competencies and fake news management skills within the social media environment.

Accordingly, the present study seeks to address this gap by analyzing the role of digital media literacy in developing Arab social media users’ skills to confront fake news and media disinformation across social media networks.

In general, The selection of Saudi Arabia, Egypt, and Yemen is theoretically motivated by the substantial variation in the information environments in which social media users encounter and evaluate news. The three countries differ in the organization and regulation of their media systems, the degree of digital transformation, and the broader political and informational conditions surrounding news circulation. Saudi Arabia represents a rapidly digitalizing but strongly regulated media environment, Egypt offers a large and historically significant media market undergoing substantial digital transformation, while Yemen represents a conflict-affected and fragmented information environment in which access to reliable information and trust in media are particularly challenging. Examining users across these contexts provides a useful basis for assessing whether the relationships between digital media literacy, cognitive information processing, and fake news handling are observable across differing information environments.

2 Literature review

The conceptual framework of this study is grounded in the fundamental assumption that Arab social media users’ ability to engage critically and consciously with fake news and media disinformation is not formed randomly. Rather, it is the product of a complex interaction between acquired digital media knowledge and skills on the one hand, and internal cognitive information-processing mechanisms on the other.

Within a digital media environment characterized by an overwhelming flow of information and the ease with which news can be produced and disseminated outside traditional professional frameworks, audiences—particularly young people—are increasingly required to play an active role in analyzing media messages, evaluating their sources, and making informed decisions regarding their credibility and circulation (Areej, 2022; Wardle and Derakhshan, 2017).

2.1 Theoretical framework: information processing theory

Information Processing Theory is based on the premise that the human mind functions as a system that receives information, processes it, stores it, and subsequently uses it in decision-making (Fayez, 2024; Karakatsani and Pliogou, 2026), The present study adopts Information Processing Theory as a cognitive framework for explaining how individuals receive, interpret, evaluate, and respond to news information in digital environments. Rather than treating information processing as a single cognitive act, the study conceptualizes it as a sequence of interrelated cognitive operations through which individuals attend to information, comprehend and interpret its meaning, integrate it with existing knowledge, evaluate its credibility, and ultimately form judgments and behavioral responses. In the context of social media news consumption, this framework is particularly relevant because users encounter information under conditions characterized by high information volume, limited attention, rapid message circulation, emotionally salient content, and varying degrees of source credibility (Abdelmaksoud et al., 2026; Al-Makhzoomy et al., 2026). Accordingly, the processing of potentially misleading information requires more than exposure to a news item; it involves cognitive engagement with its content, comparison with alternative information, evaluation of internal and external consistency, and judgment concerning its credibility (Al-Ababneh et a., 2026).

Based on this perspective, the present study operationalizes cognitive information processing through five interrelated indicators: (1) logical analysis of news content, (2) comparison of news across multiple sources, (3) integration of new information with prior knowledge, (4) detection of contradictions and logical fallacies, and (5) regulation of emotional responses when evaluating news. These indicators represent the higher-order cognitive activities that are particularly relevant to the evaluation of potentially misleading information in social media environments.

The theoretical model therefore does not conceptualize information processing as merely receiving and storing information. Rather, it emphasizes the transition from initial exposure and comprehension to analytical evaluation and judgment, which provides the cognitive basis for distinguishing credible information from misleading or fabricated content.

2.2 Digital media literacy as a multidimensional competency framework

Digital media literacy is defined as the body of knowledge and skills that enables individuals to access digital information, analyze and evaluate it, distinguish between authentic and misleading content, and ultimately make informed media-related decisions (Mrah, 2022). These competencies constitute an essential tool for confronting fake news on social media networks, as they equip individuals with the ability to use digital channels critically and effectively.

The significance of digital media literacy stems from the transformations brought about by digitization in the production and distribution of news, which have created a complex information environment in which any user can easily produce and disseminate content. Consequently, advanced critical skills have become a social and professional necessity.

The study conceptualizes digital media literacy as a multidimensional set of cognitive, analytical, evaluative, technical, and ethical competencies that enable individuals to access, interpret, critically evaluate, and responsibly engage with digital media content. This conceptualization builds on the broader media literacy tradition, particularly the view that media literacy involves not only the technical ability to access media but also the cognitive capacity to critically interpret media messages and understand their contexts of production and circulation (Hobbs, 2010; Potter, 2013).

For the purposes of the present study, digital media literacy is operationalized through five interrelated dimensions: (1) awareness of digital news sources, (2) analysis of media messages, (3) credibility assessment and verification, (4) recognition of media deception and manipulation, and (5) ethical participation and responsible content sharing (Alquaary et al., 2026; Fayez et al., 2026b). These dimensions were selected because they collectively capture the competencies required to navigate and evaluate news in social media environments, where the distinction between professionally produced journalism, user-generated content, opinion, manipulated material, and fabricated information may be difficult to establish (Guess et al., 2020).

The five dimensions should not be interpreted as independent forms of technical proficiency. Rather, they constitute complementary components of a broader literacy competency. Source awareness helps users identify and contextualize information providers; message analysis enables them to examine claims and framing; credibility assessment and verification facilitate the examination of evidence and sources; recognition of deception and manipulation enhances sensitivity to misleading communicative strategies; and ethical participation concerns the responsible use and redistribution of information after evaluation.

This multidimensional conceptualization provides the media-literacy component of the integrated framework used in this study.

2.3 Integrated theoretical model

To clarify the conceptual foundation of this study, we draw upon an integrative approach combining Digital Media Literacy (DML) frameworks (Hobbs, 2010; Potter, 2013) with Cognitive Information Processing (CIP) paradigms (Kahneman, 2011; Pennycook and Rand, 2019). Within this integrated framework, DML is conceptualized as an antecedent repository of critical skills and knowledge, while CIP represents a complementary cognitive process—comprising internal cognitive mechanisms such as logical reasoning, source comparison, and emotional regulation, involved in information (Fayez, 2022).

Because this study employs a cross-sectional correlational design, all hypothesized relationships are formulated in terms of statistical associations rather than direct causal or deterministic effects. Accordingly, higher competencies in digital media literacy are expected to correlate positively with deeper cognitive processing and higher self-reported skills in handling fake news on social media environments.

The central theoretical proposition of this study is that digital media literacy and cognitive information processing represent complementary rather than competing explanatory perspectives. Digital media literacy specifies the competencies and evaluative resources that users bring to their encounters with digital news, whereas Information Processing Theory explains the cognitive operations through which these resources are activated when users encounter, interpret, evaluate, and respond to information. Accordingly, the integrated framework proposed in this study conceptualizes digital media literacy as a competency domain that is theoretically relevant to systematic and critical information processing. Source awareness, message analysis, credibility assessment, recognition of manipulation, and responsible sharing provide users with the knowledge and evaluative resources necessary to engage more critically with news content. These competencies, in turn, are reflected in cognitive processing activities such as comparing information across sources, integrating new information with prior knowledge, identifying contradictions and logical fallacies, and regulating affective reactions during evaluation.

The integrated model can therefore be represented conceptually as a non-directional framework of associations: follows:Within this framework, digital media literacy represents the competency-based component, cognitive information processing represents the cognitive mechanism through which information is evaluated, and fake news handling represents the applied outcome reflected in users’ abilities to identify, verify, evaluate, and appropriately respond to potentially misleading information (Kahneman, 2011).

DigitalMediaLiteracyCompetencies↔CognitiveInformationProcessing↔FakeNewsHandlingSkills

The model is used in the present study as an explanatory conceptual framework for examining the relationships among the study variables. Given the cross-sectional survey design, these relationships are interpreted as theoretically informed associations rather than evidence of causal effects.

This integrated conceptualization also provides the basis for the operationalization of the study variables and for examining the hypothesized relationships between digital media literacy, cognitive information processing, and fake news handling skills among Arab Social Media Users.

2.4 Skills for handling fake news and media disinformation

Skills for handling fake news and media disinformation refer to the cognitive, evaluative, and behavioral competencies that enable individuals to identify potentially misleading content, assess the credibility of information and its sources, understand common disinformation strategies, and make informed decisions about whether to believe, verify, share, ignore, or report questionable content. These skills are particularly important in social media environments, where information is encountered rapidly and where professionally produced journalism, user-generated content, opinion, manipulated material, and fabricated information may appear within the same information stream.

The ability to handle potentially misleading information involves more than simply recognizing whether a news item is true or false. It requires individuals to examine the characteristics of the content, consider the reliability of its source, compare claims with information available elsewhere, and evaluate whether the evidence provided adequately supports the claim. Such practices are consistent with the broader understanding of disinformation literacy, which emphasizes the ability to recognize the processes through which misleading information is produced, framed, and circulated (Wardle and Derakhshan, 2017).

Previous research suggests that individuals’ responses to false or misleading information are influenced by both their knowledge and their capacity to engage in analytical evaluation. Accordingly, the present study conceptualizes fake news handling as a multidimensional construct encompassing detection, verification, critical evaluation, emotional regulation, and responsible media decision-making. These dimensions reflect complementary aspects of how users can respond to potentially misleading information rather than representing a single act of recognition.

Within this conceptualization, fake news detection refers to the ability to identify indicators of fabricated or misleading information, including inconsistencies in claims, weak or absent sourcing, exaggerated presentation, and sensationalist framing. Source and content verification involves examining the identity and credibility of the publisher, comparing claims across multiple sources, and, where appropriate, using digital verification practices and fact-checking resources. These activities require users to move beyond the immediate presentation of a social media post and examine the evidentiary basis of the information.

Critical content analysis represents a further level of evaluation through which individuals distinguish news from opinion, examine the framing of information, and consider possible biases, interests, or persuasive intentions underlying a message. Such analytical engagement is consistent with the view that media literacy involves the critical interpretation of media messages rather than merely the technical ability to access them (Potter, 2013).

The handling of misleading information also involves an affective component. Emotional response management refers to the ability to recognize and regulate immediate emotional reactions, delay premature judgments, and consider information carefully before sharing or reposting it. This dimension is particularly relevant to social media environments, where emotionally evocative content may attract attention and encourage rapid engagement. Research on the relationship between analytic reasoning and susceptibility to misinformation similarly suggests that the quality of reflective evaluation can influence how individuals assess questionable claims (Martel et al., 2020).

Finally, responsible media decision-making concerns the behavioral choices individuals make after evaluating information. These choices may include refraining from sharing questionable content, reporting suspected misinformation, seeking additional evidence, or directing others toward more credible sources. Accordingly, fake news handling should be understood as extending beyond detection to include verification, evaluation, emotional regulation, and responsible action (Karakatsani and Pliogou, 2026; Vraga and Tully, 2019).

Taken together, these dimensions provide a multidimensional conceptualization of fake news handling skills that is consistent with the theoretical framework of the present study. They also provide the basis for examining the associations between these skills and the study's measures of digital media literacy and cognitive information processing.

2.5 Critical synthesis of previous research

Previous research provides substantial evidence that digital media literacy is relevant to individuals’ ability to access, evaluate, and respond to information in digital environments. Hobbs (2010, 2017) and Potter (2013, 2019) conceptualize media literacy as extending beyond technical access to include critical interpretation, evaluation, and responsible engagement with media content. However, this literature also suggests that the development of media literacy is not necessarily uniform across all competencies. Accessing and verifying information may be more readily developed than deeper analytical and critical skills, indicating that media literacy should not be treated as a single, homogeneous capability. This distinction is particularly relevant in social media environments, where users are exposed to large volumes of rapidly circulating and emotionally framed information (Wardle and Derakhshan, 2017).

Research on misinformation further indicates that the ability to recognize or evaluate misleading information is related to cognitive processing. Kahneman (2011) emphasizes the distinction between rapid, intuitive processing and more deliberate analytical processing, while Pennycook and Rand (2019) demonstrate the relevance of analytical reasoning to judgments about false and accurate information. At the same time, Guess et al. (2020) indicate that media literacy and awareness do not necessarily translate directly into consistent behavioral responses to misinformation. Similarly, Vraga and Tully (2021) highlight the relevance of news literacy to skepticism and information-related behaviors on social media. Taken together, these studies suggest that possessing media-related knowledge does not automatically ensure the consistent application of critical evaluation skills in real-world digital environments.

The Arab literature provides further contextual evidence for this distinction. Mrah (2022) examined digital media literacy among Moroccan university students, while the present literature review also identifies evidence suggesting that intensive internet or social media use does not necessarily correspond to higher levels of critical awareness or analytical competence among Arab social media users. These findings point to an important distinction between frequency of digital engagement and the quality of information processing. Thus, frequent exposure to digital media may increase familiarity with digital environments without necessarily producing deeper critical evaluation or effective misinformation detection.

Despite these contributions, the existing literature remains fragmented in at least three respects. First, digital media literacy, cognitive information processing, and misinformation-related competencies have often been examined as related but distinct domains, with relatively limited empirical integration of the three constructs within a single comparative framework. Second, much of the available evidence relies on self-reported perceptions or general measures of media literacy rather than distinguishing clearly between perceived competence and demonstrated performance. Third, comparative evidence concerning Arab social media users across different national media and digital environments remains limited. The present study therefore seeks to extend this literature by examining the statistical associations among digital media literacy, cognitive information processing, and perceived fake news handling skills among Arab Social Media Users’ in Saudi Arabia, Egypt, and Yemen. Importantly, these associations are interpreted within the limits of the cross-sectional design and are not treated as evidence of causal relationships.

2.6 Contribution of the study

The present study contributes to the growing body of research on journalism, digital media, and misinformation by examining the role of digital media literacy in shaping Arab Social Media Users’ ability to identify, verify, critically evaluate, and respond to fake news and media disinformation in social media environments. Its contribution can be understood across theoretical, empirical, and practical dimensions.

At the theoretical level, the study extends the application of digital media literacy within an Arab context by examining its relationship with cognitive information processing and fake news handling skills. By bringing these perspectives together within an integrated conceptual framework, the study provides a theoretically informed account of how media-literacy competencies and cognitive evaluative practices are related to individuals’ engagement with potentially misleading news content. This application is particularly relevant to social media environments, where users are exposed to high volumes of rapidly circulating and heterogeneous information.

At the empirical level, the study provides cross-national survey evidence from young participants in Saudi Arabia, Egypt, and Yemen. The findings contribute to the limited empirical literature examining digital media literacy and the ability to deal with misinformation across different Arab national contexts. In particular, the study examines the relationships among digital media literacy, cognitive information processing, and fake news handling skills, thereby providing empirical evidence that can be compared with findings from other cultural and media environments.

At the practical level, the findings may inform educational institutions, media organizations, and policymakers seeking to strengthen young people's ability to navigate increasingly complex digital information environments. The results can be used to inform media and digital literacy initiatives that place greater emphasis on source verification, cross-checking information, critical analysis, recognition of manipulative content, and responsible decisions concerning the sharing and circulation of news. Such efforts may contribute to strengthening digital media citizenship and promoting more critical and responsible engagement with news on social media platforms. Taken together, these contributions position the study within the broader effort to understand how media-literacy competencies and cognitive evaluative practices are associated with individuals’ capacity to navigate misinformation in contemporary digital environments, while also providing evidence from Arab societies that remain comparatively underrepresented in this area of research.

3 Research procedure and methodology

3.1 Type and methodology of the study

The present study falls within the category of descriptive research, which relies on surveying a sample of the public and analyzing patterns of media exposure and forms of audience response. Accordingly, the researchers adopted the quantitative survey method, as it is considered the most appropriate approach for examining correlational relationships among the study variables. More specifically, the study investigates the relationship between the role played by digital media literacy in developing Arab Social Media Users’ skills and their ability to confront fake news and media disinformation across social media networks.

3.2 Population and sample of the study

The study population consisted of Arab Social Media Users’ from Egypt, Saudi Arabia, and Yemen. The study sample included 800 respondents from the three countries. In order to ensure adequate representation of sample diversity, the participants were distributed according to the following variables, Table 1 show the distribution of study sample:

Table 1. Distribution of participants (study sample) according to different demographic variables.

The study employed a non-probability voluntary response sampling approach, whereby participants were invited to participate voluntarily through an electronic questionnaire disseminated across social media platforms. The target population consisted of social media users in Saudi Arabia, Egypt, and Yemen. Eligibility was limited to individuals who used social media platforms; no additional demographic, institutional, or country-specific eligibility criteria were imposed.

The questionnaire was distributed through multiple widely used social media platforms, including Facebook, X (formerly Twitter), Telegram, WhatsApp, LinkedIn, and other relevant social networking platforms. The same recruitment approach was applied across the three countries. No quotas were established for demographic categories or other participant characteristics, and no predetermined number of participants was assigned to particular groups. Participation was entirely voluntary, and no financial, material, academic, or other incentives were offered.

Data collection took place from the beginning of March 2026 through mid-May 2026, with recruitment conducted during the same general fieldwork period across Saudi Arabia, Egypt, and Yemen. The same Arabic-language questionnaire was used in all three countries, and the same recruitment procedures and eligibility criterion were maintained across the national samples.

Several procedures were implemented to improve the integrity and completeness of the dataset. The electronic questionnaire was configured to allow only one response per participant, thereby preventing multiple submissions through the survey interface. In addition, all substantive questions were required fields, meaning that participants could not submit the questionnaire with unanswered items. Consequently, no incomplete questionnaires were included in the dataset. The final sample consisted of 800 valid and complete responses, and no cases were excluded from the final analysis on the basis of missing data or incomplete responses.

Given the voluntary online recruitment strategy, the sample is appropriately considered a non-probability sample and should not be interpreted as statistically representative of all social media users in the three countries. The findings are therefore interpreted in relation to the participating sample, and this sampling limitation is acknowledged when discussing the generalizability of the results.

The demographic composition of the final sample was not established through educational quotas or an institutional sampling frame. Participants were recruited voluntarily through open dissemination of the electronic questionnaire across social media platforms. Consequently, the educational composition of the sample reflects the characteristics of those who chose to participate rather than a predefined attempt to recruit participants from particular educational groups.

The sample was characterized by a relatively high level of educational attainment, with approximately 92% of participants reporting university or postgraduate qualifications. This distribution represents an important characteristic of the obtained sample and is taken into account when interpreting the findings. Because the study employed voluntary, non-probability online sampling, the sample should not be regarded as statistically representative of the broader Arab population or of all social media users in the three countries.

3.2.1 Rationale for country selection

The selection of Saudi Arabia, Egypt, and Yemen was purposive and theoretically informed rather than intended to provide a statistically representative sample of the Arab region. The three countries were selected because they provide meaningful variation in the media and information environments within which social media users encounter, evaluate, and circulate news. Such contextual variation is relevant to the present study because the evaluation of potentially misleading information takes place within broader information environments that differ in their institutional structures, patterns of digital transformation, and conditions of information circulation.

Saudi Arabia represents a rapidly transforming digital media environment characterized by extensive development of social media alongside a strongly regulated media system. Research has characterized Saudi Arabia as an important regional media actor with modern and far-reaching media systems, while highlighting the role of state regulation and security considerations in shaping the domestic media environment (Zayani, 2012). Research on the development of social media networks in Saudi journalism has also documented the growing influence of platforms such as Twitter and YouTube on journalistic practices, audience engagement, and the production of news content (Althiabi, 2017).

Egypt was selected because of its large and historically significant media environment and the substantial transformation of journalism through digital media and social networking platforms. Sakr (2013) documents the transformation of Egyptian journalism and the growing significance of social media, online journalism, and participatory forms of communication. Her analysis also shows how restrictions affecting mainstream media contributed to the movement of political communication into digital spaces, while national political and economic conditions continued to shape the interaction between online and offline media (Sakr, 2013). Research examining Egyptian newsrooms likewise documents the adoption of Internet-based technologies and their growing role in journalistic production (Abdulla, 2014).

Yemen provides a substantially different information environment because of the prolonged political and armed conflict and its consequences for journalism, media trust, access to information, and the circulation of competing narratives. Research on the online information environment during the Yemen conflict found that Twitter networks were strongly organized along political lines, with different factions relying heavily on their own sources and using social media to circulate content reflecting competing political frames (Noman et al., 2018). In addition, a UNESCO and RNW Media survey documented substantial distrust in local media among Yemeni respondents and identified serious challenges related to journalists’ safety, freedom of expression, and access to reliable information (UNESCO, 2020).

Taken together, these cases provide theoretically relevant contextual variation: Saudi Arabia represents a rapidly digitalizing but strongly regulated media environment; Egypt represents a large and complex media environment undergoing substantial digital transformation; and Yemen represents a conflict-affected information environment characterized by fragmented information flows and competing political narratives. The purpose of this cross-national selection is not to classify these countries as representative types of Arab media systems or to claim that national context causes differences in individual media literacy. Rather, the three contexts provide meaningful variation within which the relationships among digital media literacy, cognitive information processing, and fake news handling skills can be examined.

3.3 Hypotheses

The study seeks to test the following hypotheses:

• Hypothesis One: There is a statistically significant correlational relationship between the level of digital media literacy among Arab social media users’ and the level of cognitive processing of news information across social media networks.
• Hypothesis Two: There is a statistically significant correlational relationship between the level of cognitive processing of news information and the level of fake news detection and handling skills among Arab social media users.
• Hypothesis Three: There is a statistically significant correlational relationship between the level of digital media literacy and the level of fake news handling skills across social media networks.
• Hypothesis Four: There are statistically significant differences in the level of digital media literacy among Arab social media users’ attributable to the following variables: gender, educational level, socioeconomic status, country, and intensity of social media use.
• Hypothesis Five: There are statistically significant differences in fake news handling skills attributable to the following variables: gender, educational level, socioeconomic status, country, and intensity of social media use.
• Hypothesis Six: There are statistically significant differences in the level of cognitive processing of news information attributable to the following variables: gender, educational level, socioeconomic status, country, and intensity of social media use.

3.4 Research tool

For the purpose of collecting the data necessary for the study, the researchers designed an electronic questionnaire consisting of a number of questions and scales focusing on the rate of exposure to social media sites, the most frequently used platforms, the respondents’ acquired digital media literacy values, and their levels of exposure to fake news and media disinformation.

3.41 Clarification of the measurement of fake news handling skills

It is important to clarify that the present study assesses participants’ perceived skills in handling fake news and misinformation on social media rather than their objectively demonstrated ability or actual behavioral performance in detecting fake news. These skills were measured through participants’ self-reported responses to questionnaire items addressing their perceived ability to identify potentially misleading content, evaluate and verify information sources, critically assess news content, manage emotional responses to potentially misleading information, and make informed decisions when encountering news on social media platforms.

Accordingly, the findings should be interpreted as reflecting participants’ perceived competence and confidence in applying fake news handling skills, rather than as direct evidence of their actual performance or verified ability to distinguish between true and false information. This distinction is particularly important given the cross-sectional, self-report nature of the research design. Therefore, throughout the manuscript, terminology has been revised to refer to “perceived skills in handling fake news,” “perceived competence,” or related non-causal formulations, rather than implying objectively demonstrated “capacity to combat fake news.” This clarification ensures that the interpretation of the findings remains consistent with the measurement approach and does not extend beyond what can be empirically supported by the collected data.

3.5 Statistical procedures used in the study

After verifying the validity and reliability of the questionnaire, the instrument was reviewed and administered. Following data collection, responses were coded according to the type of questionnaire item. The data were then entered, analyzed, and statistically processed using the Statistical Package for the Social Sciences (SPSS Version 28).

Initially, a normality test was conducted, revealing that the data followed a normal distribution. Pearson Correlation Coefficients were used to assess internal consistency validity and examine relationships among variables. Cronbach's Alpha was employed to measure reliability. Frequencies and percentages were also used to describe the characteristics of the research sample and present descriptive data.

In addition, arithmetic means and standard deviations were calculated to determine the ranking of items. The Independent Samples T-Test was used to test the significance of differences between two groups, while One-Way ANOVA was employed to examine differences among multiple groups. Post hoc comparisons were conducted using the Tukey's HSD test.‏

3.5.1 Reliability coefficients

Table 2 indicate A Cronbach's Alpha coefficient of 0.70 or above is generally considered good, while values between 0.60 and 0.70 may be acceptable in exploratory research (Hair et al., 2019), the Cronbach's Alpha coefficient for all variables and dimensions combined reached 0.98, which is considered extremely high and reflects excellent internal consistency among the data.

Table 2. Reliability coefficients (Cronbach's Alpha).

At the level of the main variables, the Cronbach's Alpha values for digital media literacy, cognitive information processing, and fake news handling on social media networks were also very high, reaching 0.93, 0.93, and 0.97, respectively. At the level of the subdimensions of each variable, the coefficients ranged from good to very good, as illustrated in the table above.

The same questionnaire was administered in Arabic to participants in Saudi Arabia, Egypt, and Yemen. The instrument was originally developed in Arabic; therefore, no translation or back-translation procedure was required. The questionnaire items, response categories, and measurement procedures were kept identical across the three countries to ensure procedural consistency.

Data collection was conducted simultaneously and within a closely comparable time period across the three countries using an electronic questionnaire. The recruitment procedure was standardized across the three settings. In particular, the questionnaire was disseminated through the same types of social media platforms and under comparable recruitment procedures in each country. The primary eligibility criterion was that participants were active users of social media platforms. No additional country-specific eligibility criteria were applied.

To assess the internal consistency of the study measures, reliability coefficients were calculated for the scales used in the study. The coefficients indicated satisfactory to high internal consistency across the measures.

In addition, given the cross-national nature of the study, the revised analysis examined whether the measurement structure was comparable across the three country groups through measurement invariance testing.

“To address item redundancy transparently, corrected item-total correlation coefficients were calculated (ranging between.48 and.75), along with McDonald's omega coefficient (.98). Furthermore, a theoretical note was added to clarify that high alpha values in broad, multidimensional constructs may reflect content overlap. To present accurate diagnostic results, Kolmogorov–Smirnov and Shapiro–Wilk tests were evaluated alongside skewness values (ranging from −.334 to −1.103) and kurtosis values (ranging from −1.103 to 1.210). Visual inspection of Q-Q plots confirmed acceptable normality for parametric analyses.”

4 Results

The results presented in Table 3 indicate that the intensity of exposure to social media networks among social media users is relatively high. The largest proportion of respondents (31.8%) reported using social media for more than five hours daily, followed by those using it for one to less than three hours and three to less than five hours. Respondents who used social media for less than one hour constituted the smallest group (10%). These findings indicate that intensive and regular social media use was common within the study sample.

Table 3. Daily frequency of social media use among the study sample.

Table 4 further demonstrates a high frequency of daily access to social networks. Approximately 33.8% of respondents reported accessing these platforms between four and seven times daily, while 22% accessed them more than fifteen times per day. This indicates a pattern of repeated interactive engagement rather than passive exposure.

Table 4. Frequency of daily access to social media networks.

Table 5 reveals that nearly two-thirds of the sample (68%) had more than six years of experience using social media networks. This suggests that the participants possess substantial accumulated digital experience that qualifies them to engage in advanced forms of interaction and analysis. These findings indicate that the users’ included in the study can be characterized as “digitally experienced users.”

Table 5. Duration of social media use according to number of years.

Table 6 indicates that overall social media use is distributed across three main categories. Moderate use represented the largest proportion (41.5%), followed by high use (39.5%), while low use accounted for 19% of the sample. This distribution reflects a relatively balanced pattern leaning toward intensive usage, as more than 80% of respondents fall within the moderate and high usage categories. Such findings indicate that social media use was well integrated into the daily routines of the respondents, with the majority reporting moderate to high levels of engagement.

Table 6. Levels of social media use according to country.

Table 7 illustrates the extent to which respondents were exposed to various social media networks. The findings show that WhatsApp ranked first with a very high level of continuous use (Mean = 4.53), followed by YouTube (Mean = 3.46) and Instagram (Mean = 3.42). LinkedIn recorded the lowest level of use (Mean = 2.15).

Table 7. Most frequently used social media networks.

The dominance of WhatsApp reflects the interpersonal and group-oriented communicative nature of Arab societies, as well as a preference for closed communication networks over open platforms. The moderate use of platforms such as Facebook and TikTok indicates a balance between entertainment and informational functions.

The relatively low use of LinkedIn suggests limited professional engagement compared to the predominant social use patterns among participants from the three countries.

The data presented in Table 8 indicate that the overall level of digital media literacy among the participating in the study was high. The overall mean score reached 4.19 with a standard deviation of 0.886 and a percentage weight of 83.5%, reflecting a high degree of agreement regarding respondents’ possession of media literacy skills related to engaging with social media content.

Table 8. Distribution of the study sample according to dimensions of digital media literacy toward social media content.

Table 9 demonstrates the level of cognitive information processing among the study participants when exposed to social media content through several dimensions associated with critical thinking processes.

Table 9. Dimensions of cognitive processing of social media content among the study sample.

Table 10 shows the level at which young people engage with fake news on social media networks. The data reveal that the overall mean score reached 4.17 with a standard deviation of 0.884 and a percentage weight of approximately 83.9%, suggesting that the participants adopt a relatively high level of conscious and responsible behavior in handling fake news, with a general tendency toward agreement.

Table 10. Distribution of the study sample according to level of engagement with fake news encountered through social media platforms.

4.1 Hypotheses testing

H1: There is a statistically significant correlation between the level of digital media literacy among Arab social media users’ and the level of cognitive processing of news information on social media platforms.

Table 11 indicates a strong statistically significant correlation between the level of digital media literacy and the level of cognitive processing of news information on social media networks. Pearson's correlation coefficient reached (0.784) at a significance level of (p = < 0.001).

Table 11. Relationship between level of digital media literacy and level of cognitive processing of news information.

The first hypothesis proposed the existence of a statistically significant correlation between the level of digital media literacy and the level of cognitive processing of news information on social media networks. The findings revealed a high Pearson correlation coefficient (r = 0.784) at a significance level of (p = < 0.001), indicating a strong positive statistical relationship between the two variables.

This result suggests that higher levels of digital media literacy among Arab social media users’ in the three societies are associated with greater capacity for in-depth cognitive processing of news content, including comprehension, analysis, and critical evaluation The strength of the correlation indicates that higher levels of digital media literacy were strongly associated with higher reported levels of cognitive information processing among the participants. This finding suggests a meaningful statistical relationship between the two constructs, rather than providing evidence of a causal effect of digital media literacy on cognitive processing.

H2: There is a statistically significant correlation between the level of cognitive processing of news information and the level of fake news detection and management skills among Arab social media users.

Table 12 indicates a strong statistically significant correlation between the level of cognitive processing of news information on social media platforms and the level of fake news detection and management skills among Arab social media users, with Pearson's correlation coefficient reaching (0.782) at a significance level of (p = < 0.001).

Table 12. Relationship between the level of cognitive processing of news information and the level of fake news detection and handling.

The second hypothesis assumed the existence of a correlation between the level of cognitive processing of news information and fake news detection and management skills. The results revealed a strong correlation coefficient (r = 0.782) at a significance level of (p = < 0.001), thereby confirming the validity of the hypothesis.

This finding indicates that individuals with higher levels of cognitive processing are more capable of distinguishing between accurate and misleading information, reflecting the role of higher-order cognitive processes, such as critical analysis and logical reasoning, in confronting the phenomenon of fake news. It also demonstrates that skills related to detecting media misinformation are not isolated competencies but are closely linked to the depth of cognitive processing.

This relationship is consistent with theoretical perspectives asserting that detecting fake news requires systematic content processing, including source verification, contextual analysis, and bias detection. Accordingly, the observed association is consistent with the theoretical view that systematic cognitive processing is relevant to the evaluation of potentially misleading information.

H3: There is a statistically significant correlation between the level of digital media literacy and the skills of handling fake news on social media platforms.

Table 13 demonstrates a strong statistically significant correlation between the level of digital media literacy and the skills of handling fake news among Arab social media users, as Pearson's correlation coefficient reached (0.763) at a significance level of (p = < 0.001).

Table 13. Relationship between digital media literacy and perceived fake news handling skills.

This result indicates that higher levels of digital media literacy were strongly associated with higher perceived fake news handing skills among social media users. The strength of the association suggests that participants with higher reported levels of digital media literacy also tended to report stronger perceived abilities to identify and critically evaluate potentially misleading content. However, given the cross-sectional design and correlational analysis, this finding should be interpreted as a statistical association rather than evidence that digital media literacy causes improvements in fake news handling skills.

H4: There are statistically significant differences in the level of digital media literacy among Arab social media users’ attributable to the following variables: gender, educational level, socioeconomic status, country, and intensity of social media use.

The data presented in Table 14 reveal variation in the influence of demographic variables on the level of digital media literacy among Arab social media users’ across the three societies.

• No statistically significant differences were found between male and female respondents regarding the level of digital media literacy. The calculated T-value was (−1.495), which is not statistically significant at a significance level of (0.135). This indicates that the level of media literacy is not affected by respondents’ gender.
• Conversely, the table reveals statistically significant differences based on the respondents’ educational level regarding their digital media literacy, favoring the “Postgraduate” group. The mean score for postgraduates was $M = 4.46$, compared to $M = 4.06$ for the pre-university group and $M = 4.38$ for the undergraduate group. The difference among means was statistically significant, with an $F$-value of $F = 8.068$ ($p <.001, df = 797$).
• To determine the source of these differences, post-hoc comparisons were conducted using the Tukey HSD test. The results indicated statistically significant differences between the pre-university group and both the undergraduate group (mean difference $MD = −0.322, p = .001$) and the postgraduate group (mean difference $MD = −0.399, p < .001$). The differences in both comparisons favored the undergraduate and postgraduate levels, respectively. However, no statistically significant differences were found between undergraduates and postgraduates, with a mean difference of $MD = −0.077$ ($p = .439$).
• These findings indicate that the digital media literacy level among pre-university participants was significantly lower than that of undergraduate and postgraduate participants, whereas no significant difference was observed between the latter two groups. Despite the presence of statistically significant differences across educational levels, the overall effect size for educational level appeared relatively small, with an Eta Squared ($\eta^2$) value of approximately $0.020$.
• In contrast, the findings revealed no statistically significant differences according to respondents’ socioeconomic status with respect to digital media literacy. The F-value reached (0.936) at a significance level of (0.393) and a degree of freedom of (797) indicating that socioeconomic status does not affect respondents’ level of digital media literacy.
• Similarly, no statistically significant differences were found according to country. The F-value reached (0.922), which is not statistically significant at a significance level of (0.398) and a degree of freedom of (797), indicating that respondents’ country does not influence their level of digital media literacy.
• On the other hand, the table revealed statistically significant differences based on the intensity of social network usage regarding digital media literacy levels, favoring the “High Intensity” group (M = 4.53), followed by the medium-use group (M = 4.37) and low-use group (M = 4.04). The difference among means was statistically significant, with F = 27.707 (p < .001, df = 797).
• To identify the sources of these differences, post-hoc comparisons were performed using the Tukey HSD test. The results showed statistically significant differences between low-intensity and medium-intensity users (MD = −0.328, p < .001) in favor of medium-intensity users. Significant differences were also found between low-intensity and high-intensity users (MD = −0.492, p < .001) in favor of high-intensity users. Additionally, significant differences emerged between medium-intensity and high-intensity users (MD = −0.164, p = .005) in favor of high-intensity users. The effect size was calculated at eta^2 = .065, indicating that usage intensity accounts for approximately 6.5% of the variance in digital media literacy levels.

Table 14. Differences in the level of digital media literacy among arab social media users’ according to demographic and other variables.

This finding is consistent with the study conducted by Alshohaib (2022), which confirmed that intensive exposure to digital media contributes to the accumulation of technical and critical experience among users, thereby increasing their awareness of the risks associated with false content.

H5: There are statistically significant differences in perceived fake news handling skills among Arab social media users according to gender, educational level, socioeconomic status, country, and intensity of social media use.

The Table 15 shows no statistically significant differences in fake news handling skills according to respondents’ gender, where the calculated T-value (−0.722) was not significant at the 0. 471 significance level. This indicates that no statistically significant gender differences were observed in perceived fake news handling skills.

Table 15. Differences in handling fake news skills according to demographic and other variables.

Similarly, no statistically significant differences were found according to educational level, as the F-value (2.614) at a significance level of 0.071 indicates that no statistically significant differences were observed in perceived fake news handling skills across educational-level groups.

Likewise, socio-economic status shows no statistically significant differences (F = 0.183, Sig = 0.833, indicating that No statistically significant differences were observed in perceived fake news handling skills across socioeconomic groups.

The same applies to country of residence, where no statistically significant differences were found (F = 0.555, Sig = 0.574, meaning No statistically significant differences were observed in perceived fake news handling skills across the three countries.

On the other hand, the table demonstrates statistically significant differences based on the intensity of social network usage regarding fake news handling skills, favoring the “High Intensity” group (M = 4.51), followed by the medium-use group (M = 4.36) and the low-use group (M = 4.00). The mean difference was statistically significant, with F = 26.169 (p < .001, df = 797).

To identify the sources of these differences, post-hoc comparisons were conducted using the Tukey HSD test. The results revealed statistically significant differences between low-intensity and medium-intensity users (MD = −0.355, p < .001), between low-intensity and high-intensity users (MD = −0.513, p < .001), and between medium-intensity and high-intensity users (MD = −0.157, p = .015). Across all comparisons, differences favored the group with higher usage intensity. These findings indicate that participants in the high-intensity use group reported higher perceived fake news handling skills than those in the medium- and low-intensity groups, with a moderate effect size (η² = .062).

H6: There are statistically significant differences in the level of cognitive processing of news information attributable to the following variables (gender, educational level, socio-economic status, country, and intensity of use).

The data presented in the Table 16 indicate that there are statistically significant differences between respondents regarding the level of cognitive processing of news information based on gender. The calculated t-value was −2.882, which is statistically significant at a significance level of p = 0.004 and degrees of freedom df = 798. Despite the statistical significance of the difference, the effect size was small, with Cohen's d approx 0.20, indicating that gender differences are limited in terms of practical magnitude.

Table 16. Differences in cognitive processing of news information according to demographic and other variables.

The data also reveal statistically significant differences based on respondents’ educational level regarding cognitive processing of news information, favoring the “Postgraduate” category (M = 4.39), compared to the “Pre-University” category (M = 4.06) and the “Undergraduate” category (M = 4.37). The difference among means was statistically significant, with an F-value of F = 5.409 (p = .005, df = 797), while the effect size was small (Eta Squared eta^2 = .013).

To determine the sources of these differences, post-hoc comparisons were conducted using the Tukey HSD test. The results demonstrated statistically significant differences between pre-university and undergraduate participants (MD = −0.305, p = .004), as well as between pre-university and postgraduate participants (MD = −0.331, p = .007). The differences in both comparisons favored the undergraduate and postgraduate levels, respectively. In contrast, no statistically significant differences were observed between undergraduate and postgraduate participants (MD = −0.027, p = .918). These results suggest that differences in cognitive information processing are concentrated between the pre-university group and both undergraduate and postgraduate groups, whereas cognitive processing levels remain relatively comparable between undergraduates and postgraduates.

Furthermore, the data showed no statistically significant differences across respondents’ socioeconomic status regarding cognitive processing of news information (F = 2.229, p = .108, df = 797), implying that socioeconomic status does not affect their level of cognitive processing.

Similarly, no statistically significant differences were found based on the country of residence regarding respondents’ level of cognitive processing of news information (F = 1.114, p = .329, df = 797), indicating that the country of residence does not impact cognitive news processing levels.

On the other hand, the table demonstrates statistically significant differences based on the intensity of social network usage regarding cognitive processing of news information, favoring the “High Intensity” group (M = 4.51), followed by the medium-use group (M = 4.32) and the low-use group (M = 4.07). The mean difference was statistically significant, with F = 19.233$ (p < .001, df = 797).

The effect size was eta^2 = .046, indicating that usage intensity accounts for approximately 4.6% of the variance in cognitive information processing, which represents a small-to-medium effect size.

To identify the sources of these differences, post-hoc comparisons were conducted using the Tukey HSD test. The results revealed statistically significant differences between low-intensity and medium-intensity users (MD = −0.253, p = .001), between low-intensity and high-intensity users (MD = −0.441, p < .001), and between medium-intensity and high-intensity users (MD = −0.187, p = .003). These findings indicate that the level of cognitive information processing increases significantly as usage intensity rises, with differences in all pairwise comparisons favoring the group with higher usage intensity.

5 Discussion

The present study examined the relationships among digital media literacy, cognitive processing of news information, and perceived fake news handling skills among Arab social media users, while also considering differences associated with social media use intensity and selected demographic characteristics. Overall, the findings provide evidence of meaningful associations among the study constructs, but they also reveal an uneven distribution of digital media competencies. The results indicate relatively stronger self-reported competencies in information verification, source comparison, and connecting new information with prior knowledge, whereas higher-order competencies such as critical content analysis, fallacy detection, and emotional response regulation were comparatively less pronounced. This pattern suggests that digital media literacy should be understood as a multidimensional competency rather than as a uniform ability. At the same time, because the study employed a cross-sectional self-report design, the findings should be interpreted as associations and perceived competencies rather than as evidence of causal effects or objectively demonstrated behavioral performance.

The relatively stronger performance reported for verification-related competencies is consistent with the conceptualization of digital media literacy proposed by Hobbs (2010, 2017) and Potter (2013, 2019), which extends beyond technical access to include the ability to analyze, evaluate, and responsibly engage with media content. The present findings, however, suggest that these competencies may not develop equally. Respondents may find observable and procedural activities, such as comparing sources or checking information, more accessible than cognitively demanding activities involving the analysis of persuasive content, identification of logical fallacies, or regulation of emotional responses. This distinction is particularly relevant in contemporary digital environments, where information is frequently presented in rapid, emotionally salient, and highly shareable forms. Wardle and Derakhshan (2017) emphasize the complexity of the information disorder environment, while the cognitive perspective developed by Kahneman (2011) highlights the distinction between rapid intuitive processing and more deliberate analytical processing. The findings therefore support the importance of distinguishing procedural verification from deeper critical evaluation when conceptualizing digital media literacy (Al-Hazwarah and Haidar, 2020).

The relationship between cognitive processing and perceived fake news handling skills provides further support for the relevance of cognitive evaluation in digital information environments. The results suggest that respondents reporting stronger cognitive processing competencies also tended to report stronger abilities to deal with potentially misleading news. This finding is broadly consistent with research emphasizing the importance of analytical reasoning in evaluating misinformation, including the work of Pennycook and Rand (2019). However, the observed association should not be interpreted as evidence that cognitive processing causes improved fake news handling. Because the relevant constructs were measured through self-reported responses collected within the same cross-sectional survey, the association may partly reflect shared perceptions of competence. Nevertheless, the finding reinforces the theoretical value of considering cognitive evaluation processes alongside media-literacy competencies when examining how young people engage with questionable or misleading information.

The strong association observed between digital media literacy and perceived fake news handling skills is also theoretically and practically relevant. Participants who reported greater digital media literacy also tended to report higher perceived competence in dealing with potentially misleading news. This pattern is consistent with the broader literature suggesting that media literacy and news literacy are relevant to individuals’ responses to misinformation (Guess et al., 2020; Vraga and Tully, 2021). However, the magnitude of this association should be interpreted cautiously. The present study measures perceived competencies rather than independently verified performance, and therefore cannot determine whether participants who report higher media literacy are objectively more successful in identifying misinformation. The finding is consequently best interpreted as evidence of a strong statistical association between perceived digital media literacy and perceived fake news handling skills within the study sample.

The findings concerning social media use intensity require a similarly cautious interpretation. Participants reporting higher levels of social media use also reported higher levels of digital media literacy, cognitive processing, and fake news handling skills. One possible interpretation is that more frequent engagement with digital environments may provide users with greater familiarity with online information practices and more opportunities to encounter verification activities. Nevertheless, frequency or duration of use should not be equated with quality of engagement or critical competence. Heavy users may encounter more information without necessarily evaluating it more critically. Thus, the present findings do not establish that intensive social media use produces higher digital media competencies. Rather, they indicate an association between reported use intensity and perceived competencies at the time of data collection. This distinction is important because frequent exposure may increase familiarity and confidence without necessarily corresponding to deeper analytical engagement.

The gender-related findings should also be interpreted conservatively. Although a statistically significant difference was observed for cognitive processing, the magnitude of the difference was relatively limited. Accordingly, the finding should not be interpreted as evidence of inherent differences between male and female respondents in their ability to process news information or in their media consumption styles. The present study did not directly measure gender-specific patterns of news consumption or the cognitive mechanisms that might account for the observed difference. The result is therefore better understood as a sample-specific statistical association. Sampling variation and the possibility of Type I error should also be considered when interpreting relatively small group differences. Future studies using larger and more diverse samples and direct measures of cognitive processing and news consumption behavior would be better positioned to determine whether such differences are robust and theoretically meaningful.

The absence of statistically significant differences across Saudi Arabia, Egypt, and Yemen is also important to interpret carefully. This finding does not imply that the three national contexts are identical in terms of media systems, internet infrastructure, political communication environments, or digital inequalities. Rather, it indicates that country membership did not account for statistically significant variation in the measured constructs within the present sample. Differences in access to digital technologies, patterns of news consumption, media institutions, and opportunities for political communication may nevertheless shape how young people encounter, evaluate, and share information. The absence of a significant country effect should therefore not be interpreted as evidence that contextual differences are irrelevant. Instead, it suggests that these contextual differences were not sufficiently reflected in the observed differences between the participating groups. Future comparative studies using larger probability-based samples and measurement-invariance procedures could provide a more rigorous assessment of whether national contexts shape these relationships.

The findings also have implications for the theoretical integration of Digital Media Literacy and Cognitive Information Processing. Digital media literacy can be understood as the broader competency domain encompassing access, analysis, evaluation, and responsible engagement with digital information, whereas cognitive information processing provides a complementary perspective for understanding how individuals interpret and evaluate information. The present findings support the conceptual relevance of considering these dimensions together, but they do not establish a causal or mediating pathway between them. In this respect, the results provide empirical support for a complementary rather than deterministic relationship between media-related competencies and cognitive evaluation processes. This interpretation is consistent with the multidimensional character of media literacy described by Hobbs and Potter and with the emphasis on analytical reasoning within cognitive approaches to misinformation evaluation.

From a practical perspective, the findings suggest that digital media literacy interventions should move beyond teaching basic access and verification procedures alone. Training programs targeting young people may benefit from placing greater emphasis on higher-order competencies, including critical content analysis, identification of logical fallacies, evaluation of emotionally framed claims, comparison of competing sources, and deliberate reflection before sharing information. Such interventions should also distinguish between frequent digital engagement and effective digital engagement. High levels of social media use should not automatically be treated as evidence of digital competence. Instead, educational initiatives should focus on the quality of information practices and on the ability to apply critical evaluation consistently across different digital contexts.

The cross-national findings further suggest that interventions should be sensitive to contextual conditions rather than assuming that a single national or cultural explanation accounts for differences in digital media competencies. Although no statistically significant country differences were identified in the present study, differences in digital infrastructure, media environments, internet accessibility, and opportunities for political communication remain relevant contextual considerations. Accordingly, media-literacy programs may benefit from maintaining common core competencies while adapting examples, training materials, and misinformation scenarios to the specific information environments in which young people operate.

Several limitations should be considered when interpreting the findings. First, the cross-sectional design prevents conclusions about temporal change, developmental trajectories, or causal relationships among the study variables. Second, the reliance on self-reported measures means that the findings primarily capture perceived competencies and may be influenced by response tendencies, confidence, or social desirability. Third, the absence of an objective misinformation-identification task limits the extent to which perceived fake news handling skills can be equated with demonstrated accuracy. Fourth, although the study provides a comparative perspective across three Arab countries, the findings should not be generalized to all Arab social media users because of differences in sampling, social conditions, and digital environments across the wider region. Finally, the possibility of common-method bias cannot be completely excluded because the principal constructs were assessed using the same survey-based method. Future research should therefore incorporate behavioral or performance-based measures, multiple data sources, longitudinal designs, and probability-based samples where feasible. Formal mediation or longitudinal models could also be employed in future studies to determine whether the associations identified here reflect theoretically meaningful temporal pathways.

Overall, the study contributes to the literature by examining digital media literacy, cognitive processing of news information, and perceived fake news handling skills within an integrated comparative framework involving Arab social media users. The findings suggest that digital media competence is multidimensional, that cognitive evaluation is relevant to the way young people perceive their ability to handle misleading information, and that frequency of social media use should not be equated with the quality of digital engagement. At the same time, the results underscore the importance of avoiding causal or developmental interpretations that cannot be established from a cross-sectional design. The study therefore provides a contextually grounded account of the associations among these constructs while identifying clear directions for future research based on objective performance measures, longitudinal evidence, and more rigorous cross-national comparisons.

6 Limitations

Several limitations should be considered when interpreting the findings of this study.

First, the cross-sectional design limits the ability to establish temporal precedence, causal direction, or developmental change among digital media literacy, cognitive processing of news information, and perceived fake news handling skills. Second, the study relied on self-reported questionnaire measures. perceived competencies, rather than objectively verified.

Third, the study relied on a non-probability voluntary online sampling approach with a relatively high level of educational attainment (92% university/postgraduate), which may limit generalizability to broader populations.

Fourth, no formal mediation or longitudinal models were conducted to establish causal pathways.

Fifth, although the study considered gender, educational level, socioeconomic status, country, and social media use intensity, other potentially relevant factors were not examined. These may include the type of platforms used, patterns of news exposure, frequency of fact-checking, trust in different information sources, political interest, and prior training in media or digital literacy. Their omission means that the present findings should not be interpreted as providing an exhaustive explanation of variations in digital media literacy or fake news handling skills. Also Sampling and Generalizability. The study relied on a voluntary non-probability online sampling approach, which may have resulted in self-selection bias. In particular, the sample was characterized by a relatively high level of educational attainment, with approximately 92% of participants reporting university or postgraduate qualifications. Although no educational quotas or institutional sampling frames were used in recruitment, the overrepresentation of highly educated participants may limit the generalizability of the findings to social media users with different educational backgrounds. In addition, the online recruitment strategy may have favored individuals who are more digitally engaged or more willing to participate in online surveys. Accordingly, the findings should be interpreted as reflecting the participants included in the present cross-national sample rather than as statistically representative estimates of Arab social media users or the broader Arab population. Future research should employ probability-based, stratified, or quota-based sampling designs to achieve greater demographic balance and strengthen the generalizability of findings.

Sixth, the study did not employ a formal mediation model or longitudinal analysis. Accordingly, the observed associations among digital media literacy, cognitive processing, and perceived fake news handling skills should not be interpreted as evidence of a mediating mechanism or indirect causal pathway. The conceptual integration of these constructs provides a theoretical basis for examining their relationships, but formal tests of mediation and temporal ordering require appropriate analytical designs and additional empirical evidence.

Finally, common-method variance cannot be completely excluded because the principal constructs were assessed through the same questionnaire and from the same respondents. Although procedural and statistical measures were considered to reduce this concern, the possibility of shared measurement effects remains. Future studies could address this limitation by combining multiple data sources, separating measurements temporally, and incorporating objective behavioral indicators. Taken together, these limitations do not invalidate the findings, but they define the appropriate scope of their interpretation and identify important directions for future research.

7 Recommendations

Based on the empirical findings of the present study, several recommendations can be proposed.

First, digital media literacy initiatives targeting Arab social media users should place greater emphasis on higher-order critical competencies, particularly critical analysis of media content, identification of logical fallacies, evaluation of source credibility, and systematic verification of information. This recommendation is directly supported by the finding that competencies related to critical content analysis, fallacy detection, and emotional response regulation were comparatively less pronounced than some procedural verification skills.

Second, educational and media-literacy programs should move beyond basic access and information-search skills toward structured practices for evaluating the credibility and evidentiary basis of digital news. Training activities may include comparing multiple sources, identifying unsupported claims, distinguishing evidence from opinion, recognizing emotionally framed or misleading content, and verifying information before sharing it. Such practices are consistent with the study's finding that verification-related competencies are relevant to perceived fake news handling skills.

Third, the significant association observed between social media use intensity and the study variables suggests that interventions should focus not simply on the amount of social media use but on the quality of users’ engagement with information. Accordingly, media-literacy programs should encourage deliberate evaluation, source comparison, and verification practices among frequent social media users rather than assuming that frequent exposure itself represents higher digital competence.

Fourth, because the study identified significant associations among digital media literacy, cognitive processing of news information, and perceived fake news handling skills, educational interventions should integrate media-literacy activities with cognitive evaluation strategies. Such activities may encourage users to pause before accepting or sharing information, consider alternative interpretations, examine the supporting evidence, and distinguish rapid intuitive judgments from more deliberate evaluation. These recommendations are intended to strengthen the competencies examined in the present study rather than to imply a demonstrated causal effect of one construct on another.

Fifth, the absence of statistically significant differences across Saudi Arabia, Egypt, and Yemen suggests that the core competencies identified in this study may provide a useful basis for broadly applicable media-literacy initiatives across the participating contexts. However, the absence of significant country differences should not be interpreted as evidence that the three media environments are identical. Accordingly, implementation should retain common competency objectives while allowing educational content and examples to be adapted to the specific media, technological, and information environments of each country.

Finally, future research should strengthen the empirical basis for intervention design by incorporating objective misinformation-identification tasks, behavioral measures, longitudinal designs, and larger probability-based samples. Such approaches would help determine whether improvements in the competencies identified in the present study are associated with measurable improvements in actual misinformation detection and evaluation performance.

8 Specific Future Research Directions:

• Empirical and Objective Performance Tasks: Recommending the inclusion of objective tasks for detecting misinformation, behavioral measures, and fact-checking performance indicators to go beyond self-reported measures.
• Longitudinal and Causal Modeling: Proposing longitudinal designs and formal structural equation modeling (SEM)/mediation analyses to test trend relationships and cognitive pathways over time.
• Expanding Comparative and Probability Sampling: Proposing probability-based stratified sampling across broader demographic, socioeconomic, and social contexts in the Arab region, as well as testing the reliability of formal measurement across national media environments.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Temple Uwalaka, University of Canberra, Australia

Reviewed by: Iryna Mihus, Scientific Center of Innovative Research, Estonia

Abd Allah Sultan Aljalabneh, Zarqa University, Jordan