Educenter : Jurnal Ilmiah Pendidikan Vol 4 No 3 2025 ISSN: 2827-8542 (Prin.
ISSN: 2827-7988 (Electroni.
Open Access: https://jurnal.
id/index.
php/educenter/index The role of character education and social media literacy in strengthening national resilience among Generation Z in Timor-Leste Albino Mauzino1*.
Muhammad NursaAoban2.
Agus Martins3.
Herianto4
1,2,3
Universitas Negeri Yogyakarta.
Indonesia Institut Turatea Indonesia.
Indonesia 2024@student.
id*, 2m_nursaban@uny.
id, 3agus9001pasca.
2023@student.
antoherianto47@gmail.
Article Info
ABSTRACT
Article history:
Amidst globalization and digital disruption, a nation's resilience is tested, especially among Generation Z.
This study aims to analyze the role of character education and social media literacy in strengthening national resilience among Generation Z in TimorLeste.
A quantitative survey with a cross-sectional design was employed with 256 high school students (Generation Z) in Dili, selected through purposive sampling.
Data were collected using the Character Values Scale.
Social Media Literacy Scale, and National Resilience Scale.
Data analysis included descriptive statistics.
Pearson correlation, and multiple linear regression.
The findings indicate that both character education .
c = 23, ycy < 0.
and social media literacy .
c = 0.
13, ycy < 0.
have a significant positive relationship with national resilience.
However, in a joint regression model that 7% of the variance in national resilience .
= 7.
32, ycy < 0.
character education emerged as the sole significant predictor ( = 0.
23, p = .
The implication for the field of education is the critical importance of prioritizing a strong character-based curriculum, complemented by digital media literacy, within the national education system to fortify the younger generation against ideological threats.
Received October 3rd 2025 Revised October 30th 2025 Accepted November 14th 2025 Keyword:
National resilience.
Character Social media Generation Z.
TimorLeste A2025 Albino Mauzino.
Muhammad NursaAoban.
Agus Martins.
Herianto.
Published by Arka Institute.
This work is licensed under a Creative Commons AttributionNonCommercial 4.
0 International License.
ttps://creativecommons.
org/licenses/by-nc/4.
INTRODUCTION
As a relatively young nation engaged in the ongoing process of nation-building.
Timor-Leste faces unique challenges in forging and maintaining a unified national identity and resilience (Stead.
Generation Z, born and raised in the post-independence era (Twenge & Park, 2.
and fully integrated with digital technology, represents a pivotal demographic that will determine the nation's future trajectory (Panagiotou et al.
, 2022.
Roblek et al.
, 2.
Their daily lives are inextricably linked to social media, which functions as a distinct double-edged sword: it serves as a primary window to the global community for educational and economic progress, yet it is also the principal arena where their self-presentation and social identity are formed, making them highly susceptible to external ideological and cultural influences (Milkman, 2017.
Pyrez-Escoda et al.
, 2.
This deep digital integration presents a direct and complex challenge for the national education system, which is tasked with preparing students for a reality where the very process of creating an Auimagined communityAy (Albrow et al.
, 1994.
Tolstenko et al.
, 2.
is now either strengthened or fragmented by interactions in the digital space.
The boundless and algorithmically curated flow of information on social media brings with it serious and immediate challenges to national resilience.
The constant exposure to Auinformation disorderAy .
toxic mix of hoaxes, targeted disinformation, and emotionally charged mal-informatio.
can systematically erode patriotism and trust in national institutions (Wardle & Derakhshan, 2.
Furthermore, the architecture of these platforms can create Aufilter bubblesAy and Auecho chambersAy that algorithmically reinforce pre-existing beliefs, thereby accelerating social polarization along ethnic, political, or religious lines (Modgil et al.
, 2021.
Sunstein, 2.
and degrading the shared values that PAG Journal Homepage : https://jurnal.
id/index.
php/educenter/index Educenter : Jurnal Ilmiah Pendidikan Vol 4 No 3 2025 form the bedrock of the nation (Benkler et al.
, 2018.
Halili, 2023.
Nye, 2.
The central problem, therefore, is the vulnerability of digitally-native youth who may lack the internal values and external skills to navigate this complex information environment, posing a direct threat to national cohesion.
Despite the clear urgency of this issue, empirical research that specifically examines the synergistic relationship between character education .
s an internal, value-based fortres.
and social media literacy .
s an external, skill-based shiel.
in the context of Generation Z in Timor-Leste remains markedly limited.
This study aims to fill this critical gap by analyzing how an internal character foundation .
and external digital competencies .
complement each other to form a more comprehensive and robust form of resilience.
This study posits that comprehensive national resilience requires both a strong internal value system and adept external skills.
Therefore, the following hypotheses were formulated to be tested:
H1: There is a positive and significant relationship between the internalization of character values and national resilience.
H2: There is a positive and significant relationship between social media literacy and national H3: The internalization of character values and social media literacy together have a significant effect on national resilience.
National Resilience in the Digital Era The concept of national resilience has traditionally focused on a state's ability to protect its territorial integrity and sovereignty from military aggression.
However, in the digital age, this definition has undergone a critical expansion to include ideological, social, and psychological resilience (Skidmore, 1.
Threats are no longer merely physical but have become informational, targeting the cognitive 'center of gravity' of a society, which includes the shared perceptions, trust in institutions, and social cohesion that bind a nation together (Nye, 2.
In this new landscape, social media platforms transform into a primary battlefield where national narratives can be reinforced or systematically deconstructed with unprecedented speed and scale.
Consequently, the national resilience of Generation Z, who rely heavily on these platforms for information and identity formation, is contingent upon their ability to maintain a positive and robust national identity, exhibit healthy skepticism towards polarizing content, and actively reject divisive narratives that are designed to exploit societal fault lines amidst a constant flood of information (Benkler et al.
, 2.
Character Education as an Internal Foundation Character education serves as an internal moral compass that guides an individual's thoughts, feelings, and actions, particularly when faced with complex ethical dilemmas (Kohlberg & Power.
In the context of nationhood, this educational approach is not merely about teaching facts but about instilling essential civic virtues such as patriotism, tolerance, and integrity, which collectively form the foundation of a strong national identity (Ben-Porath, 2007.
Lickona, 1.
The ultimate goal of this character development is to shape citizens who are not only academically proficient but also ethically and socially mature .
mart and good citizen.
, which is the core of effective civic education (Westheimer & Kahne, 2.
An individual with a strong and well-formed character possesses a Aupsychological immune systemAy that allows them to intuitively filter and resist negative narratives that conflict with their core values, thus making them less susceptible to manipulation and ideological Social Media Literacy as an External Competency If character provides the moral framework, social media literacy provides the set of external skills needed to navigate the digital landscape intelligently and safely.
This ability extends far beyond mere technical proficiency, such as posting content, and delves into higher-order cognitive These include the critical skills needed to identify subtle forms of misinformation, understand the persuasive power of algorithms, and protect personal data from exploitation (Foster.
Livingstone, 2.
The comprehensive media and information literacy framework developed by UNESCO emphasizes the complete cycle of being able to access, critically evaluate, and responsibly create information (Alcolea-Dyaz et al.
, 2.
Without adequate social media literacy.
Generation Z, despite often being labeled as Audigital nativesAy (Prensky, 2.
, remains paradoxically vulnerable to PAG Journal Homepage : https://jurnal.
id/index.
php/educenter/index Educenter : Jurnal Ilmiah Pendidikan Vol 4 No 3 2025 Their technical fluency does not automatically translate into critical understanding, leaving them susceptible to becoming unintentional agents in the spread of harmful narratives (Kahne & Bowyer, 2.
RESEARCH METHODS
Design This study adopts a cross-sectional survey design with a quantitative approach.
This approach was chosen for its ability to measure research variables numerically, allowing for objective statistical analysis to test for correlational and predictive relationships between variables (Creswell.
Creswell, 2.
The cross-sectional design is specifically utilized for its efficiency in gathering data from a large sample at a single point in time, thereby providing a 'snapshot' of the prevailing relationships among character education, social media literacy, and national resilience, although it does not permit causal inferences (Cao et al.
, 2.
Participants The target population was Generation Z in Timor-Leste .
ges 16-.
The final sample consisted of 256 high school students from Dili.
Participants were selected using purposive sampling based on a key inclusion criterion: being an active social media user for a minimum of one hour per The demographic profile of the participants is detailed in Table 1.
Table 1.
Demographic Characteristics of Participants (N=.
Characteristic Gender Age Category Male Female 16 Years 17 Years 18 Years 19 Years Frequency .
Percentage (%) Procedure After securing official consent from the headmasters of the selected schools, an online questionnaire was distributed to students via trusted school channels over a four-week period.
The first page of the questionnaire detailed the informed consent information, emphasizing the voluntary nature of participation and guaranteeing the confidentiality and anonymity of all responses to encourage candid Instruments Three self-report scales were used, with all items measured on a 5-point Likert scale .
= Strongly Disagree to 5 = Strongly Agre.
, and all were adapted from established instruments for revalidation in the local context.
The first, the Internalization of Character Values Scale, was adapted from the Character Strengths and Virtues (CSV) framework (Peterson.
Seligman, 2.
, focusing on citizenship virtues like patriotism and social responsibility, and demonstrated good internal consistency .
aycycuycuycaycaycaEaAoyc yu = 0.
The second, the Social Media Literacy Scale, was adapted from the UNESCO MIL framework (Grizzle, 2.
to assess competencies in information evaluation, privacy management, and digital ethics, showing acceptable reliability .
aycycuycuycaycaycaEaAoyc yu = 0.
Finally, the National Resilience Scale was constructed as a composite measure from multiple instruments, including the Nationalism and Patriotism Scale (Kosterman & Feshbach, 1.
and relevant items from the World Values Survey (WVS), to measure national identity, social cohesion, and vigilance, which also proved to be reliable .
aycycuycuycaycaycaEaAoyc yu = 0.
Data Analysis Data were analyzed using R Studio (Version 4.
The analysis began with data screening for missing values and outliers, followed by descriptive statistics to summarize the variables.
To test the hypotheses, a two-step inferential analysis was conducted at a significance level of p < 0.
First.
Pearson's product-moment correlation was used to examine the bivariate relationships between each PAG Journal Homepage : https://jurnal.
id/index.
php/educenter/index Educenter : Jurnal Ilmiah Pendidikan Vol 4 No 3 2025 independent variable (CE and SML) and the dependent variable (NR), addressing H1 and H2 (Schober et al.
, 2.
Second, a multiple linear regression was conducted to assess the simultaneous predictive power of both independent variables on national resilience, addressing H3 (Cohen et al.
, 2.
The validity of the regression model was confirmed by examining diagnostic plots to ensure key assumptions .
, linearity, normality of residuals, and homoscedasticit.
were met.
RESULTS AND DISCUSSION
The analysis was conducted on a sample of 256 Generation Z participants from Dili.
TimorLeste.
Descriptive statistics for the key variables are presented in Table 2.
On average, participants reported high scores across all three scales, with means clustering near the maximum possible scores.
Table 2.
Descriptive Statistics of Key Variables Variable Mean Min Character Education (CE) Social Media Literacy (SML) 256 National Resilience (NR) Max The distribution of these scores, visualized in Figure 1, shows a strong left-skew for all three This indicates a possible ceiling effect in the self-report measures, where most participants selected the highest possible scores, potentially due to social desirability bias.
This restriction of variance is a limitation that may have suppressed the strength of the observed relationships.
Figure 1.
Score Distributions for Key Variables The first and most striking descriptive finding is the significant ceiling effect in the data.
The mean scores for Character Education .
55 out of .
Social Media Literacy .
56 out of .
, and National Resilience .
31 out of .
are all heavily concentrated at the highest end of the scale.
This extreme skewness indicates that most respondents selected the highest possible response options .
AuStrongly AgreeA.
for nearly all items.
This phenomenon can likely be attributed to social desirability bias, a common issue in selfreport surveys where participants provide answers they deem more socially acceptable rather than answers that reflect their true beliefs or behaviors (Paulhus, 1.
In the context of this study, topics such as Aucharacter,Ay Aupatriotism,Ay and Aunational resilienceAy are highly normative and value-laden.
The participants, being high school students, may have felt implicit pressure to present themselves as good, virtuous, and patriotic citizens, regardless of their actual private feelings.
It is critical to understand that this ceiling effect has significant statistical implications.
When variance in variables is severely restricted .
, almost everyone has the same scor.
, the ability to detect statistical relationships between them is substantially diminished (Schober et al.
, 2.
Correlation coefficients, by their mathematical nature, become attenuated or suppressed.
Therefore, the correlation and regression findings reported in this study .
hile statistically significan.
are highly likely to PAG Journal Homepage : https://jurnal.
id/index.
php/educenter/index Educenter : Jurnal Ilmiah Pendidikan Vol 4 No 3 2025 underestimate the true strength of the real-world relationships.
The AuactualAy relationship between character, literacy, and national resilience is probably much stronger than the instruments were able to capture in this study.
The Fundamental Role of Character Education As hypothesized (H.
, the correlation analysis found a significant positive relationship between Character Education (CE) and National Resilience (NR) .
c = 0.
23, ycy < 0.
This finding robustly aligns with the theoretical pillars of civic education, which posit that the internalization of virtues .
uch as patriotism, tolerance, and social responsibilit.
forms the very bedrock of a stable and resilient citizenry (Lickona, 1996.
Westheimer & Kahne, 2.
This finding can be discussed along three dimensions.
First, character education provides the AuwhyAy: moral reasoning and intrinsic motivation to remain loyal and contribute to the nation.
In Lickona .
model, character has cognitive .
nowing the goo.
, affective .
esiring the goo.
, and conative .
oing the goo.
This finding suggests the affective domain .
sense of patriotism and social responsibilit.
is core to resilience.
This internalized value system functions as a Aupsychological immune system,Ay acting as a cognitive and emotional filter that allows individuals to intuitively reject divisive or anti-national online narratives that contradict their core identity, often before a conscious critical analysis is even performed.
Second, this relationship becomes especially critical in the specific context of Timor-Leste's nation-building process.
As a relatively young nation, instilling a collective identity and a shared set of values is a national priority (Stead, 2.
Here, character education is more than individual moral it is a fundamental tool for forging social cohesion.
In a digital environment where global narratives can easily erode local identity, a strong character foundation instilled by the education system becomes the anchor that holds Generation Z from the currents of de-nationalization.
Third, character serves as the first line of defense in the digital age.
Individuals are constantly bombarded by Auinformation disorderAy (Wardle & Derakhshan, 2.
In the face of deepfakes, sophisticated disinformation, and emotionally charged malinformation, literacy skills alone may be Character Education (CE) acts as the initial filter.
A student with a strong national character will instinctively feel that a post denigrating national symbols or inciting ethnic hatred is AuwrongAy .
he affective dimensio.
, even if the post is presented in a professionally deceptive manner .
literacy The Supportive (Yet Weake.
Role of Social Media Literacy Hypothesis 2 was also supported, revealing a significant, albeit weaker, positive relationship between Social Media Literacy (SML) and National Resilience (NR) .
c = 0.
13, ycy = 0.
This result is consistent with contemporary media studies literature, which argues that the ability to critically evaluate information is a crucial competency for navigating the modern information ecosystem (Allison, 2010.
Kahne & Bowyer, 2.
This weaker correlation, however, is far more interesting than its significance and requires deeper discussion.
This finding may be empirical evidence of the Audigital native paradox.
Ay Generation Z, often labeled Audigital nativesAy (Prensky, 2.
, is undoubtedly technically fluent.
they know how to post, share, and use platforms.
However, this technical fluency often does not translate to critical fluency (Pyrez-Escoda et al.
, 2.
They may still struggle to identify subtle manipulation, understand algorithmic bias, or recognize emotional persuasion.
The SML scale used may have inadvertently captured more technical confidence than true evaluative skill.
Furthermore, this weak correlation may be the first indication of the study's main finding: SML is a tool, and tools are passive.
Possessing a set of skills .
does not guarantee those skills will be used, or used for a pro-national purpose.
This suggests SML is likely a moderating or mediating variable, but not the primary driver.
Finally, the most effective disinformation is often designed to bypass critical analysis and appeal directly to emotional responses (Modgil et al.
, 2.
Content that triggers outrage, fear, or in-group belonging can cause even AuliterateAy users to share it, as their emotional response overrides their rational filter (Sunstein, 2.
In this case.
SML .
rational skillse.
fails against an emotional assault.
PAG
Journal Homepage : https://jurnal.
id/index.
php/educenter/index Educenter : Jurnal Ilmiah Pendidikan Vol 4 No 3 2025 Therefore, the finding for H2 is important not because it shows a relationship, but because it hints at the limitations of SML as a standalone variable.
It provides the AuhowAy .
he tool.
, but this appears to be less critical than the AuwhyAy .
he motivatio.
provided by character.
Figure 2.
Correlation Matrix of Variables The Primacy of Character The most compelling and theoretically significant finding emerged from the multiple linear regression analysis (H.
, detailed in Table 3.
Table 3.
Multiple Linear Regression Predicting National Resilience Std.
Variable p-value Error (Intercep.
< 0.
Character Education (CE) Social Media Literacy (SML) Note.
B = unstandardized coefficient.
= standardized coefficient.
The combined model was significant, ya.
= 7.
32, ycy < 0.
01, which supports H3.
However, the model only explained 4.
7% of the variance in national resilience.
Most importantly, when both variables were entered simultaneously, only Character Education remained a statistically significant predictor .
u = 0.
23, ycy = 0.
The influence of Social Media Literacy became nonsignificant .
cy = 0.
This result does not mean SML is unimportant.
Rather, it suggests a far more nuanced and hierarchical relationship.
It points to the primacy of character.
This finding strongly resonates with social cognitive theory (Bandura, 2.
and media literacy models (Cho et al.
, 2.
that place values and motivation at the core of action.
SML provides the AuhowAy .
he skills to deconstruct propaganda, identify a hoa.
, but CE provides the AuwhyAy .
he motivation to want to defend the natio.
This finding suggests the AuwhyAy is a necessary antecedent to the Auhow.
Ay Consider a student with high SML but low CE.
They might see a piece of disinformation designed to incite ethnic unrest.
Because of their high SML, they know it is a hoax.
However, due to their low CE .
ack of social responsibility or patriotis.
, they may feel no motivation or obligation to counter it.
They may simply Auscroll past,Ay allowing the poison to spread.
Their skill remains passive.
A more dangerous scenario involves motivated reasoning (Sunstein, 2.
A student with high SML but negative CE .
, cynicism towards the stat.
could use their literacy skills to actively seek out and amplify narratives that confirm their negative bias, effectively weaponizing their literacy against national resilience.
This suggests that SML taught in a value-neutral vacuum .
ithout characte.
is not only insufficient but potentially dangerous.
PAG
Journal Homepage : https://jurnal.
id/index.
php/educenter/index Educenter : Jurnal Ilmiah Pendidikan Vol 4 No 3 2025 Character, therefore, appears to function as the catalyst that transforms latent digital competency into active, pro-national behavior.
The internal value system (CE) is what activates SML skills and directs them toward a constructive purpose.
This is why, in the regression model.
CE AuabsorbsAy all the predictive power.
The small effect that SML had in the H2 correlation was likely because that effect is entirely mediated by the student's character.
Addressing the Low Explained Variance It is important to explicitly address the model's low ycIA value .
7%).
While the model is statistically significant .
ndicating the relationship is real and not due to chanc.
, it explains a small fraction of the variance in national resilience.
A critic might see this as a weakness, but in this context, it is an explicable and logical finding.
First, as discussed in detail, the ceiling effect and restricted variance across all three variables mathematically caps the maximum possible RA.
When nearly everyone has the same score on the dependent variable (NR), it becomes statistically impossible for any independent variable to explain much of the AuvarianceAy .
s there is almost no variance to explai.
The low RA here is likely a statistical artifact of the measurement instruments, not a reflection of a weak theoretical relationship.
Second, national resilience is an immensely complex, multi-faceted construct.
It is influenced by a vast number of factors beyond education .
conomic stability, trust in political institution.
(Nye, 2.
, family values, local community cohesion (Benkler et al.
, 2.
, and personal experiences with justice and governance.
It would be unrealistic to expect two educational variables (CE and SML) alone to explain a large portion of the variance in such a broad phenomenon.
Therefore, the value of this regression model is not in its total predictive power .
he ycIA valu.
, but in its ability to clarify the relative importance of the predictors.
The finding that CE statistically outperforms SML within the model is its key theoretical contribution.
Limitations and Future Research Directions This study has several limitations that must be acknowledged and used to inform future The cross-sectional design, while efficient, provides only a snapshot in time and does not permit inferences of causality.
We cannot definitively say that character causes resilience.
Longitudinal studies are needed to track how changes in character and literacy over time affect resilience.
Furthermore, an experimental design .
or instance, testing an integrated CE-SML curriculum versus an SML-only curriculu.
would provide stronger causal evidence.
The sample was also drawn from an urban setting (Dil.
The views, social media habits, and core values of youth in rural Timor-Leste, where digital access and social structures may differ, are not represented.
Future research must include a more geographically diverse sample.
The greatest limitation was the reliance on self-report instruments that led to the ceiling effect.
Future research must employ more robust measures.
For SML, this could include performance-based tasks where students are asked to identify disinformation in a simulated social media feed.
For character and resilience, locally developed scales with greater granularity or scenario/vignette methods may be necessary to mitigate social desirability bias.
Finally, as indicated by the low RA, the model is Future research should build more comprehensive models that incorporate other variables, such as trust in government, media exposure .
ypes of platforms use.
, and parental/peer influence.
CONCLUSION
This research provides empirical evidence of the vital role of character education and social media literacy as twin pillars in building national resilience among Generation Z in Timor-Leste.
Character education acts as the foundational driver, providing the AuwhyAy .
he reason to love and defend the countr.
, while media literacy provides the AuhowAy .
he means to do so effectively in the digital ag.
The findings strongly suggest that character is the more dominant predictor, highlighting its primacy in educational and national security strategies.
Based on these findings, we recommend that policymakers and educational institutions in Timor-Leste design curriculum policies that systematically integrate character education based on PAG Journal Homepage : https://jurnal.
id/index.
php/educenter/index Educenter : Jurnal Ilmiah Pendidikan Vol 4 No 3 2025 national values with digital media literacy training.
The pedagogical approach should not treat these as separate subjects but should embed literacy skills within a character-driven framework to ensure that technical skills are guided by a strong moral compass.
ACKNOWLEDGEMENT
This research is an excerpt from the final project thesis.
The researcher would like to thank the Master's Program in Educational Research & Evaluation.
Universitas Negeri Yogyakarta Postgraduate Program, and the supporting institution.
Comando Geral da Policia Nacional de Timor Leste.
REFERENCES