Report on Media Literacy Index

Project “Strengthening Freedom of Expression and Freedom of Media in Serbia” SERBIA 2025 MEDIA LITERACY INDEX - Pilot Version – Belgrade, 2025 Authors: Dr Marijana Matović, Faculty of Political Sciences, University of Belgrade Dr Ana Milojević, Faculty of Political Sciences, University of Belgrade Dr Snežana Milin Perković, media literacy expert Dr Aleksandar Tomašević, Institute of Physics, University of Belgrade This project is funded by the European Union Disclaimer: This report was published with financial support from the European Union. The views expressed in this report belong exclusively to the authors and their collaborators and do not necessarily represent the official position of the OSCE Mission in Serbia and do not necessarily reflect the views of the European Union.

Table of Contents Introduction ............................................................................................................................................. 1 Index Construction Process ................................................................................................................. 2 Basic Premises ........................................................................................................................................ 3 Definition of Media Literacy .............................................................................................................. 3 Media Literacy in Context .................................................................................................................. 5 Serbia’s Media Literacy Index ................................................................................................................ 7 Basic Framework and Operational Definition of Media Literacy ...................................................... 7 I MEDIA LITERACY INDEX (MLI), based on individual competences ......................................... 8 Instrument Design and Measurement Framework .......................................................................... 8 Sample and Data Collection ............................................................................................................ 9 Media Literacy Index – MLI ......................................................................................................... 10 II INTEGRATED MEDIA LITERACY INDEX (IMLI) ................................................................. 14 ENVIRONMENTAL FACTORS – ANALYSIS OF EXTERNAL FACTORS .......................... 14 III IMLI VALIDATION AND FINAL CONSTRUCTION ............................................................. 22 Delphi Process and Consensus Building ....................................................................................... 22 Key Findings ................................................................................................................................ 24 Concluding Considerations and Recommendations ............................................................................. 25 References ............................................................................................................................................. 29

1 Introduction The Serbia 2025 Media Literacy Index report responds to several requirements. One requirement was to provide an assessment of the current situation and lay the groundwork for an annual evaluation of the cumulative impact of activities in the field of media and digital literacy in relation to the Index. The second important requirement was to develop a research instrument to measure the level of media literacy of the general population. The fulfilment of these requirements also constitutes the first step in implementing Activity 5.1.1 of the national Media Strategy 2020–2025 Action Plan, which calls for “the establishment of a media literacy index in the Republic of Serbia in line with European methodology.” The third requirement was to develop an Index that is sensitive to changes in the environment (media, social, technological, etc.), as well as to the availability of the relevant data. Whilst developing the Index, we examined several key questions that shape the scope and implications of the media literacy concept in today’s context: • Which factors affect media and media content today - their quality, production, distribution, audiences and perceptions of them? • Is media literacy a progressive competence that develops linearly, or are there contextual factors that may lead to its regression? • Are there media-literate individuals - those who have access to media, are able to analyse and evaluate media messages and actively participate in the digital environment, albeit in undemocratic ways and/or in undemocratic societies? • To what extent can media literacy exist - or be meaningfully developed - in environments with limited media freedom and pluralism? • Does media literacy include a normative dimension - values and views aligned with democratic principles (e.g., trust in institutions, protection of fundamental rights, freedom of speech)? The answers to these questions shaped the dual structure of the Integrated Media Literacy Index (IMLI): a survey-based measurement of individual competences (access, critical evaluation, participation), combined with external contextual indicators (infrastructure, education system, media freedom, democratic context, trust and cybersecurity and AI

2 readiness). This approach enabled not only an assessment of the current situation, but also the identification of systemic barriers and policy priorities. This report presents the Media Literacy Index (hereinafter: MLI), which is based on a survey conducted on a representative sample of the population of the Republic of Serbia and is designed to measure the current level of media literacy among Serbia’s citizens. The report also proposes a model for the Integrated Media Literacy Index (hereinafter: IMLI), a comprehensive methodological framework for measuring the level of development of media literacy in Serbia. Unlike models that focus exclusively on either individual competences or contextual factors, the IMLI captures both levels simultaneously and their interaction - namely, the individuals’ ability to access, critically evaluate and actively participate in the media environment, as well as the structural conditions that enable or constrain the development of these competences. The IMLI presented in this report is the result of a multidisciplinary research process integrating theoretical analysis, empirical research and expert validation 1 . As a pilot study, it provides broader findings on media literacy in Serbia in 2025 and offers a methodological foundation for future longitudinal analyses and international comparisons. Index Construction Process Construction of an index is a complex, multi-phase process. As Wadhwa and Nandal (2025) note, indices aggregate various indicators relevant to a given phenomenon into a single composite measure, offering a more comprehensive overview that facilitates comparison, trend analysis and policy formulation. The process of index construction begins with the operationalisation of the concept to be measured. This is a critical phase that requires balancing the theoretical definition of the concept (in this case, media literacy) with the availability of reliable data necessary to translate the concept into measurable and mutually compatible constructs. As Wadhwa and Nandal further observe: 1 The Research Team thanks the experts who had taken part in the Delphi Process for their contribution to IMLI’s development.

3 “These abstract concepts require careful operationalization, which involves defining the concept in measurable terms, selecting appropriate indicators, and determining the methodology for combining these indicators into a cohesive index. The selection of indicators is a critical step that often involves balancing theoretical relevance with data availability and reliability.” (Wadhwa & Nandal, 2025, p. 3927) Departing from this definition of an index, we set the following project goal: to construct a media literacy index that incorporates environmental factors. In the preliminary stage, we conducted a review of the literature and relevant research to identify the key elements of media literacy and its relationship with the environment. Based on this review, we established a working framework of the integrated index and examined selected external data sources. We structured the implementation of the goal into three phases: The first phase involved the design and implementation of the survey: developing the questionnaire (ensuring its complementarity with the existing external data sources), conducting the survey (data collection), analysing the survey data and constructing the Media Literacy Index (MLI) at the level of individual competences. The second phase focused on preparing external data for integration with the MLI results, as well as integrating these data into the preliminary Integrated Media Literacy Index (IMLI). The third phase consisted of validating the IMLI using the Delphi method 2 (consultations with media literacy experts), resulting in the construction of the final IMLI for Serbia in 2025. Basic Premises Definition of Media Literacy Media literacy is a concept with a long theoretical tradition but heterogenous operationalisations. Most definitions, from early formulations (Aufderheide, 1993) to 2 The Delphi method, developed by the RAND Corporation in the mid-1950s, is a research technique used in social sciences to achieve consensus among experts. It is primarily applied in situations of uncertainty, where insufficient data exist to support informed decision-making. The method is consensus-based, structured and iterative, with the results of each round informing subsequent rounds until consensus is reached (more on the Delphi method in Rowe & Wright, 1999).

4 contemporary frameworks such as UNESCO (2022) and NAMLE (2024), have retained three core elements: (1) access to media and information, (2) critical evaluation of content, and (3) participation via media (creation, communication, civic action). This consensus is confirmed by meta-analytical studies. In a review of definitions, Bulger (2012) identified access/use, analysis/evaluation and content creation as the dominant components. Based on an analysis of 210 studies, Potter (2022) categorised over 400 elements of media literacy into four domains: access, analysis, evaluation and communication. The European Commission (COM 2007/833), UNESCO (Grizzle et al., 2013), regulatory bodies (BAI, 2022; Ofcom, 2025, 2023) and professional associations (NAMLE, 2024; Jolls & Thoman, 2008) likewise define media literacy by relying on three domains - access, analysis and evaluation, creation and communication. However, each of these domains has evolved conceptually in response to technological and social developments. Early definitions of access primarily focused on physical and technological availability, typically measured by whether households had internet access and whether individuals possessed the skills to use digital devices. Contemporary approaches expand the understanding of access in three directions. First, access entails not only technical availability but also the diversity of perspectives available (content pluralism), which directly depends on media freedom and structural pluralism (EC, 2007). Second, individuals may have technical access but lack the skills to navigate, search or assess the relevance of information; access is therefore also understood as cognitive availability (Grizzle et al., 2013). Third, access is understood as awareness of one’s information needs. Access is meaningful only if individuals are aware of their needs for specific media content and understand how to satisfy them (ibid.). Paradoxically, access is both the easiest and the most difficult dimension to measure: while infrastructure availability can be easily quantified, cognitive availability and awareness of information needs require more sophisticated instruments. The ability to analyse and evaluate media content is the most stable component in definitions of media literacy (Pereira & Moura, 2022). This dimension includes identification of the type of content (news, advertisement, opinion), the source (public service media, private platform, unidentified actor) and the context of creation. It also involves understanding the media ecosystem, the basic regulatory framework, ownership structures, the economic models underlying content creation and the technological constraints shaping content. Furthermore, it includes application of criteria for assessing source credibility, as well as content accuracy and

5 balance. In digital environments, critical evaluation necessarily extends to new domains, such as AI-generated content, disinformation campaigns and coordinated manipulation of public opinion. This dimension reflects the educational capacity of the system - the extent to which critical thinking is developed through formal education and the availability of media education. The conceptualisation of the participatory dimension of media literacy is the most diverse. Some authors focus on content creation - writing blogs, producing videos, editing images (Jolls & Thoman, 2008). Others put emphasis on communication - the ability to express oneself effectively across different media formats (Pereira & Moura, 2022). Others yet link participation to civic action - use of digital tools for social engagement, petitions, campaigns and public deliberation (Hobbs, 2010; NAMLE, 2024). The EAVI framework (Celot & Pérez Tornero, 2009) subdivides communicative and participative skills into three categories: first, as social relations, involving networking, connections, communities and peer communication; second, as citizen participation, encompassing interaction with institutions, expression of opinions and mobilisation of interest groups; and third, as content creation, involving content production, collaborative creation and problem-solving via the media. The Australian regulatory authority (ACMA, 2009) also includes digital safety among participatory competences (managing privacy, protection from fraud, responsible data sharing), arguing that meaningful participation is not possible without security (Perez-Escoda et al., 2016). Media Literacy in Context Our model departs from the key premise that media literacy is not an individual attribute in a vacuum, but rather a relational phenomenon - interaction between competences and the environment. As Erstad and Amdam (2013) note, literacy is not only developed but also applied within specific contexts that may be enabling or constraining. This approach is explicitly articulated in European policy frameworks. The Audiovisual Media Services Directive (Directive 2007/65/EC: 3) sets out that “[M]edia literacy refers to skills, knowledge and understanding that allow consumers to use media effectively and safely. Media-literate people are able to exercise informed choices, understand the nature of content and services and take advantage of the full range of opportunities offered by new communications technologies. They are better able to protect themselves and their families from harmful or offensive material. Therefore, the development of media literacy in all sections of society should be promoted and its progress followed closely” (italics ours). Subsequent documents (EC, 2009; EC Expert

6 Group, 2009) further elaborate this dimension, emphasising that the same individual competences may produce different outcomes depending on media freedom, educational systems, democratic institutions and technological infrastructure. UNESCO (Grizzle et al., 2013a; UNESCO, 2022) explicitly links media and media and information literacy (MIL) to a democratic society and intercultural dialogue, qualifying it as a prerequisite for informed participation in digital society, thereby implicitly acknowledging that media literacy loses its emancipatory function in the absence of a democratic context. The Irish regulatory authority (BAI, 2022) operationalises the participatory dimension of media literacy via individual-level skill indicators and “success indicators”, translating individual competences into collective capacities at the level of communities and systems. Based on the understanding of media literacy as a relational phenomenon (the interaction between competences and the environment), the European Association for Viewers’ Interests (EAVI) framework was developed through a series of studies for the European Commission (Celot, 2009; Celot & Pérez Tornero, 2009; DTI & EAVI, 2011). Initiated by efforts to establish a methodology for measuring media literacy at the European level, the EAVI framework introduces a dual measurement architecture that distinguishes between individual competences and environmental factors. Individual competences are defined in this framework as “an individual capacity to apply certain skills” (Celot, 2009, p. 7), and they encompass both personal competences (media use, critical understanding of the content) and social competences (communication via the media, participation in the public sphere). Environmental factors are defined as a “set of contextual factors” that engender or endanger the development and application of individual competences. These include media availability (technological infrastructure, content pluralism), media policies (regulatory frameworks, user protection), education (formal and non-formal media education), media industries (ownership structures, economic models, professional standards) and civil society (organisations and media literacy initiatives). The EAVI framework rests on the key premise that identical individual competences may yield different outcomes depending on structural conditions, wherefore different scenarios can be envisaged. For example, individuals may possess critical evaluation skills, but they will lack access to diverse sources and their competences will remain underutilised in an environment with low media pluralism. In another scenario, infrastructure is well developed and access is high, but the education system fails to foster critical thinking, resulting in the people’s passive use of media and limited evaluation of content. A third scenario may involve

7 an active civil society that encourages participation, while the broader democratic context is restrictive (e.g. low level of freedom of expression), rendering civic engagement via the media potentially risky. Media literacy thus emerges from the interplay between competences and the environment. The degree of their convergence or divergence determines the individuals’ actual capacity to participate effectively in the mediated public sphere. Therefore, we developed an index for assessing the state of media literacy in Serbia departing from the understanding of media literacy as a relational phenomenon. Serbia’s Media Literacy Index Basic Framework and Operational Definition of Media Literacy The construction of the Integrated Media Literacy Index (IMLI) is based on the EAVI conceptual framework applied in the context of Serbia. The EAVI framework offers three key advantages for our study. First, its core structure remains relevant despite rapid technological change - the distinction between what individuals are able to do and the context in which they do it has not been rendered obsolete by the emergence of new technologies, such as artificial intelligence, or new threats, such as coordinated disinformation campaigns. On the contrary, these developments underscore the importance of the contextual approach, as the same individual skills can produce different outcomes depending on media freedom, educational capacities and democratic institutions in which they are applied. The EAVI framework’s second advantage lies in its comparative applicability. The framework was developed with the aim of enabling cross-national comparisons despite different political traditions, media systems and educational models across European countries. Incorporation of the EAVI logic into the IMLI enables the application of the methodology in other countries in the region and thus the gradual creation of a base of comparative data that would allow a deeper analysis of the relationship between structural factors and the people’s actual competences. Third, the integrated approach has direct implications for diagnosis and public policy design. If the index indicates a low level of media literacy, it is necessary to determine whether it is the result of a lack of individual skills, systemic barriers or a misalignment between the

8 two. Each of these scenarios requires different types of intervention. Lack of skills points to the need for media education programmes; systemic barriers call for structural reforms (improving media freedom, the education system and democratic institutions); while situations in which both competences and opportunities exist but do not align suggest the need for interventions related to motivation, trust and perceived safety. The integrated index enables this kind of diagnosis; we will demonstrate its analytical potential in the concluding sections following the presentation of the empirical findings for Serbia. Therefore, taking into account these three key dimensions and building on the EAVI framework, we based the construction of the index on the following definition of media literacy: Media literacy is the ability to access, critically understand (analyse and evaluate) and participate via media (as content and technologies), with these competences developed and applied within specific social contexts that include technological infrastructure, the education system, media freedom, democratic institutions, social trust and preparedness for emerging technologies, including artificial intelligence. Such a definition enables the construction of an index that simultaneously measures what people are capable of doing (individual competences - MLI), the environment in which they do it (structural factors - external indicators) and the form of media literacy emerging from their interaction (the IMLI). The following sections explain in detail how individual competences were operationalised through the survey, how external contextual indicators were selected and harmonised and how the Delphi process involving expert consultations informed the final structure and weighting of the integrated index. I MEDIA LITERACY INDEX (MLI), based on individual competences Instrument Design and Measurement Framework The design of the survey instrument followed the EAVI conceptual framework, focusing on individual competences and measuring three key dimensions: the extent to which the citizens are connected to the media ecosystem through the diversity and intensity of their use of different platforms; their capacity to critically evaluate the content to which they are exposed; and the extent to which they use media for citizen participation and the protection of their privacy in the digital environment. However, when we operationalised these dimensions

9 for Serbia in 2025, we also took into account the technological and social changes that did not exist at the time the original model was developed. Adaptation of the EAVI framework required three key extensions. First, we introduced the dimension of artificial intelligence (AI), which was not included in traditional conceptualisations of media literacy. This dimension captures not only people’s awareness of the presence of AI in the media environment, but also their patterns of using AI tools for accessing information and creating content. Second, we expanded the dimension of critical evaluation to include verification behaviours - measuring not only the respondents’ ability to recognise problematic content, but also the concrete actions they take when they come across suspicious information (e.g., cross-checking it across multiple sources, consulting fact-checking platforms and searching for the original context). These extensions respond to the realities of the contemporary media environment, in which it is no longer sufficient to think critically, but it is also necessary to verify information actively and systematically. Third, recognising that people cannot meaningfully participate in the digital public sphere if they lack a sense of safety and control over their personal data, we extended the dimension of participation to include protection skills in the digital environment - such as blocking tracking, safeguarding privacy and reporting hate speech. The final instrument comprises sixty-three questions organised into three sections corresponding to the theoretical structure of media literacy. 3 The first section (P1–P16) measures access to and use of media, including traditional media (television, radio, print) and digital platforms (social media, news portals, video platforms and messaging applications), as well as patterns of AI use. The second section (P17–P42) focuses on the critical understanding of media content and includes the identification of various types of problematic content (disinformation, sensationalism, biased reporting), the ability to manage emotional responses to media messages and concrete verification behaviours. The third section (P43–P56) measures communication, content creation, participation and protection practices in the digital environment. The additional seven questions (P57–P63) capture the respondents’ sociodemographic characteristics, enabling stratified analyses by gender, age, education, type of settlement and other relevant variables. Sample and Data Collection The survey was conducted on a two-stage stratified representative sample of 800 adult citizens of Serbia. Stratification was carried out at two levels: the first level consisted of 3 The full questionnaire is available upon reasonable request.

10 territorial units (Belgrade, Vojvodina and Central Serbia), while the second level comprised types of settlement (urban and rural) within each territorial unit. This structure allows the disaggregation of the data by the key sociodemographic characteristics, including gender, age group, level of education, income, region and type of settlement. The distribution of the sample by gender and region is broadly aligned with estimates of the Statistical Office of the Republic of Serbia (SORS), with slight skew toward older and rural segments of the population, which is typical of telephone surveys. Despite this skew, the sample ensures satisfactory representativeness of Serbia’s adult population. Field data collection was conducted between 15 May and 15 June 2025, using the CATI (Computer-Assisted Telephone Interviewing) method. This approach enabled the rapid collection of data from a geographically dispersed sample while ensuring quality control through call supervision and real-time checks of logical consistency during interviews. The interviews lasted twenty-five minutes on average. The data were coded and analysed using SPSS, following standard procedures for data cleaning and consistency checks. Media Literacy Index – MLI Data cleaning and reliability testing were followed by the normalisation of seven validated scales to enable their aggregation into composite scores. Normalisation was performed while preserving the original response metrics. The normalised scales were then combined into three distinct dimensions of media literacy, based on both theoretical considerations (i.e., which aspects are conceptually central to each dimension) and empirical findings from factor analysis (which questions genuinely measure a common latent construct and the extent of variance they explain). The first dimension measures the diversity and intensity of use of various media platforms, as well as readiness to master new tools. It consists of two components: one measuring the diversity of sources people use for information and entertainment and patterns of AI use (P10–P16), which capture whether and how the respondents use AI tools. Greater weight is assigned to media diversity, given that it is a more fundamental aspect of media literacy – people must first use media before they can use more advanced tools based on AI. The second dimension comprises three components of critical understanding of media content. The identification of problematic content (P17–P21) measures whether citizens can recognise disinformation and biased reporting. Managing the impact of information (P22–P24) focuses on emotional regulation - whether the respondents allow media messages to quickly

11 influence them or whether they are capable of maintaining a more distanced, reflective stance. Verification behaviours (P25–P30) capture the concrete actions people take when encountering suspicious information, such as cross-checking across multiple sources or consulting fact- checking platforms. The greatest weight is assigned to recognition, as it is prerequisite for the other two components - citizens must first identify a problem before they can address it. The third dimension integrates two components: digital safety and protection skills (P43–P50) measure the application of concrete privacy-protection practices, such as blocking tracking, managing shared data and reporting hate speech. Civic engagement (P51–P56) measures the extent to which the citizens use digital platforms for public participation - sharing socially relevant information, holding actors accountable, advocating ideas and organising events. Greater weight is assigned to civic engagement, as it represents the ultimate goal of media literacy in a democratic society, while protection skills serve as a prerequisite enabling safe participation. The results reveal significant differences across dimensions. The dimension related to access and use shows the lowest mean value (M = 25.5, SD = 19.7) and the widest range (0– 79.9), indicating substantial variation among respondents in access to media and AI use. Critical understanding emerges as the most stable dimension (M = 42.1, SD = 15.1), suggesting that strategies of critical understanding and verification are more evenly distributed across the population and constitute the pillar of individual media literacy. Participation shows a lower average (M = 27.6, SD = 20.2) and the highest level of heterogeneity, indicating that participatory behaviours and protection practices have not yet become widespread - a small segment of the population is highly active, while the majority remains passive or engages only occasionally. The final composite Media Literacy Index (MLI) was constructed as a weighted average of the three dimensions. The dimension of critical evaluation carries the greatest weight, as the most stable and conceptually central domain - neither access nor participation can be meaningful without the ability to critically understand and verify information. The dimension capturing media and AI use, as well as the dimension related to digital citizenship and participation, are assigned lower weights, as they exhibit greater variability and dependence on structural factors (such as infrastructure and the democratic context), which are captured through external indicators. The distribution of weights was determined after testing several alternative formulas, based on both theoretical considerations and empirical stability.

12 The final formula for calculating the average level of media literacy among citizens of Serbia is as follows: MLI = 0.20 × D1 + 0.60 × D2 + 0.20 × D3 Table 1: MEDIA LITERACY INDEX Structure by Media Literacy Dimensions MEDIA LITERACY DIMENSION WHAT IT MEASURES Questionnaire Items SHARE VALUE D1. ACCESS AND USE P1–P16 Diversity of Media Use Frequency of use of nine types of media, ranging from television and radio to social media and AI (P1–P9) 20% 25.5 AI Use Patterns Ways in which people rely on AI tools (P10–P16) D2. CRITICAL UNDERSTANDING P17–P42 Recognition of problematic content How often respondents notice fake news, conspiracy theories, hate speech and similar phenomena (P17–P21) 60% 42.1 Managing the impact of information How people experience the pressure of information and their emotional responses to media content (P22–P24) Verification behaviours Strategies respondents use to check suspicious information (P25–P30) D3. PARTICIPATION P43–P56 Digital safety and protection skills Eight activities checking the respondents’ knowledge of ways to protect themselves and others online (P43–P50) 20% 27.6 Civic Engagement Ways in which people use the media to inform themselves, (P51–P56)

13 share their opinions and organise Average MLI value for Serbia in 2025 stood at 35.9 points (SD = 14.9), with a range from 3 to 77.2. None of the respondents reached the upper quartile of the scale (above 75 points), which indicates that very high levels of media literacy are virtually absent in the general population. The distribution is skewed toward the lower ranges: 16% of respondents fell into the very low category (0–20 points), 45% into the low category (21–40), 32% made the medium category (41–60) and only 7% exceeded the high category threshold (61–80). The very high category (81–100) remained unrepresented. Such distribution clearly points to the need for systemic interventions. The majority of citizens (61%) exhibit low or very low levels of competence, indicating significant gaps in knowledge and suggesting that a substantial share of the population is inadequately equipped to navigate the contemporary media environment safely and effectively. At the same time, the existence of a small but identifiable segment with high levels of competence (7%) suggests that structural conditions do allow for the development of media literacy among those with access to education, resources and motivation - the question is how to make these conditions more widely accessible to the broader population.

14 Distribution of the Overall Media Literacy Index The survey findings provide the basis for calculating the Media Literacy Index at the level of individual competences. The Index can thus be used to regularly monitor increases or declines in media literacy levels in Serbia, enabling also indirect tracking of the efficiency of public policies and other related activities. Monitoring is also possible across the media literacy dimensions, offering insights into the effectiveness of campaigns targeting specific competences, such as verification behaviours, digital skills, et al. The Media Literacy Index (MLI) addresses the second requirement concerning the development of an instrument for assessing the level of media literacy among the general population. The MLI also indirectly addresses the third requirement, namely the Index’s sensitivity to changes in the broader environment. In any case, the survey instrument should be reviewed on an annual basis in consultation with experts from the academic community. II INTEGRATED MEDIA LITERACY INDEX (IMLI) One of the requirements was to provide an assessment of the current situation and to lay the groundwork for the annual evaluation of the effects of cumulative activities in the fields of media and digital literacy in relation to the Index. Prior findings offer insight into the level of media literacy among Serbia’s citizens, but they cannot, on their own, explain why competences are at this level. The answers lie in the structural environmental factors - namely, the quality of the media system, the capacity of the education system and the level of democratic freedoms and social trust. The ensuing analysis of these factors helps us understand not only what citizens are capable of, but also the conditions under which they exercise these capabilities, thus more directly addressing the requirement that the Index be sensitive to changes in the environment. One possible extension of the MLI, through the inclusion of environmental factors, is the proposed Integrated Media Literacy Index (IMLI), which is elaborated in detail in the text below. ENVIRONMENTAL FACTORS – ANALYSIS OF EXTERNAL FACTORS While the survey measures what citizens are currently capable of doing, the analysis of environmental factors examines the systemic conditions under which they can exercise these capabilities. Media literacy is not merely an individual attribute, but emerges from the interaction between competences and structural context. The selected environmental factors were grouped into six categories: (1) infrastructure and access, (2) educational foundations, (3)

15 media freedom and pluralism, (4) the democratic context, (5) social trust, privacy and cybersecurity and (6) AI preparedness. The selection of categories was guided by a combination of theoretical considerations and pragmatic constraints. Theoretically, the categories cover dimensions identified by the EAVI framework as relevant environmental factors - media availability, education, the media system and social capital. Pragmatically, the selection was limited by the availability of reliable and internationally comparable data. Certain dimensions that would be ideally relevant - such as systematic monitoring of the implementation of media literacy policies or the evaluation of non-formal education programmes - were not included due to the lack of regularly updated data sources. In addition, we introduced two new dimensions that were not part of the original EAVI framework but respond to transformations in the contemporary media environment: cybersecurity and AI preparedness. Publicly available data from reliable sources were used as indicators to assess the values of these categories. External data were collected from national sources (Statistical Office of the Republic of Serbia – SORS, Gemius and the SHARE Foundation) and international databases (OECD PISA, Eurostat, European Social Survey – ESS, UN E-Government Survey, Reporters Without Borders Press Freedom Index, Centre for Media Pluralism and Media Freedom, Freedom House and the Global AI Index). A total of thirty-one indicators, organised into six categories, were used to evaluate the key environmental aspects relevant to media literacy. The data cover the 2022–2025 period, with the most recent available data used for each indicator. All indicators were normalised to a common 0–100 scale to enable comparison and aggregation into composite scores by category. The following sections provide a detailed account of each of the six categories, explain the indicators they include and how they are interrelated and analyse their relationship with the survey findings on individual competences. II.1. INFRASTRUCTURE AND ACCESS Technological availability is a fundamental prerequisite for the development of media literacy - people cannot use digital media without access to the internet, devices and platforms. This category includes indicators such as household internet penetration (SORS, 2023), the frequency of use of social media and digital services (SORS, 2023), population coverage by domestic media portals (Gemius, 2024) and a composite Digital Connectivity Index measuring

16 the quality of network infrastructure and access speed (Digital Connectivity Index, Global AI Index 2024). The data indicate that Serbia has a well-developed digital infrastructure. Household internet penetration has reached 85%, meaning that the vast majority of the population has technical access to digital media. Data show that 82% of the adult population use social media on a daily basis, while 69% of internet users visit domestic media portals on a monthly basis. The Digital Connectivity Index (64.8/100) confirms that the network infrastructure supports not only access but also continuous and relatively fast data exchange necessary for a contemporary media experience. The average score for the infrastructure and access category is 78.98/100, the highest among all environmental dimensions, positioning Serbia high on the scale of infrastructure availability and quality. However, comparison with the survey findings reveals a significant mismatch. While external data indicate high technological availability, the survey dimension of access and use suggests that technological capacity does not automatically translate into effective competence. For example, citizens can have access to the internet and devices but do not necessarily use a diverse range of sources - they rely on a limited number of platforms, predominantly social media as their primary source of information, and use traditional media and specialised portals only sporadically. This finding indicates that the issue lies not in the infrastructure but in motivation, media habits, or lack of awareness of the benefits of diverse information sources. Therefore, rather than focusing on infrastructure development, systemic interventions should concentrate on media literacy programmes helping citizens make better use of already available resources. II.2. EDUCATIONAL CAPACITIES The education system plays a key role in fostering critical thinking and media literacy among younger generations. The ability to analyse texts, understand arguments, evaluate sources and use information to solve problems - competences developed through formal education - underpins capacities for the critical understanding of media. This category combines objective indicators of educational attainment with expert assessments of systemic support for media literacy. The primary indicator is the PISA assessment (OECD Programme for International Student Assessment, 2022), which measures 15-year-olds’ ability to use their reading, mathematics and science knowledge and skills to meet real-life challenges. Reading literacy is

17 particularly relevant, as it involves not only understanding texts, but also evaluating and reflecting on them, including recognising perspectives and using information to draw well- reasoned conclusions. Conversion of Serbia’s PISA 2022 results to a 0–100 scale yields an average score of 49.33/100. This result suggests that the education system does not sufficiently develop competences for critical understanding and reasoning among young people. In addition to the objective measurement of student achievement, this category includes an expert assessment of systemic support for media literacy within the education system. During the Delphi process, an expert panel (comprising teachers, education professionals and education and media literacy researchers) assessed the extent to which the education system in Serbia supports the development of media literacy. The average score was 3.21/10, which, after normalisation, corresponds to 32.1/100. This low score indicates a gap between objective measures of student competences (PISA) and expert assessments of systemic support for media literacy, pointing to a structural issue: the education system as a whole does not provide sufficient targeted support for the development of critical media literacy through curricula, teacher training and pedagogical approaches. The experts’ assessment is based on insights into curricula, teaching practices and institutional capacities, thereby complementing the evaluation of the education system with insights not captured by PISA. The link with the survey is indirect but important: the survey dimension measures critical skills among adults, while PISA reflects the educational foundations acquired by younger generations. The survey dimension related to critical understanding stands at 42.10/100 and represents the most stable and highest individual domain among adults. This is slightly lower than the PISA average for younger cohorts (49.33/100), but substantially higher than the expert assessment of systemic support (32.1/100). This suggests that people develop certain critical skills despite the lack of systemic support in education, probably through informal learning channels and life experience. However, neither group reaches higher levels of competence, which reinforces the need for systemic interventions in the education system aimed at fostering critical thinking from the earliest stages of schooling. II.3. MEDIA FREEDOM AND PLURALISM The quality of the media system via which people access information fundamentally shapes the conditions for the development of media literacy. Even where people possess well- developed critical competences, their application is constrained if the media environment offers limited pluralism of perspectives, if journalists operate under pressures compromising the independence of reporting, or if ownership structures result in a concentration of control over information. Media literacy entails the ability to access diverse sources, evaluate their

18 reliability and form informed opinions - however, all of this depends on whether such sources exist in the first place and whether they are able to operate freely. This category integrates data from two internationally recognised sources. The Reporters Without Borders Press Freedom Index (RSF, 2025), after inversion (46.5/100), indicates that Serbia falls among countries facing moderately high risks to media freedom. The second source comprises four dimensions of the Media Pluralism Monitor (Centre for Media Pluralism and Media Freedom, 2024), which systematically assesses risks to media pluralism: fundamental protection (constitutional and legal safeguards of media freedom), market plurality (ownership concentration and transparency), political independence (government influence over media and public service funding) and social inclusiveness. The average score for Serbia is 47.75/100, with political independence and market plurality identified as the most problematic dimensions. The average score for the category of media freedom and plurality - 47.17/100 – indicates serious structural constraints. Low plurality means that even people seeking to consult diverse perspectives encounter a media market that is structurally homogenised - certain viewpoints dominate the public sphere, while others are marginalised not due to lack of audience demand, but because of economic and political barriers to entry. Low level of media freedom suggests that even journalists committed to independent reporting operate under pressures compromising their ability to investigate and publish critical content. The link with the survey findings is particularly revealing. A scale measuring trust in various sources of information was tested during the construction of the MLI. This scale was excluded from the final index due to low reliability, i.e. because the respondents did not demonstrate a consistent pattern in evaluating sources - some expressed high levels of trust across all sources, while others were broadly distrustful, without clear differentiation between types of sources. Such inconsistency may be a direct consequence of the ambiguous quality of the media environment. Where people cannot rely on stable quality criteria, due to wide variations or shifts in media standards, their assessments become unsystematic and tend to reflect general levels of trust or distrust rather than evaluations of specific sources. Therefore, structural deficiencies of the media system not only limit access to high-quality information, but also impede the development of stable criteria for assessing reliability. II.4. DEMOCRATIC CONTEXT

19 Media literacy entails the active use of media for civic engagement - expressing opinions, mobilising around public issues, connecting with others and organising collective action. However, such use depends not only on individual competences, but also on the institutional framework, which may either encourage or discourage citizen participation. In environments with limited democratic freedoms, citizens may possess the necessary competences and have access to platforms, yet refrain from public engagement due to fear of repercussions or the realisation that their participation will have no real impact on decision- making. This category combines indicators of democratic freedoms and opportunities for digital participation. Freedom House classifies Serbia as “partly free,” with a score of 56.0/100 in its 2024 Freedom in the World report. Internet Freedom is rated somewhat more highly, at 70/100, indicating that the digital space is relatively freer than traditional media. The average score for the democratic context category stands at 64.5/100. The UN E-Participation Index (UN E-Government Survey, 2024) measures the extent to which governments use digital platforms to facilitate citizen participation. This indicator is particularly relevant for the digital dimension of media literacy, as it reflects whether citizens who wish to engage have access to resources. Serbia scored 73.0/100 on this index, which is relatively high, especially compared to its overall level of democratic freedoms. This mismatch suggests that e-participation platforms formally exist and that institutions have established online mechanisms, but do not substantively respond to citizens’ demands, which may discourage participation over time. While the average score for the democratic context category is 64.5/100, the survey- based participation dimension stands at a mere 27.56/100. This gap between the existence of platforms and their actual use suggests that technological availability does not guarantee use when people presume their participation will not produce results. Addressing this problem requires not only merely improving the digital platforms, but also a shift in institutional culture toward meaningful responsiveness to citizen participation. II.5. DOMAIN OF TRUST, PRIVACY AND CYBER-SECURITY Social trust, data protection culture and cyber-security together form the foundation for meaningful media participation. People lacking even a minimum level of trust in others and in institutions tend to withdraw from the public sphere and limit their activities to private circles.

20 People unfamiliar with privacy protection measures are exposed to data abuse, which may deter them from public engagement. The assessment of this category draws on three types of data. First, we used Eurostat data on interpersonal trust from the European Social Survey (ESS, 2023), which measures trust in others, fairness and helpfulness. According to these data, the level of interpersonal trust in Serbia stands at 44.2/100, which is below the European average. Second, we included data on privacy protection practices from the national ICT survey (SORS, 2023). SORS data show that only 21.9% of citizens on average implement privacy protection measures (such as disabling location tracking, restricting the visibility of social media profiles, or limiting the use of personal data for advertising), which is a very low share, particularly given the widespread use of digital platforms in Serbia. Third, the category includes data on cybersecurity in the media sector (SHARE Foundation, Cybersecurity Monitoring, 2024), which tracks incidents such as hacking attacks on media websites, direct threats and attacks against journalists via digital channels, abuse, discrimination and hate speech. According to these data, the media cybersecurity indicator appears relatively high (71.94/100). However, this measure reflects only reported incidents (with actual figures likely higher) and does not capture subjective perceptions of security. The average score for the category of trust, privacy and cybersecurity stands at 38.52/100 (combining moderate interpersonal trust at 44.2/100, very low prevalence of protection practices at 21.9/100 and moderate cybersecurity at 71.9/100). While the relatively high cybersecurity indicator raises the overall average, it cannot compensate for low levels of social trust and individual privacy protection. The link with the survey reveals multiple gaps between knowledge, practices and systemic security. The survey’s participation dimension (P43–P50) measures whether respondents know how to apply eight different protection measures (e.g., report hate speech, disable tracking, protect against fraud), while external data show that only around 22% of them actually implement these measures in practice. At the same time, the cybersecurity indicator (71.94/100) suggests that systemic security is stronger than individual preparedness. This indicates that citizens are insufficiently equipped to protect themselves, even in an environment where objective risks are not extremely high. The presence of knowledge among certain segments of the population does not necessarily manifest itself in everyday practice. These findings point to the need for interventions that go beyond skills training to also reshape risk perceptions and foster a culture of digital safety. At the same time, the low level of social trust (44/100) is closely linked to low participation levels. People who lack trust in others are less

21 likely to engage publicly, as they perceive themselves to be exposed to risks without reciprocal support or protection. Therefore, interventions are needed not only to promote protective behaviours, but also to strengthen the foundations for proactive engagement – to establish relationships of trust both among individuals and between citizens and institutions. II.6. AI PREPAREDNESS Artificial intelligence is increasingly becoming an integral part of the media ecosystem - from automated news generation to personalised recommendations that shape the users’ information experiences. Media literacy in the contemporary environment requires both an understanding of traditional media processes and the ability to navigate content mediated or generated by artificial intelligence. This includes recognising AI-generated content, as well as critically evaluating the authenticity and reliability of such content. The development of these competences depends not only on institutional capacities to regulate and ensure the transparent use of AI technologies, but also on public trust in AI. In this category, we incorporated the IMF AI Preparedness Index (2024), which assigned Serbia a score of 54.0/100, indicating a moderate level of institutional readiness. This index captures regulatory frameworks, technical infrastructure, human capital and innovation capacity in the field of AI. We also used the Digital Connectivity data from the Global AI Index, which confirm that the technological foundation for AI adoption is relatively strong (64.8/100). However, data on public trust from the same source reveal a gap between infrastructural readiness and trust in AI technologies (36.9/100), pointing to limited understanding of how AI functions, lack of transparency in its use, or fears of potential abuse. The average score of the AI preparedness category stands at 51.9/100. Low levels of trust are clearly reflected in the survey data, which show that only 23.9% of the respondents (191 out of 800) use AI platforms. Low trust in AI and low levels of usage reinforce one another: citizens refrain from using AI due to distrust, while trust does not increase due to lack of experience. A comprehensive assessment of this category points to the need for a proactive approach that integrates several dimensions. First, transparent policies on use of AI by public institutions and media organisations - which algorithms curate news, how models are trained and who is accountable for errors. Second, education on AI technologies that goes beyond technical skills to include an understanding of their societal implications - how AI is transforming the media, the ethical dilemmas AI raises and how users’ rights can be protected.

22 Third, civic empowerment through programmes that foster the critical evaluation of AI-generated content. Without such interventions, there is a risk of creating a new digital divide - between the minority that uses AI tools and understands their underlying logic and the majority that remains passive and distrustful – which will further deepen inequalities in media literacy. III IMLI VALIDATION AND FINAL CONSTRUCTION Following the survey and the analysis of external indicators, which produced two complementary profiles - individual competences (MLI: 35.9/100) and structural conditions (six environmental categories) - the remaining methodological question was how to integrate them into a single metric. To address this, we applied the Delphi method with an expert panel in order to empirically determine the relative importance of the individual dimensions. Delphi Process and Consensus Building The Delphi process involved thirty experts in education, media, media and education policies and from civil society. The panel included teachers, journalists with experience in fact- checking and verification, researchers specialising in digital competences and disinformation, representatives of state institutions and civil society activists conducting media literacy programmes. The panel’s geographic and institutional diversity ensured that the resulting consensus did not reflect the perspective of just one group of actors, but instead provided a synthesis of perspectives across the media literacy field. The Delphi method was conducted in two rounds during October 2025. In the first round, the experts independently assigned weights to seven proposed dimensions of media literacy: Connectivity and Access, Critical Evaluation, Democratic Participation, Media Freedom, Trust and Privacy, Cybersecurity, and AI Preparedness. Each dimension was accompanied by a brief explanation of what it measures and how it derives from survey and external data. In the second round, the experts were provided with aggregated results from the first round (median and interquartile ranges for each dimension), insight in qualitative explanations of other participants and the opportunity to revise their weights in light of collective assessments. Agreement was achieved on all dimensions based on two predefined criteria of statistical stability and consensus. Two methodological adjustments were made during the construction of the final index.

23 First, the dimensions of Trust and Privacy and Cybersecurity were merged into a single dimension - Trust, Privacy and Cybersecurity - with a combined weight. This decision was based on the conceptual proximity of these dimensions and the fact that they rely on external indicators capturing social trust, users’ protection measures and systemic cybersecurity. Second, following the merger, all the weights were normalised through proportional scaling. The IMLI’s final structure comprises six dimensions: (1) Connectivity and Access – 15.71%, (2) Critical Evaluation – 20.94%, (3) Democratic Participation – 10.47%, (4) Media Freedom – 20.94%, (5) Trust, Privacy and Cybersecurity – 20.94%, and (6) AI Preparedness – 11.00%. This distribution indicates that experts identified Critical Evaluation and Media Freedom as central and equally important dimensions of media literacy. Application of the Delphi-validated weights to the dimension scores yielded the final Integrated Media Literacy Index for Serbia (2025): IMLI = 43.11/100. Table 2 shows the contribution of each dimension to the overall index. Serbia’s 2025 score on the Integrated Media Literacy Index (IMLI) stands at 43.11 out of 100. Table 2: Final IMLI Score for Serbia: Dimension Score Weight Contribution CONNECTIVITY AND ACCESS 52.23 15.71% 8.2 CRITICAL EVALUATION 45.72 20.94% 9.57 DEMOCRATIC PARTICIPATION 46.03 10.47% 4.82 MEDIA FREEDOM 47.17 20.94% 9.88 TRUST, PRIVACY AND CYBERSECURITY 38.52 20.94% 8.07 AI PREPAREDNESS 23.42 (60/40) 11.00% 2.58 IMLI Serbia 100% 43.11

24 Key Findings Serbia’s score of 43.11 out of 100 in 2025 can be interpreted as a medium–low level of integrated media literacy. Serbia performs better on the Integrated Media Literacy Index (IMLI: 43.11) than on the survey-based Media Literacy Index (MLI: 35.9/100), due to high scores on the external infrastructure (78.98) and the democratic context (64.50) indicators. However, the mismatch between structural opportunities and individual use remains evident also in the integrated index. Three dimensions contribute almost equally to the IMLI. Critical Evaluation (9.57), Media Freedom (9.88) and Trust, Privacy and Cybersecurity (8.07) together account for nearly two-thirds of the index. These dimensions form the core of media literacy from a theoretical perspective as well: the capacity for critical reasoning, a media environment offering pluralism of perspectives and a level of security enabling public engagement. Contributions of Connectivity and Access (8.20) and Democratic Participation (4.82) are moderate, while AI Preparedness (2.58) contributes the least, due to both low usage levels reflected in the survey and low levels of trust indicated by external data. The value of the Connectivity and Access dimension (52.23) masks the gap between infrastructure (78.98) and use (25.48). Although external conditions enable diverse media use, the citizens do not fully take advantage of these opportunities. This finding has direct policy implications, suggesting that further investment in infrastructure is unlikely to significantly improve media literacy. Interventions that enable and motivate citizens to make better use of the existing resources are likely to make greater impact. Similarly, within the Democratic Participation dimension (46.03), high levels of e-participation measured by external indicators contrast with low levels of civic engagement observed in the survey, again pointing to the underutilisation of the available resources. The lowest-scoring dimension, AI Preparedness (23.42), signals an area where both structural conditions and individual practices remain underdeveloped. Low trust and low usage reinforce one another, creating a vicious cycle. Without proactive interventions - such as transparent policies, AI education and ethical standards - there is a risk that this gap will widen, creating a new digital divide between the minority that uses AI tools and understands their logic and the majority that remains passive. In any case, the Media Literacy Index (MLI) and the Integrated Media Literacy Index (IMLI) provide a much-needed foundation for future analysis and trend monitoring. First, annual tracking will show whether interventions (media literacy programmes, education reforms, changes in the media system) produce genuine change. Second, application of the

25 methodology in other countries will enable cross-country comparisons and the identification of best practices. Third, analysis of the IMLI across its dimensions will help identify priorities for systemic interventions, both in the current and future moments and contexts in which the index is applied. Concluding Considerations and Recommendations Policy Recommendations: Four Priority Areas for Intervention Based on the IMLI findings, four priority areas for systemic intervention can be identified that are likely to produce measurable improvements in media literacy in Serbia. These areas are not isolated. Rather they are mutually reinforcing - progress in one domain strengthens progress in others, and vice versa. I. The development of comprehensive media literacy programmes that enable citizens to make better use of the already available technological infrastructure is the primary priority. The findings clearly indicate that the problem does not lie in lack of technology, but rather in how it is used. Media literacy programmes should focus on strategies for using diverse information sources and verification behaviours requiring systemic support. Such programmes should teach citizens how to identify and actively use a variety of information sources, while also raising awareness of the importance of source diversity - recognising that different platforms offer different perspectives, that monopolies over information can lead to bias and that consulting multiple independent sources supports informed judgment. In addition, citizens should learn the practical steps for responding to questionable information – they should, for example, learn to verify information across multiple independent sources before sharing it, consult fact-checking organisations when encountering sensationalist claims, seek the original context of images or statements circulating on social media and recognise clickbait headlines promising more than the text actually delivers. Special attention should be paid to older demographic groups and less-educated segments of the population, who often lack access to informal learning channels available to younger and more educated groups. Programmes targeting these groups should be tailored to their prior knowledge, learning styles and practical needs. This may include adaptation of both content and format - slower pacing, repetition of key concepts, use of everyday examples and

26 practical demonstrations instead of abstract explanations. Participatory formats are particularly effective: citizens can practice on real-life examples, analyse news content, use tools for verifying images and videos, engage in group discussions on how to distinguish news from comment and practice source and context verification. Such programmes can be delivered through workshops using the existing infrastructure, such as libraries or pensioners’ associations. It is essential that the programmes are free of charge, geographically inclusive (covering both urban and rural areas) and sustained over time rather than implemented as one- off initiatives that quickly fade. II. Experts assess systemic support for media literacy in education as low, while PISA results confirm that the education system does not sufficiently develop critical understanding competences among young people. These findings point to the need for a comprehensive reform of the education system that positions media literacy as a transversal competence across all levels of formal education, from primary and secondary education to higher education, particularly in teacher training programmes. Such reform should begin with curriculum revision to ensure that media literacy is both introduced as a standalone subject - ideally taught by graduates in media and communication studies - and systematically integrated across other relevant subjects. Integration of media literacy as a cross-curricular competence entails embedding critical media analysis skills within existing subjects where they naturally fit. For example, in history classes, students can critically analyse how media represent historical events and social groups. This approach allows students to develop media competences across multiple contexts and subjects, which reinforces their knowledge and enhances its practical applicability. It goes without saying that curriculum reform can only be effective if media literacy is first taught by those specifically trained in the field of media and communications and if teachers of other subjects are additionally empowered to guide students in critically evaluating the content they are exposed to. This requires continuous professional development programmes providing the teachers not only with a theoretical understanding of media literacy, but also with practical pedagogical strategies, activities and materials they can use in class. One-off seminars have limited impact, since teachers quickly forget the new knowledge they do not use regularly. Long-term programmes involving mentoring, peer exchange and ongoing support, would be more effective. Finally, the effects of integrating media literacy into curricula should be systematically evaluated through studies assessing whether students exposed to reformed programmes demonstrate higher competences than previous cohorts. While educational reform does not

27 yield immediate results, it produces systemic change by raising the overall level of media literacy of the population over time. III. Structural constraints within the media system directly affect people’s ability to develop and apply media literacy skills. Even where individuals possess critical evaluation skills, limited pluralism of perspectives and concentrated ownership structures constrain access to diverse sources that can be meaningfully compared. Reform of the media system is therefore a necessary parallel intervention; without it, individual competences will remain only partially applicable. Strengthening the independence of public service media should be a strategic priority because public broadcasters operating in line with their mandate should serve as models of high-quality, pluralistic journalism, accessible to all citizens regardless of economic status. This requires a fundamental reform of the mechanisms for appointing governing bodies to ensure that the public broadcasters serve the public rather than the interests of the government of the day, as well as transparent and predictable forms of funding enabling planning and safeguarding their independence from political pressure. Public service media must have the resources and institutional autonomy to produce high-quality news, engage in investigative journalism and create educational content that elevates the level of public debate. Public service media reform should be accompanied by systemic support for independent and local media providing alternative perspectives and covering issues neglected by commercially driven or pro-government media. These outlets frequently operate on the edge of economic sustainability. Transparent public funding mechanisms - based on clear criteria and professional standards - can support their viability without compromising editorial independence. At the same time, support for local media covering regional and common issues ensures access to information relevant to the people’s everyday lives, which national outlets often overlook. Transparency of media ownership should become a binding standard since citizens have the right to know who owns the media they follow and how ownership structures may influence the content they consume. Regulatory frameworks should ensure public access to clear and comprehensible information on media ownership, business interests, funding sources and any conflicts of interest. For examples, individuals are better equipped to assess whether an outlet’s reporting is biased if they are aware that it is owned by a company with government contracts. Finally, the protection of journalists from pressure, threats, physical attacks and lawsuits aimed at silencing critical reporting is a prerequisite for free journalism. Without

28 journalists who can safely investigate and publish their reports, citizens cannot access high- quality information, regardless of how developed their individual competences are. This requires not only appropriate legal frameworks adequately penalising attacks against journalists, but also effective law enforcement and strengthening rapid legal and institutional response mechanisms when journalists are targeted. Although such measures require political will and may face resistance because they are undermining the status quo, they are indispensable for creating a media environment in which media literacy has substantive purpose. IV. Low social trust, extremely limited use of privacy protection measures and low trust in artificial intelligence together create an environment in which citizens - even when equipped with knowledge and skills - do not actively participate in the public sphere or adopt emerging technologies that are becoming part of the media ecosystem. This problem is related to people’s views on privacy, their trust in institutions and other people, and their understanding of how the new technologies are reshaping the media landscape. Interventions in this area must be multidimensional and target both individual practices and systemic safeguards. With regard to the domain of privacy and security, most citizens do not implement basic protection measures, even when they are aware of them. This suggests that they do not perceive the risks – how valuable their data are or the ways in which they can be abused, leading them to treat privacy protection as unnecessary. Awareness campaigns should focus on these issues, combining practical guidance (e.g., disabling tracking, managing app permissions) with clear explanations of concrete ways in which data can be abused. People should understand that privacy protection is not a luxury for those who have something to hide but a fundamental right and a reasonable precaution in the digital environment. Systemic security must be embedded into platforms through binding data protection standards applicable to all digital services - from global platforms to local e-government systems. This includes clear rules on data collection and use, mandatory encryption of sensitive data, automatic deletion of unnecessary data and transparency regarding data sharing. Citizens must have simple tools to control their data, monitor access and withdraw consent. Regulatory frameworks must prescribe effective sanctions for violations of standards, rather than symbolic penalties that large platforms treat as a cost of doing business. With regard to artificial intelligence, low trust and low usage reinforce one another, creating a vicious circle. Breaking this cycle requires transparent policies governing the use of AI in public institutions and media organisations. Citizens should know when they are

29 interacting with automated systems not real people, which algorithms select the news they see on social media, how image- and content-generating models are trained and who is accountable for AI errors. Such transparency fosters both trust and understanding. At the same time, education on AI should go beyond technical knowledge to include its societal implications - how AI transforms news and content production, the ethical dilemmas of automated content moderation, how citizens can protect their rights when algorithmic decisions affect them and how to recognise AI-generated content. At the same time, media organisations that use AI for generating or editing content must also ensure transparency and accountability. Ethical standards of AI use in the media should prevent the erosion of human accountability and the spread of disinformation. References Aufderheide, P. (1993). Media literacy: A report of the National Leadership Conference on Media Literacy. Aspen Institute. Broadcasting Authority of Ireland – BAI (2022). Media Literacy Policy. https://www.medialiteracyireland.ie/wp-content/uploads/2022/12/BAI-ML-Policy-1.pdf. Bulger, M. (2012). Measuring media literacy in a national context: challenges of definition, method and implementation. Media Studies, 3(6), 83-104. http://hrcak.srce.hr/96376. Celot, P. (2015). Assessing media literacy levels and the European Commission pilot initiative. European Association for Viewers Interests (EAVI). https://eavi.eu/wp- content/uploads/2017/08/assessing.pdf Celot, P. and Pérez Tornero, J. M. (2009). Study on assessment criteria for media literacy levels: A comprehensive view of the concept of media literacy and an understanding of how media literacy levels in Europe should be assessed (Final report). European Commission, Information Society and Media Directorate-General, Media Literacy Unit. https://ec.europa.eu/assets/eac/culture/library/studies/literacy-criteria-report_en.pdf Centre for Media Pluralism and Media Freedom (2024). Media Pluralism Monitor 2024: Serbia. European University Institute. https://cmpf.eui.eu/media-pluralism-monitor-2024/ Directive 2007/65/EC of the European Parliament and of the Council of 11 December 2007 amending Council Directive 89/552/EEC on the coordination of certain provisions laid down by law, regulation or administrative action in Member States concerning the pursuit of television broadcasting activities. Official Journal of the European Union, L 332, 27–45. DTI & EAVI (2011). Testing and refining criteria to assess media literacy levels in Europe: Final report. European Commission, Directorate-General for Information Society and Media, Media Literacy Unit. https://eavi.eu/wp- content/uploads/2017/08/study_testing_and_refining_ml_levels_in_europe.pdf EAVI (2011). Assessing Media Literacy Levels and the European Commission. https://eavi.eu/wp- content/uploads/2017/08/assessing.pdf. Erstad, O., & Amdam, S. (2013). From Protection to Public Participation: A Review of Research Literature on Media Literacy. Javnost - The Public, 20(2), 83-98. European Commission (2009). Commission Recommendation of 20 August 2009 on media literacy in the digital environment for a more competitive audiovisual and content industry and an

30 inclusive knowledge society. Official Journal of the European Union, C 257, 1-6. https://eur- lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32009H0625. European Commission (2009). Study on Assessment Criteria for Media Literacy Levels. Brussels: European Commission - Expert Group on Media Literacy. https://ec.europa.eu/assets/eac/culture/library/studies/literacy-criteria-report_en.pdf European Commission (2011). Testing and Refining Criteria to Assess Media Literacy Levels in Europe Final Report. Commissioned by the European Commission Directorate-General for Information Society and Media; Media Literacy Unit. Danish Technological Institute & European Association for Viewers’ Interests. https://eavi.eu/wp- content/uploads/2017/08/study_testing_and_refining_ml_levels_in_europe.pdf European Commission (2007). A European approach to media literacy in the digital environment (Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM(2007) 833 final). Eurostat (2024). Trust in others by sex, age and educational attainment. Eurostat. https://ec.europa.eu/eurostat/databrowser/view/ilc_pw03/default/table FarrPoint (2024). International Digital Connectivity Readiness Index (DCRI), September 2024. FarrPoint. https://www.farrpoint.com/uploads/store/mediaupload/1591/file/FarrPoint_Internat ional_DCRI_Sept_2024.pdf Freedom House (2024). Freedom in the World 2024: Serbia. Freedom House. https://freedomhouse.org/country/serbia/freedom-world/2024 Gemius (2024). Gemius Audience Serbia 2024: Online news portal reach. Gemius. https://e- public.gemius.com/rs/rankings/12730 Government of the Republic of Serbia (2020). Strategija razvoja sistema javnog informisanja u Republici Srbiji za period 2020-2025. godina. Službeni glasnik RS 11/2020. https://www.media.srbija.gov.rs/medsrp/dokumenti/medijska_strategija210_cyr.pdf. Grizzle, A., Moore, P., Dezuanni, M., Asthana, S., Wilson, C., Banda, F., Onumah, C. (2013). Media and information literacy: policy and strategy guidelines. UNESCO. https://unesdoc.unesco.org/ark:/48223/pf0000225606 Grizzle, Alton, Torrent, Jordi & Perez Tornero, Jose Mauel (2013a). MIL as a Tool to Reinforce Intercultural Dialogue: Principles and Aims of the UNESCO-UNAOC UNITWIN Cooperation Programme on Media and Information Literacy and Intercultural Dialogue. In: Carlsson, Ula & Hope Culver, Sherri (eds.) Media and Information Literacy and Intercultural Dialogue, (pp. 9-16). MILID Yearbook, The International Clearinghouse on Children, Youth and Media Nordicom. Sweden: University of Gothenburg. https://milunesco.unaoc.org/wp- content/uploads/2013/04/Media_and_Information_Literacy_and_Intercultural_Dialogue.pdf Hobbs, R. (2010). Digital and media literacy: A plan of action. Aspen Institute. Jolls, T. & Thoman, E. (2008). Literacy for the 21 st Century. An Overview & Orientation Guide to Media Literacy Education. Part I: Theory CML MediaLit Kit™ A Framework for Learning and Teaching in a Media Age. Centre for Media Literacy. https://webspace.ship.edu/hliu/etextbook/theory/doc/media%20literacy_v02.pdf Ofcom (2023). Future Technology and Media Literacy: Anchor document. https://www.ofcom.org.uk/siteassets/resources/documents/research-and-data/media-literacy- research/making-sense-of-media/future-technology-trends-and-media-literacy/anchor- document.pdf?v=329832 Ofcom. (2025). Adults' media use and attitudes report 2025. https://www.ofcom.org.uk/media-use- and-attitudes/media-habits-adults/adults-media-use-and-attitudes

31 Organisation for Economic Co-operation and Development. (2023). PISA 2022 results (Volume I): The state of learning worldwide – Reading. OECD Publishing. https://www.oecd.org/content/dam/oecd/en/publications/reports/2023/12/pisa- 2022-results-volume-i_76772a36/53f23881-en.pdf OSCE (2022) Media users over the age of 65 in Serbia. https://www.osce.org/files/f/documents/3/2/541704.pdf OSCE (2025). 2025 Media Literacy of Serbia’s citizens, Survey Research Report. Belgrade. 2025 Media Literacy of Serbia’s citizens, Survey Research Report.pdf Pereira, S., & Moura, P. (2022). Assessing media literacy competences: Reflections and recommendations from a quantitative study. Journal of Media Literacy Education, 14(3), 79- 93. https://doi.org/10.23860/JMLE-2022-14-3-7 Perez-Escoda, Ana, Garcia-Ruiz, Rosa, Aquaded, Ignacio. 2016. International dimensions of media literacy in connected world. ATI – Appliens Technologies and Innovations, 12 (2), 95-106. Potter, James W. (2022). Analysis of definitions of media literacy. Journal of Media Literacy Education. 14(2), 27-43. https://doi.org/10.23860/JMLE-2022-14-2-3 Reporters Without Borders (2024). 2024 World Press Freedom Index. Reporters Without Borders. https://rsf.org/en/index?year=2024 Rowe, G. & Wright, G. (1999). “The Delphi Technique as a Forecasting Tool: Issues and Analysis,” International Journal of Forecasting, Vol. 15, No. 3, pp. 53-375. SHARE Foundation (2024). Cyber-safety monitoring report 2024: Serbia. SHARE Foundation. https://monitoring.labs.rs/en Statistical Office of the Republic of Serbia (2023). USAGE OF INFORMATION AND COMMUNICATION TECHNOLOGIES IN THE REPUBLIC OF SERBIA, 2023: Households/Individuals, Enterprises (ISSN 1820-9084). SORS. G202316018.pdf The National Association for Media Literacy Education (NAMLE). 2024. Media Literacy Framework. https://namle.org/resources/media-literacy-defined/ Tortoise Media (2025). The Global AI Index. Tortoise Media. https://www.tortoisemedia.com/data/global-ai UNESCO (2022). Global Standards for Media and Information Literacy Curriculum Development. Paris: UNESCO. https://www.unesco.org/sites/default/files/medias/files/2022/02/Global%20Standards%20for %20Media%20and%20Information%20Literacy%20Curricula%20Development%20Guidelin es_EN.pdf United Nations Department of Economic and Social Affairs (2024). United Nations E-Government Survey 2024: E-Participation index – Serbia. UN. https://publicadministration.un.org/egovkb/en-us/Data/Country-Information/id/151- Serbia Wadhwa, A., & Nandal, S. (2025). Development of an Index in Social Science: A Systematic Literature Review. The Scientific Temper, 16(03), 3927–3942. Doi: 10.58414/SCIENTIFICTEMPER.2025.16.3.09

Projekat “Jačanje slobode izražavanja i slobode medija u Srbiji“, Misija OEBS-a u Srbiji INDEKS MEDIJSKE PISMENOSTI U SRBIJI 2025 - pilot verzija – Beograd, 2025 Autori izveštaja: dr Marijana Matović, Fakultet političkih nauka, Univerzitet u Beogradu dr Ana Milojević, Fakultet političkih nauka, Univerzitet u Beogradu dr Snežana Milin Perković, ekspertkinja za medijsku pismenost dr Aleksandar Tomašević, Institut za fiziku, Univerzitet u Beogradu Ovaj projekat finansira Evropska unija Napomena: Ovaj izveštaj objavljen je uz finansijsku pomoć Evropske unije. Stavovi izrečeni u ovom izveštaju pripadaju isključivo autorima i njihovim saradnicima i ne predstavljaju nužno zvaničan stav Misije OEBS-a u Srbiji i nužno ne odražavaju stavove Evropske unije.

Sadržaj Uvod........................................................................................................................................................ 1 Proces konstruisanja indeksa ........................................................................................................... 2 Osnovna polazišta ................................................................................................................................. 3 Definisanje medijske pismenosti ..................................................................................................... 3 Medijska pismenost u kontekstu ..................................................................................................... 5 Indeks medijske pismenosti Srbije ...................................................................................................... 7 Osnovni okvir i operativna definicija medijske pismenosti ........................................................... 7 I INDEKS MEDIJSKE PISMENOSTI (IMP), zasnovan na individualnim kompetencijama ............ 8 Dizajn instrumenta i dimenzije merenja ..................................................................................... 8 Uzorak i prikupljanje podataka .................................................................................................... 9 Indeks medijske pismenosti - IMP ............................................................................................. 10 II INTEGRISANI INDEKS MEDIJSKE PISMENOSTI IIMP ............................................................... 14 FAKTORI OKRUŽENJA - ANALIZA EKSTERNIH INDIKATORA ................................................. 14 III VALIDACIJA I FINALNA KONSTRUKCIJA IIMP ......................................................................... 22 Delphi proces i postizanje konsenzusa ...................................................................................... 22 Glavna saznanja ........................................................................................................................... 24 Zaključna razmatranja i preporuke ................................................................................................... 26 Literatura ............................................................................................................................................. 30

1 Uvod Izveštaj Indeks medijske pismenosti u Srbiji 2025 odgovara na nekoliko zahteva. Jedan od zahteva je bio da se pruži procena trenutne situacije i pripremi teren za godišnju evaluaciju efekata kumulativnih aktivnosti u oblasti medija i digitalne pismenosti u odnosu na Indeks. Drugi važan zahtev odnosio se na razvijanje instrumenta za istraživanje razvijenosti medijske pismenosti opšte populacije. Ostvarenje ovih zahteva predstavlja ujedno i prve korake u realizaciji Aktivnosti 5.1.1 Akcionog plana Medijske strategije 2020–2025 gde piše da je potrebno „uspostaviti indeks medijske pismenosti u Republici Srbiji koji je u skladu sa evropskom metodologijom“. Treći zahtev je bio da se napravi Indeks koji bi bio osetljiv na promene u okruženju (medijskom, društvenom, tehnološkom, itd.) kao i na dostupnost relevantnih podataka. U procesu konstruisanja indeksa razmatrali smo ključna pitanja koja definišu granice i implikacije pojma medijske pismenosti u današnjem kontekstu: • Šta sve utiče na medije i medijske sadržaje danas - na njihov kvalitet, kreiranje, distribuciju, publiku, percipiranje? • Da li je medijska pismenost progresivna kompetencija koja se razvija linearno, ili postoje kontekstualni faktori koji mogu dovesti do regresije? • Da li postoje medijski pismene osobe, koje imaju pristup, umeju da analiziraju i evaluiraju medijske poruke i aktivno participiraju u digitalnom okruženju, ali na nedemokratski način i/ili u nedemokratskom društvu? • Može li se govoriti o medijskoj pismenosti u okruženjima sa ograničenim medijskim slobodama i pluralizmom? • Da li medijska pismenost podrazumeva normativnu dimenziju – vrednosti i stavove usklađene sa demokratskim principima (poverenje u institucije, zaštita osnovnih prava, sloboda govora)? Odgovori na ova pitanja oblikovali su dvostruku strukturu Integrisanog indeksa medijske pismenosti (IIMP): anketno merenje individualnih kompetencija (pristup, kritička evaluacija, participacija) kombinovano sa eksternim indikatorima okruženja (infrastruktura, obrazovni sistem, medijske slobode, demokratski kontekst, poverenje i sajber-bezbednost, pripremljenost

2 za veštačku inteligenciju). Ovaj pristup omogućava ne samo dijagnozu trenutnog stanja, već i identifikaciju sistemskih prepreka i prioriteta za javne politike. U ovom izveštaju ponuđen je Indeks medijske pismenosti (IMP) koji je zasnovan na anketnom istraživanju sprovedenom na reprezentativnom uzorku populacije Republike Srbije i služi da izmeri aktuelni nivo medijske pismenosti građana Srbije. Nakon njega predložen je i model za Integrisani indeks medijske pismenosti (u daljem tekstu: IIMP) koji predstavlja sveobuhvatan metodološki pristup merenju razvijenosti medijske pismenosti u Srbiji. Za razliku od modela koji se fokusiraju isključivo na individualne kompetencije ili kontekstualne faktore, IIMP istovremeno meri oba nivoa i njihovu interakciju – sposobnosti građana da koriste, kritički evaluiraju i aktivno učestvuju u medijskom okruženju, kao i strukturalne uslove koji omogućavaju ili ograničavaju razvoj tih kompetencija. IIMP predstavljen u ovom izveštaju rezultat je multidisciplinarnog istraživačkog procesa koji integriše teorijsku analizu, empirijsko istraživanje i ekspertsku validaciju 1 . Kao pilot istraživanje, pruža šire rezultate indeksa medijske pismenosti za Srbiju 2025. i moguću metodološku osnovu za buduće longitudinalne analize i međunarodne komparacije. Proces konstruisanja indeksa Konstruisanje indeksa je kompleksan proces koji se odvija u više faza. Kako ističu Vadva i Nandal (Wadhwa i Nandal, 2025), indeksi služe da agregiraju različite, za fenomen važne indikatore u jedinstvenu kompozitnu meru kako bi omogućili sveobuhvatniji pregled konkretnog fenomena. Kao takvi, indeksi omogućavaju praćenje trendova kroz vreme, komparaciju sa drugim zemljama i formulisanje javnih politika. Proces konstruisanja indeksa počinje sa operacionalizacijom koncepta koji želimo da merimo. Ovo je ključna faza, koja zahteva pronalaženje balansa između teorijskog određenja nekog koncepta (u ovom slučaju medijske pismenosti) i dostupnosti pouzdanih podataka potrebnih za prevođenje koncepta u merljive i međusobno kompatibilne konstrukte. Koristeći ponovo reči Vadve i Nandala: „Apstraktni koncepti zahtevaju pažljivu operacionalizaciju, koja podrazumeva definisanje koncepta kroz merljive termine, izbor odgovarajućih indikatora i određivanje metodologije za kombinovanje tih indikatora u koherentan indeks. Izbor indikatora je kritičan korak koji često 1 Istraživački tim zahvaljuje ekspertima koji su učestvovali u Delphi procesu na doprinosu razvoju IIMP.

3 zahteva balanisiranje između teorijske relevantnosti i dostupnosti, odnosno pouzdanosti podataka.“ (Wadhwa & Nandal, 2025: 3927). Polazeći od ovakve definicije indeksa, postavili smo kao glavni cilj projekta da konstruišemo indeks medijske pismenosti koji uključuje faktore okruženja. U preliminarnoj fazi realizacije cilja, sproveli smo pregled literature i relevantnih istraživanja kako bismo utvrdili ključne elemente medijske pismenosti i njenu vezu sa okruženjem. Na osnovu ovog pregleda, uspostavili smo radni okvir integrisanog indeksa i pregledali odabrane eksterne podatake. Dalju realizaciju glavnog cilja podelili smo u tri faze: Prvu fazu sačinjavala je priprema i realizacija anketnog istraživanja: konstruisanje upitnika (koji će biti komplementaran sa postojećim eksternim izvorima podataka), sprovođenje anketnog istraživanja (prikupljanje podataka), analiza anketnih podataka i konstruisanje Indeksa medijske pismenosti - IMP na nivou individualnih kompetencija. Druga faza je bila posvećena pripremi eksternih podataka za integraciju sa rezultatima anketnog istraživanja (IMP), kao i integraciju podataka u preliminarni Integrisani indeks medijske pismenosti - IIMP. Treća faza je obuhvatila validaciju IIMP primenom Delfi metoda 2 (konsultacije sa ekspertima za medijsku pismenost) koji je rezultirao uspostavljanjem finalnog IIMP Integrisanog indeksa medijske pismenosti za Srbiju 2025. godine. Osnovna polazišta Definisanje medijske pismenosti Medijska pismenost je koncept sa dugom teorijskom tradicijom ali heterogenim operacionalizacijama. Većina definicija, od početnih (Aufderheide 1993) do savremenih okvira UNESCO (2022) i NAMLE (2024), je zadržala tri ključna elementa: (1) pristup medijima i informacijama, (2) kritička evaluacija sadržaja, i (3) participacija kroz medije (kreiranje, komunikacija, građanska akcija). 2 Delfi metod je razvijen od strane RAND korporacije sredinom 1950-tih godina kao istraživačka tehnika u okviru društvenih nauka čiji je cilj postizanje konsenzusa među ekspertima. Primenjuje se uglavnom onda kada postoji neizvesnost, odnosno kada ne postoji dovoljno podataka potrebnih za donošenje informisanih odluka. Ovaj metod je zasnovan na konsenzusu, strukturiran je i koristi iterativni pristup gde se rezultat svake iteracije koristi kao početna tačka sledeće, do postizanja konsenzusa. (više o Delfi metodu videti kod Rowe & Wright, 1999.)

4 Ovaj konsenzus potvrđuju meta-analitičke studije. Bulger (2012) je u pregledu definicija identifikovala pristup/korišćenje, analizu/evaluaciju i kreiranje sadržaja kao dominantne komponente. Potter (2022) je na osnovu analize 210 radova, kategorizovao preko 400 elemenata medijske pismenosti u četiri domena: pristup, analiza, evaluacija i komunikacija. Evropska komisija (COM 2007/833), UNESCO (Grizzle et al., 2013), regulatorna tela (BAI, 2022; Ofcom, 2025, 2023) i profesionalna udruženja (NAMLE, 2024; Jolls & Thoman, 2008) takođe definišu medijsku pismenost oslanjajući se na tri domena – pristup, analizu i evaluaciju, kreaciju i komunikaciju. Međutim, svaki od ovih domena je konceptualno evoluirao u odgovoru na tehnološke i društvene transformacije. U ranim konceptualizacijama, pristup se primarno odnosio na fizičku i tehnološku dostupnost i proveravano je da li domaćinstva imaju internet, a građani znanje da koriste uređaje. Savremena određenja proširuju razumevanje pristupa u tri pravca. Prvo, pristup podrazumeva ne samo tehničku mogućnost već i raznovrsnost dostupnih gledišta (pluralizam sadržaja), što direktno zavisi od medijskih sloboda i strukturnog pluralizma (EC, 2007). Drugo, građani mogu imati pristup tehnički, a da im nedostaju veštine navigacije, pretraživanja ili evaluacije relevantnosti. Prema tome, pristup se posmatra i kao kognitivna dostupnost (Grizzle et al., 2013). Treće je shvatanje pristupa kao samosvesti o informacionim potrebama. Pristup je smislen ako je korisnik svestan svojih potreba za određenim medijskim sadržajima i poznaje način da te potrebe adekvatno zadovolji (ibid). Paradoksalno, pristup je i najlakše i najteže merljiv aspekt, jer se dostupnost infrastrukture može lako kvantifikovati, dok kognitivna dostupnost i svest o potrebama zahtevaju sofisticiranije instrumente. Sposobnost analize i evaluacije medijskih sadržaja predstavlja najstabilniju komponentu u definicijama medijske pismenosti (Pereira & Moura, 2022). Ova dimenzija obuhvata prepoznavanje tipa sadržaja (vest, reklama, mišljenje), izvora (javni medij, privatna platforma, neidentifikovan akter) i konteksta proizvodnje. Zatim, razumevanje medijskog ekosistema, osnove regulatornog okvira, zastupljenost vlasničkih struktura, ekonomske modele na kojima počiva kreiranje sadržaja, i tehnološka ograničenja koje oblikuju sadržaj. Potom, primenu kriterijuma za procenu verodostojnosti izvora, tačnosti i uravnoteženosti sadržaja. U digitalnom okruženju, kritička evaluacija obuhvata nužno i nove domene, poput sadržaja generisanih veštačkom inteligencijom, kampanje dezinformisanja i koordiniranih manipulacija javnim mnenjem. Ova dimenzija odražava obrazovne kapacitete sistema – kvalitet razvijanja kritičkog mišljenja kroz formalno školovanje i dostupnost medijskog obrazovanja.

5 Konceptualizacija participativne dimenzije medijske pismenosti je najraznolikija. Neki autori fokusiraju se na kreiranje sadržaja – pisanje blogova, snimanje videa, uređivanje slika (Jolls & Thoman, 2008). Drugi naglašavaju komunikaciju – sposobnost efektivnog izražavanja kroz različite medijske formate (Pereira & Moura, 2022). Treći povezuju participaciju sa građanskom akcijom – korišćenjem digitalnih alata za društveni angažman, peticije, kampanje, javnu deliberaciju (Hobbs, 2010; NAMLE, 2024). EAVI okvir (Celot & Pérez Tornero, 2009) sagledava participativne veštine trostruko: prvo, kao socijalnu participaciju, koja podrazumeva umrežavanje, kontakte, zajednice, vršnjačku komunikaciju; potom, kao javnu participaciju kroz koju se ostvaruje interakcija sa institucijama, izražavanje mišljenja, mobilizacija interesnih grupa; najzad, kao kreativnu participaciju, odnosno učešće u produkciji sadržaja, kreiranje kroz saradnju, rešavanje problema putem medija. Australijsko regulatorno telo (ACMA, 2009) dodatno uključuje digitalnu sigurnost kao participativnu kompetenciju (upravljanje privatnošću, zaštitu od prevara, odgovorno deljenje podataka) sa argumentom da bez bezbednosti nema smislenog učešća (Perez-Escoda et al., 2016). Medijska pismenost u kontekstu Ključna pretpostavka našeg modela jeste da medijska pismenost nije atribut individue u vakuumu već relacijski fenomen – interakcija između kompetencija i okruženja. Kako ističu Erstad i Amdam (2013), pismenost se ne samo razvija već i praktikuje u specifičnim kontekstima koji mogu biti podsticajni ili ograničavajući. Ovaj pristup eksplicitno je artikulisan u evropskim politikama. Direktiva o audiovizuelnim medijskim uslugama (Directive 2007/65/EC: 3) definiše medijsku pismenost kao sposobnost „pristupa medijima, razumevanja i kritičke procene različitih aspekata medija i medijskih sadržaja i stvaranja komunikacija“ uz ključno dodavanje: „u različitim kontekstima“ (naglasak autora). Naredni dokumenti (EC, 2009; EC Expert Group, 2009) razrađuju ovu dimenziju, naglašavajući da iste individualne kompetencije proizvode različite ishode zavisno od medijskih sloboda, obrazovnih sistema, demokratskih institucija i tehnološke infrastrukture. UNESCO (Grizzle et al., 2013a; UNESCO, 2022) eksplicitno vezuje medijsku i informacionu pismenost (MIL) za demokratsko društvo i međukulturnu saradnju, pozicionirajući je kao „preduslov za informisano učešće u digitalnom društvu“ – čime implicitno priznaje da bez demokratskog konteksta, medijska pismenost gubi svoju

6 emancipatornu funkciju. Regulatorno telo Irske (BAI, 2022) operacionalizuje participativni domen medijske pismenosti kroz indikatore individualnih veština i indikatore društvenog uticaja, prevodeći individualne kompetencije na nivo kapaciteta zajednica i sistema. Na razumevanju medijske pismenosti kao relacijskog fenomena (interakcija kompetencija i okruženje) razvijen je okvir Evropske asocijacije za ostvarivanje interesa gledalaca (EAVI) kroz seriju studija za Evropsku komisiju (Celot, 2009; Celot & Pérez Tornero, 2009; DTI & EAVI, 2011). Iniciran naporima za uspostavljanje metodologije za merenje medijske pismenosti na evropskom nivou EAVI okvir uvodi dvostruku arhitekturu merenja koja razlikuje individualne kompetencije i faktore okruženja. Individualne kompetencije, u ovom okviru predstavljaju „kapacitet da se primene određene veštine“ (Celot, 2009: 7) koje se granaju na lične kompetencije (korišćenje medija, kritičko čitanje sadržaja) i socijalne kompetencije (komunikacija kroz medije, participacija u javnoj sferi). Faktore okruženja su definisali kao „skup kontekstualnih faktora“ koji omogućavaju ili ograničavaju razvoj i praktikovanje individualnih kompetencija. Ovi faktori obuhvataju, dostupnost medija (tehnološka infrastruktura, pluralizam sadržaja), medijske politike (regulatorni okviri, zaštita korisnika), obrazovanje (formalno i neformalno medijsko obrazovanje), medijske industrije (vlasništvo, ekonomski modeli, profesionalni standardi) kao i civilno društvo (organizacije, inicijative za medijsku pismenost). Ključna premisa EAVI okvira jeste da iste individualne kompetencije mogu da proizvode različite ishode zavisno od strukturalnih uslova. Prema tome možemo zamisliti različite scenarije. Na primer, građani mogu posedovati veštine kritičke evaluacije, ali u okruženju sa niskim medijskim pluralizmom nemaju pristup raznovrsnim izvorima i njihove kompetencije ostaju nerealizovane. U drugom scenariju, infrastruktura je razvijena i pristup visok, ali obrazovni sistem ne razvija kritičko mišljenje pa građani koriste medije pasivno i ne evaluiraju sadržaj. Treći scenario može biti, aktivno civilno društvo koje podstiče participaciju, ali demokratski kontekst je ograničavajući (nizak nivo slobode govora) pa građanska upotreba medija nosi rizike. Medijska pismenost, prema tome, proizilazi iz sadejstva kompetencija i okruženja – od njihove konvergencije ili divergencije zavisi stvarna sposobnost građana da efektivno

7 participiraju u medijskoj javnoj sferi. Dakle, polazeći od sagledavanja medijske pismenosti kao relacionog fenomena kreiramo indeks za procenu stanja medijske pismenosti u Srbiji. Indeks medijske pismenosti Srbije Osnovni okvir i operativna definicija medijske pismenosti Konstrukcija Integrisanog Indeksa Medijske Pismenosti - IIMP zasniva se na EAVI konceptualnom okviru koji primenjujemo u kontekstu Srbije. EAVI okvir pruža tri ključne prednosti za naše istraživanje. Najpre, njegova osnovna struktura ostaje relevantna uprkos brzim tehnološkim promenama – razlikovanje između onoga što pojedinci mogu i konteksta u kojem to čine nije zastarelo činjenicom da su se pojavile nove tehnologije poput veštačke inteligencije ili nove pretnje poput koordiniranih dezinformacionih kampanja. Naprotiv, upravo te promene naglašavaju važnost kontekstualnog pristupa jer iste individualne veštine proizvode različite ishode zavisno od medijskih sloboda, obrazovnih kapaciteta i demokratskih institucija u kojima se primenjuju. Druga prednost leži u komparativnoj primenljivosti. EAVI je razvijen sa namerom da omogući međunarodne komparacije uprkos različitim političkim tradicijama, medijskim sistemima i obrazovnim modelima evropskih zemalja. Ugrađivanje EAVI logike u IIMP otvara mogućnost da se metodologija primeni u drugim zemljama regiona, čime bi se vremenom stvorila baza komparativnih podataka koja bi omogućila dublju analizu odnosa između strukturalnih faktora i stvarnih kompetencija građana. Treće, integrisani pristup ima direktne implikacije za dijagnostiku i kreiranje javnih politika. Kada indeks pokazuje nizak nivo medijske pismenosti, potrebno je utvrditi da li je to rezultat nedostatka individualnih veština, sistemskih barijera, ili njihove neusklađenosti. Svaki od ovih scenarija zahteva različite vrste intervencija. Nedostatak veština upućuje na potrebu za programima medijskog obrazovanja, sistemske barijere zahtevaju strukturalne reforme (medijskih sloboda, obrazovnog sistema, demokratskih institucija), dok situacija gde postoje i kompetencije i mogućnosti ali se ne susreću sugeriše potrebu za intervencijama koje se odnose

8 na motivaciju, poverenje, percipiranu sigurnost. Integrisani indeks omogućava ovakvu vrstu dijagnostike i njegove analitičke mogućnosti demonstriramo u zaključnim odeljcima nakon što predstavimo empirijske nalaze za Srbiju. Dakle, imajući u vidu tri ključne dimenzije, i polazeći od EAVI okvira, u kreiranju indeksa polazimo od sledeće definicije medijske pismenosti: Medijska pismenost predstavlja sposobnost pristupa, kritičkog čitanja (analize i evaluacije) i participacije kroz medije (kao sadržaje i tehnologije), pri čemu se ove kompetencije razvijaju i praktikuju u specifičnim društvenim kontekstima koji obuhvataju tehnološku infrastrukturu, obrazovni sistem, medijske slobode, demokratske institucije, društveno poverenje i pripremljenost za nove tehnologije (veštačka inteligencija). Ovakva definicija omogućava konstrukciju indeksa koji istovremeno meri, šta građani mogu (individualne kompetencije – IMP), u kakvom okruženju to čine (strukturalni faktori – eksterni indikatori), i kakvu medijsku pismenost proizvodi njihova interakcija (integrisani IIMP). U narednim odeljcima detaljno objašnjavamo kako smo operacionalizovali individualne kompetencije kroz anketno istraživanje, kako smo odabrali i harmonizovali eksterne indikatore okruženja, i na koji način smo kroz Delphi proces konsultacije sa ekspertima došli do finalne strukture i težina integrisanog indeksa. I INDEKS MEDIJSKE PISMENOSTI (IMP), zasnovan na individualnim kompetencijama Dizajn instrumenta i dimenzije merenja Dizajn anketnog instrumenta sledi EAVI konceptualni okvir, fokusiran je na individualne kompetencije građana i meri tri ključne dimenzije: koliko su građani povezani sa medijskim ekosistemom kroz raznovrsnost i intenzitet korišćenja različitih platformi, koliko su sposobni za kritičku evaluaciju sadržaja kojima su izloženi, i u kojoj meri koriste medije za građansku participaciju i zaštitu sopstvene privatnosti u digitalnom prostoru. Međutim, u operacionalizaciji ovih dimenzija za kontekst Srbije 2025. godine uzeli smo u obzir tehnološke i društvene promene koje nisu postojale u vreme koncipiranja originalnog modela. Prilagođavanje EAVI okvira zahtevalo je tri ključna proširenja. Najpre, uvedena je dimenzija veštačke inteligencije (VI) koja ranije nije bila prisutna u tradicionalnim konceptualizacijama medijske pismenosti. Ova dimenzija obuhvata ne samo svest građana o prisustvu VI u medijskom okruženju već i njihove obrasce korišćenja VI alata za pristup informacijama i

9 kreiranje sadržaja. Drugo, dimenzija kritičke evaluacije proširena je da uključi verifikaciona ponašanja – ne meri se samo sposobnost prepoznavanja problematičnog sadržaja već i konkretne radnje koje građani preduzimaju kada naiđu na sumnjive informacije (proveravanje kroz više izvora, konsultovanje platformi za proveru činjenica, traganje za originalnim kontekstom). Ova proširenja odgovaraju na stvarnost savremenog medijskog okruženja gde nije dovoljno samo kritički misliti već aktivno i sistematski proveravati ono što se prezentuje kao informacija. Treće, dimenzija participacije dopunjena je veštinama zaštite u digitalnom okruženju – blokiranjem praćenja, zaštitom privatnosti, prijavljivanjem govora mržnje – prepoznajući da bez osećaja sigurnosti i kontrole nad ličnim podacima, građani ne mogu smisleno participirati u digitalnoj javnoj sferi. Finalni instrument obuhvata šezdeset tri pitanja organizovana u tri sekcije koje prate teorijsku strukturu medijske pismenosti 3 . Prva sekcija (P1–P16) meri pristup i korišćenje medija, uključujući tradicionalne medije (televizija, radio, štampa) i digitalne platforme (društvene mreže, portali, video platforme, aplikacije za razmenu poruka), kao i obrasce korišćenja veštačke inteligencije. Druga sekcija (P17–P42) fokusirana je na kritičko čitanje medijskih sadržaja i obuhvata prepoznavanje različitih tipova problematičnog sadržaja (dezinformacije, senzacionalizam, pristrasno izveštavanje), sposobnost upravljanja emocionalnim reakcijama na medijske poruke, i konkretna ponašanja verifikacije. Treća sekcija (P43–P56) meri komunikaciju, kreiranje sadržaja, participaciju i zaštitne prakse u digitalnom okruženju. Dodatnih sedam pitanja (P57–P63) obuhvata sociodemografske karakteristike ispitanika koje omogućavaju stratifikovane analize prema polu, uzrastu, obrazovanju, tipu naselja i drugim relevantnim varijablama. Uzorak i prikupljanje podataka Anketno istraživanje je sprovedeno na dvostepeno stratifikovanom reprezentativnom uzorku od 800 punoletnih građana Srbije. Stratifikacija je izvršena na dva nivoa: prvi nivo činile su teritorijalne celine (Beograd, Vojvodina, centralna Srbija), a drugi nivo tip naselja (urbano, ruralno) unutar svake teritorijalne celine. Ovakva struktura omogućava da se podaci prikazuju prema ključnim sociodemografskim obeležjima – polu, starosnim grupama, nivou obrazovanja, prihodima, regionu i tipu naselja. Distribucija uzorka po polu i regionu u dobroj je meri usaglašena sa procenama Republičkog zavoda za statistiku, uz blagi pomak ka starijim 3 Kompletan upitnik je dostupan na svaki osnovan zahtev.

10 i ruralnim segmentima populacije što je karakteristično za telefonska istraživanja. Uprkos ovom pomaku, uzorak obezbeđuje zadovoljavajuću reprezentativnost odrasle populacije Srbije. Terensko prikupljanje podataka realizovano je u periodu od 15. maja do 15. juna 2025. godine primenom CATI (Computer Assisted Telephone Interviewing) metoda. Telefonsko anketiranje omogućilo je brzo prikupljanje podataka na geografski disperznom uzorku uz istovremenu kontrolu kvaliteta kroz superviziju poziva i automatsku proveru logičkih konzistencija u toku razgovora. Prosečno trajanje intervjua bilo je dvadeset pet minuta. Podaci su kodirani i obrađeni u programu SPSS, pri čemu su primenjene standardne procedure čišćenja podataka i provere konzistencije odgovora. Indeks medijske pismenosti - IMP Nakon čišćenja podataka i testiranja pouzdanosti, sedam validiranih skala normalizovano je kako bi se omogućila njihova agregacija u kompozitne skorove. Normalizacija je izvršena uz poštovanje originalne metrike odgovora. Normalizovane skale objedinjene su u tri različite dimenzije medijske pismenosti na osnovu teorijskih razmatranja (koji aspekti su konceptualno centralni za svaku dimenziju) i empirijskih nalaza faktorske analize (koja pitanja stvarno mere zajedničku latentnu dimenziju i koliko objašnjavaju ukupnu varijansu). Prva dimenzija meri raznolikost i intenzitet upotrebe različitih medijskih platformi i spremnost za usvajanje novih alata. Sastoji se iz dve komponente: koja meri koliko različitih izvora građani koriste za informisanje i zabavu, i obrasci upotrebe veštačke inteligencije (P10– P16) koja meri da li i kako građani koriste VI alate. Veća težina data je raznolikosti medija jer je to fundamentalniji aspekt medijske pismenosti – građani moraju najpre koristiti medije pre nego što mogu koristiti naprednije alate koji se baziraju na VI. Druga dimenzija obuhvata tri komponente kritičkog čitanja medijskih sadržaja. Prepoznavanje problematičnog sadržaja (P17–P21) meri da li građani mogu identifikovati dezinformacije i pristrasno izveštavanje. Upravljanje uticajem informacija (P22–P24) fokusira se na emotivnu regulaciju – da li građani dozvoljavaju da ih medijske poruke brzo uznemire ili su sposobni za distanciran pristup. Verifikaciona ponašanja (P25–P30) mere konkretne radnje

11 koje građani preduzimaju kada naiđu na sumnjive informacije – da li proveravaju kroz više izvora, konsultuju fact-checking sajtove na primer. Najveća težina data je prepoznavanju jer je to preduslov za ostale dve komponente – građani najpre moraju uočiti problem pre nego što mogu na njega odgovoriti. Treća dimenzija integriše dve: Digitalne bezbednosne i zaštitne veštine (P43–P50) mere primenu konkretnih mera zaštite privatnosti – blokiranje praćenja, upravljanje podacima koje građani dele, prijavljivanje govora mržnje. Građanski angažman (P51–P56) meri koliko građani koriste digitalne platforme za javnu participaciju – deljenje društveno relevantnih informacija, pozivanje na odgovornost, zagovaranje ideja, organizovanje događaja. Veća težina data je građanskom angažmanu jer je to finalni cilj medijske pismenosti u demokratskom društvu, dok su zaštitne veštine preduslov koji omogućava sigurno učešće. Rezultati otkrivaju značajne razlike među dimenzijama. Dimenzija koja se odnosi na pristup i korišćenje pokazuje najnižu prosečnu vrednost (M = 25.5, SD = 19.7) uz najširi raspon (0–79.9), što potvrđuje velike razlike između ispitanika u pristupu medijima i korišćenju veštačke inteligencije. Kritičko čitanje se pokazala kao najstabilnija dimenzija (M = 42.1, SD = 15.1), što ukazuje da su strategije kritičkog čitanja i verifikacije relativno ravnomernije zastupljene u populaciji i predstavljaju noseći stub individualne medijske pismenosti. Participacija je prosečno niža (M = 27.6, SD = 20.2) i pokazuje najveću heterogenost, što sugeriše da participativna ponašanja i zaštitne prakse još nisu postali ujednačena praksa – mali deo populacije je veoma aktivan dok većina ostaje pasivna ili retko angažovana. Finalni kompozitni Indeks medijske pismenosti (IMP) konstruisan je kao ponderisani prosek tri dimenzije. Dimenzija kritičke evaluacije nosi dominantnu težinu kao najstabilniji i konceptualno najvažniji domen – sposobnost kritičkog razumevanja i provere informacija predstavlja srž medijske pismenosti bez koje ni pristup ni participacija ne mogu biti smisleni. Dimenzija koja obuhvata korišćenje medija i veštačke inteligencije kao i dimenzija koja se odnosi na digitalno građanstvo i participaciju nose niže težine, zato što pokazuju veću varijabilnost i zavisnost od strukturalnih faktora (infrastrukture, demokratskog konteksta) koji će biti obuhvaćeni u eksternim indikatorima. Raspodela težina određena je nakon testiranja nekoliko mogućih formula na osnovu teorijske logike i empirijske stabilnosti. Finalna formula za izračunavanje prosečne medijske pismenosti građana Srbije izgleda ovako:

12 IMP = 0,20 × D1 + 0,60 × D2 + 0,20 × D3 Tabela br. 1: Struktura INDEKSA MEDIJSKE PISMENOSTI – IMP prema dimenzijama medijske pismenosti DIMENZIJA MEDIJSKE PISMENOSTI ŠTA MERI Pitanja iz upitnika UDEO DIMENZIJE U IMP VREDNOSTI DIMENZIJA U IMP D1. PRISTUP I KORIŠĆENJE P1–P16 Raznolikost korišćenja medija Učestalost korišćenja devet vrsta medija, od televizije i radija do društvenih mreža i VI (P1–P9) 20% 25.5 Obrasci upotrebe veštačke inteligencije Načine na koje se ljudi oslanjaju na alate veštačke inteligencije (P10–P16) D2.KRITIČKO PROMIŠLJANJE P17–P42 Prepoznavanje problematičnog sadržaja Koliko često ispitanici primećuju lažne vesti, teorije zavere, govor mržnje i slične pojave (P17–P21) 60% 42.1 Upravljanje uticajem informacija Kako ljudi doživljavaju pritisak informacija i emotivne reakcije na medijske sadržaje (P22–P24) Verifikaciona ponašanja Strategije koje građani koriste kada proveravaju sumnjive informacije (P25–P30) D3.PARTICIPACIJA P43–P56 Digitalne bezbednosne i zaštitne veštine Osam aktivnosti kojima se proverava koliko građani znaju da zaštite sebe i druge onlajn (P43–P50) 20% 27.6 Građanski angažman Načine na koje ljudi koriste medije za informisanje, deljenje stavova i organizovanje (P51–P56) Prosečna vrednost IMP za Srbiju 2025. iznosi 35.9 poena (SD = 14.9), sa rasponom od 3 do 77.2.

13 Nijedan ispitanik ne dostiže gornju četvrtinu spektra (preko 75 poena), što ukazuje da veoma visoki nivoi medijske pismenosti praktično ne postoje u opštoj populaciji. Distribucija je asimetrična sa težištem u nižim zonama: 16% ispitanika se nalazi u kategoriji veoma niskog nivoa (0–20 poena), 45% u kategoriji niskog nivoa (21–40), 32% dostiže srednji nivo (41–60), a samo 7% prelazi prag visokog nivoa (61–80). Kategorija veoma visokog nivoa (81–100) ostaje nepopunjena. Ovakva distribucija jasno ukazuje na potrebu za sistemskim intervencijama. Većina građana (61%) ima niske ili veoma niske kompetencije, što znači da postoje praznine u znanju i da značajan deo populacije nije osposobljen za sigurnu i efektivnu navigaciju savremenim medijskim okruženjem. Istovremeno, postojanje malog ali prepoznatljivog segmenta sa visokim kompetencijama (7%) sugeriše da strukturalni uslovi dozvoljavaju razvoj medijske pismenosti kod onih koji imaju pristup obrazovanju, resurse i motivaciju – pitanje je kako te uslove učiniti dostupnijim široj populaciji. Distribucija ukupnog indeksa medijske pismenosti Anketni nalazi predstavljaju osnovu za izračunavanje indeksa medijske pismenosti na nivou individualnih kompetencija. Kao takav, ovaj Indeks se može korisititi za redovno praćenje porasta ili pada nivoa medijske pismenosti u Srbiji čime se posredno može pratiti i efikasnost javnih politika i drugih aktivnosti u ovoj oblasti. Praćenje je moguće i prema

14 dimenzijama medijske pismenosti, čime se pruža uvid u delotvornost kampanja koje ciljaju na konkretne kompetencije poput verifikacionih ponašanja, digitalnih veština i sl. Indeksom medijske pismenosti – IMP odgovoreno je na drugi zahtev koji se odnosio na razvijanje instrumenta za istraživanje razvijenosti medijske pismenosti opšte populacije. Ovaj Indeks (IMP) posredno odgovara i na treći zahtev da bude osetljiv na promene u okruženju. Svakako, anketni instrument je potrebno revidirati na godišnjem nivou u konsultaciji sa ekspertima iz akademske zajednice. II INTEGRISANI INDEKS MEDIJSKE PISMENOSTI IIMP Jedan od zahteva je bio da se pruži procena trenutne situacije i pripremi teren za godišnju evaluaciju efekata kumulativnih aktivnosti u oblasti medija i digitalne pismenosti u odnosu na Indeks. Prethodni nalazi pružaju uvid u medijsku pismenost građana Srbije, ali ne mogu sami objasniti zašto su kompetencije na ovom nivou. Odgovori leže u strukturalnim faktorima okruženja – u kvalitetu medijskog sistema, u kapacitetima obrazovanja, u nivou demokratskih sloboda i društvenog poverenja. Analiza tih faktora, koja sledi u narednom odeljku, omogućava da razumemo ne samo šta građani mogu već i u kakvim uslovima to mogu. Time se direktnije odgovara na zahtev osetljivosti indeksa na promene u okruženju. Jedna od mogućih nadogradnji IMP faktorima okruženja je predlog Integrisanog indeksa medijske pismenosti – IIMP koji podrobno obrazložemo u tekstu koju sledi. FAKTORI OKRUŽENJA - ANALIZA EKSTERNIH INDIKATORA Dok anketno istraživanje meri šta građani trenutno mogu, analiza faktora okruženja meri u kakvim sistemskim uslovima to čine. Medijska pismenost nije samo individualni atribut već proizilazi iz interakcije između kompetencija i strukturalnog konteksta. Odabrane faktore okruženja grupisali smo u šest kategorija: (1) infrastruktura i pristup, (2) obrazovni temelji, (3) medijske slobode i pluralizam, (4) demokratski kontekst, (5) društveno poverenje, privatnost i sajber-bezbednost i (6) pripremljenost za veštačku inteligenciju. Izbor kategorija vođen je kombinacijom teorijskih razmatranja i pragmatičnih ograničenja. Teorijski, kategorije pokrivaju dimenzije koje EAVI okvir prepoznaje kao relevantne faktore okruženja – dostupnost medija, obrazovanje, medijski sistem, društveni kapital. Pragmatično, izbor je ograničen dostupnošću pouzdanih i međunarodno uporedivih podataka. Određene dimenzije koje bi bile idealno relevantne – poput sistematskog praćenja implementacije politika

15 medijske pismenosti ili evaluacije neformalnih obrazovnih programa – nisu uključene zbog nedostatka redovno ažuriranih izvora. Takođe, uključene su dve nove dimenzije koje nisu bile deo originalnog EAVI okvira ali odgovaraju na transformacije savremenog medijskog okruženja: sajber-bezbednost i pripremljenost za veštačku inteligenciju. Za procenu vrednosti ovih kategorija korišćeni su javno dostupni podaci iz pouzdanih izvora kao indikatori. Eksterni podaci prikupljeni su iz nacionalnih izvora (Republički zavod za statistiku - RZS, Gemius, SHARE Fondacija) i međunarodnih baza podataka (OECD PISA, Eurostat, European Social Survey - ESS, UN E-Government Survey, Reporters Without Borders Press Freedom Index, Centre for Media Pluralism and Media Freedom, Freedom House, Global AI Index). Ukupno trideset jedan indikator, organizovan u šest kategorija, upotrebljen je za evaluaciju ključnih aspekata okruženja relevantnih za medijsku pismenost. Podaci se odnose na period 2022–2025, pri čemu su korišćeni najnoviji dostupni podaci za svaki indikator. Svi indikatori normalizovani su na jedinstvenu skalu 0–100 kako bi se omogućila njihova komparacija i agregacija u kompozitne skorove po kategorijama. U narednim odeljcima detaljno obrazlažemo svaku od šest kategorija, objašnjavamo koje indikatore uključuju i kako se međusobno povezuju, i analiziramo njihov odnos sa anketnim nalazima o individualnim kompetencijama. II.1. INFRASTRUKTURA I PRISTUP Tehnološka dostupnost predstavlja fundamentalni preduslov za razvoj medijske pismenosti – građani ne mogu koristiti digitalne medije ako nemaju pristup internetu, uređajima i platformama. Ova kategorija obuhvata pokazatelje penetracije interneta u domaćinstvima (RZS, 2023), učestalost korišćenja društvenih mreža i digitalnih servisa (RZS, 2023), pokrivenost populacije domaćim medijskim portalima (Gemius, 2024), i kompozitni indeks digitalne povezanosti koji meri kvalitet mrežne infrastrukture i brzinu pristupa (Digital Connectivity Index, Globalni indeks VI 2024). Podaci pokazuju da Srbija ima razvijenu digitalnu infrastrukturu. Penetracija interneta u domaćinstvima dostigla je osamdeset pet procenata, što znači da velika većina građana ima tehnički pristup digitalnim medijima. Svakodnevno korišćenje društvenih mreža registrovano je kod osamdeset dva procenta odrasle populacije, dok domaće medijske portale mesečno

16 posećuje šezdeset devet procenata korisnika interneta. Indeks digitalne povezanosti (64.8/100) potvrđuje da mrežna infrastruktura omogućava ne samo pristup već i kontinuiranu, relativno brzu razmenu podataka neophodnu za savremeno medijsko iskustvo. Prosečna vrednost kategorije infrastrukture i pristupa iznosi 78.98/100, što je najviši skor među svim dimenzijama okruženja i pozicionira Srbiju visoko na skali dostupnosti i kvaliteta infrastrukture. Međutim, komparacija sa anketnim nalazima otkriva značajnu neusklađenost. Dok eksterni podaci pokazuju visoku tehnološku dostupnost, anketna dimenzija pristupa i korišćenja sugeriše da se tehnološka mogućnost ne pretvara automatski u efektivnu kompetenciju. Na primer, građani imaju internet i uređaje ali ne koriste raznovrsne izvore – oslanjaju se na ograničen broj platformi, dominantno na društvene mreže kao primarni izvor informacija, dok tradicionalne medije i specijalizovane portale koriste sporadično. Ovaj nalaz ukazuje da problem nije u infrastrukturi već u motivaciji, medijskim navikama, ili nedostatku svesti o prednostima raznovrsnih izvora. Sistemske intervencije, prema tome, ne bi trebalo da se fokusiraju na dalji razvoj infrastrukture već na programe medijske pismenosti koji će građanima pomoći da bolje iskoriste već dostupne resurse. II.2. OBRAZOVNI KAPACITETI Obrazovni sistem predstavlja ključnu instituciju za razvoj kritičkog mišljenja i medijske pismenosti kod mladih generacija. Sposobnost analize tekstova, razumevanja argumenata, evaluacije izvora i korišćenja informacija za rešavanje problema – kompetencije koje se razvijaju kroz formalno školovanje se reflektuju na kapacitete za kritičko čitanje medija. Ova kategorija integriše objektivne pokazatelje obrazovnih postignuća i ekspertsku procenu sistemske podrške medijskoj pismenosti. Osnovni indikator su rezultati PISA testiranja (OECD Programme for International Student Assessment, 2022) koji meri funkcionalne kompetencije petnaestogodišnjaka u čitalačkoj, matematičkoj i naučnoj pismenosti. Čitalačka pismenost posebno je relevantna jer obuhvata ne samo razumevanje tekstova već i sposobnost njihove evaluacije, prepoznavanja autorske perspektive i korišćenja informacija za argumentovane zaključke. Prevođenjem rezultata PISA 2022 testiranja u Srbiji na 0-100 skalu dobili smo prosečnu vrednost 49.33/100. Ovaj rezultat sugeriše da obrazovni sistem ne razvija dovoljno kompetencije za kritičko čitanje i rezonovanje kod mladih.

17 Pored objektivnog merenja postignuća učenika, kategorija uključuje ekspertsku procenu sistemske podrške medijskoj pismenosti u obrazovanju. Tokom Delphi procesa, ekspertski panel (koji je uključivao nastavnike, pedagoge, istraživače obrazovanja i medijske pismenosti) procenjivao je u kojoj meri obrazovni sistem Srbije podržava razvoj medijske pismenosti. Prosečna ocena bila je 3.21/10, što nakon normalizacije daje 32.1/100. Ova niska ocena pokazuje da postoji jaz između objektivnih merenja kompetencija učenika (PISA) i ekspertske procene sistemske podrške obrazovnog sistema za medijsku pismenost i on ukazuje na strukturalni problem: obrazovni sistem kao celina ne pruža dovoljno usmerenu podršku za razvoj kritičke medijske pismenosti kroz nastavne programe, obuke nastavnika i pedagoške pristupe. Ocena eksperata proističe iz uvida u kurikulume, nastavne prakse i institucionalne kapacitete, pa dopunjuju procenu obrazovnog sistema uvidima koje PISA testiranje ne uključuje. Veza sa anketom je indirektna ali važna: dimenzija u okviru ankete meri kritičke veštine kod odraslih, dok PISA pokazuje kakve obrazovne temelje stiču mlađe generacije. Anketna dimenzija koja se odnosi na kritičko čitanje iznosi 42.10/100 i predstavlja najstabilniji i najviši individualni domen među odraslima, što je neznatno niže od PISA proseka mladih (49.33/100), ali znatno iznad ekspertske procene sistemske podrške (32.1/100). Ovo sugeriše da građani razvijaju određene kritičke veštine uprkos nedostatku sistemske podrške u obrazovanju, verovatno kroz neformalne kanale učenja i životno iskustvo. Međutim, ni jedni ni drugi ne dostižu viši nivo kompetencija, što potvrđuje potrebu za sistemskim intervencijama u obrazovnom sistemu usmerenim na razvoj kritičkog mišljenja od najranijih nivoa školovanja. II.3. MEDIJSKE SLOBODE I PLURALIZAM Kvalitet medijskog sistema u kojem se građani informišu fundamentalno određuje mogućnosti za razvoj medijske pismenosti. Čak i ako građani poseduju razvijene kritičke kompetencije, njihova primena je ograničena ako medijsko okruženje nudi nizak pluralizam perspektiva, ako novinari rade pod pritiscima koji kompromituju nezavisnost izveštavanja, ili ako vlasničke strukture koncentrišu kontrolu nad informacijama. Medijska pismenost podrazumeva sposobnost pristupa raznovrsnim izvorima, evaluacije njihove pouzdanosti i formiranja informisanih mišljenja – ali sve to zavisi od toga da li takvi izvori uopšte postoje i da li mogu raditi slobodno.

18 Ova kategorija integriše dva međunarodno priznata izvora podataka. Press Freedom Index organizacije Reporters Without Borders (RSF, 2025) nakon invertovanja (46.5/100) pokazuje da je Srbija među zemljama sa umereno visokim rizicima po slobodu medija. Drugi izvor su četiri dimenzije Media Pluralism Monitor (Centre for Media Pluralism and Media Freedom, 2024) koji sistematski evaluira rizike po medijski pluralizam: fundamentalnu zaštitu (ustavna i zakonska osnova za slobodu medija), tržišnu raznovrsnost (koncentracija vlasništva, transparentnost), političku nezavisnost (uticaj vlasti na medije, finansiranje javnog servisa), i socijalnu inkluzivnost. Prosečan skor za Srbiju iznosi 47.75/100, pri čemu su najproblematičnije dimenzije politička nezavisnost i tržišna raznovrsnost. Prosečna vrednost kategorije medijskih sloboda i pluralizma iznosi 47.17/100, što signalizira ozbiljna sistemska ograničenja. Nizak pluralizam znači da građani, čak i kada žele da konsultuju različite perspektive, nailaze na medijsko tržište koje je strukturalno homogenizovano – određena gledišta dominiraju javnom sferom dok su druga marginalizovana ne zbog nedostatka publike već zbog ekonomskih i političkih barijera ulaska. Nizak nivo slobode medija znači da čak i novinari koji teže nezavisnom izveštavanju rade pod pritiscima koji kompromituju njihovu sposobnost da istražuju i objavljuju kritičke sadržaje. Veza sa anketnim nalazima posebno je indikativna. Tokom konstrukcije IMP, testirana je skala poverenja u različite izvore informacija. Ova skala isključena je iz finalnog indeksa zbog niske pouzdanosti, što znači da ispitanici nisu pokazali konzistentan obrazac vrednovanja izvora – neki su istovremeno veoma poverljivi prema svim izvorima, drugi veoma nepoverljivi, bez jasne diferencijacije između tipa izvora. Ovakva nedoslednost može biti direktna posledica nejasnog kvaliteta medijskog okruženja. Kada građani ne mogu da se oslone na stabilne kriterijume kvaliteta jer se medijski standardi toliko razlikuju ili menjaju, njihove procene postaju nesistematične i više odražavaju opšti nivo poverenja ili nepoverenja nego evaluaciju konkretnih izvora. Strukturalni nedostaci medijskog sistema, dakle, ne samo da ograničavaju pristup kvalitetnim informacijama već i otežavaju razvoj stabilnih kriterijuma za procenu pouzdanosti. II.4. DEMOKRATSKI KONTEKST Medijska pismenost podrazumeva aktivno korišćenje medija za građansko delovanje – izražavanje mišljenja, mobilizaciju oko javnih pitanja, povezivanje i organizovanje. Međutim,

19 takvo korišćenje zavisi ne samo od individualnih kompetencija već i od institucionalnog okvira koji može da podstiče ili da obeshrabruje građansku participaciju. U okruženjima sa ograničenim demokratskim slobodama, građani mogu imati kompetencije i dostupne platforme ali se mogu suzdržavati od javnog angažovanja zbog straha od posledica, ili spoznaje da njihovo učešće neće imati stvarni uticaj na odlučivanje. U ovoj kategoriji kombinujemo indikatore demokratskih sloboda i mogućnosti za digitalnu participaciju. Prema izveštaju Freedom House (Freedom in the World, 2024) Srbija je klasifikovana kao "delimično slobodna" sa skorom 56.0/100, Sloboda interneta (Internet Freedom) je nešto viša i iznosi 70/100, što ukazuje da je digitalni prostor relativno slobodniji od tradicionalnih medija. Prosečna vrednost kategorije demokratskog konteksta iznosi 64.5/100. UN E-Participation Index (UN E-Government Survey, 2024) meri stepen u kojem vlade koriste digitalne platforme za građansko učešće. Ovaj indikator posebno je relevantan za digitalnu dimenziju medijske pismenosti jer pokazuje da li građani koji žele da se angažuju imaju dostupne resurse. Srbija na ovom indeksu beleži 73.0/100, što je visok skor, posebno u odnosu na opšti nivo demokratskih sloboda. Ova neusklađenost ukazuje da platforme za e- participaciju formalno postoje, institucije kreiraju onlajn mehanizme, ali ne odgovaraju suštinski na građanske zahteve, što vremenom može da demotiviše participaciju. Prosečna vrednost kategorije demokratskog konteksta iznosi 64.5/100, a anketna dimenzija participacije je samo 27.56/100. Ova razlika između postojanja platformi i njihovog korišćenja sugeriše da tehnološka mogućnost ne garantuje upotrebu kada građani pretpostavljaju da njihovo učešće neće proizvesti rezultate. Ovaj problem zahteva kulturnu promenu u odnosu institucija prema građanskoj participaciji ne samo unapređivanje platformi. II.5. DOMEN POVERENJA, PRIVATNOSTI I SAJBER-BEZBEDNOSTI Društveno poverenje, kultura zaštite podataka i bezbednost digitalnog okruženja zajedno čine osnovu za smislenu medijsku participaciju. Bez minimalnog nivoa poverenja u druge ljude i institucije, građani se povlače iz javne sfere i ograničavaju svoje aktivnosti na privatne krugove. Bez svesti o privatnosti i poznavanja zaštitnih mera, građani su izloženi

20 zloupotrebama podataka koje ih mogu odvratiti od javnog angažovanja. U proceni ove kategorije oslanjamo se na tri tipa podataka. Prvo, koristimo Eurostat podatak o poverenju u ljude, iz European Social Survey (ESS, 2023) poverenje u ljude, fer ponašanje i spremnost na pomoć. Prema ovim podacima nivo međuljudskog poverenja u Srbiji iznosi 44.2/100, što je ispod evropskog proseka. Drugo, uključujemo podatke o praksi zaštite privatnosti iz nacionalne ICT ankete (RZS, 2023). Prema podacima RZS, u proseku samo 21.9% građana primenjuje mere zaštite privatnosti (poput isključivanja praćenja lokacije, ograničavanja vidljivosti profila na društvenim mrežama, blokiranja ličnih podataka za oglašavanje), što je veoma nizak procenat, posebno jer se radi o populaciji koja intenzivno koristi digitalne platforme. Treće, kategorija obuhvata podatke o sajber-bezbednosti medijskog sektora (SHARE Fondacija, Cybersecurity Monitoring, 2024) koji prate incidente poput hakerskih napada na medijske sajtove, direktne pretnje i napade na novinare putem digitalnih kanala, zloupotreba, diskriminaciju i govor mržnje. Prema ovim podacima indikator sajber-bezbednosti medija deluje visoko (71.94/100), ali treba imati u vidu da ovaj indikator ukazuje samo na prijavljene incidente (stvaran broj je verovatno viši) i ne meri subjektivnu percepciju bezbednosti. Prosečna vrednost kategorije poverenja, privatnosti i sajber-bezbednosti iznosi 38.52/100 (kombinacija umerenog poverenja 44.2/100, veoma niske rasprostranjenosti zaštitnih praksi 21.9/100, i srednje sajber-bezbednosti 71.9/100). Relativno visok pokazatelj sajber-bezbednosti podiže ukupan prosek ali ne može kompenzovati niske nivoe društvenog poverenja i individualne zaštite privatnosti. Veza sa anketom otkriva višestruki jaz između znanja, prakse i sistemske bezbednosti. Anketna dimenzija participacije (P43-P50) meri da li ispitanici znaju kako da primene osam različitih zaštitnih mera (prijava govora mržnje, isključivanje praćenja, zaštita od prevara itd.), dok eksterni podaci pokazuju da samo oko 22% građana zaista primenjuje te mere u praksi. Istovremeno, pokazatelj sajber-bezbednosti (71.94/100) sugeriše da je sistemska bezbednost bolja od individualne spremnosti. To nam govori da građani nisu dovoljno osposobljeni za zaštitu, čak i u okruženju gde objektivni rizici nisu ekstremno visoki. Iako kod određenog dela populacije znanje postoji, ono se ne manifestuje nužno u svakodnevnoj praksi. Ovi nalazi nam

21 ukazuju na potrebu za intervencijama koje ne podučavaju samo veštinama, već i menjaju percepciju rizika, grade kulturu digitalne bezbednosti. Istovremeno, nizak nivo društvenog poverenja (44/100) u vezi je sa niskom participacijom. Građani koji ne veruju drugim ljudima manje su skloni javnom angažovanju jer percipiraju da će biti izloženi rizicima bez recipročne podrške ili zaštite. Prema tome, potrebne su intervencije koje pospešuju protektivna ponašanja, ali i poboljšanje osnova za proaktivno delovanje - uspostavljanje odnosa poverenja između ljudi, zatim građana i institucija. II.6. PRIPREMLJENOST ZA VEŠTAČKU INTELIGENCIJU Veštačka inteligencija sve više postaje integralni deo medijskog ekosistema – od automatizovanog generisanja vesti do personalizovanih preporuka koje oblikuju informaciono iskustvo korisnika. Medijska pismenost u savremenom okruženju zahteva razumevanje tradicionalnih medijskih procesa i sposobnost navigacije sadržajima posredovanim ili kreiranim veštačkom inteligencijom. To uključuje prepoznavanje sadržaja generisanog VI, kao i kritičku evaluaciju autentičnosti i pouzdanosti takvog sadržaja. Razvoj ovih kompetencija zavisi i od institucionalnih kapaciteta društva da reguliše i transparentno koristi VI tehnologije ali i poverenja javnosti u VI. U ovoj kategoriji integrišemo Indeks spremnosti za VI 2024 Međunarodnog Monetarnog Fonda (IMF AI Preparedness Index) koji za Srbiju iznosi 54.0/100 i ukazuje na umerenu institucionalnu spremnost. Ovaj indeks obuhvata regulatorne okvire, tehničku infrastrukturu, ljudske resurse i kapacitete za inovacije u oblasti VI. Iz Globalnog indeksa VI koristimo podatke o Digitalnoj povezanosti (Global AI Index – Digital Connectivity) koji potvrđuju da je tehnološka osnova za primenu VI dobra 64.8/100. Međutim podatak o poverenju javnosti iz istog izvora podataka pokazuje jaz između infrastrukturne osnove i nivoa poverenja u tehnologije VI - 36.9/100, upućujući na nedovoljno razumevanje načina funkcionisanja VI, nedostatka transparentnosti u primeni VI, ili straha od potencijalne zloupotrebe. Prosečna vrednost kategorije pripremljenosti za VI iznosi 51.9/100. Nisko poverenje se jasno nadovezuje na podatke iz ankete koji pokazuju da samo 23,9% (191 od 800) ispitanika koristi VI platforme. Nisko poverenje u VI i niska stopa upotrebe se međusobno pojačavaju jer građani zbog nepoverenja ne koriste VI, a poverenje ne raste zbog

22 manjka iskustva u upotrebi. Celovito sagledavanje ove kategorije ukazuje na potrebu za proaktivnim pristupom koji integriše nekoliko dimenzija. Prvo, transparentne politike o tome kako javne institucije i medijske organizacije koriste VI – koji algoritmi biraju vesti, kako se treniraju modeli, ko je odgovoran za greške. Drugo, obrazovanje o VI tehnologijama ne samo na tehničkom nivou već i na nivou razumevanja društvenih implikacija – kako VI menja medije, koje su etičke dileme, kako zaštititi prava korisnika. Treće, građansko osnaživanje kroz programe koji podučavaju kritičku evaluaciju AI- generisanih sadržaja. Bez ovakvih intervencija, postoji rizik da se stvori novi digitalni jaz – između manjine koja koristi VI alate i razume njihovu logiku, i većine koja ostaje pasivna i nepoverljiva, što dalje produbljuje nejednakosti u medijskoj pismenosti. III VALIDACIJA I FINALNA KONSTRUKCIJA IIMP Nakon što su anketno istraživanje i analiza eksternih indikatora proizveli dva komplementarna profila – individualne kompetencije (IMP 35.9/100) i strukturalne uslove (šest kategorija okruženja) – preostalo je metodološko pitanje kako ih integrisati u jedinstvenu metriku. Za tu svrhu primenjen je Delphi metod konsenzusa sa ekspertskom grupom kako bi se empirijski utvrdile relativne važnosti pojedinih dimenzija. Delphi proces i postizanje konsenzusa U Delphi procesu učestvovalo je trideset eksperata iz oblasti obrazovanja, medija, medijskih i obrazovnih politika, i civilnog društva. Panel je uključivao nastavnike, novinare sa iskustvom u fact-checkingu i verifikaciji, istraživače koji se bave digitalnim kompetencijama i dezinformacijama, predstavnike državnih institucija, i aktiviste organizacija civilnog društva koje sprovode programe medijske pismenosti. Geografski i institucionalno raznovrsna struktura panela osigurala je da konsenzus ne odražava perspektivu samo jedne grupe aktera već integriše uvide iz različitih pozicija u sferi medijske pismenosti.

23 Delphi metod sproveden je u dve runde tokom oktobra 2025. godine. U prvoj rundi, eksperti su individualno dodeljivali težine na sedam ponuđenih dimenzija medijske pismenosti: Povezanost i pristup, Kritička evaluacija, Demokratska participacija, Medijske slobode, Poverenje i privatnost, Sajber-bezbednost, i Spremnost za veštačku inteligenciju. Svaka dimenzija bila je praćena kratkim objašnjenjem šta meri i kako proističe iz anketnih i eksternih podataka. U drugoj rundi, eksperti su dobili agregirane rezultate prve runde (medijana i interkvartilni raspon za svaku dimenziju), uvid u kvalitativna obrazloženja drugih učesnika, i mogućnost da revidiraju svoje težine u svetlu kolektivnih procena. Postignuta je saglasnost za sve dimenzije prema dvostrukim, unapred definisanim kriterijumima statističke stabilnosti i konsenzusa. U konstrukciji finalnog indeksa izvršena su dva metodološka prilagođavanja. Prvo, dimenzije Poverenje i privatnost i Sajber-bezbednost spojene su u jednu dimenziju Poverenje, privatnost i sajber-bezbednost sa kombinovanom težinom. Ova odluka zasnovana je na konceptualnoj bliskosti ovih dimenzija, činjenici da počivaju na eksternim indikatorima koji mere društveno poverenje, zaštitne mere korisnika i sistemsku sajber- bezbednost. Drugo, nakon spajanja izvršena je normalizacija svih težina proporcionalnim skaliranjem. Finalna struktura IIMP zasniva se na šest dimenzija: (1) Povezanost i pristup – 15.71%, (2) Kritička evaluacija – 20.94%, (3) Demokratska participacija – 10.47%, (4) Medijske slobode – 20.94%, (5) Poverenje, privatnost i sajber-bezbednost – 20.94%, (6) Spremnost za veštačku inteligenciju – 11.00%. Ovaj raspon pokazuje da su ekspertima centralne i podjednako važne za medijsku pismenost - kritička evaluacija i medijske slobode. Primena Delphi-validiranih težina na skorove dimenzija proizvodi finalni Integrisani indeks medijske pismenosti za Srbiju 2025: IIMP = 43.11/100. Tabela br. 2 prikazuje doprinos svake dimenzije ukupnom indeksu.

24 U okviru Integrisanog indeksa medijske pismenosti - IIMP ukupan skor za Srbiju za 2025. godinu iznosi 43.11 od maksimalnih 100. Tabela br. 2: Konačan obračun IIMP za Srbiju: Dimenzija Skor dimenzije Težina u IIMP Doprinos IIMP POVEZANOST I PRISTUP 52.23 15.71% 8.2 KRITIČKA EVALUACIJA 45.72 20.94% 9.57 DEMOKRATSKA PARTICIPACIJA 46.03 10.47% 4.82 MEDIJSKE SLOBODE 47.17 20.94% 9.88 POVERENJE, PRIVATNOST I SAJBER- BEZBEDNOST 38.52 20.94% 8.07 PRIPREMLJENOST ZA VEŠTAČKU INTELIGENCIJU 23.42 (60/40) 11.00% 2.58 IIMP Srbija 100% 43.11 Glavna saznanja Broj poena koje je Srbija ostvarila u 2025. godini 43.11 može se interpretirati kao srednje-nizak nivo integrisane medijske pismenosti. Takođe, posmatrano prema integrisanom indeksu IIMP (43.11), Srbija bolje stoji u medijskoj pismenosti nego prema indeksu baziranom samo na anketnim podacima IMP (35.9/100). Ovakav rezultat zasnovan je na visokim procenama Srbije prema eksternim pokazateljima infrastrukture (78.98) i demokratskog konteksta (64.50). Međutim, neusklađenost između strukturalnih mogućnosti i individualne upotrebe ostaje vidljiva i u integrisanom indeksu.

25 Tri dimenzije približno podjednako doprinose IIMP. Kritička evaluacija (9.57), Medijske slobode (9.88), Poverenje, privatnost i sajber bezbednost (8.07), zajedno nose skoro dve trećine indeksa. Ove tri dimenzije i prema teorijskim pretpostavkama predstavljaju osnov medijske pismenosti – sposobnost kritičkog rasuđivanja, okruženje koje pruža pluralizam perspektiva, i sigurnost koja omogućava javno angažovanje. Povezanost i pristup (8.20) i Demokratska participacija (4.82) doprinose umereno, dok Spremnost za VI (2.58) najslabije doprinosi, zbog niske upotrebe koju pokazuje anketa kao i niskog poverenja na koje ukazuju eksterni podaci. Vrednost dimenzije Povezanost i pristup (52.23) maskira jaz između infrastrukture (78.98) i upotrebe (25.48). Iako spoljašnji uslovi omogućavaju raznovrsno korišćenje medija, građani to ne čine. Ovaj nalaz ima direktnu implikaciju za javne politike, ukazujući donosiocima odluka da dalje investiranje u infrastrukturu najverovatnije neće značajno poboljšati medijsku pismenost. Veće pomake mogu doneti intervencije koje će građane osposobiti i motivisati da koriste već dostupne resurse. Slično tome, unutar dimenzije Demokratska participacija (46.03) visoka e-participacija merena prema spoljnim indikatorima je u nesaglasju sa niskim građanskim angažmanom prema podacima iz anketne, ponovo ukazujući na neiskorišćenost postojećih resursa. Najniža dimenzija, Spremnost za VI (23.42), signalizira da je ovo oblast gde su i strukturalni uslovi i individualne prakse nerazvijeni. Nisko poverenje i upotreba međusobno se pojačavaju kreirajući začarani krug. Bez proaktivnih intervencija – transparentnih politika, obrazovanja o VI, etičkih standarda – postoji rizik da se ovaj jaz produbljuje i kreira novi digitalni jaz između manjine koja koristi VI alate i većine koja ostaje pasivna. U svakom slučaju, Indeks medijske pismenosti (IMP) i Integrisani indeks medijske pismenosti (IIMP) postavljaju preko potreban temelj za buduće analize i praćenje trendova. Prvo, godišnje praćenje pokazaće da li intervencije (programi medijske pismenosti, reforme obrazovanja, promene u medijskom sistemu) proizvode stvarne promene. Drugo, kada se metodologija primeni u drugim zemljama, omogućiće rangiranje i identifikaciju najboljih praksi. Treće, posmatranjem IIMP kroz dimenzije pokazuje gde su prioriteti za sistemske intervencije, za sadašnje i buduće trenutke i kontekste u kojima se indeks primenjuje.

26 Zaključna razmatranja i preporuke Preporuke za javne politike: Četiri prioritetna pravca intervencija Na osnovu rezultata IIMP, možemo izdvojiti četiri pravca za sistemske intervencije koje mogu proizvesti merljivo poboljšanje medijske pismenosti u Srbiji. Ovi pravci nisu izolovani već međusobno povezani – poboljšanje u jednom domenu podržava napredak u drugima i obratno. I. Centralni prioritet jeste razvoj sveobuhvatnih programa medijske pismenosti koji će građanima pomoći da bolje iskoriste već dostupnu tehnološku infrastrukturu. Nalazi jasno pokazuju da problem nije u nedostatku tehnologije već u tome kako ih koriste. Programi medijske pismenosti treba da budu fokusirani na strategije raznovrsnog informisanja i verifikaciona ponašanja koja zahtevaju sistemsku podršku. Takvi programi bi trebalo da podučavaju građane kako da identifikuju i aktivno koriste raznovrsne izvore informacija, kao i da podignu svest o važnosti raznovrsnosti izvora – da različite platforme nude različite perspektive, da monopol nad informacijama može voditi pristrasnosti, da konsultovanje više nezavisnih izvora pomaže u formiranju informisanog mišljenja. Pored toga, za građane je značajno da nauče konkretne korake koje mogu primeniti kada naiđu na sumnjive informacije. Na primer, da provere informacije kroz više nezavisnih izvora pre nego što je podele, konsultuju fact-checking organizacije kada naiđu na senzacionalističke tvrdnje, potraže originalni kontekst fotografije ili izjave koja kruži društvenim mrežama, prepoznaju clickbait naslov koji obećava više nego što tekst zapravo pruža. Posebnu pažnju bi valjalo usmeriti na starije demografske grupe i niže obrazovane segmente populacije koje često nemaju pristup neformalnim kanalima učenja za razliku od mlađe i obrazovanije populacije. Programe za starije i niže obrazovane građane treba koncipirati u skladu sa prethodnim znanjem, stilovima učenja, i različitim praktičnim potrebama. Pored prilagođavanja sadržaja može se prilagoditi i format – sporiji tempo, ponavljanje ključnih koncepata, upotreba primera iz svakodnevnog života, praktične demonstracije umesto apstraktnih objašnjenja. Forma programa može biti participativna. Na primer, građani mogu da vežbaju na realnim primerima, analiziraju konkretne vesti, koriste

27 alate za proveru fotografija i videa, diskutuju u grupama kako razlikovati vest od komentara, vežbaju proveru izvora i konteksta. Ovakvi programi mogu biti realizovani kroz radionice, oslanjajući se na postojeću infrastrukturu, poput biblioteka ili udruženja penzionera. Ključno je da programi budu besplatni, geografski rasprostranjeni kako bi pokrivali ne samo urbane već i ruralne oblasti, i kontinuirani kroz vreme, a ne jednokratne akcije koje brzo padaju u zaborav. II. Ekspertska procena sistemske podrške obrazovanja medijskoj pismenosti pokazuje nizak nivo, dok PISA rezultati potvrđuju da obrazovni sistem ne razvija dovoljno kompetencije kritičkog čitanja kod mladih. Ovi nalazi ukazuju da je potrebna sveobuhvatna reforma obrazovnog sistema koja medijsku pismenost pozicionira kao transverzalnu kompetenciju kroz sve nivoe formalnog obrazovanja, od osnovne škole preko srednjeg obrazovanja do visokog obrazovanja, gde bi se kompetencije mogle koristiti, što je naročito važno kod fakulteta koji obrazuju nastavike. Takva reforma bi trebalo da počne revizijom nastavnih planova i programa kako bi medijska pismenost bila i poseban predmet koji predaju svršeni studenti medijskih studija, ali i sistematski integrisana u sve druge relevantne predmete. Integrisanje medijske pismenosti kao međupredmetnu kompetenciju znači ugrađivanje veština kritičkog čitanja medija u postojeće predmete gde se to prirodno uklapa. Na primer, na časovima istorije đaci mogu kritički analizirati kako mediji reprezentuju istorijske događaje i društvene grupe. Ovakva integracija omogućava da učenici razvijaju medijske kompetencije kroz različite kontekste i predmete, čime se znanje učvršćuje i postaje primenljivo. Naravno, reforma kurikuluma može postići pun efekat ako medijsku pismenost najpre predaju oni koji su za to obrazovani u okviru svojih sudija medija i komunikacija, a onda i kada se nastavnici drugih predmeta dodatno osnaže da podučavaju učenike kako da kritički evaluiraju sadržaje kojima su izloženi. To zahteva kontinuirane programe profesionalnog razvoja koji nastavnicima pružaju ne samo teorijsko razumevanje medijske pismenosti već i konkretne pedagoške strategije, aktivnosti i materijale koje mogu primeniti u učionici. Jednokratni seminari pokazuju ograničene efekte jer nastavnici brzo zaboravljaju sadržaj koji ne koriste redovno. Značajniji bi bili dugoročni programi sa mentorskim vođenjem, razmenom iskustava između kolega, i kontinuiranom podrškom. Na kraju, efekti integrisanja medijske pismenosti u kurikulume trebalo bi sistemski evaluirati, pratiti kroz studije da li učenici koji su prošli kroz reformisane programe pokazuju bolje kompetencije od prethodnih generacija. Reforma obrazovnog sistema ne daje trenutne

28 rezultate, ali stvara sistemsku promenu jer mlađe generacije koje razviju kritičke veštine vremenom podižu ukupan nivo medijske pismenosti u populaciji. III. Strukturalna ograničenja medijskog sistema direktno utiču na mogućnosti građana da razvijaju i primenjuju medijsku pismenost. Čak i kada građani poseduju kompetencije kritičke evaluacije, nizak pluralizam perspektiva i koncentrisana vlasnička struktura ograničavaju pristup raznovrsnim izvorima koje mogu kritički porediti. Reforma medijskog sistema je nužna paralelna intervencija bez koje individualne kompetencije ostaju ograničeno primenljive. Jačanje nezavisnosti javnog medijskog servisa treba da predstavlja strateški prioritet jer javni mediji, kada funkcionišu prema svom mandatu, treba da budu model kvalitetnog novinarstva i raznovrsnosti perspektiva dostupan svim građanima bez obzira na ekonomski status. To zahteva fundamentalnu reformu mehanizama imenovanja upravljačkih tela kako bi se osiguralo da javni servis služi građanima a ne trenutnim vladinim interesima, kao i transparentne i predvidive forme finansiranja koje omogućavaju planiranje i nezavisnost od političkih pritisaka. Javni servis mora imati resurse i institucionalnu autonomiju da proizvodi kvalitetne informativne programe, istraživačko novinarstvo i edukativan sadržaj koji podiže nivo javne debate. Paralelno sa reformom javnog servisa, potrebna je sistemska podrška nezavisnim i lokalnim medijima koji nude alternativne perspektive i pokrivaju teme koje mediji koji imaju usko-komercijalne interese i prodržavni mediji zanemaruju. Ovi mediji često opstaju na ivici ekonomske održivosti, što ih čini ranjivim na pritiske i korupciju. Transparentnost javnog finansiranja medija – zasnovane na jasnim kriterijumima i profesionalnim standardima – mogu osigurati njihovu ekonomsku održivost bez ugrožavanja uredničke nezavisnosti. Istovremeno, podrška lokalnim medijima koji pokrivaju regionalne i zajedničke teme omogućava građanima pristup informacijama relevantnim za njihov neposredni život koje nacionalni mediji često ignorišu. Transparentnost vlasništva medija mora postati obavezujući standard jer građani imaju pravo da znaju ko stoji iza medijskih kuća koje prate i kako to potencijalno utiče na sadržaj koji konzumiraju. Regulatorni okvir treba da obavezuje javnu dostupnost podataka o vlasnicima medija, njihovim drugim poslovnim interesima, izvorima finansiranja, i

29 eventualnim konfliktima interesa. Ove informacije moraju biti lako dostupne i razumljive, ne sakrivene u nejasnim pravnim dokumentima. Kada građani znaju da određeni medij poseduje firma koja ima ugovore sa vladom, mogu svesnije proceniti da li izveštavanje tog medija o vladinom radu pokazuje bilo kakvu pristrasnost. Konačno, zaštita novinara od pritisaka, pretnji, fizičkih napada i tužbi koje imaju za cilj da ućutkaju kritičko izveštavanje predstavlja neophodan uslov za slobodno novinarstvo. Bez novinara koji mogu bezbedno istraživati i objavljivati istraživačke priče, građani ne mogu imati pristup kvalitetnim informacijama bez obzira koliko razvijene kompetencije poseduju. To zahteva ne samo pravni okvir koji adekvatno sankcioniše napade na novinare već i stvarnu primenu zakona, kao i jačanje mehanizama brzog pravnog i institucionalnog odgovora kada su novinari mete napada. Ove mere zahtevaju političku volju i mogu nailaziti na otpor jer ugrožavaju status quo, ali su neophodne za kreiranje medijskog okruženja u kojem medijska pismenost građana ima stvarni smisao. IV. Nizak nivo društvenog poverenja, veoma niska primena zaštitnih mera privatnosti, i nisko poverenje u veštačku inteligenciju zajedno kreiraju okruženje u kojem građani, čak i kada poseduju znanje i veštine, ne participiraju aktivno u javnoj sferi ili ne koriste nove tehnologije koje postaju deo medijskog ekosistema. Ovaj problem je povezan sa odnosom prema privatnosti, poverenjem u institucije i druge ljude, i razumevanje kako nove tehnologije menjaju medijsko okruženje. Intervencije u ovoj oblasti moraju biti višedimenzionalne i adresirati ne samo individualne prakse već i sistemsku bezbednost. U domenu privatnosti i digitalne bezbednosti, većina građana ne primenjuje osnovne zaštitne mere uprkos tome što neke od njih poznaju. To ukazuje da ne uviđaju rizike – koliko su njihovi podaci vredni i kako mogu biti zloupotrebljeni, pa zaštitu privatnosti doživljavaju kao nepotrebnu. Kampanje podizanja svesti bi trebalo usmeriti na ova pitanja, podučavajući tehničkim merama poput blokiranja praćenja ili kontrole dozvola aplikacija, i objašnjavajući konkretne načine na koje podaci mogu biti zloupotrebljeni. Građani treba da razumeju da zaštita privatnosti nije luksuz za one koji imaju nešto da sakriju već osnovno pravo i razumna predostrožnost u digitalnom okruženju. Sistemska bezbednost mora biti ugrađena u platforme kroz obavezujuće standarde zaštite podataka koji važe za sve digitalne servise koji rade sa građanima – od globalnih društvenih mreža do lokalne e-uprave. To znači jasna pravila o tome koji podaci smeju biti prikupljani i u koje svrhe, obaveznu enkripciju osetljivih podataka,

30 automatsko brisanje podataka koji više nisu potrebni, i transparentnost o tome sa kim se podaci dele. Građani moraju imati jednostavan način da kontrolišu svoje podatke, vide ko im pristupa, i povuku dozvole koje su prethodno dale. Regulatorni okvir mora uključivati efektivne sankcije za kršenje standarda, ne samo simbolične kazne koje velike platforme tretiraju kao trošak poslovanja. Kada je reč o veštačkoj inteligenciji, nizak nivo poverenja i niska stopa upotrebe međusobno se pojačavaju kreirajući začarani krug. Prekidanje ovog kruga zahteva transparentne politike upotrebe veštačke inteligencije u javnim institucijama i medijskim organizacijama. Građani treba da znaju kada komuniciraju sa automatizovanim sistemom a ne sa osobom, koji algoritmi biraju koje vesti vide na društvenim mrežama, kako se treniraju modeli koji generišu tekst ili slike i ko je odgovoran kada automatizovani sistem napravi grešku. Ova transparentnost gradi razumevanje i poverenje. Paralelno sa transparentnošću, potrebno je obrazovanje o veštačkoj inteligenciji koje nije fokusirano samo na tehničke detalje već na razumevanje društvenih implikacija. Građani treba da razumeju kako veštačka inteligencija menja produkciju vesti i sadržaja, koje su etičke dileme automatizovane moderacije sadržaja, kako zaštititi svoja prava kada algoritmi donose odluke koje utiču na njihov život, kako prepoznati sadržaj koji je generisan veštačkom inteligencijom. Istovremeno, medijske organizacije koje koriste veštačku inteligenciju za generisanje ili uređivanje sadržaja moraju biti transparentne o tome i odgovorne za ishode. Etički standardi primene veštačke inteligencije u medijima treba da osiguraju da ne vodi gubitku ljudske odgovornosti ili širenju dezinformacija. Literatura Aufderheide, P. (1993). Media literacy: A report of the National Leadership Conference on Media Literacy. Aspen Institute. Broadcasting Authority of Ireland - BAI. (2022). Media Literacy Policy. https://www.medialiteracyireland.ie/wp-content/uploads/2022/12/BAI-ML-Policy-1.pdf. Bulger, M. (2012). Measuring media literacy in a national context: challenges of definition, method and implementation. Media Studies, 3(6), 83-104. http://hrcak.srce.hr/96376. Celot, P. (2015). Assessing media literacy levels and the European Commission pilot initiative. European Association for Viewers Interests (EAVI). https://eavi.eu/wp- content/uploads/2017/08/assessing.pdf

31 Celot, P., & Pérez Tornero, J. M. (2009). Study on assessment criteria for media literacy levels: A comprehensive view of the concept of media literacy and an understanding of how media literacy levels in Europe should be assessed (Final report). European Commission, Information Society and Media Directorate-General, Media Literacy Unit. https://ec.europa.eu/assets/eac/culture/library/studies/literacy-criteria-report_en.pdf Directive 2007/65/EC of the European Parliament and of the Council of 11 December 2007 amending Council Directive 89/552/EEC on the coordination of certain provisions laid down by law, regulation or administrative action in Member States concerning the pursuit of television broadcasting activities. Official Journal of the European Union, L 332, 27–45. DTI & EAVI (2011). Testing and refining criteria to assess media literacy levels in Europe: Final report. European Commission, Directorate-General for Information Society and Media, Media Literacy Unit. https://eavi.eu/wp- content/uploads/2017/08/study_testing_and_refining_ml_levels_in_europe.pdf Centre for Media Pluralism and Media Freedom. (2024). Media Pluralism Monitor 2024: Serbia. European University Institute. https://cmpf.eui.eu/media-pluralism-monitor-2024/ EAVI (2011). Assessing Media Literacy Levels and the European Commission. https://eavi.eu/wp- content/uploads/2017/08/assessing.pdf. European Commission. 2011. Testing and Refining Criteria to Assess Media Literacy Levels in Europe Final Report. Commissioned by the European Commission Directorate-General for Information Society and Media; Media Literacy Unit. Danish Technological Institute & European Association for Viewers’ Interests. https://eavi.eu/wp- content/uploads/2017/08/study_testing_and_refining_ml_levels_in_europe.pdf European Commission (2009). Commission Recommendation of 20 August 2009 on media literacy in the digital environment for a more competitive audiovisual and content industry and an inclusive knowledge society. Official Journal of the European Union, C 257, 1-6. Dostupno na: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32009H0625. European Commission (2009). Study on Assessment Criteria for Media Literacy Levels. Brussels: European Commission - Expert Group on Media Literacy . https://ec.europa.eu/culture/library/studies/literacy-criteria-report_en.pdf. European Commission. (2007). A European approach to media literacy in the digital environment (Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM(2007) 833 final). Erstad, O., & Amdam, S. (2013). From Protection to Public Participation: A Review of Research Literature on Media Literacy. Javnost - The Public, 20(2), 83-98. Eurostat. (2024). Social trust (ilc_pw03). Eurostat. https://ec.europa.eu/eurostat/databrowser/view/ilc_pw03/default/table FarrPoint. (2024). International Digital Connectivity Readiness Index (DCRI), September 2024. FarrPoint. https://www.farrpoint.com/uploads/store/mediaupload/1591/file/FarrPoint_Intern ational_DCRI_Sept_2024.pdf. Freedom House. (2024). Freedom in the World 2024: Serbia. Freedom House. https://freedomhouse.org/country/serbia/freedom-world/2024 Gemius. (2024). Gemius Audience Serbia 2024: Online news portal reach. Gemius. https://e- public.gemius.com/rs/rankings/12730 Grizzle, Alton, Torrent, Jordi & Perez Tornero, Jose Mauel (2013a) MIL as a Tool to Reinforce Intercultural Dialogue: Principles and Aims of the UNESCO-UNAOC UNITWIN Cooperation Programme on Media and Information Literacy and Intercultural Dialogue. In: Carlsson, Ula & Hope Culver, Sherri (eds.) Media and Information Literacy and

32 Intercultural Dialogue, (pp. 9-16). MILID Yearbook, The Interantional Clearinghouse on Children, Youth and Media Nordicom. Sweden: University of Gothenburg. Grizzle, A., Moore, P., Dezuanni, M., Asthana, S., Wilson, C., Banda, F., Onumah, C. (2013). Media and information literacy: policy and strategy guidelines. UNESCO Hobbs, R. (2010). Digital and media literacy: A plan of action. Aspen Institute. Jolls, T. & Thoman, E. (2008). Literacy for the 21st Century. An Overview & Orientation Guide To Media Literacy Education. Part I: Theory CML MediaLit Kit™ A Framework for Learning and Teaching in a Media Age. Centre for Media Literacy. https://webspace.ship.edu/hliu/etextbook/theory/doc/media%20literacy_v02.pdf. Republički zavod za statistiku. (2023). Upotreba informaciono-komunikacionih tehnologija u Republici Srbiji, 2023: Domaćinstva/pojedinci, preduzeća (ISSN 1820-9084). Republički zavod za statistiku. https://publikacije.stat.gov.rs/G2023/Pdf/G202316018.pdf The National Association for Media Literacy Education (NAMLE). 2024. Media Literacy Framework. https://namle.org/resources/media-literacy-defined/ OEBS (2025) Medijska pismenost građana Srbije. Beograd. Ofcom. (2025). Adults' media use and attitudes report 2025. https://www.ofcom.org.uk/media-use- and-attitudes/media-habits-adults/adults-media-use-and-attitudes. Ofcom. (2023). Future Technology and Media Literacy: Anchor document. https://www.ofcom.org.uk/siteassets/resources/documents/research-and-data/media-literacy- research/making-sense-of-media/future-technology-trends-and-media-literacy/anchor- document.pdf?v=329832. Organisation for Economic Co-operation and Development. (2023). PISA 2022 results (Volume I): The state of learning worldwide – Reading. OECD Publishing. https://www.oecd.org/content/dam/oecd/en/publications/reports/2023/12/pisa- 2022-results-volume-i_76772a36/53f23881-en.pdf OSCE (2022) Media users over the age of 65 in Serbia. https://www.osce.org/files/f/documents/3/2/541704.pdf. Pereira, S., & Moura, P. (2022). Assessing media literacy competences: Reflections and recommendations from a quantitative study. Journal of Media Literacy Education, 14(3), 79- 93. https://doi.org/10.23860/JMLE-2022-14-3-7. Perez-Escoda, Ana, Garcia-Ruiz, Rosa, Aquaded, Ignacio. 2016. International dimensions of media literacy in connected world. ATI – Appliens Technologies and Innovations, 12 (2), 95-106. Potter, James W. (2022). Analysis of definitions of media literacy. Journal of Media Literacy Education. 14(2), 27-43. https://doi.org/10.23860/JMLE-2022-14-2-3. Reporters Without Borders. (2024). 2024 World Press Freedom Index. Reporters Without Borders. https://rsf.org/en/index Rowe, G. & Wright, G. (1999). “The Delphi Technique as a Forecasting Tool: Issues and Analysis,” International Journal of Forecasting, Vol. 15, No. 3, pp. 53-375. SHARE Foundation. (2024). Cyber-safety monitoring report 2024: Serbia. SHARE Foundation. https://monitoring.labs.rs/en. Tortoise Media. (2025). The Global AI Index. Tortoise Media. https://www.tortoisemedia.com/data/global-ai UNESCO (2022). Global Standards for Media and Information Literacy Curriculum Development. Paris: UNESCO. https://unesdoc.unesco.org/ark:/48223/pf0000383043 United Nations Department of Economic and Social Affairs. (2024). United Nations E-Government Survey 2024: E-Participation index – Serbia. United

33 Nations. https://publicadministration.un.org/egovkb/en-us/Data/Country-Information/id/151- Serbia Vlada Republike Srbije. (2020). Strategija razvoja sistema javnog informisanja u Republici Srbiji za period 2020-2025. godina. Službeni glasnik RS, br. 11/2020. https://www.media.srbija.gov.rs/medsrp/dokumenti/medijska_strategija210_cyr.pdf. Wadhwa, A., & Nandal, S. (2025). Development of an Index in Social Science: A Systematic Literature Review. The Scientific Temper, 16(03), 3927–3942. DOI - 10.58414/SCIENTIFICTEMPER.2025.16.3.09