War, what is it good for? Propaganda, value-instantiating beliefs, war support and resistance in Russia

Introduction

What role does propaganda play in decisions to resist or comply in an authoritarian society? Beyond its practical relevance, the question also provides a test case for theories of political influence. Existing approaches focus on persuasion, domination, and misinformation (Rosenfeld & Wallace, 2024), yet less attention is given to how propaganda structures and anchors the moral interpretation of events. This study advances this frontier by introducing a new theoretical framework for capturing how citizens construe events in value-laden terms and examining how these construals relate to exposure to state media and behavioural intentions, using the Russian invasion of Ukraine as a case study.

The role of propaganda in political action is widely debated, particularly in the case of Russia. The polls relay high levels of support for Vladimir Putin and the invasion, but opinions vary on the trustworthiness of the polls and whether propaganda is an important factor (Alyukov, 2022; Frye et al., 2023; Snegovaya, 2024).

The invasion of Ukraine, launched on February 24, 2022, was a deeply consequential event for Russian society. As the largest land war in Europe since 1945, it has resulted in substantial casualties and widespread societal involvement, with hundreds of thousands of Russians directly participating in the fighting and many more affected through family members or close acquaintances (Jones & McCabe, 2025).

Despite its scale, public reaction within Russian society has been relatively muted. While protests did occur, with over 20,000 arrests between 2022 and 2025 (OVD-Info, 2025), they did not reach the magnitude of protests against unfair elections in 2012 (Robertson, 2013) or against the arrest of an opposition politician Alexey Navalny. Why?

Along with fear and feelings of inefficacy, propaganda is one of the go-to explanations: people do not protest because they accept the interpretation of the war provided by state-controlled media. Authoritarian governments themselves seem to share this view, investing heavily in propaganda. In Russia, federal spending on state media rose despite fiscal strain, reaching $1.7 billion in 2025, up from $1.1 billion the year prior (Kommersant, 2024).

Authoritarian information ecosystems do not consist of monolithic propaganda. In Russia, the state controls national television, but citizens still access alternative outlets, especially online. The television, online news, and social media create interacting systems where the repetition of official narratives may boost their credibility – or, alternatively, challenge them (Alyukov, 2024). Distinguishing between pro-state and independent media sources is essential, as they differ in content, tone, and audience self-selection.

Russian state media closely follow the propaganda model of journalism and adopt an unequivocally pro-state perspective (Nordenstreng et al., 2023). Within this environment, the war is portrayed as a defensive necessity in response to NATO expansion, as a collective national effort, continuation of a historical struggle with the ‘collective West’, and a rerun of the Great Patriotic War of 1941-1945 (Brusylovska & Maksymenko, 2023; Khlevniuk et al., 2025).

Existing research conceptualises propaganda in several ways. One line of work treats it as exposure, showing that access to propaganda alone can influence behaviour (e.g., the Rwandan radio effect; Yanagizawa-Drott, 2014). Another views it as a type of political persuasion that shapes attitudes and disparages opposing opinions (Guess & Lyons, 2020; Horz, 2024; Hyzen, 2021; Jowett & O’Donnell, 2015). A third perspective emphasizes that regimes use “hard” or even absurd propaganda to signal strength, demobilize, and create expectations, even when the messages are not fully believed (Huang, 2015; Rosenfeld & Wallace, 2024).

Although conceptually values are placed at the core of persuasion effects of propaganda (Hyzen, 2021, p. 3483: “values + beliefs = opinion”), empirical work on the role of values in persuasion has been limited. One notable exception is research on value and moral framing, which shows that messages resonate particularly well when they match the recipients’ values and moral foundations (Dennison, 2020; Feinberg & Willer, 2019).

Prior work addresses values at the message design side (what the elites say) or receiver side (what does the audience value) (DeMora et al., 2021; Dennison, 2020; Schultz & Zelezny, 2003). We argue that the missing link is in how propaganda shapes the perceived meaning of concrete actions and events. Rather than simply aligning messages with pre-existing moral concerns, authoritarian persuasion redefines how these concerns can be addressed by portraying the same act (such as supporting the war) as an act of self-defence, expression of solidarity with fellow citizens, or continuity with historic tradition. In this sense, propaganda operates not only through exposure or alignment, but it also provides audiences with interpretations that make state-sanctioned actions consistent with their own value priorities, offering people answers to the question “What is the war good for?”

This perspective contributes to both persuasion and domination accounts of propaganda. From the persuasion side, it specifies which specific motives are engaged by propaganda rather than assuming a uniform belief change. From the domination side, it clarifies how coercive or absurd messages can succeed by breaking down value-behaviour ties to make action meaningless – or by offering emotionally coherent motivations for compliance, rather than by simply instilling fear. Understanding propaganda in this way requires measuring not only what people hear and whether they agree, but how they come to understand the moral significance of events.

We test this meaning-making account by (a) distinguishing pro-state and independent media consumption and measuring use and trust for each; (b) measuring how people construe the war in terms of its perceived consequences for basic human values; and (c) relating these construals to media use, attitudes, and intentions to resist or support the war. To operationalise construals of the war, we rely on the theory of basic human values (Schwartz, 1992, 2017) and value instantiation approach (Hanel et al., 2017; Maio, 2010; Ponizovskiy et al., 2019) for a conceptual framework of the study.

Values and Value-Instantiating Beliefs

Basic values (e.g., security, benevolence, achievement; Table 1) are abstract goals that guide evaluation and selection of behaviour (Schwartz & Bilsky, 1990). People universally agree that these goals are important (Davidov, 2010; Sagiv & Schwartz, 2022) but disagree on which actions reflect them in practice (Hanel et al., 2018; Ponizovskiy, 2022).

Value-Instantiating Beliefs (VIBs; Ponizovskiy et al., 2019) capture these act-by-value links: beliefs about whether a particular event or action benefits or threatens a particular value, for example, “the war protects me and my country” (a positive link to Security) or “the war harms the weak and vulnerable” (a negative link to Universalism). Each VIB therefore integrates a target (action or event), a value, and a direction (benefit vs. threat). Measuring VIBs across the ten basic value types produces a snapshot of the motivational meaning of the event: does the person see the war as primarily protective of their security, thwarting pro-social motives, or an invigorating challenge?

Importantly, prior work has shown that VIBs explain how values affect behaviour: people act on a value (e.g., universalism) only if they believe the action has consequences for that value (Ponizovskiy, 2022; Ponizovskiy et al., 2019). VIBs also systematically covary with partisan media exposure, supporting the view that media preferences are reflected, either causally or through self-selection, in how events are construed and not simply in whether they are supported (Ponizovskiy et al., 2023).

Table 1

Basic Human Values, their Definitions, and corresponding VIB items

Current Study

In this study, we apply the value instantiation framework to the Russian context. We ask the following exploratory research questions:

RQ1: Are there coherent construals of the war which emerge from the patterns of reported VIBs?

RQ2: If so, are these construals associated with intentions to resist or support the war effort?

RQ3: Do these construals merely reflect broader ideology, or do they account for additional variance in support for government-sanctioned violence?

Please note that we, given the inherent limitations of correlational mediation and in response to the reviews, tested preregistered hypotheses regarding mediations and other processes in the Supplementary Information. We provide a brief summary of these results in the Results section.

To address these questions, we first conduct variable-based analyses: (1) examine the relationships between individual VIBs towards the war and consumption of independent/state media; (2) examine the relationships between VIBs and intentions for political action. We then proceed to profile-based analyses: (3) examine the structure of construals by conducting a latent profile analysis (LPA) on VIB variables; and (4) test the resulting profiles as statistical predictors of attitudes towards the war, intentions to act in support of the war, and intentions to act in opposition while controlling for ideological covariates.

The study extends the current literature by proposing and testing a new covariate of propaganda exposure and political (de)mobilisation: the motivational meaning of events. Instead of viewing propaganda solely as persuasive or coercive, we conceptualise it as a part of the system that provides interpretative frames for events and actions. This approach builds a new link between theories of propaganda, collective action, and the psychology of values. Empirically, we introduce an approach to measuring the motivational meaning of a political event for individuals that can be applied to other socially significant events beyond the war in Ukraine.

Deviations from Preregistration

The study was preregistered prior to data collection (see Data Availability and Preregistration). While the preregistration specified a series of regression and path-analytic models, several deviations were introduced. First, instead of selecting a subset of VIBs as mediators, we included all VIBs in the mediation model to avoid selection effects. Second, the preregistered exploratory clustering approach (k-means clustering) was complemented with latent profile analysis, following the reviewers’ suggestions. Although profile-based analyses were preregistered as exploratory, they are emphasised in the main text because they provide a more concise and interpretable summary of associations between VIBs, media use, and behavioural intentions; the preregistered variable-based analyses are reported in the Supplementary Materials. Third, additional analyses such as confirmatory factor analyses and regression models with covariates were conducted to strengthen measurement and provide clearer tests of associations. Full details of all deviations and the outcomes of hypothesis tests are reported in Table S0 in the Supplementary Materials.

Methods

The data were collected in August 2022, five months after the invasion of Ukraine but before the first round of military mobilisation in Russia. Participants were invited to a study entitled “Research on attitudes and political behaviour in Russia” via the Yandex Toloka crowdsourcing platform. The study description and the participant information explained that the study will assess attitudes towards “the special military operation in Ukraine” and political behaviour, and that it is conducted by a group of researchers from the Higher School of Economics (Moscow) and Ruhr University Bochum. In all materials, the Russian invasion of Ukraine was referred to as “The special military operation”, as at that time the use of the word “war” was criminally prosecuted. We used quota sampling to match the sample to the age and gender distributions of the Russian population. After agreeing to participate in the study, participants were taken to an online questionnaire that assessed media consumption, trust towards the media, value-instantiating beliefs about the war with Ukraine, attitudes towards political action, and behavioural intentions. The questionnaire included additional variables that are not used in the present study but are detailed in the data repository and study registration materials. The study took on average 26 minutes to complete. Participants were reimbursed $0.50 for their time. The study received ethical approval from the Ethics Committee of Ruhr University Bochum. Demographic variables were removed from the publicly shared dataset to make reidentification of participants less likely.

Participants

The target sample size for the study of 1,000 was determined for the following multiple regression: G*Power 3.1 (Erdfelder et al., 1996; multiple linear regression, fixed model, deviation from zero), assuming 12 predictors (VIBs and covariates), 80% power, α = .05, directional tests and a small partial effect of f² = 0.015 (corresponding approximately to partial R² of 0.015) indicated a required sample of 963. A total of 1011 people gave informed consent to participate in the study, 981 completed the study, and 973 did so while successfully passing both attention checks. The gender distribution (49.9% women, 49.6% men, and 0.4% non-binary) underrepresented women compared to the national population (53.5% women, UN DESA, 2024). On other demographic variables, the sample moderately deviated from population parameters. The median age was 37 years (vs. 39.1 nationally; UN DESA, 2024). Ninety-one percent identified as ethnically Russian (vs. 81% in the 2021 Census; Federal State Statistics Service, 2022), and 51% had a university degree (vs. 57% nationally; OECD, 2021). In psychological terms, participants expressed strong attachment to their country (M = 3.94 on a five-point scale). For comparison, the analogous ISSP item produced means of 3.18 (1995), 2.83 (2003), and 2.77 (2013) on a four-point scale (ISSP, 1998; 2012; 2015). While ISSP data showed a decline in attachment to the country over that period alongside increases in nationalism and political patriotism (Grigoryan & Ponizovskiy, 2018), the present wartime data show stronger affective identification. Since the quota sampling by age and gender resulted in a sample composition that, despite some deviations, is broadly comparable to population averages on other demographic variables and shows plausible affective identification levels, we consider the sample suitable for the present analyses. Prior work on survey research in Russia after the 2022 invasion has not identified major change in sampling composition, response rate to the polls, length of interviews, and rate of interview interruptions after the 2022 invasion of Ukraine (Rosenfeld, 2022; Snegovaya, 2024; Zvonovskii & Khodykin, 2024). At the same time, the possibility of self-selection and response bias in the present data cannot be ruled out.

Measures

Media Consumption

We measured the frequency of use of pro-state and independent media sources using a scale from 1 (“almost never”) to 5 (“every day”). The media sources were selected in consultation with Russian media experts. Pro-state media sources included state television channels (e.g., Channel One, NTV), federal news agencies (e.g., RIA, TASS), Russia-based online newspapers (e.g., Kommersant, Gazeta.ru), and Russia-based social networks (e.g., Vkontakte, Odnoklassniki). Although particular forms of state control over these media channels vary, the high degree of control is broadly acknowledged (Pleines et al., 2020). Independent news sources included foreign-based Russophone online newspapers (e.g., Medusa, Radio Liberty), foreign social media (e.g., Facebook, Twitter), and Youtube.

The measurement model was tested using Confirmatory Factor Analysis (CFA). Base model with all pro-state sources and all independent sources loading on corresponding factors showed unsatisfactory fit to the data (χ²(13) = 236.58, p < .001, CFI = .870, TLI = .790, RMSEA = .133, SRMR = .100). Model fit improved after allowing the residuals of the use of Russia-based and independent online newspapers to correlate and setting a small negative variance for state news agencies to zero (χ²(13) = 138.56, p < .001, CFI = .928, TLI = .884, RMSEA = .099, SRMR = .069). The frequencies of use for pro-state and independent use scores were averaged to produce media use scores by type (pro-state: α = .73 and independent: α = .66, respectively).

Media Trust

To assess media trust, we asked participants to indicate how much they trust the aforementioned sources, with responses ranging from 1 (“Distrust completely”) to 5 (“Trust completely”). A CFA with the same two adjustments (correlation for online newspapers and zero variance for state news agencies) showed acceptable fit (χ²(13) = 120.63, p < .001, CFI = .916, TLI = .864, RMSEA = .092, SRMR = .079). As with consumption, the trust towards media was averaged by type (pro-state: α = .87 and independent: α = .79, respectively).

Value-Instantiating Beliefs

VIBs towards the war were assessed using the value instantiating beliefs questionnaire (Ponizovskiy, 2022). The scale parallels the structure of the PVQ-21 (Schwartz, 2003): it includes two items for each value type, except for universalism, which was measured with three items capturing its Concern, Nature, and Tolerance facets. The prompt presented the target event (“The special military operation will…”) followed by 21 semantic differential items describing positive and negative value consequences. Participants indicated their agreement on a seven-point scale ranging from −3 to +3. For example, a Security item asked participants to rate whether the “special military operation” will “make the country less stable and secure” (−3) or “more stable and secure” (+3). The items are presented in Table 1. Internal consistency estimates (Person’s correlations for two-item scales and Cronbach’s α for multi-item scales) were: Security (r = .49, p < .001), Conformity (r = .62, p < .001), Tradition (r = .88, p < .001), Benevolence (r = .49, p < .001), Universalism (α = .74), Self-Direction (r = .63, p < .001), Stimulation (r = .66, p < .001), Hedonism (r = .57, p < .001), and Achievement-Power (α = .85). Due to high overlap between Achievement and Power VIBs, they were combined into a single scale (see Measurement Model).

Measurement Model. The expected ten-factor model could not be estimated due to a non-positive definite covariance matrix caused by near-perfect correlation of Achievement and Power VIBs. These two were merged into a single Achievement-Power factor. The resulting nine-VIB model showed acceptable fit, χ²(153) = 701.61, p < .001, CFI = .940, TLI = .918, RMSEA = .061, SRMR = .045 (See Tables S1–S2 for model selection and loadings).

Ipsatisation and Interpretation. VIB scores can be analysed either raw or person-centred, and the two versions have distinct meanings.

Raw VIB Scores reflect absolute agreement with each belief statement. For example, a Security score of +2 indicates moderate agreement that “the war will be safe for me personally” and “will make the country more stable and secure”. As shown in Ponizovskiy (2022), respondents rarely treat values as competing when evaluating a single target. Instead of saying “the war is good for Security but bad for Universalism”, people tend to say that it is good for both values, though for Security more than for Universalism. This “positive across the board” pattern suggests that raw VIBs combine the general evaluative stance towards the object with the differentiated motivational meaning. In raw form, VIBs can be thought of as attitudinal beliefs, or a cognitive component of the attitude (Fishbein & Ajzen, 1975).

Ipsatised VIB scores remove the overall evaluative tendency by centring each participant’s VIB on their personal mean. What remains is the relative evaluation of consequences: given that the event is generally good (or bad), which value consequences make it especially good? And which contribute the least? Ipsatised VIB scores therefore, capture the motivational meaning of the event – the specific value-based interpretation of what the war is good for. Since the present analysis focuses on value-based meaning, all subsequent models use person-centred (ipsatised) VIB scores.

Attitudes Towards the War

Attitudes were assessed using a three-item Osgood-type five-point semantic differential scale with anchors “good-bad”, “harmful-beneficial”, and “useless-useful” (α = .90).

Behavioural Intentions

Intentions for political action were measured separately for action in support of the war and anti-war action using three items: “I want to / I intend to / I plan to take part in political action against the war in Ukraine” and “…in support of the war in Ukraine”, with response options ranging from 1 (“Absolutely disagree”) to 5 (“Absolutely agree”) (support of the war: α = .97 and against the war: α = .95, respectively).

National Identity

Identification with the nation was measured using a four-item Likert-type scale adapted from Becker and Wagner’s (2009) gender identity measure. An example item is “Being Russian is important to me”, with response options ranging from 1 (“Absolutely disagree”) to 5 (“Absolutely agree”) (α = .92).

Right-Wing Authoritarianism (RWA)

We used the Authoritarianism short scale (KSA-3; Nießen et al., 2019). The nine items assess aggression, submissiveness, and conventionalism, e.g., “We should take strong action against misfits and slackers in society”, with response options ranging from 1 (“Absolutely disagree”) to 5 (“Absolutely agree”) (α = .90).

Social Dominance Orientation (SDO)

SDO was measured with a Russian adaptation of Pratto et al. (1994) by Gulevich et al. (2018). The ten-item scale includes items such as “Low status groups should know their place”, with response options from 1 (“Absolutely disagree”) to 7 (“Absolutely agree”) (α = .80).

Other variables (e.g., conspiracy beliefs, political cynicism) were included but are outside the scope of the present analysis. Full details are available in the OSF materials.

Other variables (e.g., conspiracy beliefs, political cynicism) were included but are outside the scope of the present analysis. Full details are available in the OSF materials.

All analyses were conducted in R (Version 4.4.1; R Core Team, 2024). Data management and visualisation used packages from the tidyverse (Wickham et al., 2019), including dplyr 1.1.4, tidyr 1.3.1, scales 1.4.0, and ggplot2 3.5.2. Confirmatory factor analysis was performed with lavaan 0.6-19 (Rosseel, 2012). Regressions were estimated using the base R function lm, with robust confidence intervals estimated using lmtest 0.9-40 and sandwich 3.1-1 (Zeileis & Hothorn, 2002; Zeileis et al., 2020). Latent profile analysis was conducted using mclust 6.1.1 (Scrucca et al., 2023).

Results

The means, standard deviations, and correlations among key variables are presented in Table 2. The intercorrelations between the VIB variables are presented in Tables S3–S5. State media were both used more than independent media, t(972) = 6.05, p < .001, d = 0.19, and trusted more, t(972) = 7.65, p < .001, d = 0.25. However, mean trust for both types of sources fell below the scale’s midpoint, indicating broad scepticism. Attitudes towards the war centred slightly below the midpoint of the scale. Intentions for political action were low, with the intention to act in support of the war higher than to act in opposition. The raw value-instantiating beliefs towards the war were broadly negative (Figure 1), with the only positive mean for Stimulation (“be a new, stimulating experience”, “be exciting”). The most negative means were observed for Universalism (e.g., “help the weak and vulnerable in the society”), Hedonism (e.g., “be pleasant to my senses…”), and Power (e.g., “allow me to gain money and status”).

Table 2

Means, Standard Deviations, and Intercorrelations Among Key Variables

Figure 1

Distributions of Raw Value-Instantiating Beliefs Scores.

Correlational Analyses

First, we estimated the associations between VIBs and media consumption. Because ipsatised VIBs are linearly dependent (sum to zero within individuals), we estimated separate regression models for each value type, regressing pro-state and independent media consumption on VIB for each value type, controlling for sociodemographic covariates: age, gender, and level of education. The resulting models, corrected for False Discovery Rate (FDR), reflect the relative positivity of consequences for that value compared with other values (Tables S6).

As shown in Figure 2, consumers of independent media perceived the war as having less positive consequences for Conservation values (Security, Conformity, and Tradition), and as having more positive consequences for Self-Enhancement values (Achievement-Power) and Stimulation. This pattern was mostly mirrored for consumers of pro-state media, apart from Security VIB: people who consumed more state-affiliated media did not perceive the war as more safe.

Figure 2

The Associations of Independent and Pro-State Media Consumption with Ipsatised VIBs.

Next, we evaluated the associations between VIBs and attitude towards the war as well as behavioural intention to act in support and opposition to the war (Figure 3). The models were controlled for sociodemographic as well as ideological covariates (RWA, SDO, and National Identity), and p-values were FDR corrected (Tables S7 and S8).

As shown in Figure 3, seven out of nine VIBs were significant predictors after ideological controls and FDR correction. People with more positive attitudes towards the war perceived it as more aligned with conservation values (Security, Conformity, Tradition) and less strongly aligned with Universalism, Stimulation, Hedonism, and Achievement-Power.

Behavioural intentions to act in support of the war were positively associated with perceiving the war as more strongly aligned with Tradition and less strongly with Achievement-Power and Hedonism. Intention to oppose the war was weakly related to VIBs, with a small negative association with perceived alignment with Tradition and a small positive association with perceived alignment with Benevolence.

Although our data are cross-sectional and cannot establish causal direction, we tested whether the associations were consistent with the preregistered mediation structure linking media consumption, VIBs, attitudes and behavioural intentions. The model fit the data well (Figure S4) and showed the expected pattern: VIBs, especially for Security, Conformity, and Tradition, mediated the relationship between pro-state media consumption and attitudes and then intentions. The paths stemming from independent media consumption showed opposite effects. These results are correlational and cannot establish a causal direction, but they are compatible with the theorized pathway where VIBs connect propaganda exposure to attitudes and intentions.

Figure 3

Associations Between Value-Instantiating Beliefs and Attitudes and Behavioural Intentions

Person-Centred Analyses

Ipsatisation already captures the relative evaluation of value consequences by centring each individual’s responses on their own mean. However, this transformation makes VIB scores linearly interdependent, complicating their interpretation. For example, negative associations of high Achievement-Power scores may be interpretable on their own: people who see the war as serving their interests may be less willing to join in pro-war political action. Or, alternatively, high Achievement-Power scores might simply mean that the consequences for other (perhaps more relevant) values are more negative.

The patterns of differences between the VIBs, rather than the absolute scores, describe how people make sense of the war. To examine these profiles directly, we moved from variable-centred regressions to person-centred analyses.

We conducted LPA on the nine ipsatised VIBs, which ensured that profiles reflected differences in the content of evaluations rather than overall positivity. We estimated LPA models with one through six classes under a constrained covariance structure (EEI; equal volume and shape, spherical within class; Table 3). More flexible covariance structures (e.g. VVV, EEE) did not yield admissible solutions beyond a single profile and were not considered further.

Table 3

Model fit of LPA models with 1 to 6 classes

Fit indices suggested improvement of fit as more classes were added: the Bayesian Information Criterion monotonically decreased, and the bootstrap likelihood ratio test (BLRT) remained significant across comparisons. However, solutions with three or more profiles yielded small classes (<15% of participants) and did not improve classification quality. To balance statistical fit with parsimony and interpretability, we retained the two-profile solution. Solutions with additional classes added nuance but were consistent with the presented results (Figures S1–S2).

Figure 4 shows posterior-weighted means and 95% confidence intervals for each class; the left panel shows ipsatised VIBs that were used for profile identification, and the right panel shows raw participant responses. Class 1 (31% of respondents) saw the most positive consequences of the war for Conservation values (Security, Conformity and Tradition) and the least positive consequences for Self-Enhancement values (Achievement-Power and Hedonism). Class 2 (69% of respondents) showed the reverse pattern. When mapped onto non-ipsatised VIB scores, it becomes clear that the pattern differed not because one group saw the war as self-serving, but because Class 1 viewed it as securing the social order, and Class 2 – as undermining it. In absolute terms, Class 1 reported more positive consequences for Conservation values: Security, t(972) = -18.2, p < .001, d = 1.09, Conformity, t(972) = -27.6, p < .001, d = 1.62, and Tradition, t(972) = -22.2, p < .001, d = 1.28. In contrast, perceived consequences were similar across groups for Hedonism, t(972) = -.26, p = .792, d = 0.02 and Achievement-Power, t(972) = – 1.92, p = .055, d = 0.11

Although the analysis used ipsatised scores to define profiles based on their shape, moderate mean-level differences remained: Class 1 was, on average, more positive on most values than Class 2. This suggests that while profiles were differentiated based on structural differences in meaning – whether the war is construed as preserving the social order or disrupting it – certain construal “shapes” also co-occur with more generally positive or negative evaluations of the war. In the following analyses, we examined whether this content of construal merely reflects broader ideological orientation or uniquely explains variation in attitudes and behaviour.

We regressed the attitude towards the war, intentions to act in support for or in opposition to the war on posterior probability of belonging to VIB Class 1 (vs. 2) alongside RWA, SDO, and national identification: constructs proposed to explain endorsement of violence by one’s own group (Abou-Ismail et al., 2025; Varmann et al., 2024). The VIB class emerged as a significant predictor of the attitude, contributing explained variance over and above RWA, SDO, and national identity (Table 4). The probability of belonging to VIB Class 1 was also predictive of intention to act in support of the war, although not in opposition to it. Results were substantively similar using k-means clustering (see Supplementary Materials).

Taken together, these findings show that people’s answers to the question “What is war good for?” – whether they see it as promoting or thwarting a particular set of values – capture a psychological dimension that is distinct from broader ideology and helpful in explaining people’s attitudes and actions.

Figure 4

Latent Profile Means of Ipsatised (Left Panel) and Raw (Right Panel) Value-Instantiating Belief Scores (EEI, 2-class model).

Table 4

Multiple Regressions Predicting Attitudes and Behavioural Intention

Discussion

We found systematic differences in how Russians construed the war in value terms, and these construals were associated with media use, attitudes towards the war, and intentions to engage in pro/anti-war political action. While Right-Wing Authoritarianism was the closest correlate of pro-war attitudes and actions in our analyses, VIBs made a unique and sizeable contribution beyond RWA, SDO, and national identification.

In August 2022, our Russian participants held varying attitudes towards the war, but largely agreed on some of its expected value consequences: the war will harm the vulnerable and will limit opportunities to achieve wealth, influence, and recognition. However, some Russians (31% of participants) saw the invasion of Ukraine as beneficial for their own and societal security, as an expression of their traditional values, and as something that their compatriots expect or approve – while others did not (69% of participants). These estimates are broadly consistent with prior research showing that strongly pro-Putin and pro-war attitudes are concentrated in the minority of respondents, alongside a small but meaningful group that strongly opposes the war, and a larger group expressing ambivalence and conformity-driven responses (Frye et al., 2023; Snegovaya, 2024). At the same time, the sensitive political nature of the survey could have led to higher rates of participation from individuals who are more politically engaged and more approving of the war.

Beyond these distributions, the profiles capture how the war is construed in value terms. Notably, we did not find a profile differentiated by stronger perceptions of the war as violating universalistic values. Instead, responses were differentiated along conservation-related values – Security, Tradition, and Conformity – suggesting that the variation in attitudes is structured less by universalistic moral concerns and more by perceptions of social order. Conservation values prominently feature in Russian wartime propaganda, and the relevant VIBs covaried with the consumption of pro-state and independent media. The difference between the two types of construals was related to behavioural intentions. Including VIB profiles in the same models with RWA, SDO, and national identity showed that the pattern of VIB responses accounted for sizeable variance over and above these powerful explanatory constructs.

Contribution

From a theoretical standpoint, we introduce value instantiations as a novel correlate of media use and political intentions and demonstrate that they are distinct from broader political ideology. This extends research on influence and propaganda by incorporating value-based meaning, capturing not only how the events are described or evaluated, but how they are understood vis-à-vis fundamental life principles.

Research on media effects and propaganda has typically focused on communication content and individuals’ attitudes towards the topic of the messages. While these approaches capture what is being said and how the topic is evaluated, they do not specify the structure of meaning underlying these evaluations. We address this gap by examining how Russian individuals understand the same event – the invasion of Ukraine – in terms of its implications for basic human values.

Value instantiations capture the motivational meaning of actions and events: do they support or threaten basic human values such as security, benevolence, or tradition. They are assessed using a standardised measure grounded in value theory, and allow for comparisons across topics, individuals, and contexts. These beliefs are more proximal to specific situations than ideological orientations, but may serve as building blocks of ideology. Value instantiations have previously been shown to impact behaviour (Maio et al., 2009) and moderate value-intention links (Ponizovskiy et al., 2019; Ponizovskiy, 2022), linking to the theory of planned behaviour (Ajzen, 1991) and the theory of basic human values (Schwartz, 1992) and allowing for grounded predictions.

Empirically, using a quota sample from an authoritarian country at war, we observe substantial variation in construals even in a repressive context with a heavily funded propaganda machine. Furthermore, this variation is practically and predictively meaningful – accounting for it adds substantive power in modelling political behavioural intentions.

Limitations

Survey research in repressive contexts raises concerns about self-selection and response bias (Tannenberg, 2022). The reliability of polls in repressive contexts such as wartime Russia has been questioned (Alyukov, 2022). Participants who strongly oppose the war, distrust the safety of online surveys, or are especially sensitive to surveillance might have been less likely to participate and could have moderated their answers in the direction of greater conformity or ambiguity. These processes could have affected both the composition of the sample as well as reported VIB, attitudes, and behavioural intentions.

The nature of repression in Russia complicates expectations about such biases, as the behavioural consequences of repression may depend on its form, visibility, and patterns of enforcement (Ayanian et al., 2024). Prior to 2022, the regime had been described as an ‘informational autocracy’, relying on control of narratives and selective punishment of political action rather than control through fear and systematic punishment of privately expressed views (Guriev & Treisman, 2019). Following the invasion, repression intensified, including increased penalties for public and online dissent. At the same time, enforcement has remained focused primarily on public, visible acts of opposition. The combination of real sanctions and uncertain enforcement can elicit different responses: while some people may withdraw from participation and self-censor, others may continue to express their views in settings perceived as private or low-risk.

Consistent with this complexity, work on survey research in Russia has documented falsification of preferences (Chapkovski & Schaub, 2022) and reduced participation from more dissenting regions prior to the war, alongside evidence that participation did not uniformly decline following the invasion and may, in some contexts, have made surveys a channel for expressing disagreement (Vyrskaia et al., 2025). At the same time, other recent research has not identified major changes in sampling composition, response biases, or interview dynamics following the invasion (Rosenfeld, 2022; Snegovaya, 2024; Zvonovskii & Khodykin, 2024).

In the present study, such biases cannot be ruled out. Demographic quota sampling, while ensuring approximate balance on age and gender, does not guarantee representativeness on psychological variables such as risk tolerance or trust. Consequently, the results should not be interpreted as an unbiased population estimate of opinions in Russia. At the same time, to the degree that the selection and response biases may be present in the data, they would be expected to constrain the distribution of VIBs, attitudes, and behavioural intentions, making profile differentiation more difficult and biasing the estimates of effects downwards.

The present study is correlational and cannot test the theorised causal relationships that drove hypothesis development. While past experimental research has shown that message content impacts VIBs (Ponizovskiy et al., 2019), in real-world scenarios the relationship is almost certainly bidirectional: on the one hand, the media shape the meaning of events (Chong & Druckman, 2007) and on the other, people choose the media that are aligned with their views (Stroud, 2008). The correlational nature of the study did not allow us to tease the two influences apart.

This research was possible thanks to a slow reaction of Russian survey companies to the beginning of the war – participation in foreign academic surveys was not regulated before the war. It came under increasing scrutiny only at the end of 2022 when Yandex.Toloka introduced pre-screening of surveys in late 2022, and in July 2023, the service was replaced with state-controlled Yandex.Tasks (Kommersant, 2023). This means, however, that we make inferences based on a snapshot of Russian society at an early stage of the war. While the patterns that we present here are robust, they do not directly test the proposed causal mechanism and only show that observational data is consistent with it.

Future Directions

The value instantiation framework can potentially provide a fine-grained account of real-world propaganda effects, both in its persuasion and domination facets. While a longitudinal study in Russia is not feasible at this time, similar research could be undertaken in an authoritarian context that is less repressive than today’s Russia, such as Hungary or China.

The effects of propaganda are also not limited by state borders. Beyond spending on state media at home, Russia invests heavily in its foreign-facing media and disruptive politicians abroad (Gerber & Zavisca, 2016; Yang et al., 2024). The VIB framework described here can be applied to quantify the relationship between exposure to Russian, Chinese, or American propaganda and the construals of important events across the world.

Finally, further theoretical work could move us closer to a unified theory of ideological uptake. The application of VIBs to propaganda hypothesises a cognitive-motivational mechanism, but at present it lacks boundary conditions. When is propaganda successful and when not? When is it motivating and when demotivating? What situational, personality, content variables influence uptake? This content-first approach to propaganda effects could provide an interface between communications studies, political science, and moral psychology.

Conclusion

This study examined how people in Russia construe the invasion of Ukraine, and how these construals relate to political attitudes and intended action. Using a large quota sample collected in Russia during the early stages of the war, we identified distinct profiles of understanding of the war’s consequences to people’s core values.

Drawing on value instantiation framework, we showed that some Russians viewed the war as enhancing conservation values such as Security, Conformity, and Tradition, and others saw it as undermining these and other values. These value-instantiating belief profiles were associated with media use and were unique statistical predictors of attitudes and intentions to act in support for the war, even after controlling for national identity, right-wing authoritarianism and social dominance orientation. Although our data are correlational and cannot establish causality, the findings are consistent with the account where citizens’ interpretations of political events mediate between media exposure and intentions for action.

By introducing a method to quantify these interpretations, we contribute a new variable – the motivational meaning that people attach to actions and events – that can complement traditional predictors such as identity and ideology. This approach helps explain variation in political (in)action in authoritarian contexts and bridges research on values, propaganda, and collective action.

Conflicts of Interest

The authors declare no competing interests.

Data Availability Statement

Data, materials, and analysis scripts are available at https://osf.io/v7dwh. The preregistration form is available at https://aspredicted.org/8z2q-4psz.pdf. All deviations from the preregistrations are detailed in Table S0 in the Supplementary Materials.

Supplementary Materials

The supplementary materials can be found here.

Author Contributions

V.P., L.G., and E.O designed the study, E.O. collected the data, V.P. and M.W. analysed the data, V.P. drafted the manuscript. All authors provided critical revisions.

References

Abou-Ismail, R., Cichocka, A., Phillips, J., & Sengupta, N. K. (2025). Authoritarianism and social dominance as differential predictors of individuals’ support for collective violence. Psychology of Violence, 15(5), 634–644. https://doi.org/10.1037/vio0000580

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211. https://doi.org/10.1016/0749-5978(91)90020-T

Alyukov, M. (2022). Propaganda, authoritarianism and Russia’s invasion of Ukraine. Nature Human Behaviour, 6(6), 763–765. https://doi.org/10.1038/s41562-022-01375-x

Alyukov, M. (2024). News reception and authoritarian control in a hybrid media system: Russian TV viewers and the Russia-Ukraine conflict. Politics, 44(3), 400–419. https://doi.org/10.1177/02633957211041440

Ayanian, A. H., Tausch, N., & Saab, R. (2024). Social psychological processes underlying collective action in repressive contexts: What we know and ways forward for future research. In F. Bou Zeineddine & J. R. Vollhardt (Eds.), Resistance to repression and violence: Global psychological perspectives (pp. 25–50). Oxford University Press. https://doi.org/10.1093/9780197687703.003.0002

Becker, J. C., & Wagner, U. (2009). Doing gender differently—The interplay of strength of gender identification and content of gender identity in predicting women’s endorsement of sexist beliefs. European Journal of Social Psychology, 39(4), 487–508. https://doi.org/10.1002/ejsp.551

Brusylovska, O., & Maksymenko, I. (2023). Analysis of the media discourse on the 2022 war in Ukraine: The case of Russia. Regional Science Policy & Practice, 15(1), 222–236. https://doi.org/10.1111/rsp3.12579

Chapkovski, P., & Schaub, M. (2022). Solid support or secret dissent? A list experiment on preference falsification during the Russian war against Ukraine. Research & Politics, 9(2). https://doi.org/10.1177/20531680221108328

Chong, D., & Druckman, J. N. (2007). Framing theory. Annual Review of Political Science, 10(1), 103–126. https://doi.org/10.1146/annurev.polisci.10.072805.103054

Davidov, E. (2010). Testing for comparability of human values across countries and time with the third round of the European Social Survey. International Journal of Comparative Sociology, 51(3), 171–191. https://doi.org/10.1177/0020715210363534

DeMora, S. L., Merolla, J. L., Newman, B., & Zechmeister, E. J. (2021). Reducing mask resistance among White evangelical Christians with value-consistent messages. Proceedings of the National Academy of Sciences, 118(21), e2101723118. https://doi.org/10.1073/pnas.2101723118

Dennison, J. (2020). A basic human values approach to migration policy communication. Data & Policy, 2, e18. https://doi.org/10.1017/dap.2020.17

Erdfelder, E., Faul, F., & Buchner, A. (1996). Gpower: A general power analysis program. Behavior Research Methods, Instruments & Computers, 28(1), 1–11. https://doi.org/10.3758/BF03203630

Federal State Statistics Service. (2022). All-Russian Population Census. Volume 5: National (ethnic group) composition and language proficiency [Data set]. https://rosstat.gov.ru/storage/mediabank/Tom5_tab1_VPN-2020.xlsx

Feinberg, M., & Willer, R. (2019). Moral reframing: A technique for effective and persuasive communication across political divides. Social and Personality Psychology Compass, 13(12), e12501. https://doi.org/10.1111/spc3.12501

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Addison-Wesley.

Frye, T., Gehlbach​, S., Marquardt, K. L., & Reuter, O. J. (2023). Is Putin’s popularity (still) real? A cautionary note on using list experiments to measure popularity in authoritarian regimes. Post-Soviet Affairs, 39(3), 213–222. https://doi.org/10.1080/1060586X.2023.2187195

Gerber, T. P., & Zavisca, J. (2016). Does Russian propaganda work? The Washington Quarterly, 39(2), 79–98. https://doi.org/10.1080/0163660X.2016.1204398

Grigoryan, L. K., & Ponizovskiy, V. (2018). The three facets of national identity: Identity dynamics and attitudes toward immigrants in Russia. International Journal of Comparative Sociology, 59(5–6), 403–427. https://doi.org/10.1177/0020715218806037

Guess, A. M., & Lyons, B. A. (2020). Misinformation, disinformation, and online propaganda. In N. Persily & J. A. Tucker (Eds.), Social media and democracy (pp. 10–33). Cambridge University Press. https://doi.org/10.1017/9781108890960.003

Gulevich, O., Agadullina, E., & Khukhlaev, O. (2018). Approval of group hierarchy: Russian version of social dominance orientation scale. Psychology. Journal of Higher School of Economics, 15(3), 407–426.

Guriev, S., & Treisman, D. (2019). Informational autocrats. Journal of Economic Perspectives, 33(4), 100–127. https://doi.org/10.1257/jep.33.4.100  

Hanel, P. H. P., Vione, K. C., Hahn, U., & Maio, G. R. (2017). Value instantiations: The missing link between values and behavior? In S. Roccas & L. Sagiv (Eds.), Values and behavior: Taking a cross cultural perspective (pp. 175–190). Springer International Publishing. https://doi.org/10.1007/978-3-319-56352-7_8

Hanel, P. H. P., Maio, G. R., Soares, A. K. S., Vione, K. C., de Holanda Coelho, G. L., Gouveia, V. V., Patil, A. C., Kamble, S. V., & Manstead, A. S. R. (2018). Cross-cultural differences and similarities in human value instantiation. Frontiers in Psychology, 9. https://doi.org/10.3389/fpsyg.2018.00849

Horz, C. M. (2024). Identity propaganda. British Journal of Political Science, 54(2), 313–338. https://doi.org/10.1017/S0007123423000182

Huang, H. (2015). Propaganda as signaling. Comparative Politics, 47(4), 419–444. https://doi.org/10.5129/001041515816103220

Hyzen, A. (2021). Revisiting the theoretical foundations of propaganda. International Journal of Communication, 15, 3479–3496.

ISSP. (1998). International Social Survey Programme: National Identity I – ISSP 1995. [Data set]. GESIS Data Archive. ZA2880 Data file Version 1.0.0, https://doi.org/10.4232/1.2880

ISSP. (2012). International Social Survey Programme: National Identity II – ISSP 2003. [Data set]. GESIS Data Archive. ZA3910 Data file Version 2.1.0, https://doi.org/10.4232/1.11449

ISSP. (2015). International Social Survey Programme: National Identity III – ISSP 2013. [Data set]. GESIS Data Archive. ZA5950 Data file Version 2.0.0, https://doi.org/10.4232/1.12312

Jones, S. G., & McCabe, R. (2025). Russia’s Battlefield Woes in Ukraine. Center for Strategic & International Studies. https://www.csis.org/analysis/russias-battlefield-woes-ukraine

Jowett, G. S., & O’Donnell, V. (2015). Propaganda and persuasion. Sage Publications.

Khlevniuk, D., GN, & Noordenbos, B. (2025). The temporality of memory politics: An analysis of Russian state media narratives on the war in Ukraine. The British Journal of Sociology, 76(2), 390–406. https://doi.org/10.1111/1468-4446.13171

Kommersant. (2023, July 26). Razdelu «Jandeksa» zadali napravlenie [The Yandex demerger gains direction]. https://www.kommersant.ru/doc/6124071

Kommersant. (2024, October 12). Bjudzhet transliruet den’gi [The budget channels money]. https://www.kommersant.ru/doc/7363945

Maio, G. R., Hahn, U., Frost, J.-M., & Cheung, W.-Y. (2009). Applying the value of equality unequally: Effects of value instantiations that vary in typicality. Journal of Personality and Social Psychology, 97(4), 598–614. https://doi.org/10.1037/a0016683

Maio, G. R. (2010). Mental representations of social values. Advances in Experimental Social Psychology, 42, 1–43. https://doi.org/10.1016/S0065-2601(10)42001-8

Nießen, D., Schmidt, I., Beierlein, C., & Lechner, C. M. (2019). Authoritarianism Short Scale (KSA-3). Open Access Repositorium für Messinstrumente. https://doi.org/10.6102/zis272

Nordenstreng, K., Pasti, S., Zhang, T., Jain, S., Bobba, G., Wolgast, H., Hyzen, A., Guazina, L., Patnaik, S., & Ndlovu, M. (2023). Coverage of the Russia–Ukraine war by television news. International Journal of Communication, 17, 6857–6873.

Organisation for Economic Cooperation and Development. (2021). Education at a glance 2021: OECD indicators. OECD Publishing. https://doi.org/10.1787/b35a14e5-en

OVD-Info. (2025). Persecution of the anti-war movement report: Three Years into Russia’s Full-Scale Invasion of Ukraine. https://ovd.info/en/antiwar_3_years

Pleines, H., Somfalvy, E., & Litvinenko, A. (2020). Media control as source of political power: Differentiating reach and impact. Russian Analytical Digest, 258, 2–5. https://doi.org/10.3929/ethz-b-000450633

Ponizovskiy, V., Grigoryan, L., Kühnen, U., & Boehnke, K. (2019). Social construction of the value–behavior relation. Frontiers in Psychology, 10. https://doi.org/10.3389/fpsyg.2019.00934

Ponizovskiy, V. (2022). On Sanders, Trump, and rhinoceroses: Quantifying subjective construals helps predict political attitudes. British Journal of Social Psychology, 61(3), 842–860. https://doi.org/10.1111/bjso.12516

Ponizovskiy, V., Grigoryan, L., & Hofmann, W. (2023). Why is right-wing media consumption associated with lower compliance with COVID-19 measures? Journal of Media Psychology, 35(1), 3–16. https://doi.org/10.1027/1864-1105/a000337

Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social dominance orientation: A personality variable predicting social and political attitudes. Journal of Personality and Social Psychology, 67(4), 741–763. https://psycnet.apa.org/doi/10.1037/0022-3514.67.4.741

R Core Team. (2024). R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.r-project.org

Robertson, G. (2013). Protesting Putinism: The election protests of 2011–2012 in broader perspective. Problems of Post-Communism, 60(2), 11–23. https://doi.org/10.2753/PPC1075-8216600202

Rosenfeld, B. (2022). Survey research in Russia: in the shadow of war. Post-Soviet Affairs, 39, 38–48. https://doi.org/10.1080/1060586X.2022.2151767

Rosenfeld, B., & Wallace, J. (2024). Information politics and propaganda in authoritarian societies. Annual Review of Political Science, 27, 263–281. https://doi.org/10.1146/annurev-polisci-041322-035951

Rosseel, Y. (2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1–36. https://doi.org/10.18637/jss.v048.i02

Sagiv, L., & Schwartz, S. H. (2022). Personal values across cultures. Annual Review of Psychology, 73(1), 517–546. https://doi.org/10.1146/annurev-psych-020821-125100

Schultz, P. W., & Zelezny, L. (2003). Reframing environmental messages to be congruent with American values. Human Ecology Review, 10(2), 126–136.

Schwartz, S., & Bilsky, W. (1990). Toward a theory of the universal content and structure of values: Extensions and cross-cultural replications. Journal of Personality and Social Psychology, 58(5), 878–891. https://doi.org/10.1037/0022-3514.58.5.878

Schwartz, S. H. (1992). Universals in the content and structure of values: Theoretical advances and empirical tests in 20 countries. In Advances in experimental social psychology, Vol. 25 (pp. 1–65). Academic Press. https://doi.org/10.1016/S0065-2601(08)60281-6

Schwartz, S. H. (2003). A proposal for measuring value orientations across nations. Questionnaire Development Package of the European Social Survey. https://www.europeansocialsurvey.org/sites/default/files/2023-06/ESS_core_questionnaire_human_values.pdf  

Schwartz, S. H. (2012). An overview of the Schwartz theory of basic values. Online Readings in Psychology and Culture, 2(1). https://doi.org/10.9707/2307-0919.1116

Schwartz, S. H. (2017). The refined theory of basic values. In S. Roccas & L. Sagiv (Eds.), Values and behavior: Taking a cross cultural perspective (pp. 51–72). Springer International Publishing. https://doi.org/10.1007/978-3-319-56352-7_3

Scrucca, L., Fraley, C., Murphy, T. B., & Raftery, A. E. (2023). Model-based clustering, classification, and density estimation using mclust in R. Chapman and Hall/CRC. https://doi.org/10.1201/9781003277965

Snegovaya, M. (2024). The reluctant consensus: War and Russia’s public opinion. Atlantic Council. https://www.atlanticcouncil.org/content-series/russia-tomorrow/reluctant-consensuswar-and-russias-public-opinion/

Stroud, N. J. (2008). Media use and political predispositions: Revisiting the concept of selective exposure. Political Behavior, 30(3), 341–366. https://doi.org/10.1007/s11109-007-9050-9

Tannenberg, M. (2022). The autocratic bias: Self-censorship of regime support. Democratization, 29(4), 591–610. https://doi.org/10.1080/13510347.2021.1981867

United Nations Department of Economic and Social Affairs. (2024). World population prospects 2024: Online edition [Data set]. https://population.un.org/wpp

Varmann, A. H., Kruse, L., Bierwiaczonek, K., Gómez, Á., Vázquez, A., & Kunst, J. R. (2024). How identity fusion predicts extreme pro-group orientations: A meta-analysis. European Review of Social Psychology, 35(1), 162–197. https://doi.org/10.1080/10463283.2023.2190267

Vyrskaia, M., Tkachenko, A., & Martynova, N. (2025). Want to be heard: Survey participation in Russia before and during the war. Post-Soviet Affairs, 41(6), 593–610. https://doi.org/10.1080/1060586X.2025.2548622

Wickham, H., Averick, M., Bryan, J., Chang, W., D’Agostino McGowan, L., François, R., Grolemund, G., Hayes, A., Henry, L., Hester, J., Kuhn, M., Pedersen, T. L., Miller, E., Bache, S. M., Müller, K., Ooms, J., Robinson, D., Seidel, D. P., Spinu, V., Takahashi, K., Vaughan, D., Wilke, C., Woo, K., & Yutani, H. (2019). Welcome to the tidyverse. Journal of Open Source Software, 4(43), 1686. https://doi.org/10.21105/joss.01686

Yanagizawa-Drott, D. (2014). Propaganda and conflict: Evidence from the Rwandan genocide. The Quarterly Journal of Economics, 129(4), 1947–1994. https://doi.org/10.1093/qje/qju020

Yang, Y., McCabe, S., & Hindman, M. (2024). Does Russian propaganda lead or follow? Topic coverage, user engagement, and RT and Sputnik’s agenda influence on US media. The International Journal of Press/Politics, 19401612241271074. https://doi.org/10.1177/19401612241271074

Zeileis, A., & Hothorn, T. (2002). Diagnostic checking in regression relationships. R News, 2(3), 7–10.

Zeileis, A., Köll, S., & Graham, N. (2020). Various versatile variances: An object-oriented implementation of clustered covariances in R. Journal of Statistical Software, 95(1), 1–36. https://doi.org/10.18637/jss.v095.i01

Zvonovskii, V., & Khodykin, A. (2024). Rossiyskoe obshchestvennoe mnenie v usloviyah voennogo konflikta, 2022–2023 [Public opinion in Russia during the military conflict, 2022–2023]. Historical Expertise, Chișinău.