Moral Desensitization in Digital Contexts: Instagram exposure alters moral perception and emotional reactivity

Moral Desensitization in Digital Contexts: Instagram exposure alters moral perception and emotional reactivity | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Moral Desensitization in Digital Contexts: Instagram exposure alters moral perception and emotional reactivity Nicola Chinchella, Aldo Gangemi, Alessia Gazineo, Chiara Lucifora This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8328737/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Social media thrive on virality amplifying content that maximizes engagement by triggering strong emotional and moral responses. As users’ feeds become saturated with affect-laden material, it becomes crucial to examine whether this exposure shapes online judgments and moral norms. In this work, we examine how engaging with moral dilemmas on Instagram affects users’ responses, focusing on perceived moral acceptability, arousal and valence. Our findings demonstrate that participants interacting with dilemmas on Instagram, compared to those using Qualtrics, exhibit increased moral acceptability ratings alongside attenuated arousal and valence responses. Additionally, the Instagram cohort required more time to complete the task. Our results suggest that simply being on social media platforms like Instagram can induce a moral desensitization posing significant implications given the widespread daily engagement with these platforms and the potential erosion of sensitivity to moral issues. Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 1. Introduction Over the past decade, the proliferation of social media platforms has significantly reshaped global communication patterns, with Facebook, Twitter, Instagram and TikTok collectively engaging billions of users. As for 2025, there are 5 billion active social media accounts around the globe, with an average use of 2 hours and 27 minutes per day (Statista 2025 ) and continuously evolving usage patterns (Marino et al., 2025 ). Social media success can be ascribed to a mix of technological wit and human dispositions. Platforms are explicitly designed to exert a persuasive pull (Lupinacci, 2021 ; Rixen et al., 2023 ) “pulsing with colourful notification badges, whoosh sounds and gentle vibrations” (Fisher, 2022 , p. 33). Furthermore, they construct detailed models of users’ preferences to maximize engagement by tailoring content (Stöcker & Preuss, 2020 ; Bruineberg, 2023 ). In doing so, social media have tapped into, and made increasingly visible, the human attraction to affect-laden content (Berger & Milkman, 2012 ), unveiling a pronounced preference for extreme material that can evoke, for example, heightened emotional arousal, moral outrage and out-group animosity (Rathje et al., 2021 ; Pandey et al., 2023 ; Van Bavel et al., 2024 ). Algorithms quickly learnt that the best way to increase engagement is to leverage this human disposition, amplifying affect-laden and extreme content (Stöcker & Preuss, 2020 ; Whittaker et al., 2021 ; Konovalova et al., 2023 ; Pandey et al., 2023 ). Mounting evidence indicates that recommendation systems are steering users towards such types of content even when users do not seek it (Whittaker et al., 2021 ; Pandey et al., 2023 ) and would prefer to avoid it (Milli et al., 2025 ). A particularly worrisome prospect of social media enabling and enhancing environments where affect-laden content can proliferate in unpredictable ways relates to what users are learning from their continuous interactions with the platforms, especially in relation to social influence. People are influenced by the behaviours of others around them and rely on social learning to navigate the world (Rimal & Lapinski 2015 ; Chung & Rimal, 2016; Carpenter & Amaravadi, 2019), balancing descriptive evidence (many people do this) and validation (people approve of this). Never before were people subject to such an easily reachable and massive social influence as with social media, whose consequences are now within glimpse. Consistently with reinforcement learning models (Lindström et al., 2021 ), research shows that likes, shares and other metrics successfully predict first person perception and behaviour (Chung, 2019 ; Brady et al., 2021 ). At the same time, social media afford environments where descriptive information is easily observable, e.g., in extreme networks, dampening the sensitivity to critical social feedback (Brady et al., 2021 ). Both create powerful normative pressure toward behavioural adoption. In principle learning could cut both ways, nudging ethical and responsible actions (Pan et al., 2023 ; Pastor et al., 2024 ) as well as forming bad habits and discouraging rationality (Shareef et al., 2020 ; Savolainen et al., 2021 ). However, there is a widespread suspicion in the literature that social norms generated on social media are often more extreme than offline ones (Robertson et al., 2024 ), tilting the needle to the latter. Many argue that algorithm-mediated learning on social media is effectively shaping new more extreme ways for individuals to think and feel about social norms, opening interesting research venues (Acerbi, 2019 ; Brady et al., 2023; Brady & Crockett, 2024 ). Given the context, moral learning is of particular concern. Moral emotions serve as powerful motivators for social behaviour and norm enforcement (Haidt, 2003 ), making their potential amplification or distortion through social media algorithms especially consequential. Naturally the question arises: are users regulating moral judgments differently between social media and offline environments? In other words, is the social media context altering the way individuals experience and carry out moral judgments? Building on these concerns our work investigates whether presenting moral dilemmas on Instagram versus a traditional platform like Qualtrics influences participants’ moral judgments. According to the dual process theory of moral judgments (Greene et al., 2001 ; Greene et al., 2004 ), moral decisions are the product of either fast automatic processes, relying on intuitive and effective responses, or of slow effortful processes, driven by the evaluation of the potential outcomes. It is well known that contextual factors play a crucial role in determining which process dominates (Greene et al., 2001 ; Greene et al., 2004 ; Costa et al., 2014 ; Cao et al., 2017 ; Pilcher & Smith, 2024 ), including digital contexts. For example, Barque-Duran et al. (2017) showed that the manipulation of the digital context in which moral dilemmas are presented (PC vs Smartphone) affects the outcome, resulting in an increase in utilitarian (vs. deontological) responses when using smartphones. In line with this, we investigate whether similar effects can be found by manipulating the smartphone platform through which dilemmas are presented (Instagram vs Qualtrics). However, given the growing body of evidence on how social media shape emotional engagement and perceived extremity, we deemed of primary importance to move beyond the traditional focus on utilitarian-deontological choices. To capture a fuller picture of modal cognition in digital settings, we integrated affective and moral ratings, enabling us to assess how the Instagram environment modulates not only moral outcomes but also the felt and perceived dimensions of moral judgement. Unlike previous studies that have focused on the post-use effects of social media (Korte, 2020 ; Kross et al., 2021 ; Hancock et al., 2022 ; Valkenburg, 2022 ; Verduyn et al., 2022 ), in this study, we took a different perspective, focusing on the effects during use, with the aim of contributing to a more complete understanding of the impact of these platforms on human moral judgment. 2. Materials and Methods 2.1 Participants A total of 70 participants were recruited at the University of Bologna. Our sample is made up of 30 males and 40 females, with a mean age of M = 22.44, SD = 2.0 (Men: M = 23.19, SD = 2.71, women: M = 21.87, SD = 1.08). No compensation was provided to participants. Informed consent was obtained from all participants, and the study was approved by the local ethics committee of the University of Bologna (protocol number 0390134). All experiments were conducted in accordance with relevant ethical guidelines and regulations. 2.2 Instruments Moral Dilemmas We used the moral dilemmas validated by Lotto et al., ( 2014 ) that are divided into three main categories and two subcategories. The three main categories are: incidental, instrumental and filler dilemmas. Incidental dilemmas (trolley-like dilemmas, Foot 1967) are those where the death of one person is a foreseen but unintended consequence of the action aimed at saving more people. Instrumental dilemmas (Footbridge-like dilemmas, Thomson 1984) are those in which the death of one person is a means to save more people. Filler dilemmas are those which describe moral issues such as stealing, lying and being dishonest and never involving killing. Of the incidental and instrumental types, there are two subtypes: those with self-involvement and those without self-involvement. The difference is that the decision to kill one individual contributes to saving one’s own and other people's lives in the self-involvement dilemmas, or just other people's lives in the other-involvement dilemmas. In our study, we used a total of 12 moral dilemmas, divided into 4 incidental dilemmas (two with self-involvement and two without), 4 instrumental dilemmas (two with self-involvement and two without) and 4 filler dilemmas. Each dilemma is accompanied by questions regarding moral decision (yes or no), moral acceptability (7-point Likert scale), valence and arousal (9-point Likert scale). Standardized Questionnaires To assess participants’ social media behaviour across different dimensions, we used three questionnaires: the Social Media Mindset Scale (SMMS) (Lee & Hancock, 2024 ), the social media use scale (SMUS) (Tuck & Thompson, 2024 ), and the Social Media Activity Questionnaire (SMAQ) (Ozimek et al., 2023 ). The Social Media Mindset Scale (SMMS) (Lee & Hancock, 2024 ) is designed to measure individuals’ core beliefs about social media’s role in their lives, focusing specifically on the dimensions of agency and valence. Agency reflects the degree of control the individuals feel they have over their social media use (e.g., “I’m in control of how I use social media”). Valence captures whether individuals perceive social media’s effects as positively enhancing relationships, learning, and enjoyment or negatively wasting time or manipulation (e.g., “Using social media is a waste of time for me”). The SMMS test consists of 12 items based on a 5-point Likert scale (from 1:Strongly disagree to 5:Strongly Agree). The Social Media Use Scale (SMUS) (Tuck & Thompson, 2024 ), was developed to address limitations in existing measures of Social Media Use test (SMU), which often focus narrowly on frequency or passive vs active use, producing inconsistent results (Valkenburg, van Driel & Beyens, 2022 ). The SMUS is instead intended to prove a nuanced and empirically validated classification of SMU behaviours across multiple platforms. It consists of 17 items on a 9-point Likert scale (Never-Hourly or more). The SMUS is designed to capture the frequency of four distinct factors: image-based SMU, namely, activities related to managing one’s social image, for example, editing/deleting content, monitoring likes/comments (e.g., ‘Played with photo filtering/photo editing’). Secondly, comparison-based SMU, behaviours involve social comparison (e.g., ‘Compared my life or experiences to others’). Thirdly, belief-based SMU, expressing or engaging with contentious beliefs such as posting negative opinions, seeking morally conflicting content (e.g., ‘Commented unsupportively or disliked/’’reacted’’ unsupportively on others’ post(s). Lastly, consumption-based SMU, namely, passive content consumption such as scrolling feeds, watching videos (e.g., ‘Scrolled aimlessly through my feed(s)’). The Social Media Activity Questionnaire (SMAQ) (Ozimek et al., 2023 )was developed to assess social media use, distinguishing between active and passive behaviours, which we still decided to maintain. It consists of 18 items on a 5-point Likert Scale (never-very often), where eleven items measure passive use (e.g., ‘I look at the newsfeed to see the latest activities of other users’) and eight measure active use (e.g., ‘I post photos’). 2.3. Procedure Of the whole participant pool, half did the questionnaire and the dilemmas on Qualtrics in class and half of them on Instagram in the lab. Participants who did the experiment on Qualtrics performed it on their own smartphones. The pool that performed it on Instagram was furnished with a lab smartphone where Instagram was preinstalled. We created two private empty accounts, namely, that have no posts and follow no one but each other, with the university email of one of the experimenters. One of the accounts was not accessible to the participants and was used to post the dilemmas and the questions as Instagram stories, and to receive the participants' answers. The researchers had access to this account from their own smartphones. The second account was logged into the lab’s smartphone and was given to participants. After instructing the participants, the phone is given to them already showing the first story, that is an introduction to the experiment (Fig. 1 ). Then follows each one of the dilemmas and the corresponding questions, until all 12 dilemmas were presented. The participants had to answer the question stories with a direct message to the posting account, under the control of the experimenter. We then received the answer on our own smartphone, dialed on that account and recorded it accordingly in a spreadsheet. This method allowed us to maintain the highest similarity of stimulus presentation and answering between Qualtrics and Instagram. In both cases, the dilemmas and following questions were presented as text, and the answer was given as a text entry question, while maintaining a high grade of ecological validity in how subjects usually interact with Instagram stories. The order of presentation of the dilemmas has been randomized in both conditions. Data about duration was automatically collected in Qualtrics, while for the Instagram group a screen-recording was started before presenting them the first story and was ended at the end of the session. Afterwards, through the free version of DaVinci Resolve 19, an editing software, the experimenters placed 74 markers to identify the moments of start and end of each story. Given that the moral dilemmas were 12 and for each of them there were 5 questions, each consisting of a total of 6 stories. 12 x 6 = 72, plus one marker for the beginning of the experiment and one for the end, hence a total of 74. The markers were then exported in a CSV file and converted into a specific duration through a tailor-made code in Python. They were then assigned automatically to each of the dilemmas, thanks to the order registered and saved in a CSV file with the subject number. 3. Data Analysis and Results A power analysis (G*Power 3.1.9.7 Faul et al., 2007) was conducted for type of dilemmas, risk and group main interactions and consequent interactions. Assuming a conservative medium effect size (f = 0.25) and a correlation of 0.50 among repeated measures, using a α = 0.10 at 80% power, a total sample size of 64 participants was needed. We recruited a total of 70 to accommodate potential exclusions and drop-outs. Based on the Shapiro-Wilk tests ( p <.001) and on Levene’s test revealing a significant difference in variances across groups only for the percentage of affirmative responses [ F (7, 552) = 3.22, p = .002], we performed four robust three-way ANOVAs (Wilcox, 2012) using 10% trimmed means to examine the effects of dilemma type, group and risk on the percentage of affirmative responses, on moral acceptability, on valence and arousal ratings (Paragraph 3.1). For all variables, robust post-hoc were also performed. For filler dilemmas, only the group served as the variable of interest; we therefore performed the Wilcoxon rank-sum test (Paragraph 3.2). Shapiro-Wilk and Levene’s yielded significant results also for duration times, W = 0.681, p .0002, respectively. We therefore performed a robust three-way ANOVA (Wilcox, 2012) using 10% trimmed means and robust post-hoc (Paragraph 3.3). Finally, we tested with a robust two-way ANOVA whether the groups differed in questionnaire scores, and then performed Spearman’s correlation (Paragraph 3.4). 3.1 Moral Decision Our results show a significant main effect of dilemma type (Q = 251.37, p < .0001) and a main effect of risk (Q = 18.53, p 0.05). To provide context for the size of these effects, a classic (non-robust) ANOVA was also performed. This indicated a large effect of dilemma type [η² = .29 (partial η² = .30)], a small effect of risk [η² = .022 (partial η² = .031)] and a negligible effect for group and all the interactions. Following the initial ANOVA, we conducted robust post-hoc with 10%trimmed means (Wilcox, 2012) analysis to further explore the nature of these effects within each dilemma type. Specifically, we examined the influence of group and risk. For Incidental dilemmas neither risk nor group were significant, ( psihat = -0.018, 95% CI [–0.26, 0.22] p = 0.885 and psihat 0.23, 95% CI [–0.01, 0.47] p = 0.061), nor a significant interaction between group and risk ( psihat = -0.054, 95% CI [–0.29, 0.19] p = 0.665). In contrast, for the instrumental dilemmas, a significant main effect of risk was observed, ( psihat = 0.50, 95% CI [0.500, 0.27] p < 0.001), indicating that participants’ responses differed robustly between risk conditions. Specifically, the proportion of positive responses is higher in the involvement dilemmas than in the non-involvement ones. Neither the group effect, ( psihat = 0.179, 95% CI [–0.05, 0.40], p = 0.128), nor the group-by-risk interaction, ( psihat = -0.036, p = 0.760), reached significance in this dilemma type. 3.1.1. Moral Acceptability Moral acceptability ratings yielded a significant effect of group (Q = 26.69, p < .0001) and a main effect of dilemma type (Q = 51.11, p < .0001). No significant effects were found for Risk or the interaction terms. To provide context for the size of these effects, a classic (non-robust) ANOVA was also performed. This indicated a small effect of group η² = .040 (partial η² = .042) and a small effect of dilemma type [η² = .07 (partial η² = .074)]. For the Moral Acceptability outcome, robust post-hoc analyses with 10% trimmed means (Wilcox, 20212) were conducted separately for dilemma types, examining the effects of group, risk, and their interaction. For incidental dilemmas, a significant main effect of group was revealed ( psihat = 1.68, 95% CI [0.75, 2.61], p = .00048) with Instagram participants responding with higher moral acceptability than Qualtrics’s. There was no significant main effect of risk framing ( psihat = 0.05, 95% CI [–0.88, 0.98], p = .915). However, a significant group per risk interaction was observed ( psihat = –0.98, 95% CI [–1.91, –0.05], p = .039), suggesting that the effect of group differed depending on the risk framing condition. For instrumental dilemmas, the robust post-hoc test again revealed a significant main effect of group ( psihat = 1.96, 95% CI [1.03, 2.88], p = .00004), indicating higher moral acceptability from Instagram participants compared to Qualtrics participants. The effect of risk framing was non-significant ( psihat = 0.09, 95% CI [–0.83, 1.02], p = .840). Similarly, the group × risk interaction was not significant ( psihat = 0.02, 95% CI [–0.90, 0.94], p = .965), c.f. Figure 5 3.1.2. Valence Valence ratings indicated a significant effect of group (Q = 13.22, p < .0004). No significant effects were found for other main factors or any of the interaction terms. To provide context for the size of these effects, a classic (non-robust) ANOVA was also performed. This indicated a small effect of group [η² = .031 (partial η² = .033)]. Robust post-hoc analyses with 10% trimmed means (Wilcox, 20212) were conducted separately for dilemma types, examining the effects of group, risk, and their interaction. For the Incidental dilemmas, a significant effect of group was found ( psihat = 1.27, 95% CI [0.28, 2.25], p = .011), indicating that the Instagram group answered higher on valence than the Qualtrics group. Risk ( psihat = 0.71, 95% CI [-0.27, 1.69], p = .154), and interaction were insignificant ( psihat = -0.03, 95% CI [-1.01, 0.94], p = .944). For the instrumental dilemmas the situation was the same, with the Instagram group scoring higher than Qualtrics’s on valence ( psihat = 1.27, 95% CI [0.31, 2.22], p = .009), and without any significance for risk ( psihat = 0.44, 95% CI [-0.50, 1.39], p = .354), or risk group interactions ( psihat = -0.14, 95% CI [-1.09, 0.80], p = .762). 3.1.3. Arousal Arousal ratings analysis yielded a significant effect of group (Q = 38.1, p < .0001). No significant effects were found for other main factors or any interaction terms. To provide context for the size of these effects, a classic (non-robust) ANOVA was also performed. This indicated a medium effect of group [η² = .062 (partial η² = .063)]. Robust post-hoc analyses with 10% trimmed means (Wilcox, 20212) were conducted separately for dilemma types, examining the effects of group, risk, and their interaction. For the Incidental dilemmas, a significant effect of group was found ( psihat = -3.19, 95% CI [-4.61, -1.77], p < .0000), indicating that the Instagram group answered lower on arousal than the Qualtrics group. Risk ( psihat = 1.11, 95% CI [-0.30, 2.52], p = .124), and interaction were insignificant ( psihat = 0.39, 95% CI [-1.02, 1.81], p = .579). For the instrumental dilemmas the situation was the same, with the Instagram group scoring lower than Qualtrics’s on arousal ( psihat = -2.99, 95% CI [-4.49, -1.49], p = .0001), and without any significance for risk ( psihat = 0.19, 95% CI [-1.30, 1.69], p = .798) or risk group interactions ( psihat = 0.30, 95% CI [-1.19, 1.81], p = .688). 3.2. Filler Dilemmas For the filler items, we did not include multiple independent variables—only group membership served as the variable of interest. Since the assumption of normality was violated for all dependent variables (p0.05). However, a significant difference emerged for arousal (W = 7885, p = .015), with the Instagram group reporting lower arousal levels on average compared to the Qualtrics group (Figure 6), aligning with the overall trend observed in the other analyses. 3.3 Duration Time Shapiro-Wilk tests yielded significant results (W = 0.681, p .0002). We therefore opted for a robust three-way ANOVA (Wilcox, 2012) using 10% trimmed means to examine the effect of group and dilemma type on duration. The ANOVA yielded a main effect of dilemma type (Q = 7.00, p = .031) and a main effect of group (Q = 42.93, p = .001). The interaction between dilemma type and group was not significant (Q = 5.88, p = 0.05). We therefore conducted a post-hoc linear contrast analysis for both group and dilemma type on duration. As expected, the contrast revealed a significant difference ( psihat =1.97, 95% CI [1.39, 2.56], p < .0001), indicating that the groups differed significantly in their duration. Post hoc contrast comparing the levels of dilemma type on duration indicated the following: comparing Filler dilemmas with Incidental dilemmas yielded a marginally non-significant trend toward shorter durations in Filler compared to Incidental dilemmas ( psihat = -0.81, 95% CI [-1.66, 0.05]), with a p-value of 0.053. When comparing Filler with instrumental dilemmas, the estimated difference was again a non-significant trend ( psihat = -0.78 (95% CI [-1.63, 0.06], p = 0.053). When comparing Instrumental with Incidental, the difference was negligible ( psihat = 0.02, 95% CI [-0.93, 0.97], p = 0.96), indicating no significant difference between these dilemma types. To provide context for the size of these effects, a classic (non-robust) ANOVA was also performed. This indicated a small effect of group, [η² = .0027 (partial η² = .0027)] and a small effect of dilemma type [η² = .004 (partial η² = .004)]. The interaction between dilemma type and group was negligible (η² = .0003). 3.4. Standardized Questionnaires To rule out the possibility that observed differences in moral judgments were due to distinct patterns of Instagram use between groups, we analyzed responses across multiple validated social media questionnaires. A robust two-way ANOVA on trimmed means revealed a significant main effect of questionnaire type ( Q = 625.50, p = .001), indicating that the measures captured distinct dimensions of social media engagement. Crucially, there was no significant main effect of the group ( Q = 1.08, p = .300), suggesting that overall usage patterns did not differ meaningfully between participants. While a significant interaction was observed ( Q = 18.68, p = .014), post hoc robust comparisons (with Holm correction) found no statistically significant group differences on any individual subscale ( all adjusted p >0.05). This pattern supports the interpretation that group-related effects in the main experiment are not attributable to differences in how participants engage with social media platforms, but rather to the context in which moral decisions were presented. Correlational analysis were performed with Spearman’s correlation coefficient in three different ways. Firstly, a correlational analysis has been run between the questionnaire subsets and the answers to the dilemmas without considering either the type of dilemma or the risk associated with it. Spearman’s rank-order correlation revealed significant positive associations: - Arousal ratings were positively correlated with both image scores ( ρ = .20, p < .001, Bonferroni-adjusted p = .004) and valence scores ( ρ = .23, p < .001, Bonferroni-adjusted p < .001). -Moral acceptability judgments were positively associated with belief scores ( ρ = .23, p < .001, Bonferroni-adjusted p < .001). In other words, subjects scoring high in how much they manage their social media image (image scores) and perceiving social media effects as positive (valence scores) were also more agitated in making the decision (higher arousal). Lastly, subjects scoring higher in expressing or engaging with contentious content on social media (belief scores) reported higher moral acceptability. When we added the type of dilemma, significant positive correlations emerged between - Valence scores and I_Arousal ( ρ = .30, p < .001, Bonferroni-adjusted p = .026). - Belief scores and I_Valence ( ρ = .32, p < .001, Bonferroni-adjusted p = .014). These replicate the result above with the further specification of the Incidental dilemma. Perception of social media effects as positive correlates with more agitation in Incidental dilemmas, and expressing or engaging with contentious contents correlates with higher pleasure (valence ratings) for Incidental dilemmas. When we also added risk, no statistically significant correlations were found after applying Bonferroni correction for multiple comparisons ( all adjusted p > .05). 4. Discussion The present study examined whether presenting moral decision-making scenarios on the Instagram app affects individuals’ responses. Our hypothesis that Instagram would warp the emotions associated with moral judgments was confirmed by all three ratings, where the Instagram group reported higher moral acceptability, higher calm and lower regret. Furthermore, the Instagram group was consistently slower. Both groups showed a similar trend where filler dilemmas are the fastest, followed by instrumental and finally by incidental. We didn’t find a group difference for utilitarian/deontological decisions. Interestingly, moral acceptability yielded a significant interaction of group and risk in incidental dilemmas, and an opposite but not significant trend was found in instrumental dilemmas. In the Qualtrics group, moral acceptability decreased when the incidental dilemmas were without involvement, showing instead the opposite trend in the Instagram group (Fig 7). In the Qualtrics group, this could be explained by a simple survival instinct, for which killing to save one’s own life is considered more acceptable, therefore driving down moral acceptability in the case of lack of involvement. On the other hand, for the Instagram group, it is possible that killing one person to save others when one’s life is not at stake is perceived as a more virtuous action, hence more morally acceptable. The rationale behind these findings, we suggest, is that different platforms induce adherence to different social norms. For valence and arousal ratings, the Instagram group ranked significantly higher valence (less regret) and lower arousal (calmer) for both dilemma types. Valence followed a similar trend in both groups and dilemma types, with ratings higher for dilemmas with involvement and lower for dilemmas without involvement, signaling that subjects regret killing someone less when their own life is at stake. Furthermore, the change in mean valence (Tab 1.) from dilemmas with and without involvement is similar for both the Instagram and Qualtrics groups, suggesting a consistent pattern of emotional response. Arousal ratings followed a similar decreasing trend in the case of incidental dilemmas, where killing was less arousing (less agitating) when the participant's life was not at risk than when it was. However, the change in the mean arousal between dilemmas with and without involvement (Tab 2.) is quite different for the two groups. The Instagram group has more than half a point of difference, while the Qualtrics group has less than half. We found the same pattern for instrumental dilemmas, even though less pronounced. Furthermore, while the Instagram trend is the same across incidental and instrumental dilemmas, with dilemmas with involvement scoring higher than dilemmas without, Qualtrics’s one is the opposite for instrumental dilemmas. Dilemmas without involvement are scored as more arousing than with involvement, albeit slightly. This is in line with recent studies suggesting that Instagram's social and emotional context might enhance emotional reactivity (Ozimek et al., 2023; Yue et al., 2022) and with the aforementioned literature on enhanced extremity. Overall, our findings emphasise how Instagram enhances arousal patterns tied to personal involvement. Table 1. Descriptive statistics for valence ratings. Dilemma Type Risk Group N Mean SD Incidental With involvement Instagram 70 3.07 2.15 Incidental With involvement Qualtrics 70 2.43 1.83 Incidental Without involvement Instagram 70 2.76 2.03 Incidental Without involvement Qualtrics 70 2.16 1.74 Instrumental With involvement Instagram 70 2.93 2.27 Instrumental With involvement Qualtrics 70 2.34 1.79 Instrumental Without involvement Instagram 70 2.78 2.15 Instrumental Without involvement Qualtrics 70 2.01 1.54 Table 2 Descriptive statistics for arousal ratings. Dilemma Type Risk Group N Mean SD Incidental With involvement Instagram 70 5.65 2.85 Incidental With involvement Qualtrics 70 6.87 2.43 Incidental Without involvement Instagram 70 5.03 2.51 Incidental Without involvement Qualtrics 70 6.62 2.26 Instrumental With involvement Instagram 70 5.66 2.83 Instrumental With involvement Qualtrics 70 6.77 2.45 Instrumental Without involvement Instagram 70 5.44 2.73 Instrumental Without involvement Qualtrics 70 6.81 2.50 Turning to the temporal dimension, participants in the Instagram group consistently took longer to respond compared to those in the Qualtrics group. Despite this difference in overall response time, both groups exhibited a similar pattern, in line with the dual process theory (Greene et al., 2001; Greene et al., 2004) across dilemma types: filler dilemmas were answered the quickest, followed by instrumental dilemmas, and incidental dilemmas (Fig 6.) The longer response times observed in the Instagram group may reflect subjects being used to lingering on, idling on stories without a clear-cut willingness to reach the end of that activity. However, it might be argued that this should be true only when the content is interesting to the user. In line with this, our data (Tab 3.) shows that the Instagram group has both a longer maximum total duration as well as a shorter minimum duration. This then reflects the platform’s dual engagement styles: lingering on interesting content while rapidly skipping uninteresting content. Therefore, our data are in line with Instagram’s characteristic user behaviour of selective engagement, where attention is flexibly allocated based on personal interest. Table 3. Descriptive statistics for duration in minutes Group Instagram Instagram Instagram Instagram Qualtrics Qualtrics Qualtrics Qualtrics Dilemma Total Incidental Instrumental Filler Total Incidental Instrumental Filler Mean 14.8 15.9 15.4 13.2 13.3 14.1 13.5 12.4 Standard Deviation 4.86 5.37 4.78 4.05 3.99 4.73 4.03 2.91 Min 5.85 6.19 5.94 5.85 6.86 6.86 7.70 7.95 Max 31.2 31.2 30.5 21.8 28.5 28.5 26.1 19.1 Importantly, we didn’t find a group difference for utilitarian/deontological decisions. Usually, this difference is found when comparing two starkly different contexts, and according to the dual process theory, when one context can induce psychological distance (Costa et al., 2014) favouring utilitarian judgments (Greene et al., 2001; Greene et al., 2004). We observed a slight increase in utilitarian responses in the instrumental dilemmas when compared to Qualtrics, even though not significant. Concerning standardized questionnaires, higher scores on SMMS valence correlate with more agitation. SMMS valence is designed to measure the perception of social media effects as positive; therefore, subjects believing that social media has positive effects might have reacted with more agitation when forced to make life-or-death decisions. Overall, our results can be interpreted by considering the intrinsic characteristics of social media content. As aforementioned, morally and emotionally resonant content is the primary driver of engagement (Al-Rawi, 2019; Nontasil & Payne, 2019; Brady et al., 2020 ; Rathje et al., 2021; Pandey et al., 2023; Van Bavel, 2024), incentivizing creation of such content while shadowing moderate ones (Stöcker & Preuss, 2020; Whittaker et al., 2021; Robertson et al., 2024). In line with other work suggesting that online environments teach and reinforce behaviours (Masur et al., 2021; Cheng, 2023), we provide empirical data supporting a learning perspective. To be embedded in the Instagram context induces a more extreme mindset in the users and elicits the adherence to different social norms, enhancing moral acceptability while dampening arousal and valence. One mainstream concern about social media–driven norm warping is that behaviours learned online can seep over into everyday life. We have already seen cases where influencers, by virtue of their status, refuse to pay for services (e.g., a meal), reflecting the transfer of norms from social media into offline contexts. The danger becomes far more serious when content promoting hate or extreme actions escapes these platforms, potentially putting lives at risk. It is therefore essential to pursue further research into this phenomenon, rigorously testing both the likelihood of such seepover and the context-dependent nature of social norm adherence. 5. Conclusion In this study, we investigated whether answering moral dilemmas on different smartphones’ apps would affect the ratings of moral acceptability, valence and arousal, as well as the amount of utilitarian and deontological responses. We tested Instagram vs Qualtrics platforms, using a novel methodology by presenting stimuli directly on social media and prompting decision-making on the social platform itself. We discovered that subjects answering the dilemmas on Instagram uphold higher moral acceptability ratings, higher valence (less regret) and lower arousal (less agitation). No differences were found for utilitarian and deontological decisions. Our study provided significant empirical evidence regarding the immediate cognitive effects of embedding human moral judgment within the Instagram platform. These findings support the widespread belief that the peculiar social norms afforded by social media are shaping new ways to think and feel. Specifically, our findings point to a concerning dynamic: immersion in the Instagram environment, optimized to highlight and reward extreme content, seems to foster the normalization and reinforcement of more extreme behaviours and moral judgments. Thus, social media can be framed as a training ground to for the acceptance of extremeness foregoing moderate content, even more so when extremeness is concealed beneath content shaped by the users’ individual preferences. The implications of this research raise questions of social and ethical relevance, suggesting the need for greater user awareness of the unintended cognitive effects of social media use. Declarations Conflict of interest The authors have no financial or proprietary interests in any material discussed in this article. Funding Nicola Chinchella is supported by the European Union – NextGenerationEU through the Italian Ministry of University and Research under PNRR – Mission 4 – Component 2 – Investment 3.1 “Fund for the realization of an integrated research and innovation infrastructure system” D.M. 118/2023 CUP B83C22003950001 Data availability statement Data can be found at: https://osf.io/jfyc6 Author contribution Conceptualization: N.C., C.L.; methodology: N.C., C.L.; data collection: N.C., A.G; data analysis: N.C.; writing original draft preparation, N.C.; writing review and editing, N.C., C.L., A.G.; supervision, C.L., A.G. All authors have read and agreed to the published version of the manuscript. References Acerbi, A. (2019). Cultural Evolution in the Digital Age . Oxford University Press. https://doi.org/10.1093/oso/9780198835943.001.0001 Al-Rawi, A. (2019). Viral News on Social Media. Digital Journalism , 7 (1), 63–79. https://doi.org/10.1080/21670811.2017.1387062 Barque-Duran, A., Pothos, E. M., Yearsley, J. M., & Hampton, J. A. (s.d.). The impact of the Digital Age in Moral Judgments . Berger, J., & Milkman, K. L. (2012). What Makes Online Content Viral? Journal of Marketing Research , 49 (2), 192–205. https://doi.org/10.1509/jmr.10.0353 Brady WJ, Crockett MJ, Van Bavel JJ. (2020a). The MAD model of moral contagion: the role of motivation, attention, and design in the spread of moralized content online. Perspectives on psychological science : a journal of the Association for Psychological Science , 15 (4), 978–1010. https://doi.org/10.1177/1745691620917336 Brady WJ, Gantman AP, Van Bavel JJ.(2020b). Attentional capture helps explain why moral and emotional content go viral. J. Exp. Psychol. Gen. Brady, W. J., & Crockett, M. J. (2024). Norm Psychology in the Digital Age: How Social Media Shapes the Cultural Evolution of Normativity. Perspectives on Psychological Science , 19 (1), 62–64. https://doi.org/10.1177/17456916231187395 Brady, W. J., McLoughlin, K., Doan, T. N., & Crockett, M. J. (2021). How social learning amplifies moral outrage expression in online social networks. Science Advances , 7 (33), eabe5641. https://doi.org/10.1126/sciadv.abe5641 Bruineberg, J. (2023). Adversarial inference: Predictive minds in the attention economy. Neuroscience of Consciousness , 2023 (1), niad019. https://doi.org/10.1093/nc/niad019 Cao, F., Zhang, J., Song, L., Wang, S., Miao, D., & Peng, J. (2017). Framing Effect in the Trolley Problem and Footbridge Dilemma: Number of Saved Lives Matters. Psychological Reports , 120 (1), 88-101. https://doi.org/10.1177/0033294116685866 (Original work published 2017) Carpenter, C. J., & Amaravadi, C. S. (2019). A big data approach to assessing the impact of social norms: Reporting one's exercise to a social media audience. Communication Research , 46 (2), 236–249. https://doi.org/10.1177/0093650216657776 Chung, A., & Rimal, R.N. (2016). Social Norms: A Review. Review of Communication Research, 4, 1-28. https://doi.org/10.12840/issn.2255-4165.2016.04.01.008 Chung, M. (2019). The message influences me more than others: How and why social media metrics affect first person perception and behavioral intentions. Computers in Human Behavior , 91 , 271–278. https://doi.org/10.1016/j.chb.2018.10.011 Costa A, Foucart A, Hayakawa S, Aparici M, Apesteguia J, Heafner J, et al. (2014) Your Morals Depend on Language. PLoS ONE 9(4): e94842. https://doi.org/10.1371/journal.pone.009484 DigitalNRG (2025)https://www.digitalnrg.co.uk/top-social-media-statistic-insights-for-2025/#elementor-toc__heading-anchor-4 Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, 175–191. Fisher, M. (2022). The chaos machine: The inside story of how social media rewired our minds and our world (Hardcover ed.). Little, Brown and Company Greene JD, Nystrom LE, Engell AD, Darley JM, Cohen JD (2004). The neural bases of cognitive conflict and control in moral judgment. Neuron . 44 (2): 389–400. Greene JD, Sommerville RB, Nystrom LE, Darley JM, Cohen JD (2001). An fMRI investigation of emotional engagement in moral judgment. Science . 293 (5537): 2105–8. Haidt, J. (2003). Elevation and the positive psychology of morality. In C. L. M. Keyes & J. Haidt (Eds.), Flourishing: Positive psychology and the life well-lived (pp. 275–289). American Psychological Association. https://doi.org/10.1037/10594-012 Hancock, J., Liu, S. X., Luo, M., & Mieczkowski, H. (2022). Psychological Well-Being and Social Media Use: A Meta-Analysis of Associations between Social Media Use and Depression, Anxiety, Loneliness, Eudaimonic, Hedonic and Social Well-Being. SSRN Electronic Journal . https://doi.org/10.2139/ssrn.4053961 Konovalova, E., Mens, G. L., & Schöll, N. (2023). Social media feedback and extreme opinion expression. PLOS ONE , 18 (11), e0293805. https://doi.org/10.1371/journal.pone.0293805 Korte, M. (2020). The impact of the digital revolution on human brain and behavior: Where do we stand? Dialogues in Clinical Neuroscience , 22 (2), 101–111. https://doi.org/10.31887/DCNS.2020.22.2/mkorte Kross, E., Verduyn, P., Sheppes, G., Costello, C. K., Jonides, J., & Ybarra, O. (2021). Social Media and Well-Being: Pitfalls, Progress, and Next Steps. Trends in Cognitive Sciences , 25 (1), 55–66. https://doi.org/10.1016/j.tics.2020.10.005 Lee, A. Y., & Hancock, J. T. (2024). Social Media Mindsets: A New Approach to Understanding Social Media Use & Psychological Well-Being . https://doi.org/10.1093/jcmc/zmad048 Lindström, B., Bellander, M., Schultner, D. T., Chang, A., Tobler, P. N., & Amodio, D. M. (2021). A computational reward learning account of social media engagement. Nature Communications , 12 , 1311. https://doi.org/10.1038/s41467-020-19607-x Lotto, L., Manfrinati, A., & Sarlo, M. (2014). A New Set of Moral Dilemmas: Norms for Moral Acceptability, Decision Times, and Emotional Salience. Journal of Behavioral Decision Making , 27 (1), 57–65. https://doi.org/10.1002/bdm.1782 Lupinacci, L. (2021). ‘Absentmindedly scrolling through nothing’: Liveness and compulsory continuous connectedness in social media. Media, Culture & Society , 43 (2), 273–290. https://doi.org/10.1177/0163443720939454 Mackay D. (2023). Infinite Scrolling, Dissociation, and Boredom Spiraling as the Drivers of Habitual Social Media Use. Ph. D. Dissertation. Southern Connecticut State University. Marino, C., Bersia, M., Furstova, J., Galeotti, T., van den Eijnden, R. J. J. M., Boniel-Nissim, M., Pickett, W., Lenzi, M., Canale, N., Eriksson, C., Lahti, H., Ozolina, K., Craig, W., & Vieno, A. (2025). Global change in adolescent social media use (2018–2022): An ecological analysis across 28 countries. Computers in Human Behavior , 173 , 108789. https://doi.org/10.1016/j.chb.2025.108789 Masur, P. K., DiFranzo, D., & Bazarova, N. N. (2021). Behavioral contagion on social media: Effects of social norms, design interventions, and critical media literacy on self-disclosure. PLoS ONE , 16 (7), e0254670. https://doi.org/10.1371/journal.pone.0254670 Milli, S., Carroll, M., Wang, Y., Pandey, S., Zhao, S., & Dragan, A. D. (2025). Engagement, user satisfaction, and the amplification of divisive content on social media. PNAS Nexus , 4 (3), pgaf062. https://doi.org/10.1093/pnasnexus/pgaf062 Nontasil, P., & Payne, S. J. (2019). Emotional Utility and Recall of the Facebook News Feed. Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems , 1–9. https://doi.org/10.1145/3290605.3300252 Ozimek, P., Brandenberg, G., Rohmann, E., & Bierhoff, H.-W. (2023). The Impact of Social Comparisons More Related to Ability vs. More Related to Opinion on Well-Being: An Instagram Study. Behavioral Sciences , 13 (10), 850. https://doi.org/10.3390/bs13100850 Pan, X., Hou, Y., & Wang, Q. (2023). Are we braver in cyberspace? Social media anonymity enhances moral courage. Computers in Human Behavior , 148 , 107880. https://doi.org/10.1016/j.chb.2023.107880 Pandey, S., Cao, Y., Dong, Y., Kim, M., MacLaren, N. G., Dionne, S. D., Yammarino, F. J., & Sayama, H. (2023). Generation and influence of eccentric ideas on social networks. Scientific Reports , 13 (1), 20433. https://doi.org/10.1038/s41598-023-47823-0 Pastor, Y., Pérez-Torres, V., Thomas-Currás, H., Lobato-Rincón, L. L., López-Sáez, M. Á., & García, A. (2024). A study of the influence of altruism, social responsibility, reciprocity, and the subjective norm on online prosocial behavior in adolescence. Computers in Human Behavior , 154 , 108156. https://doi.org/10.1016/j.chb.2024.108156 Pastor, Y., Pérez-Torres, V., Thomas-Currás, H., Lobato-Rincón, L. L., López-Sáez, M. Á., & García, A. (2024). A study of the influence of altruism, social responsibility, reciprocity, and the subjective norm on online prosocial behavior in adolescence. Computers in Human Behavior , 154 , 108156. https://doi.org/10.1016/j.chb.2024.108156 Pilcher JJ, Smith PD (2024). Social context during moral decision-making impacts males more than females. Front Psychol. doi: 10.3389/fpsyg.2024.1397069. PMID: 38836238; PMCID: PMC11148431. Rathje S, Robertson C, Brady WJ, Van Bavel JJ. (2024) People Think That Social Media Platforms Do (but Should Not) Amplify Divisive Content. Perspect Psychol Sci.. doi: 10.1177/17456916231190392. Rathje, S., Van Bavel, J. J., & van der Linden, S. (2021). Out-group animosity drives engagement on social media. Proceedings of the National Academy of Sciences , 118 (26), e2024292118. https://doi.org/10.1073/pnas.2024292118 Rimal, R. N., & Lapinski, M. K. (2015). A Re-Explication of Social Norms, Ten Years Later. Communication Theory , 25 (4), 393-409. https://doi.org/10.1111/comt.12080 Rixen, J. O., Meinhardt, L.-M., Glöckler, M., Ziegenbein, M.-L., Schlothauer, A., Colley, M., Rukzio, E., & Gugenheimer, J. (2023). The Loop and Reasons to Break It: Investigating Infinite Scrolling Behaviour in Social Media Applications and Reasons to Stop. Proc. ACM Hum.-Comput. Interact. , 7 (MHCI), 228:1-228:22. https://doi.org/10.1145/3604275 Robertson, C. E., Del Rosario, K. S., & Van Bavel, J. J. (2024). Inside the funhouse mirror factory: How social media distorts perceptions of norms. Current Opinion in Psychology , 60 , 101918. https://doi.org/10.1016/j.copsyc.2024.101918 Saternus, Z., Mihale-Wilson, C. & Hinz, O. (2024). Influencer marketing on Instagram—The optimal disclosure strategy from influencers’ and marketers’ perspectives. Electron Markets https://doi.org/10.1007/s12525-024-00743-x Savolainen, I., Oksanen, A., Kaakinen, M., Sirola, A., Zych, I., & Paek, H.-J. (2021). The role of online group norms and social identity in youth problem gambling. Computers in Human Behavior , 122 , 106828. https://doi.org/10.1016/j.chb.2021.106828 Shareef, M. A., Kapoor, K. K., Mukerji, B., Dwivedi, R., & Dwivedi, Y. K. (2020). Group behavior in social media: Antecedents of initial trust formation. Computers in Human Behavior , 105 , 106225. https://doi.org/10.1016/j.chb.2019.106225 Statista. 2025. Number of social media users worldwide from 2017 to 2027. Statista. https://www.statista.com/statistics/278414/number-of-worldwide-social-network-users Stöcker, C., & Preuss, M. (2020). Riding the Wave of Misclassification: How We End up with Extreme YouTube Content. In G. Meiselwitz (A c. Di), Social Computing and Social Media. Design, Ethics, User Behavior, and Social Network Analysis (pp. 359–375). Springer International Publishing. https://doi.org/10.1007/978-3-030-49570-1_25 Tran, J. A., Yang, K. S., Davis, K., & Hiniker, A. (2019). Modeling the Engagement-Disengagement Cycle of Compulsive Phone Use. Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems , 1–14. https://doi.org/10.1145/3290605.3300542 Tuck, A. B., & Thompson, R. J. (2024). The Social Media Use Scale: Development and Validation. Assessment , 31 (3), 617–636. https://doi.org/10.1177/10731911231173080 Valkenburg, P. M. (2022). Social media use and well-being: What we know and what we need to know. Current Opinion in Psychology , 45 , 101294. https://doi.org/10.1016/j.copsyc.2021.12.006 Valkenburg, P. M., van Driel, I. I., & Beyens, I. (2022). The associations of active and passive social media use with well-being: A critical scoping review. New Media & Society , 24 (2), 530–549. https://doi.org/10.1177/14614448211065425 Van Bavel JJ, Robertson CE, Del Rosario K, Rasmussen J, Rathje S. (2024) Social Media and Morality. Annu Rev Psychol. doi: 10.1146/annurev-psych-022123-110258. Verduyn, P., Gugushvili, N., & Kross, E. (2022). Do Social Networking Sites Influence Well-Being? The Extended Active-Passive Model. Current Directions in Psychological Science , 31 (1), 62–68. https://doi.org/10.1177/09637214211053637 Whittaker, J., Looney, S., Reed, A., & Votta, F. (2021). Recommender systems and the amplification of extremist content. Internet Policy Review, 10(2). https://doi.org/10.14763/2021.2.1565 Wilcox, R. R. (2012). Introduction to Robust Estimation and Hypothesis Testing (3rd ed.). Academic Press. Yue, Z., Zhang, R., & Xiao, J. (2022). Passive social media use and psychological well-being during the COVID-19 pandemic: The role of social comparison and emotion regulation. Computers in human behavior , 127 , 107050. https://doi.org/10.1016/j.chb.2021.107050 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8328737","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":570423891,"identity":"e26669e6-91c8-4f80-922a-023e7013f037","order_by":0,"name":"Nicola Chinchella","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBUlEQVRIie2Pv2rDMBCHDwTycqnXAwX7FWQMTUvzMDJdM3jskMFgyNbOeYxOgW4yAk8OeYAEYlPIXAhkKqFqU9PFdtYO+uDgpz8fdwfgcPxHmK3almQ/+W4MXq7tUV+e+hT1pxAClupX6XNaBVqFZt9ZQ1+biceaWs0hnHij4pg+EfrZ7CTSdBeA5+su5T7nsVQlRG/5zaNYVoSk1yuxlIe4bzBpkJPioGy4ZaOFHax4XgmUJskGlXOrnAlDg4frSrJolYxQlsivKHaX5IWiV4OxwJIwqnj8gHYXzpjsVDamqT9O01BuquiI82kQ7N+bLX7uAt8v6s42F6jjjg/8dzgcDscwX/hcT1iCoQ7jAAAAAElFTkSuQmCC","orcid":"","institution":"University of Bologna","correspondingAuthor":true,"prefix":"","firstName":"Nicola","middleName":"","lastName":"Chinchella","suffix":""},{"id":570423893,"identity":"59070e67-8fcf-47fa-9271-957fa369937d","order_by":1,"name":"Aldo Gangemi","email":"","orcid":"","institution":"University of Bologna","correspondingAuthor":false,"prefix":"","firstName":"Aldo","middleName":"","lastName":"Gangemi","suffix":""},{"id":570423896,"identity":"7897f26e-9362-4464-90b2-f8c60542845b","order_by":2,"name":"Alessia Gazineo","email":"","orcid":"","institution":"University of Bologna","correspondingAuthor":false,"prefix":"","firstName":"Alessia","middleName":"","lastName":"Gazineo","suffix":""},{"id":570423899,"identity":"9481d17e-5373-40b8-b108-7e2fba15496b","order_by":3,"name":"Chiara Lucifora","email":"","orcid":"","institution":"Link Campus University","correspondingAuthor":false,"prefix":"","firstName":"Chiara","middleName":"","lastName":"Lucifora","suffix":""}],"badges":[],"createdAt":"2025-12-10 15:46:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8328737/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8328737/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":99858570,"identity":"f97f0842-5d70-4871-9dd9-5d3c3f717bdb","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"jpg","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":15985,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1ExperimentalTimeline.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/9e127c1a8723b132726f76de.jpg"},{"id":100357711,"identity":"ccc2d5ff-4556-465c-91b1-d1fa3339c0d8","added_by":"auto","created_at":"2026-01-16 07:20:14","extension":"jpg","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":27273,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2Percentageofutilitariandecisionforthetwodilemmatypesandr.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/a0a74b349ac1b40483c468fc.jpg"},{"id":100357488,"identity":"bd3e80ca-828c-4526-ab40-78e47cb6cd9d","added_by":"auto","created_at":"2026-01-16 07:19:55","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3305891,"visible":true,"origin":"","legend":"","description":"","filename":"MoralDesensitizationinDigitalContextsInstagramexposurealtersmoralperceptionandemotionalreactivity.docx","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/0f3d345b58f03ebec3b82c6a.docx"},{"id":100357643,"identity":"876f9afe-4263-42d9-8891-5674fd496423","added_by":"auto","created_at":"2026-01-16 07:20:08","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9392,"visible":true,"origin":"","legend":"","description":"","filename":"Table1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/72c5de3a6d7d80fe936cbcc7.xlsx"},{"id":99858575,"identity":"0890fdb8-ff60-4885-8e7e-5a087faab2fd","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9402,"visible":true,"origin":"","legend":"","description":"","filename":"Table2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/7ecc9a96d313b44aa8f41aef.xlsx"},{"id":99858579,"identity":"4ca0cae7-7849-402c-9743-13fbf8c30181","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"jpg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":16872,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3Moralacceptabilityratingsbygroupforincidentalandinstrument.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/2175258fe5ed829e9b78da23.jpg"},{"id":99858589,"identity":"1aa4ab75-915e-4069-b07b-89c778aa1f81","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9489,"visible":true,"origin":"","legend":"","description":"","filename":"Table3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/1906fb05f8f48ac0cef02adc.xlsx"},{"id":100357685,"identity":"69ac1ffd-7c20-442c-a6b8-190efa1400f1","added_by":"auto","created_at":"2026-01-16 07:20:10","extension":"jpg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17054,"visible":true,"origin":"","legend":"","description":"","filename":"Figure4Valenceratingsbygroupforincidentalandinstrumentaldilemmas.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/b3a073e71b512c5f12c1760b.jpg"},{"id":99858581,"identity":"a57b2aa2-1852-44c9-9ee9-b29cb83b911b","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"jpg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17706,"visible":true,"origin":"","legend":"","description":"","filename":"Figure5Arousalratingsbygroupforincidentalandinstrumentaldilemmas.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/6903a27ec329204269375b8a.jpg"},{"id":100357276,"identity":"5ae32ce5-03e4-4f21-9fc4-8bee0013f50f","added_by":"auto","created_at":"2026-01-16 07:19:33","extension":"jpg","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":21633,"visible":true,"origin":"","legend":"","description":"","filename":"Figure6Meandurationtimesinminutesfordilemmatypesandgroup.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/42460323c4fc3dbff62d067e.jpg"},{"id":99858585,"identity":"98b42c78-4c56-4caa-b337-5005ad46b16f","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"jpg","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":21261,"visible":true,"origin":"","legend":"","description":"","filename":"Figure7Meanmoralacceptabilityratingsforgroupdilemmatypeandrisk.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/d50c876a709e435ec7107b07.jpg"},{"id":100357269,"identity":"694e5e8d-254c-4e0e-a61a-39fe9f19ae3a","added_by":"auto","created_at":"2026-01-16 07:19:33","extension":"json","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5324,"visible":true,"origin":"","legend":"","description":"","filename":"6724f62a63e84e2b9ab297bffc64aa2f.json","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/a70b2e91b9ddeb6c2666c4e1.json"},{"id":99858612,"identity":"8af30d95-ba31-4852-8333-4212bddc2567","added_by":"auto","created_at":"2026-01-09 06:34:25","extension":"xml","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":153881,"visible":true,"origin":"","legend":"","description":"","filename":"6724f62a63e84e2b9ab297bffc64aa2f1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/67a83d7cae1629d3336b42ce.xml"},{"id":99858583,"identity":"fb505b3a-1ffc-4d11-9b19-896a4d91069b","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"jpg","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":15985,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1ExperimentalTimeline.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/f41f29a796ac8630f5137eb1.jpg"},{"id":100357550,"identity":"5079a49c-8e3a-4d7a-8bf6-7fc6eda3d696","added_by":"auto","created_at":"2026-01-16 07:20:02","extension":"jpg","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":27273,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2Percentageofutilitariandecisionforthetwodilemmatypesandr.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/073db17ba79dee895cc8a08d.jpg"},{"id":99858595,"identity":"699a04f6-925c-4927-bcf4-41efb9c1abe9","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"jpg","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":16872,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3Moralacceptabilityratingsbygroupforincidentalandinstrument.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/4733476527a84f091a6c9638.jpg"},{"id":99858587,"identity":"95b1e12a-6c18-4b14-aa25-a8d118176de5","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"jpg","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17054,"visible":true,"origin":"","legend":"","description":"","filename":"Figure4Valenceratingsbygroupforincidentalandinstrumentaldilemmas.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/65d2a1ec8ff6edf177ce8fbc.jpg"},{"id":100357675,"identity":"65228b17-15a5-4d35-b68a-d65bcf237d1c","added_by":"auto","created_at":"2026-01-16 07:20:10","extension":"jpg","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17706,"visible":true,"origin":"","legend":"","description":"","filename":"Figure5Arousalratingsbygroupforincidentalandinstrumentaldilemmas.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/fc33377d5bbc7b195d8b7501.jpg"},{"id":100357788,"identity":"c3dfc492-7869-4ac4-bfb4-58341869ca36","added_by":"auto","created_at":"2026-01-16 07:20:19","extension":"jpg","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":21633,"visible":true,"origin":"","legend":"","description":"","filename":"Figure6Meandurationtimesinminutesfordilemmatypesandgroup.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/d5da222daedb6ca8b90a3c4c.jpg"},{"id":100357812,"identity":"4adc6b3a-fabe-467c-b08c-3eeb3c3d0459","added_by":"auto","created_at":"2026-01-16 07:20:22","extension":"jpg","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":21261,"visible":true,"origin":"","legend":"","description":"","filename":"Figure7Meanmoralacceptabilityratingsforgroupdilemmatypeandrisk.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/c6afefc987cbcbf33bd3fc53.jpg"},{"id":100357505,"identity":"e5de3e49-90bc-4dc5-82d2-9bef9dba99d2","added_by":"auto","created_at":"2026-01-16 07:19:56","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12950,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/c3cdaf71446fdeb457385542.png"},{"id":100357794,"identity":"aaeed45b-2938-4d7f-ace3-75aabb93efc9","added_by":"auto","created_at":"2026-01-16 07:20:20","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8176,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/c759d64413afff9ab3ab0a68.png"},{"id":99858592,"identity":"da06827e-1ea4-471f-968e-74021518fb9b","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5850,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/3781c5e85980cb2a769d1cfe.png"},{"id":100357414,"identity":"6a74c3b7-562f-403b-95b8-849969a63d84","added_by":"auto","created_at":"2026-01-16 07:19:51","extension":"png","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5756,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/aa285912a9b063cc5c4337dc.png"},{"id":100357311,"identity":"7daa4e6e-ce09-4d6b-9483-5bbbccbf4660","added_by":"auto","created_at":"2026-01-16 07:19:38","extension":"png","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5698,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/d07632cc3241ceb28ec13a2a.png"},{"id":100357614,"identity":"d345650e-e34e-42d7-bc6a-2e75d4a86a5a","added_by":"auto","created_at":"2026-01-16 07:20:06","extension":"png","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9000,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/ee09d2f331b4c86576f11eb1.png"},{"id":99858610,"identity":"f7468cc5-fbe9-4742-b1b6-8e0b3d49c9cb","added_by":"auto","created_at":"2026-01-09 06:34:25","extension":"png","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8152,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/4fc7f67211cd0afc5e67323d.png"},{"id":99858591,"identity":"0e54f9a7-b3e2-4f36-91e6-a7cfddab858c","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"png","order_by":27,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":16217,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure1ExperimentalTimeline.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/b5c7f4fe2700992210a5ff3b.png"},{"id":100358045,"identity":"8398bae3-a17a-4773-b4d9-ba4d00444879","added_by":"auto","created_at":"2026-01-16 07:20:36","extension":"png","order_by":28,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13793,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure2Percentageofutilitariandecisionforthetwodilemmatypesandr.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/6ae0a724e855e024d3adc6b3.png"},{"id":99858588,"identity":"0356e9d2-cfc6-4609-a872-27b999ef1d59","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"png","order_by":29,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8444,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure3Moralacceptabilityratingsbygroupforincidentalandinstrument.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/29c1d1c49684abf68e24cc00.png"},{"id":99858590,"identity":"d1220be0-1f9d-4c63-8a74-1686aaf8231f","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"png","order_by":30,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7848,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure4Valenceratingsbygroupforincidentalandinstrumentaldilemmas.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/ddae8a32d6284473eda82b1c.png"},{"id":99858596,"identity":"75619a0f-1491-4d85-9e3c-606e6154a9c6","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"png","order_by":31,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8072,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure5Arousalratingsbygroupforincidentalandinstrumentaldilemmas.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/06e7d024a8529a6a1ee7bb6c.png"},{"id":99858598,"identity":"de6ee348-ec74-4d17-a09a-a3da1965c773","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"png","order_by":32,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":15744,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure6Meandurationtimesinminutesfordilemmatypesandgroup.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/7a2f6afbc1d029ad005a5b31.png"},{"id":99858616,"identity":"981a99e5-632d-4c2f-9798-fd2235c5dd7d","added_by":"auto","created_at":"2026-01-09 06:34:25","extension":"png","order_by":33,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13526,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure7Meanmoralacceptabilityratingsforgroupdilemmatypeandrisk.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/cfa3f6a6d96d1e1a23e7f697.png"},{"id":99858602,"identity":"48b0f878-7e41-4e0e-8fa3-7308abe6ffa8","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"png","order_by":34,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12295,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/eda6afd002f8fcbdaaa1b05e.png"},{"id":99858600,"identity":"3b78a9cd-0043-46a1-b33a-348cc81f9977","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"png","order_by":35,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6866,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/f0a32149c7b336f3b41aacdb.png"},{"id":100357209,"identity":"7bd308da-7465-4593-9480-42dfcb6f672d","added_by":"auto","created_at":"2026-01-16 07:19:19","extension":"png","order_by":36,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6056,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/913697d1fad3e95c0768d184.png"},{"id":100357773,"identity":"dcbd1736-97d4-415b-b463-5b7641bd63f1","added_by":"auto","created_at":"2026-01-16 07:20:18","extension":"png","order_by":37,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5775,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/9ffbe50a985ca593029cd6d1.png"},{"id":99858603,"identity":"419fe955-050f-4143-97b6-adfc87316d27","added_by":"auto","created_at":"2026-01-09 06:34:25","extension":"png","order_by":38,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5801,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/b03dda2e27092ed800aa3955.png"},{"id":99858607,"identity":"4476c853-5b4d-48f1-86b4-4e74aed3624c","added_by":"auto","created_at":"2026-01-09 06:34:25","extension":"png","order_by":39,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8963,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/313aae3a18dce58f7e50486c.png"},{"id":100357637,"identity":"b9524583-9998-4fe6-be7b-df091e61ac9b","added_by":"auto","created_at":"2026-01-16 07:20:07","extension":"png","order_by":40,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7291,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/e4710df8e57e5713f46052e1.png"},{"id":100357170,"identity":"f5c791bf-6165-42a5-9986-e25753bd8358","added_by":"auto","created_at":"2026-01-16 07:19:11","extension":"xml","order_by":41,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":150615,"visible":true,"origin":"","legend":"","description":"","filename":"6724f62a63e84e2b9ab297bffc64aa2f1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/a7d4262c5019a4ffd4c73444.xml"},{"id":99858615,"identity":"9b092dbd-ca25-44a2-b594-9f45a2708c19","added_by":"auto","created_at":"2026-01-09 06:34:25","extension":"html","order_by":42,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":166180,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/4e62a816b2865ac2141c865b.html"},{"id":99858568,"identity":"acad74e3-0062-4941-a188-e400dc267e28","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":15985,"visible":true,"origin":"","legend":"\u003cp\u003eExperimental Timeline\u003c/p\u003e","description":"","filename":"Figure1ExperimentalTimeline.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/4e3a0a56721717cc30d5297b.jpg"},{"id":99858572,"identity":"785ff352-8ff4-4e82-8b86-e49c07afeec9","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":27273,"visible":true,"origin":"","legend":"\u003cp\u003ePercentage of utilitarian decision for the two dilemma types and risks\u003c/p\u003e","description":"","filename":"Figure2Percentageofutilitariandecisionforthetwodilemmatypesandr.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/295e4f7b7c78ad6be049bebd.jpg"},{"id":100357534,"identity":"75732a1a-a4de-4c8a-8393-6b6fe2f6dcf3","added_by":"auto","created_at":"2026-01-16 07:20:00","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":16872,"visible":true,"origin":"","legend":"\u003cp\u003eMoral acceptability ratings by group for incidental and instrumental dilemmas.\u003c/p\u003e","description":"","filename":"Figure3Moralacceptabilityratingsbygroupforincidentalandinstrument.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/2fa753d4e12ae0379ba3770e.jpg"},{"id":100357422,"identity":"7c17b8f3-3982-452c-b0ec-ce151527ffee","added_by":"auto","created_at":"2026-01-16 07:19:52","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":17054,"visible":true,"origin":"","legend":"\u003cp\u003eValence ratings by group for incidental and instrumental dilemmas.\u003c/p\u003e","description":"","filename":"Figure4Valenceratingsbygroupforincidentalandinstrumentaldilemmas.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/4baf084a829cc1e5df4a630f.jpg"},{"id":100357491,"identity":"2f533f8a-e1f2-43e6-bf30-770dc3945cbd","added_by":"auto","created_at":"2026-01-16 07:19:55","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":17706,"visible":true,"origin":"","legend":"\u003cp\u003eArousal ratings by group for incidental and instrumental dilemmas.\u003c/p\u003e","description":"","filename":"Figure5Arousalratingsbygroupforincidentalandinstrumentaldilemmas.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/2066584ef52d5818278f87be.jpg"},{"id":100357191,"identity":"18a52813-f59b-4183-8bfa-4c4d0cedd492","added_by":"auto","created_at":"2026-01-16 07:19:14","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":21633,"visible":true,"origin":"","legend":"\u003cp\u003eMean duration times in minutes for dilemma types and group.\u003c/p\u003e","description":"","filename":"Figure6Meandurationtimesinminutesfordilemmatypesandgroup.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/9bf3ec6300b4cbd65522cb2d.jpg"},{"id":99858577,"identity":"ab9702dd-49d2-4af2-99f2-2b5f12070848","added_by":"auto","created_at":"2026-01-09 06:34:24","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":21261,"visible":true,"origin":"","legend":"\u003cp\u003eMean moral acceptability ratings for group, dilemma type and risk.\u003c/p\u003e","description":"","filename":"Figure7Meanmoralacceptabilityratingsforgroupdilemmatypeandrisk.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/e5ee7c69b72d014ac9ef798f.jpg"},{"id":103220676,"identity":"4edd3ce9-b412-4091-9bd4-78fe50176a51","added_by":"auto","created_at":"2026-02-23 10:13:19","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":907537,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8328737/v1/d092d451-e60f-4575-8674-d251ecb486b8.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Moral Desensitization in Digital Contexts: Instagram exposure alters moral perception and emotional reactivity","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eOver the past decade, the proliferation of social media platforms has significantly reshaped global communication patterns, with Facebook, Twitter, Instagram and TikTok collectively engaging billions of users. As for 2025, there are 5\u0026nbsp;billion active social media accounts around the globe, with an average use of 2 hours and 27 minutes per day (Statista \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and continuously evolving usage patterns (Marino et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Social media success can be ascribed to a mix of technological wit and human dispositions. Platforms are explicitly designed to exert a persuasive pull (Lupinacci, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Rixen et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) \u0026ldquo;pulsing with colourful notification badges, whoosh sounds and gentle vibrations\u0026rdquo; (Fisher, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e, p. 33). Furthermore, they construct detailed models of users\u0026rsquo; preferences to maximize engagement by tailoring content (St\u0026ouml;cker \u0026amp; Preuss, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Bruineberg, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). In doing so, social media have tapped into, and made increasingly visible, the human attraction to affect-laden content (Berger \u0026amp; Milkman, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2012\u003c/span\u003e), unveiling a pronounced preference for extreme material that can evoke, for example, heightened emotional arousal, moral outrage and out-group animosity (Rathje et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Pandey et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Van Bavel et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Algorithms quickly learnt that the best way to increase engagement is to leverage this human disposition, amplifying affect-laden and extreme content (St\u0026ouml;cker \u0026amp; Preuss, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Whittaker et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Konovalova et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Pandey et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Mounting evidence indicates that recommendation systems are steering users towards such types of content even when users do not seek it (Whittaker et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Pandey et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and would prefer to avoid it (Milli et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eA particularly worrisome prospect of social media enabling and enhancing environments where affect-laden content can proliferate in unpredictable ways relates to what users are learning from their continuous interactions with the platforms, especially in relation to social influence. People are influenced by the behaviours of others around them and rely on social learning to navigate the world (Rimal \u0026amp; Lapinski \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Chung \u0026amp; Rimal, 2016; Carpenter \u0026amp; Amaravadi, 2019), balancing descriptive evidence (many people do this) and validation (people approve of this). Never before were people subject to such an easily reachable and massive social influence as with social media, whose consequences are now within glimpse. Consistently with reinforcement learning models (Lindstr\u0026ouml;m et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), research shows that likes, shares and other metrics successfully predict first person perception and behaviour (Chung, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Brady et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). At the same time, social media afford environments where descriptive information is easily observable, e.g., in extreme networks, dampening the sensitivity to critical social feedback (Brady et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Both create powerful normative pressure toward behavioural adoption.\u003c/p\u003e \u003cp\u003eIn principle learning could cut both ways, nudging ethical and responsible actions (Pan et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Pastor et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) as well as forming bad habits and discouraging rationality (Shareef et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Savolainen et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). However, there is a widespread suspicion in the literature that social norms generated on social media are often more extreme than offline ones (Robertson et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), tilting the needle to the latter. Many argue that algorithm-mediated learning on social media is effectively shaping new more extreme ways for individuals to think and feel about social norms, opening interesting research venues (Acerbi, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Brady et al., 2023; Brady \u0026amp; Crockett, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Given the context, moral learning is of particular concern. Moral emotions serve as powerful motivators for social behaviour and norm enforcement (Haidt, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2003\u003c/span\u003e), making their potential amplification or distortion through social media algorithms especially consequential. Naturally the question arises: are users regulating moral judgments differently between social media and offline environments? In other words, is the social media context altering the way individuals experience and carry out moral judgments?\u003c/p\u003e \u003cp\u003eBuilding on these concerns our work investigates whether presenting moral dilemmas on Instagram versus a traditional platform like Qualtrics influences participants\u0026rsquo; moral judgments. According to the dual process theory of moral judgments (Greene et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Greene et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2004\u003c/span\u003e), moral decisions are the product of either fast automatic processes, relying on intuitive and effective responses, or of slow effortful processes, driven by the evaluation of the potential outcomes. It is well known that contextual factors play a crucial role in determining which process dominates (Greene et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Greene et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Costa et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Cao et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Pilcher \u0026amp; Smith, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), including digital contexts. For example, Barque-Duran et al. (2017) showed that the manipulation of the digital context in which moral dilemmas are presented (PC vs Smartphone) affects the outcome, resulting in an increase in utilitarian (vs. deontological) responses when using smartphones. In line with this, we investigate whether similar effects can be found by manipulating the smartphone platform through which dilemmas are presented (Instagram vs Qualtrics). However, given the growing body of evidence on how social media shape emotional engagement and perceived extremity, we deemed of primary importance to move beyond the traditional focus on utilitarian-deontological choices. To capture a fuller picture of modal cognition in digital settings, we integrated affective and moral ratings, enabling us to assess how the Instagram environment modulates not only moral outcomes but also the felt and perceived dimensions of moral judgement.\u003c/p\u003e \u003cp\u003eUnlike previous studies that have focused on the post-use effects of social media (Korte, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Kross et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Hancock et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Valkenburg, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Verduyn et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), in this study, we took a different perspective, focusing on the effects during use, with the aim of contributing to a more complete understanding of the impact of these platforms on human moral judgment.\u003c/p\u003e"},{"header":"2. Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Participants\u003c/h2\u003e \u003cp\u003eA total of 70 participants were recruited at the University of Bologna. Our sample is made up of 30 males and 40 females, with a mean age of M\u0026thinsp;=\u0026thinsp;22.44, SD\u0026thinsp;=\u0026thinsp;2.0 (Men: M\u0026thinsp;=\u0026thinsp;23.19, SD\u0026thinsp;=\u0026thinsp;2.71, women: M\u0026thinsp;=\u0026thinsp;21.87, SD\u0026thinsp;=\u0026thinsp;1.08). No compensation was provided to participants. Informed consent was obtained from all participants, and the study was approved by the local ethics committee of the University of Bologna (protocol number 0390134). All experiments were conducted in accordance with relevant ethical guidelines and regulations.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Instruments\u003c/h2\u003e \u003cp\u003e \u003cb\u003eMoral Dilemmas\u003c/b\u003e \u003c/p\u003e \u003cp\u003eWe used the moral dilemmas validated by Lotto et al., (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) that are divided into three main categories and two subcategories. The three main categories are: incidental, instrumental and filler dilemmas. Incidental dilemmas (trolley-like dilemmas, Foot 1967) are those where the death of one person is a foreseen but unintended consequence of the action aimed at saving more people. Instrumental dilemmas (Footbridge-like dilemmas, Thomson 1984) are those in which the death of one person is a means to save more people. Filler dilemmas are those which describe moral issues such as stealing, lying and being dishonest and never involving killing. Of the incidental and instrumental types, there are two subtypes: those with self-involvement and those without self-involvement. The difference is that the decision to kill one individual contributes to saving one\u0026rsquo;s own and other people's lives in the self-involvement dilemmas, or just other people's lives in the other-involvement dilemmas.\u003c/p\u003e \u003cp\u003eIn our study, we used a total of 12 moral dilemmas, divided into 4 incidental dilemmas (two with self-involvement and two without), 4 instrumental dilemmas (two with self-involvement and two without) and 4 filler dilemmas. Each dilemma is accompanied by questions regarding moral decision (yes or no), moral acceptability (7-point Likert scale), valence and arousal (9-point Likert scale).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eStandardized Questionnaires\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo assess participants\u0026rsquo; social media behaviour across different dimensions, we used three questionnaires: the Social Media Mindset Scale (SMMS) (Lee \u0026amp; Hancock, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), the social media use scale (SMUS) (Tuck \u0026amp; Thompson, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and the Social Media Activity Questionnaire (SMAQ) (Ozimek et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe Social Media Mindset Scale (SMMS) (Lee \u0026amp; Hancock, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) is designed to measure individuals\u0026rsquo; core beliefs about social media\u0026rsquo;s role in their lives, focusing specifically on the dimensions of agency and valence. Agency reflects the degree of control the individuals feel they have over their social media use (e.g., \u0026ldquo;I\u0026rsquo;m in control of how I use social media\u0026rdquo;). Valence captures whether individuals perceive social media\u0026rsquo;s effects as positively enhancing relationships, learning, and enjoyment or negatively wasting time or manipulation (e.g., \u0026ldquo;Using social media is a waste of time for me\u0026rdquo;). The SMMS test consists of 12 items based on a 5-point Likert scale (from 1:Strongly disagree to 5:Strongly Agree).\u003c/p\u003e \u003cp\u003eThe Social Media Use Scale (SMUS) (Tuck \u0026amp; Thompson, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), was developed to address limitations in existing measures of Social Media Use test (SMU), which often focus narrowly on frequency or passive vs active use, producing inconsistent results (Valkenburg, van Driel \u0026amp; Beyens, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The SMUS is instead intended to prove a nuanced and empirically validated classification of SMU behaviours across multiple platforms. It consists of 17 items on a 9-point Likert scale (Never-Hourly or more). The SMUS is designed to capture the frequency of four distinct factors: image-based SMU, namely, activities related to managing one\u0026rsquo;s social image, for example, editing/deleting content, monitoring likes/comments (e.g., \u0026lsquo;Played with photo filtering/photo editing\u0026rsquo;). Secondly, comparison-based SMU, behaviours involve social comparison (e.g., \u0026lsquo;Compared my life or experiences to others\u0026rsquo;). Thirdly, belief-based SMU, expressing or engaging with contentious beliefs such as posting negative opinions, seeking morally conflicting content (e.g., \u0026lsquo;Commented unsupportively or disliked/\u0026rsquo;\u0026rsquo;reacted\u0026rsquo;\u0026rsquo; unsupportively on others\u0026rsquo; post(s). Lastly, consumption-based SMU, namely, passive content consumption such as scrolling feeds, watching videos (e.g., \u0026lsquo;Scrolled aimlessly through my feed(s)\u0026rsquo;).\u003c/p\u003e \u003cp\u003eThe Social Media Activity Questionnaire (SMAQ) (Ozimek et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2023\u003c/span\u003e)was developed to assess social media use, distinguishing between active and passive behaviours, which we still decided to maintain. It consists of 18 items on a 5-point Likert Scale (never-very often), where eleven items measure passive use (e.g., \u0026lsquo;I look at the newsfeed to see the latest activities of other users\u0026rsquo;) and eight measure active use (e.g., \u0026lsquo;I post photos\u0026rsquo;).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Procedure\u003c/h2\u003e \u003cp\u003eOf the whole participant pool, half did the questionnaire and the dilemmas on Qualtrics in class and half of them on Instagram in the lab. Participants who did the experiment on Qualtrics performed it on their own smartphones. The pool that performed it on Instagram was furnished with a lab smartphone where Instagram was preinstalled. We created two private empty accounts, namely, that have no posts and follow no one but each other, with the university email of one of the experimenters. One of the accounts was not accessible to the participants and was used to post the dilemmas and the questions as Instagram stories, and to receive the participants' answers. The researchers had access to this account from their own smartphones. The second account was logged into the lab\u0026rsquo;s smartphone and was given to participants.\u003c/p\u003e \u003cp\u003eAfter instructing the participants, the phone is given to them already showing the first story, that is an introduction to the experiment (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Then follows each one of the dilemmas and the corresponding questions, until all 12 dilemmas were presented. The participants had to answer the question stories with a direct message to the posting account, under the control of the experimenter. We then received the answer on our own smartphone, dialed on that account and recorded it accordingly in a spreadsheet. This method allowed us to maintain the highest similarity of stimulus presentation and answering between Qualtrics and Instagram. In both cases, the dilemmas and following questions were presented as text, and the answer was given as a text entry question, while maintaining a high grade of ecological validity in how subjects usually interact with Instagram stories. The order of presentation of the dilemmas has been randomized in both conditions.\u003c/p\u003e \u003cp\u003eData about duration was automatically collected in Qualtrics, while for the Instagram group a screen-recording was started before presenting them the first story and was ended at the end of the session. Afterwards, through the free version of DaVinci Resolve 19, an editing software, the experimenters placed 74 markers to identify the moments of start and end of each story. Given that the moral dilemmas were 12 and for each of them there were 5 questions, each consisting of a total of 6 stories. 12 x 6\u0026thinsp;=\u0026thinsp;72, plus one marker for the beginning of the experiment and one for the end, hence a total of 74. The markers were then exported in a CSV file and converted into a specific duration through a tailor-made code in Python. They were then assigned automatically to each of the dilemmas, thanks to the order registered and saved in a CSV file with the subject number.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Data Analysis and Results","content":"\u003cp\u003eA power analysis (G*Power 3.1.9.7 Faul et al., 2007) was conducted for type of dilemmas, risk and group main interactions and consequent interactions. Assuming a conservative medium effect size (f = 0.25) and a correlation of 0.50 among repeated measures, using a \u0026alpha; = 0.10 at 80% power, a total sample size of 64 participants was needed. We recruited a total of 70 to accommodate potential exclusions and drop-outs.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBased on the \u0026nbsp;Shapiro-Wilk tests (\u003cem\u003ep\u003c/em\u003e\u0026lt;.001) and on Levene\u0026rsquo;s test revealing a significant difference in variances across groups only for the percentage of affirmative responses [\u003cem\u003eF\u003c/em\u003e(7, 552) = 3.22, \u003cem\u003ep\u003c/em\u003e = .002], we performed four robust three-way ANOVAs (Wilcox, 2012) using 10% trimmed means to examine the effects of dilemma type, group and risk on the percentage of affirmative responses, on moral acceptability, on valence and arousal ratings (Paragraph 3.1). For all variables, robust post-hoc were also performed. For filler dilemmas, only the group served as the variable of interest; we therefore performed the Wilcoxon rank-sum test (Paragraph 3.2). Shapiro-Wilk and Levene\u0026rsquo;s yielded significant results also for duration times, W = 0.681, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026lt; .0001, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026gt; .0002, respectively. We therefore performed a robust three-way ANOVA (Wilcox, 2012) using 10% trimmed means and robust post-hoc (Paragraph 3.3). Finally, we tested with a robust two-way ANOVA whether the groups differed in questionnaire scores, and then performed Spearman\u0026rsquo;s correlation (Paragraph 3.4).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e3.1 Moral Decision\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur results show a significant main effect of dilemma type (Q = 251.37, \u003cem\u003ep\u003c/em\u003e \u0026lt; .0001) and a main effect of risk (Q = 18.53, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026lt; .001). No significant effects were found for group or any interaction terms (\u003cem\u003ep\u003c/em\u003e\u0026gt;0.05). To provide context for the size of these effects, a classic (non-robust) ANOVA was also performed. This indicated a large effect of dilemma type [\u0026eta;\u0026sup2; = .29 (partial \u0026eta;\u0026sup2; = .30)], a small effect of risk [\u0026eta;\u0026sup2; = .022 (partial \u0026eta;\u0026sup2; = .031)] and a negligible effect for group and all the interactions. Following the initial ANOVA, we conducted robust post-hoc with 10%trimmed means (Wilcox, 2012) analysis to further explore the nature of these effects within each dilemma type. Specifically, we examined the influence of group and risk. For Incidental dilemmas neither risk nor group were significant, (\u003cem\u003epsihat\u003c/em\u003e = -0.018, 95% CI [\u0026ndash;0.26, 0.22] \u003cem\u003ep\u003c/em\u003e = 0.885 and \u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e0.23, 95% CI [\u0026ndash;0.01, 0.47]\u003cem\u003ep =\u0026nbsp;\u003c/em\u003e0.061), nor a significant interaction between group and risk (\u003cem\u003epsihat\u003c/em\u003e = -0.054, 95% CI [\u0026ndash;0.29, 0.19] \u003cem\u003ep\u003c/em\u003e = 0.665). In contrast, for the instrumental dilemmas, a significant main effect of risk was observed, (\u003cem\u003epsihat\u003c/em\u003e = 0.50, 95% CI [0.500, 0.27] \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), indicating that participants\u0026rsquo; responses differed robustly between risk conditions. Specifically, the proportion of positive responses is higher in the involvement dilemmas than in the non-involvement ones. Neither the group effect, (\u003cem\u003epsihat\u003c/em\u003e = 0.179, 95% CI [\u0026ndash;0.05, 0.40], \u003cem\u003ep\u003c/em\u003e = 0.128), nor the group-by-risk interaction, (\u003cem\u003epsihat\u003c/em\u003e = -0.036, \u003cem\u003ep\u003c/em\u003e = 0.760), reached significance in this dilemma type.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e3.1.1. Moral Acceptability\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMoral acceptability ratings yielded a significant effect of group (Q = 26.69, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026lt; .0001) and a main effect of dilemma type (Q = 51.11, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026lt; .0001). No significant effects were found for Risk or the interaction terms. To provide context for the size of these effects, a classic (non-robust) ANOVA was also performed. This indicated a small effect of group \u0026eta;\u0026sup2; = .040 (partial \u0026eta;\u0026sup2; = .042) and a small effect of dilemma type [\u0026eta;\u0026sup2; = .07 (partial \u0026eta;\u0026sup2; = .074)]. For the Moral Acceptability outcome, robust post-hoc analyses with 10% trimmed means (Wilcox, 20212) were conducted separately for dilemma types, examining the effects of group, risk, and their interaction. For incidental dilemmas, a significant main effect of group was revealed (\u003cem\u003epsihat\u003c/em\u003e = 1.68, 95% CI [0.75, 2.61], \u003cem\u003ep\u003c/em\u003e = .00048) with Instagram participants responding with higher moral acceptability than Qualtrics\u0026rsquo;s. There was no significant main effect of risk framing (\u003cem\u003epsihat\u003c/em\u003e = 0.05, 95% CI [\u0026ndash;0.88, 0.98], \u003cem\u003ep\u003c/em\u003e = .915). However, a significant group per risk interaction was observed (\u003cem\u003epsihat\u003c/em\u003e = \u0026ndash;0.98, 95% CI [\u0026ndash;1.91, \u0026ndash;0.05], \u003cem\u003ep\u003c/em\u003e = .039), suggesting that the effect of group differed depending on the risk framing condition. For instrumental dilemmas, the robust post-hoc test again revealed a significant main effect of group (\u003cem\u003epsihat\u003c/em\u003e = 1.96, 95% CI [1.03, 2.88], \u003cem\u003ep\u003c/em\u003e = .00004), indicating higher moral acceptability from Instagram participants compared to Qualtrics participants. The effect of risk framing was non-significant (\u003cem\u003epsihat\u003c/em\u003e = 0.09, 95% CI [\u0026ndash;0.83, 1.02], \u003cem\u003ep\u003c/em\u003e = .840). Similarly, the group \u0026times; risk interaction was not significant (\u003cem\u003epsihat\u003c/em\u003e = 0.02, 95% CI [\u0026ndash;0.90, 0.94], \u003cem\u003ep\u003c/em\u003e = .965), c.f. Figure 5\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e3.1.2. Valence\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eValence ratings indicated a significant effect of group (Q = 13.22, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026lt; .0004). No significant effects were found for other main factors or any of the interaction terms. To provide context for the size of these effects, a classic (non-robust) ANOVA was also performed. This indicated a small effect of group [\u0026eta;\u0026sup2; = .031 (partial \u0026eta;\u0026sup2; = .033)]. Robust post-hoc analyses with 10% trimmed means (Wilcox, 20212) were conducted separately for dilemma types, examining the effects of group, risk, and their interaction. For the Incidental dilemmas, a significant effect of group was found (\u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e= 1.27, 95% CI [0.28, 2.25], \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .011), indicating that the Instagram group answered higher on valence than the Qualtrics group. Risk (\u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e= 0.71, 95% CI [-0.27, 1.69], \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .154), and interaction were insignificant (\u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e= -0.03, 95% CI [-1.01, 0.94], \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .944). For the instrumental dilemmas the situation was the same, with the Instagram group scoring higher than Qualtrics\u0026rsquo;s on valence (\u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e= 1.27, 95% CI [0.31, 2.22], \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .009), and without any significance for risk (\u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e= 0.44, 95% CI [-0.50, 1.39], \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .354), or risk group interactions (\u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e= -0.14, 95% CI [-1.09, 0.80], \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .762).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e3.1.3. Arousal\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eArousal ratings analysis yielded a significant effect of group (Q = 38.1, \u003cem\u003ep\u003c/em\u003e \u0026lt; .0001). No significant effects were found for other main factors or any interaction terms. To provide context for the size of these effects, a classic (non-robust) ANOVA was also performed. This indicated a medium effect of group [\u0026eta;\u0026sup2; = .062 (partial \u0026eta;\u0026sup2; = .063)]. Robust post-hoc analyses with 10% trimmed means (Wilcox, 20212) were conducted separately for dilemma types, examining the effects of group, risk, and their interaction. For the Incidental dilemmas, a significant effect of group was found (\u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e= -3.19, 95% CI [-4.61, -1.77], \u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026lt; .0000), indicating that the Instagram group answered lower on arousal than the Qualtrics group. Risk (\u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e= 1.11, 95% CI [-0.30, 2.52], \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .124), and interaction were insignificant (\u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e= 0.39, 95% CI [-1.02, 1.81], \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .579). For the instrumental dilemmas the situation was the same, with the Instagram group scoring lower than Qualtrics\u0026rsquo;s on arousal (\u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e= -2.99, 95% CI [-4.49, -1.49], \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .0001), and without any significance for risk (\u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e= 0.19, 95% CI [-1.30, 1.69], \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .798) or risk group interactions (\u003cem\u003epsihat\u0026nbsp;\u003c/em\u003e= 0.30, 95% CI [-1.19, 1.81], \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .688).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e3.2. Filler Dilemmas\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFor the filler items, we did not include multiple independent variables\u0026mdash;only group membership served as the variable of interest. Since the assumption of normality was violated for all dependent variables (p\u0026lt; .0001), we employed the Wilcoxon rank-sum test. No significant differences were observed between groups for the proportion of Yes/No responses, moral acceptability, or valence (all \u003cem\u003ep\u003c/em\u003e\u0026gt;0.05). However, a significant difference emerged for arousal (W = 7885, \u003cem\u003ep\u003c/em\u003e = .015), with the Instagram group reporting lower arousal levels on average compared to the Qualtrics group (Figure 6), aligning with the overall trend observed in the other analyses.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e3.3 Duration Time\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eShapiro-Wilk tests yielded significant results (W = 0.681, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026lt; .0001, Levene\u0026rsquo;s tests\u0026rsquo; \u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026gt; .0002). We therefore opted for a robust three-way ANOVA (Wilcox, 2012) using 10% trimmed means to examine the effect of group and dilemma type on duration. The ANOVA yielded a main effect of dilemma type (Q = 7.00, \u0026nbsp;\u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .031) and a main effect of group (Q = 42.93, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= .001). The interaction between dilemma type and group was not significant (Q = 5.88, \u003cem\u003ep\u003c/em\u003e = 0.05). We therefore conducted a post-hoc linear contrast analysis for both group and dilemma type on duration. As expected, the contrast revealed a significant difference (\u003cem\u003epsihat\u003c/em\u003e =1.97, 95% CI [1.39, 2.56], \u0026nbsp;\u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026lt; .0001), indicating that the groups differed significantly in their duration. Post hoc contrast comparing the levels of dilemma type on duration indicated the following: comparing Filler dilemmas with Incidental dilemmas yielded a marginally non-significant trend toward shorter durations in \u0026nbsp;Filler compared to Incidental dilemmas (\u003cem\u003epsihat\u003c/em\u003e = -0.81, 95% CI [-1.66, 0.05]), with a p-value of 0.053. When comparing Filler with instrumental dilemmas, the estimated difference was again a non-significant trend (\u003cem\u003epsihat\u003c/em\u003e = -0.78 (95% CI [-1.63, 0.06], p = 0.053). When comparing Instrumental with Incidental, the difference was negligible (\u003cem\u003epsihat\u003c/em\u003e = 0.02, 95% CI [-0.93, 0.97], p = 0.96), indicating no significant difference between these dilemma types. To provide context for the size of these effects, a classic (non-robust) ANOVA was also performed. This indicated a small effect of group, [\u0026eta;\u0026sup2; = .0027 (partial \u0026eta;\u0026sup2; = .0027)] and a small effect of dilemma type [\u0026eta;\u0026sup2; = .004 (partial \u0026eta;\u0026sup2; = .004)]. The interaction between dilemma type and group was negligible (\u0026eta;\u0026sup2; = .0003).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e3.4. Standardized Questionnaires\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo rule out the possibility that observed differences in moral judgments were due to distinct patterns of Instagram use between groups, we analyzed responses across multiple validated social media questionnaires. A robust two-way ANOVA on trimmed means revealed a significant main effect of questionnaire type (\u003cem\u003eQ\u003c/em\u003e = 625.50, \u003cem\u003ep\u003c/em\u003e = .001), indicating that the measures captured distinct dimensions of social media engagement. Crucially, there was no significant main effect of the group (\u003cem\u003eQ\u003c/em\u003e = 1.08, \u003cem\u003ep\u003c/em\u003e = .300), suggesting that overall usage patterns did not differ meaningfully between participants. While a significant interaction was observed (\u003cem\u003eQ\u003c/em\u003e = 18.68, \u003cem\u003ep\u003c/em\u003e = .014), post hoc robust comparisons (with Holm correction) found no statistically significant group differences on any individual subscale (\u003cem\u003eall adjusted p\u003c/em\u003e\u0026gt;0.05). This pattern supports the interpretation that group-related effects in the main experiment are not attributable to differences in how participants engage with social media platforms, but rather to the context in which moral decisions were presented.\u003c/p\u003e\n\u003cp\u003eCorrelational analysis were performed with Spearman\u0026rsquo;s correlation coefficient in three different ways. Firstly, a correlational analysis has been run between the questionnaire subsets and the answers to the dilemmas without considering either the type of dilemma or the risk associated with it. Spearman\u0026rsquo;s rank-order correlation revealed significant positive associations:\u003c/p\u003e\n\u003cp\u003e- Arousal ratings were positively correlated with both image scores (\u003cem\u003e\u0026rho;\u003c/em\u003e = .20, \u003cem\u003ep\u003c/em\u003e \u0026lt; .001, Bonferroni-adjusted \u003cem\u003ep\u003c/em\u003e = .004) and valence scores (\u003cem\u003e\u0026rho;\u003c/em\u003e = .23, \u003cem\u003ep\u003c/em\u003e \u0026lt; .001, Bonferroni-adjusted \u003cem\u003ep\u003c/em\u003e \u0026lt; .001).\u003c/p\u003e\n\u003cp\u003e-Moral acceptability judgments were positively associated with belief scores (\u003cem\u003e\u0026rho;\u003c/em\u003e = .23, \u003cem\u003ep\u003c/em\u003e \u0026lt; .001, Bonferroni-adjusted \u003cem\u003ep\u003c/em\u003e \u0026lt; .001).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn other words, subjects scoring high in how much they manage their social media image (image scores) and perceiving social media effects as positive (valence scores) were also more agitated in making the decision (higher arousal). Lastly, subjects scoring higher in expressing or engaging with contentious content on social media (belief scores) reported higher moral acceptability. When we added the type of dilemma, significant positive correlations emerged between\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e- Valence scores and I_Arousal (\u003cem\u003e\u0026rho;\u003c/em\u003e = .30, \u003cem\u003ep\u003c/em\u003e \u0026lt; .001, Bonferroni-adjusted \u003cem\u003ep\u003c/em\u003e = .026).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e- Belief scores and I_Valence (\u003cem\u003e\u0026rho;\u003c/em\u003e = .32, \u003cem\u003ep\u003c/em\u003e \u0026lt; .001, Bonferroni-adjusted \u003cem\u003ep\u003c/em\u003e = .014).\u003c/p\u003e\n\u003cp\u003eThese replicate the result above with the further specification of the Incidental dilemma. Perception of social media effects as positive correlates with more agitation in Incidental dilemmas, and expressing or engaging with contentious contents correlates with higher pleasure (valence ratings) for Incidental dilemmas. When we also added risk, no statistically significant correlations were found after applying Bonferroni correction for multiple comparisons (\u003cem\u003eall adjusted p\u003c/em\u003e \u0026gt; .05).\u0026nbsp;\u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThe present study examined whether presenting moral decision-making scenarios on the Instagram app affects individuals\u0026rsquo; responses. Our hypothesis that Instagram would warp the emotions associated with moral judgments was confirmed by all three ratings, where the Instagram group reported higher moral acceptability, higher calm and lower regret. Furthermore, the Instagram group was consistently slower. Both groups showed a similar trend where filler dilemmas are the fastest, followed by instrumental and finally by incidental. We didn\u0026rsquo;t find a group difference for utilitarian/deontological decisions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eInterestingly, moral acceptability yielded a significant interaction of group and risk in incidental dilemmas, and an opposite but not significant trend was found in instrumental dilemmas. In the Qualtrics group, moral acceptability decreased when the incidental dilemmas were without involvement, showing instead the opposite trend in the Instagram group (Fig 7). In the Qualtrics group, this could be explained by a simple survival instinct, for which killing to save one\u0026rsquo;s own life is considered more acceptable, therefore driving down moral acceptability in the case of lack of involvement. On the other hand, for the Instagram group, it is possible that killing one person to save others when one\u0026rsquo;s life is not at stake is perceived as a more virtuous action, hence more morally acceptable. The rationale behind these findings, we suggest, is that different platforms induce adherence to different social norms.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor valence and arousal ratings, the Instagram group ranked significantly higher valence (less regret) and lower arousal (calmer) for both dilemma types. Valence followed a similar trend in both groups and dilemma types, with ratings higher for dilemmas with involvement and lower for dilemmas without involvement, signaling that subjects regret killing someone less when their own life is at stake. Furthermore, the change in mean valence (Tab 1.) from dilemmas with and without involvement is similar for both the Instagram and Qualtrics groups, suggesting a consistent pattern of emotional response. Arousal ratings followed a similar decreasing trend in the case of incidental dilemmas, where killing was less arousing (less agitating) when the participant\u0026apos;s life was not at risk than when it was. However, the change in the mean arousal between dilemmas with and without involvement (Tab 2.) is quite different for the two groups. The Instagram group has more than half a point of difference, while the Qualtrics group has less than half. We found the same pattern for instrumental dilemmas, even though less pronounced. Furthermore, while the Instagram trend is the same across incidental and instrumental dilemmas, with dilemmas with involvement scoring higher than dilemmas without, Qualtrics\u0026rsquo;s one is the opposite for instrumental dilemmas. Dilemmas without involvement are scored as more arousing than with involvement, albeit slightly. This is in line with recent studies suggesting that Instagram\u0026apos;s social and emotional context might enhance emotional reactivity (Ozimek et al., 2023; Yue et al., 2022) and with the aforementioned literature on enhanced extremity. Overall, our findings emphasise how Instagram enhances arousal patterns tied to personal involvement.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 1. Descriptive statistics for valence ratings.\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eDilemma Type\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eRisk\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eGroup\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eN\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eMean\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eSD\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eIncidental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWith involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstagram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.15\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eIncidental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWith involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eQualtrics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.83\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eIncidental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWithout involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstagram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.76\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.03\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eIncidental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWithout involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eQualtrics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.74\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstrumental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWith involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstagram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.93\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.27\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstrumental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWith involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eQualtrics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.79\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstrumental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWithout involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstagram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.15\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstrumental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWithout involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eQualtrics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.54\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eTable 2 Descriptive statistics for arousal ratings.\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eDilemma Type\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eRisk\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eGroup\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eN\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eMean\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eSD\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eIncidental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWith involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstagram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e5.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.85\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eIncidental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWith involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eQualtrics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e6.87\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.43\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eIncidental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWithout involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstagram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e5.03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.51\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eIncidental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWithout involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eQualtrics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e6.62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.26\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstrumental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWith involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstagram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e5.66\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.83\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstrumental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWith involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eQualtrics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e6.77\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.45\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstrumental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWithout involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstagram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e5.44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.73\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eInstrumental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWithout involvement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eQualtrics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e6.81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.50\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eTurning to the temporal dimension, participants in the Instagram group consistently took longer to respond compared to those in the Qualtrics group. Despite this difference in overall response time, both groups exhibited a similar pattern, in line with the dual process theory (Greene et al., 2001; Greene et al., 2004) across dilemma types: filler dilemmas were answered the quickest, followed by instrumental dilemmas, and incidental dilemmas (Fig 6.) The longer response times observed in the Instagram group may reflect subjects being used to lingering on, idling on stories without a clear-cut willingness to reach the end of that activity. However, it might be argued that this should be true only when the content is interesting to the user. In line with this, our data (Tab 3.) shows that the Instagram group has both a longer maximum total duration as well as a shorter minimum duration. This then reflects the platform\u0026rsquo;s dual engagement styles: lingering on interesting content while rapidly skipping uninteresting content. Therefore, our data are in line with Instagram\u0026rsquo;s characteristic user behaviour of selective engagement, where attention is flexibly allocated based on personal interest.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 3. Descriptive statistics for duration in minutes\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"598\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eGroup\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003eInstagram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003eInstagram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eInstagram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003eInstagram\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003eQualtrics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003eQualtrics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eQualtrics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003eQualtrics\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eDilemma\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003eIncidental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eInstrumental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003eFiller\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003eIncidental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003eInstrumental\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003eFiller\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e14.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e15.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e15.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e13.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e13.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e14.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e13.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e12.4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eStandard Deviation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e4.86\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e5.37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e4.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e4.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e3.99\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e4.73\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e4.03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e2.91\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eMin\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e5.85\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e6.19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e5.94\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e5.85\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e6.86\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e6.86\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e7.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e7.95\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 101px;\"\u003e\n \u003cp\u003eMax\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e31.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e31.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e30.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e21.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e28.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 60px;\"\u003e\n \u003cp\u003e\u0026nbsp;28.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 73px;\"\u003e\n \u003cp\u003e26.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 56px;\"\u003e\n \u003cp\u003e19.1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eImportantly, we didn\u0026rsquo;t find a group difference for utilitarian/deontological decisions. Usually, this difference is found when comparing two starkly different contexts, and according to the dual process theory, when one context can induce psychological distance (Costa et al., 2014) favouring utilitarian judgments (Greene et al., 2001; Greene et al., 2004). We observed a slight increase in utilitarian responses in the instrumental dilemmas when compared to Qualtrics, even though not significant. Concerning standardized questionnaires, higher scores on SMMS valence correlate with more agitation. SMMS valence is designed to measure the perception of social media effects as positive; therefore, subjects believing that social media has positive effects might have reacted with more agitation when forced to make life-or-death decisions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOverall, our results can be interpreted by considering the intrinsic characteristics of social media content. As aforementioned, morally and emotionally resonant content is the primary driver of engagement (Al-Rawi, 2019; Nontasil \u0026amp; Payne, 2019; Brady et al., 2020 ; Rathje et al., 2021; Pandey et al., 2023;\u0026nbsp;Van Bavel, 2024), incentivizing creation of such content while shadowing moderate ones (St\u0026ouml;cker \u0026amp; Preuss, 2020; Whittaker et al., 2021; Robertson et al., 2024). In line with other work suggesting that online environments teach and reinforce behaviours (Masur et al., 2021; Cheng, 2023), we provide empirical data supporting a learning perspective. To be embedded in the Instagram context induces a more extreme mindset in the users and elicits the adherence to different social norms, enhancing moral acceptability while dampening arousal and valence.\u003c/p\u003e\n\u003cp\u003eOne mainstream concern about social media\u0026ndash;driven norm warping is that behaviours learned online can seep over into everyday life. We have already seen cases where influencers, by virtue of their status, refuse to pay for services (e.g., a meal), reflecting the transfer of norms from social media into offline contexts. The danger becomes far more serious when content promoting hate or extreme actions escapes these platforms, potentially putting lives at risk. It is therefore essential to pursue further research into this phenomenon, rigorously testing both the likelihood of such seepover and the context-dependent nature of social norm adherence.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eIn this study, we investigated whether answering moral dilemmas on different smartphones\u0026rsquo; apps would affect the ratings of moral acceptability, valence and arousal, as well as the amount of utilitarian and deontological responses. We tested Instagram vs Qualtrics platforms, using a novel methodology by presenting stimuli directly on social media and prompting decision-making on the social platform itself. We discovered that subjects answering the dilemmas on Instagram uphold higher moral acceptability ratings, higher valence (less regret) and lower arousal (less agitation). No differences were found for utilitarian and deontological decisions.\u003c/p\u003e \u003cp\u003eOur study provided significant empirical evidence regarding the immediate cognitive effects of embedding human moral judgment within the Instagram platform. These findings support the widespread belief that the peculiar social norms afforded by social media are shaping new ways to think and feel. Specifically, our findings point to a concerning dynamic: immersion in the Instagram environment, optimized to highlight and reward extreme content, seems to foster the normalization and reinforcement of more extreme behaviours and moral judgments. Thus, social media can be framed as a training ground to for the acceptance of extremeness foregoing moderate content, even more so when extremeness is concealed beneath content shaped by the users\u0026rsquo; individual preferences. The implications of this research raise questions of social and ethical relevance, suggesting the need for greater user awareness of the unintended cognitive effects of social media use.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eConflict of interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;The authors have no financial or proprietary interests in any material discussed in this article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNicola Chinchella is supported by the European Union \u0026ndash; NextGenerationEU through the Italian Ministry of University and Research under PNRR \u0026ndash; Mission 4 \u0026ndash; Component 2 \u0026ndash; Investment 3.1 \u0026ldquo;Fund for the realization of an integrated research and innovation infrastructure system\u0026rdquo; D.M. 118/2023 CUP B83C22003950001\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData can be found at: https://osf.io/jfyc6\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contribution\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization: N.C., C.L.; methodology: N.C., C.L.; data collection: N.C., A.G; data analysis: N.C.; writing original draft preparation, N.C.; writing review and editing, N.C., C.L., A.G.; supervision, C.L., A.G. All authors have read and agreed to the published version of the manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAcerbi, A. (2019). \u003cem\u003eCultural Evolution in the Digital Age\u003c/em\u003e. Oxford University Press. https://doi.org/10.1093/oso/9780198835943.001.0001\u003c/li\u003e\n\u003cli\u003eAl-Rawi, A. (2019). Viral News on Social Media. \u003cem\u003eDigital Journalism\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(1), 63\u0026ndash;79. https://doi.org/10.1080/21670811.2017.1387062\u003c/li\u003e\n\u003cli\u003eBarque-Duran, A., Pothos, E. M., Yearsley, J. M., \u0026amp; Hampton, J. A. (s.d.). \u003cem\u003eThe impact of the Digital Age in Moral Judgments\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eBerger, J., \u0026amp; Milkman, K. L. (2012). What Makes Online Content Viral? \u003cem\u003eJournal of Marketing Research\u003c/em\u003e, \u003cem\u003e49\u003c/em\u003e(2), 192\u0026ndash;205. https://doi.org/10.1509/jmr.10.0353\u003c/li\u003e\n\u003cli\u003eBrady WJ, Crockett MJ, Van Bavel JJ. (2020a). The MAD model of moral contagion: the role of motivation, attention, and design in the spread of moralized content online. \u003cem\u003ePerspectives on psychological science : a journal of the Association for Psychological Science\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(4), 978\u0026ndash;1010. https://doi.org/10.1177/1745691620917336\u003c/li\u003e\n\u003cli\u003eBrady WJ, Gantman AP, Van Bavel JJ.(2020b). Attentional capture helps explain why moral and emotional content go viral. J. Exp. Psychol. Gen. \u003c/li\u003e\n\u003cli\u003eBrady, W. J., \u0026amp; Crockett, M. J. (2024). Norm Psychology in the Digital Age: How Social Media Shapes the Cultural Evolution of Normativity. \u003cem\u003ePerspectives on Psychological Science\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(1), 62\u0026ndash;64. https://doi.org/10.1177/17456916231187395\u003c/li\u003e\n\u003cli\u003eBrady, W. J., McLoughlin, K., Doan, T. N., \u0026amp; Crockett, M. J. (2021). How social learning amplifies moral outrage expression in online social networks. \u003cem\u003eScience Advances\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(33), eabe5641. https://doi.org/10.1126/sciadv.abe5641\u003c/li\u003e\n\u003cli\u003eBruineberg, J. (2023). Adversarial inference: Predictive minds in the attention economy. \u003cem\u003eNeuroscience of Consciousness\u003c/em\u003e, \u003cem\u003e2023\u003c/em\u003e(1), niad019. https://doi.org/10.1093/nc/niad019\u003c/li\u003e\n\u003cli\u003eCao, F., Zhang, J., Song, L., Wang, S., Miao, D., \u0026amp; Peng, J. (2017). Framing Effect in the Trolley Problem and Footbridge Dilemma: Number of Saved Lives Matters. \u003cem\u003ePsychological Reports\u003c/em\u003e, \u003cem\u003e120\u003c/em\u003e(1), 88-101. https://doi.org/10.1177/0033294116685866 (Original work published 2017)\u003c/li\u003e\n\u003cli\u003eCarpenter, C. J., \u0026amp; Amaravadi, C. S. (2019). A big data approach to assessing the impact of social norms: Reporting one\u0026apos;s exercise to a social media audience. \u003cem\u003eCommunication Research\u003c/em\u003e, \u003cstrong\u003e46\u003c/strong\u003e(2), 236\u0026ndash;249. \u003cstrong\u003ehttps://doi.org/10.1177/0093650216657776\u003c/strong\u003eChung, A., \u0026amp; Rimal, R.N. (2016). Social Norms: A Review. Review of Communication Research, 4, 1-28.\u003cbr\u003e https://doi.org/10.12840/issn.2255-4165.2016.04.01.008\u003c/li\u003e\n\u003cli\u003eChung, M. (2019). The message influences me more than others: How and why social media metrics affect first person perception and behavioral intentions. \u003cem\u003eComputers in Human Behavior\u003c/em\u003e, \u003cem\u003e91\u003c/em\u003e, 271\u0026ndash;278. https://doi.org/10.1016/j.chb.2018.10.011\u003c/li\u003e\n\u003cli\u003eCosta A, Foucart A, Hayakawa S, Aparici M, Apesteguia J, Heafner J, et al. (2014) Your Morals Depend on Language. PLoS ONE 9(4): e94842. https://doi.org/10.1371/journal.pone.009484\u003c/li\u003e\n\u003cli\u003eDigitalNRG (2025)https://www.digitalnrg.co.uk/top-social-media-statistic-insights-for-2025/#elementor-toc__heading-anchor-4\u003c/li\u003e\n\u003cli\u003eFaul, F., Erdfelder, E., Lang, A.-G., \u0026amp; Buchner, A. (2007). G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, 175\u0026ndash;191.\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eFisher, M.\u003c/strong\u003e (2022). \u003cem\u003eThe chaos machine: The inside story of how social media rewired our minds and our world\u003c/em\u003e (Hardcover ed.). Little, Brown and Company\u003c/li\u003e\n\u003cli\u003eGreene JD, Nystrom LE, Engell AD, Darley JM, Cohen JD (2004). The neural bases of cognitive conflict and control in moral judgment. \u003cem\u003eNeuron\u003c/em\u003e. \u003cstrong\u003e44\u003c/strong\u003e (2): 389\u0026ndash;400.\u003c/li\u003e\n\u003cli\u003eGreene JD, Sommerville RB, Nystrom LE, Darley JM, Cohen JD (2001). An fMRI investigation of emotional engagement in moral judgment. \u003cem\u003eScience\u003c/em\u003e. \u003cstrong\u003e293\u003c/strong\u003e (5537): 2105\u0026ndash;8.\u003c/li\u003e\n\u003cli\u003eHaidt, J. (2003). Elevation and the positive psychology of morality. In C. L. M. Keyes \u0026amp; J. Haidt (Eds.), \u003cem\u003eFlourishing: Positive psychology and the life well-lived\u003c/em\u003e (pp. 275\u0026ndash;289). American Psychological Association. https://doi.org/10.1037/10594-012\u003c/li\u003e\n\u003cli\u003eHancock, J., Liu, S. X., Luo, M., \u0026amp; Mieczkowski, H. (2022). Psychological Well-Being and Social Media Use: A Meta-Analysis of Associations between Social Media Use and Depression, Anxiety, Loneliness, Eudaimonic, Hedonic and Social Well-Being. \u003cem\u003eSSRN Electronic Journal\u003c/em\u003e. https://doi.org/10.2139/ssrn.4053961\u003c/li\u003e\n\u003cli\u003eKonovalova, E., Mens, G. L., \u0026amp; Sch\u0026ouml;ll, N. (2023). Social media feedback and extreme opinion expression. \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e(11), e0293805. https://doi.org/10.1371/journal.pone.0293805\u003c/li\u003e\n\u003cli\u003eKorte, M. (2020). The impact of the digital revolution on human brain and behavior: Where do we stand? \u003cem\u003eDialogues in Clinical Neuroscience\u003c/em\u003e, \u003cem\u003e22\u003c/em\u003e(2), 101\u0026ndash;111. https://doi.org/10.31887/DCNS.2020.22.2/mkorte\u003c/li\u003e\n\u003cli\u003eKross, E., Verduyn, P., Sheppes, G., Costello, C. K., Jonides, J., \u0026amp; Ybarra, O. (2021). Social Media and Well-Being: Pitfalls, Progress, and Next Steps. \u003cem\u003eTrends in Cognitive Sciences\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(1), 55\u0026ndash;66. https://doi.org/10.1016/j.tics.2020.10.005\u003c/li\u003e\n\u003cli\u003eLee, A. Y., \u0026amp; Hancock, J. T. (2024). \u003cem\u003eSocial Media Mindsets: A New Approach to Understanding Social Media Use \u0026amp; Psychological Well-Being\u003c/em\u003e. https://doi.org/10.1093/jcmc/zmad048\u003c/li\u003e\n\u003cli\u003eLindstr\u0026ouml;m, B., Bellander, M., Schultner, D. T., Chang, A., Tobler, P. N., \u0026amp; Amodio, D. M. (2021). A computational reward learning account of social media engagement. \u003cem\u003eNature Communications\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e, 1311. https://doi.org/10.1038/s41467-020-19607-x\u003c/li\u003e\n\u003cli\u003eLotto, L., Manfrinati, A., \u0026amp; Sarlo, M. (2014). A New Set of Moral Dilemmas: Norms for Moral Acceptability, Decision Times, and Emotional Salience. \u003cem\u003eJournal of Behavioral Decision Making\u003c/em\u003e, \u003cem\u003e27\u003c/em\u003e(1), 57\u0026ndash;65. https://doi.org/10.1002/bdm.1782\u003c/li\u003e\n\u003cli\u003eLupinacci, L. (2021). \u0026lsquo;Absentmindedly scrolling through nothing\u0026rsquo;: Liveness and compulsory continuous connectedness in social media. \u003cem\u003eMedia, Culture \u0026amp; Society\u003c/em\u003e, \u003cem\u003e43\u003c/em\u003e(2), 273\u0026ndash;290. https://doi.org/10.1177/0163443720939454\u003c/li\u003e\n\u003cli\u003eMackay D. (2023). Infinite Scrolling, Dissociation, and Boredom Spiraling as the Drivers of Habitual Social Media Use. Ph. D. Dissertation. Southern Connecticut State University.\u003c/li\u003e\n\u003cli\u003eMarino, C., Bersia, M., Furstova, J., Galeotti, T., van den Eijnden, R. J. J. M., Boniel-Nissim, M., Pickett, W., Lenzi, M., Canale, N., Eriksson, C., Lahti, H., Ozolina, K., Craig, W., \u0026amp; Vieno, A. (2025). Global change in adolescent social media use (2018\u0026ndash;2022): An ecological analysis across 28 countries. \u003cem\u003eComputers in Human Behavior\u003c/em\u003e, \u003cem\u003e173\u003c/em\u003e, 108789. https://doi.org/10.1016/j.chb.2025.108789\u003c/li\u003e\n\u003cli\u003eMasur, P. K., DiFranzo, D., \u0026amp; Bazarova, N. N. (2021). Behavioral contagion on social media: Effects of social norms, design interventions, and critical media literacy on self-disclosure. \u003cem\u003ePLoS ONE\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(7), e0254670. https://doi.org/10.1371/journal.pone.0254670\u003c/li\u003e\n\u003cli\u003eMilli, S., Carroll, M., Wang, Y., Pandey, S., Zhao, S., \u0026amp; Dragan, A. D. (2025). Engagement, user satisfaction, and the amplification of divisive content on social media. \u003cem\u003ePNAS Nexus\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e(3), pgaf062. https://doi.org/10.1093/pnasnexus/pgaf062\u003c/li\u003e\n\u003cli\u003eNontasil, P., \u0026amp; Payne, S. J. (2019). Emotional Utility and Recall of the Facebook News Feed. \u003cem\u003eProceedings of the 2019 CHI Conference on Human Factors in Computing Systems\u003c/em\u003e, 1\u0026ndash;9. https://doi.org/10.1145/3290605.3300252\u003c/li\u003e\n\u003cli\u003eOzimek, P., Brandenberg, G., Rohmann, E., \u0026amp; Bierhoff, H.-W. (2023). The Impact of Social Comparisons More Related to Ability vs. More Related to Opinion on Well-Being: An Instagram Study. \u003cem\u003eBehavioral Sciences\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(10), 850. https://doi.org/10.3390/bs13100850\u003c/li\u003e\n\u003cli\u003ePan, X., Hou, Y., \u0026amp; Wang, Q. (2023). Are we braver in cyberspace? Social media anonymity enhances moral courage. \u003cem\u003eComputers in Human Behavior\u003c/em\u003e, \u003cem\u003e148\u003c/em\u003e, 107880. https://doi.org/10.1016/j.chb.2023.107880\u003c/li\u003e\n\u003cli\u003ePandey, S., Cao, Y., Dong, Y., Kim, M., MacLaren, N. G., Dionne, S. D., Yammarino, F. J., \u0026amp; Sayama, H. (2023). Generation and influence of eccentric ideas on social networks. \u003cem\u003eScientific Reports\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(1), 20433. https://doi.org/10.1038/s41598-023-47823-0\u003c/li\u003e\n\u003cli\u003ePastor, Y., P\u0026eacute;rez-Torres, V., Thomas-Curr\u0026aacute;s, H., Lobato-Rinc\u0026oacute;n, L. L., L\u0026oacute;pez-S\u0026aacute;ez, M. \u0026Aacute;., \u0026amp; Garc\u0026iacute;a, A. (2024). A study of the influence of altruism, social responsibility, reciprocity, and the subjective norm on online prosocial behavior in adolescence. \u003cem\u003eComputers in Human Behavior\u003c/em\u003e, \u003cem\u003e154\u003c/em\u003e, 108156. https://doi.org/10.1016/j.chb.2024.108156\u003c/li\u003e\n\u003cli\u003ePastor, Y., P\u0026eacute;rez-Torres, V., Thomas-Curr\u0026aacute;s, H., Lobato-Rinc\u0026oacute;n, L. L., L\u0026oacute;pez-S\u0026aacute;ez, M. \u0026Aacute;., \u0026amp; Garc\u0026iacute;a, A. (2024). A study of the influence of altruism, social responsibility, reciprocity, and the subjective norm on online prosocial behavior in adolescence. \u003cem\u003eComputers in Human Behavior\u003c/em\u003e, \u003cem\u003e154\u003c/em\u003e, 108156. https://doi.org/10.1016/j.chb.2024.108156\u003c/li\u003e\n\u003cli\u003ePilcher JJ, Smith PD (2024). Social context during moral decision-making impacts males more than females. Front Psychol. doi: 10.3389/fpsyg.2024.1397069. PMID: 38836238; PMCID: PMC11148431.\u003c/li\u003e\n\u003cli\u003eRathje S, Robertson C, Brady WJ, Van Bavel JJ. (2024) People Think That Social Media Platforms Do (but Should Not) Amplify Divisive Content. Perspect Psychol Sci.. doi: 10.1177/17456916231190392. \u003c/li\u003e\n\u003cli\u003eRathje, S., Van Bavel, J. J., \u0026amp; van der Linden, S. (2021). Out-group animosity drives engagement on social media. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e, \u003cem\u003e118\u003c/em\u003e(26), e2024292118. https://doi.org/10.1073/pnas.2024292118\u003c/li\u003e\n\u003cli\u003eRimal, R. N., \u0026amp; Lapinski, M. K. (2015). A Re-Explication of Social Norms, Ten Years Later. \u003cem\u003eCommunication Theory\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(4), 393-409. https://doi.org/10.1111/comt.12080\u003c/li\u003e\n\u003cli\u003eRixen, J. O., Meinhardt, L.-M., Gl\u0026ouml;ckler, M., Ziegenbein, M.-L., Schlothauer, A., Colley, M., Rukzio, E., \u0026amp; Gugenheimer, J. (2023). The Loop and Reasons to Break It: Investigating Infinite Scrolling Behaviour in Social Media Applications and Reasons to Stop. \u003cem\u003eProc. ACM Hum.-Comput. Interact.\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(MHCI), 228:1-228:22. https://doi.org/10.1145/3604275\u003c/li\u003e\n\u003cli\u003eRobertson, C. E., Del Rosario, K. S., \u0026amp; Van Bavel, J. J. (2024). Inside the funhouse mirror factory: How social media distorts perceptions of norms. \u003cem\u003eCurrent Opinion in Psychology\u003c/em\u003e, \u003cem\u003e60\u003c/em\u003e, 101918. https://doi.org/10.1016/j.copsyc.2024.101918\u003c/li\u003e\n\u003cli\u003eSaternus, Z., Mihale-Wilson, C. \u0026amp; Hinz, O. (2024). Influencer marketing on Instagram\u0026mdash;The optimal disclosure strategy from influencers\u0026rsquo; and marketers\u0026rsquo; perspectives. \u003cem\u003eElectron Markets\u003c/em\u003e https://doi.org/10.1007/s12525-024-00743-x\u003c/li\u003e\n\u003cli\u003eSavolainen, I., Oksanen, A., Kaakinen, M., Sirola, A., Zych, I., \u0026amp; Paek, H.-J. (2021). The role of online group norms and social identity in youth problem gambling. \u003cem\u003eComputers in Human Behavior\u003c/em\u003e, \u003cem\u003e122\u003c/em\u003e, 106828. https://doi.org/10.1016/j.chb.2021.106828\u003c/li\u003e\n\u003cli\u003eShareef, M. A., Kapoor, K. K., Mukerji, B., Dwivedi, R., \u0026amp; Dwivedi, Y. K. (2020). Group behavior in social media: Antecedents of initial trust formation. \u003cem\u003eComputers in Human Behavior\u003c/em\u003e, \u003cem\u003e105\u003c/em\u003e, 106225. https://doi.org/10.1016/j.chb.2019.106225\u003c/li\u003e\n\u003cli\u003eStatista. 2025. Number of social media users worldwide from 2017 to 2027. Statista. https://www.statista.com/statistics/278414/number-of-worldwide-social-network-users\u003c/li\u003e\n\u003cli\u003eSt\u0026ouml;cker, C., \u0026amp; Preuss, M. (2020). Riding the Wave of Misclassification: How We End up with Extreme YouTube Content. In G. Meiselwitz (A c. Di), \u003cem\u003eSocial Computing and Social Media. Design, Ethics, User Behavior, and Social Network Analysis\u003c/em\u003e (pp. 359\u0026ndash;375). Springer International Publishing. https://doi.org/10.1007/978-3-030-49570-1_25\u003c/li\u003e\n\u003cli\u003eTran, J. A., Yang, K. S., Davis, K., \u0026amp; Hiniker, A. (2019). Modeling the Engagement-Disengagement Cycle of Compulsive Phone Use. \u003cem\u003eProceedings of the 2019 CHI Conference on Human Factors in Computing Systems\u003c/em\u003e, 1\u0026ndash;14. https://doi.org/10.1145/3290605.3300542\u003c/li\u003e\n\u003cli\u003eTuck, A. B., \u0026amp; Thompson, R. J. (2024). The Social Media Use Scale: Development and Validation. \u003cem\u003eAssessment\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e(3), 617\u0026ndash;636. https://doi.org/10.1177/10731911231173080\u003c/li\u003e\n\u003cli\u003eValkenburg, P. M. (2022). Social media use and well-being: What we know and what we need to know. \u003cem\u003eCurrent Opinion in Psychology\u003c/em\u003e, \u003cem\u003e45\u003c/em\u003e, 101294. https://doi.org/10.1016/j.copsyc.2021.12.006\u003c/li\u003e\n\u003cli\u003eValkenburg, P. M., van Driel, I. I., \u0026amp; Beyens, I. (2022). The associations of active and passive social media use with well-being: A critical scoping review. \u003cem\u003eNew Media \u0026amp; Society\u003c/em\u003e, \u003cem\u003e24\u003c/em\u003e(2), 530\u0026ndash;549. https://doi.org/10.1177/14614448211065425\u003c/li\u003e\n\u003cli\u003eVan Bavel JJ, Robertson CE, Del Rosario K, Rasmussen J, Rathje S. (2024) Social Media and Morality. Annu Rev Psychol. doi: 10.1146/annurev-psych-022123-110258. \u003c/li\u003e\n\u003cli\u003eVerduyn, P., Gugushvili, N., \u0026amp; Kross, E. (2022). Do Social Networking Sites Influence Well-Being? The Extended Active-Passive Model. \u003cem\u003eCurrent Directions in Psychological Science\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e(1), 62\u0026ndash;68. https://doi.org/10.1177/09637214211053637\u003c/li\u003e\n\u003cli\u003eWhittaker, J., Looney, S., Reed, A., \u0026amp; Votta, F. (2021). Recommender systems and the amplification of extremist content. Internet Policy Review, 10(2). https://doi.org/10.14763/2021.2.1565\u003c/li\u003e\n\u003cli\u003eWilcox, R. R. (2012). \u003cem\u003eIntroduction to Robust Estimation and Hypothesis Testing\u003c/em\u003e (3rd ed.). Academic Press.\u003c/li\u003e\n\u003cli\u003eYue, Z., Zhang, R., \u0026amp; Xiao, J. (2022). Passive social media use and psychological well-being during the COVID-19 pandemic: The role of social comparison and emotion regulation. \u003cem\u003eComputers in human behavior\u003c/em\u003e, \u003cem\u003e127\u003c/em\u003e, 107050. https://doi.org/10.1016/j.chb.2021.107050\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8328737/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8328737/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSocial media thrive on virality amplifying content that maximizes engagement by triggering strong emotional and moral responses. As users\u0026rsquo; feeds become saturated with affect-laden material, it becomes crucial to examine whether this exposure shapes online judgments and moral norms. In this work, we examine how engaging with moral dilemmas on Instagram affects users\u0026rsquo; responses, focusing on perceived moral acceptability, arousal and valence. Our findings demonstrate that participants interacting with dilemmas on Instagram, compared to those using Qualtrics, exhibit increased moral acceptability ratings alongside attenuated arousal and valence responses. Additionally, the Instagram cohort required more time to complete the task. Our results suggest that simply being on social media platforms like Instagram can induce a moral desensitization posing significant implications given the widespread daily engagement with these platforms and the potential erosion of sensitivity to moral issues.\u003c/p\u003e","manuscriptTitle":"Moral Desensitization in Digital Contexts: Instagram exposure alters moral perception and emotional reactivity","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-09 06:34:19","doi":"10.21203/rs.3.rs-8328737/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"1cae8a15-5024-4658-be99-330a10d6e7e9","owner":[],"postedDate":"January 9th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-23T10:13:03+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-09 06:34:19","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8328737","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8328737","identity":"rs-8328737","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}