People can disengage from their internalized moral standards and self-regulation in order to perform immoral behaviour by using different Moral Disengagement mechanisms. These mechanisms within media have a positive effect on immoral behaviour. However, Moral Identity activation is said to counter arguments of Moral Disengagement. In this study, both concepts are applied to the context of war. An additional assumption took into account in how far participants’ internalized moral standards consider war as immoral. This is important since Moral Identity and Moral Disengagement are based on internalized moral standards. To test the hypotheses, this study employed a 2 x 2 RO between-subjects factorial design. The trait variable called Moral Consideration of War was supposed to reflect participants’ internalized moral standards with regard to war. It was used to operationalize the additional assumption. Factor 1 varied the activation of Moral Identity (Moral Identity activation versus control group) and factor 2 varied the depiction of the war scenario (Permissive Scenario versus Prohibitive Scenario). Scenarios were fictive newspaper articles. A Permissive Scenario comprised a higher number of arguments based on Moral Disengagement mechanisms than a Prohibitive Scenario. Main outcome measures were the support of war and war-related Moral Disengagement (questionnaire). In total 86 participants (f=45, m=41) were randomized into four cells and completed the online experiment. The Permissive Scenario failed to increase support of war and Moral Disengagement (questionnaire), on the assumption that war is considered immoral. Moral Identity activation had a negative effect on Moral Disengagement only on the assumption that war was considered moral. Moral Identity activation had no significant effect on support of war, on the premise that war was considered immoral. The interaction term of Moral Identity activation and Permissive Scenario had no significant effect neither on support of war nor on Moral Disengagement, no matter if additional assumption was taken into account or not. Results are discussed with regard to methodological limitations measuring internalized moral standards. Their measurement implied already individual Moral Disengagement. Interaction effect failed, supposedly because Moral Identity activation was not specifically targeted at immoral behaviour and because mediating effects of Moral Identity centrality were not considered.:I. CONTENTS I. Contents 2 II. List of Figures 5 III. List of Tables 6 IV. List of Equations 7 V. Abstract 8 VI. Zusammenfassung 9 1 Introduction 10 2 Theory 12 2.1 Moral Disengagement 12 2.1.1 Socio Cognitive Theory of Morality 12 2.1.2 Mechanisms of Moral Disengagement 13 2.1.3 Moral Disengagement and War 13 2.2 Moral Disengagement and Media 17 2.2.1 Moral Disengagement in Media and Aggressive Behaviour 18 2.2.2 Moral Disengagement in Media and War 19 2.2.3 Operationalization of Scenarios 21 2.3 Moral Identity 22 2.3.1 Moral Identity Centrality 22 2.3.2 Moral Identity Activation and Moral Disengagement 24 2.3.3 Moral Identity Activation in Interaction with Moral Disengagement in Media 26 2.4 Hypotheses 28 2.5 Reasons for this study 30 3 Methods 32 3.1 Experimental Design 32 3.1.1 Participants 33 3.2 Procedures 34 3.2.1 Trait Variables 34 3.2.2 Independent Variables 36 3.2.3 Dependent Variables 38 3.3 Statistical Analysis 39 4 Results 41 4.1 Descriptive Statistics 41 4.1.1 Socio-demographic Characteristics 41 4.1.2 Moral Consideration of War 42 4.1.3 Dependent Variables 42 4.2 Manipulation Check 43 4.3 Hypotheses Tests 44 4.3.1 Hypothesis 1a 47 4.3.2 Hypotheses 2a + 3a 48 4.3.3 Hypothesis 4a 49 4.3.4 Hypotheses 5a + 6a 50 4.3.5 Hypothesis 7a 53 5 Discussion 55 5.1 Moral Disengagement and Support of War 56 5.2 Moral Identity and Support of War 57 5.3 Methodological Problems 58 5.4 Discussion Main Effects without Additional Assumption 60 5.4.1 Moral Disengagement 61 5.4.2 Moral Identity 63 5.5 Limitations 65 5.5.1 Sample 65 5.5.2 Methods 67 5.6 Conclusion 68 6 References 70 7 Annex 76 7.1 Operationalization Permissive Scenario (German) 84 7.2 Operationalization Prohibitive Scenario (German) 86 7.3 Study Description for Participants (German) 88 8 Acknowledgements 89 9 Selbstständigkeitserklärung 90 / Die Mechanismen der Moralischen Entkopplung lösen das Selbst von internalisierten moralischen Standards und verhindern damit die Selbstregulierung des moralischen Verhaltens. Diese Mechanismen kommen auch in Medien vor und tragen zu unmoralischem Verhalten bei. Die Aktivierung der Moralischen Identität wirkt jedoch den Mechanismen der Moralischen Entkopplung entgegen. In dieser Studie werden beide Konzepte auf das Thema Krieg übertragen. Dabei ist wichtig zu beachten, dass internalisierte moralische Standards Krieg als unmoralisch bewerten. Schließlich basieren sowohl Moralische Entkopplung als auch die Aktivierung der Moralischen Identität auf dieser zusätzlichen Annahme. Zur Überprüfung der Hypothesen wurde ein 2 x 2 RO Between-Subjects Design verwendet. Faktor 1 variierte die Aktivierung von Moralischer Identität (Aktivierung Moralische Identität versus Kontrollgruppe). Faktor 2 variierte die Permissivität eines Kriegsszenarios in einem Zeitungsartikel (Permissives Szenario versus Prohibitives Szenario). Dabei wurde Permissivität hinsichtlich der Anzahl der Moralischen Entkopplungsmechanismen operationalisiert (Viele versus Wenig). Als Organismusvariable ist die Moralische Bewertung von Krieg zur Überprüfung der zusätzlichen Annahme notwendig gewesen. Abhängige Variablen waren die Unterstützung von Krieg und Moralische Entkopplung (Fragebogen). Teilnehmerinnen und Teilnehmer (N=86, f=45, m=41) des online Experiments wurden in vier verschiedene Versuchsbedingungen randomisiert. Die Ergebnisse zeigten, dass die Permissivität des Kriegsszenarios keinen Effekt auf Moralische Entkopplung (Fragebogen) oder die Unterstützung von Krieg hatte, unter Berücksichtigung der Zusatzannahme. Moralische Identitätsaktivierung verringerte Moralische Entkopplung (Fragebogen) aber nur unter der Bedingung, dass Krieg als moralisch bewertet wurde. Moralische Entkopplung hatte keinen Effekt auf die Unterstützung von Krieg, unter Berücksichtigung der Zusatzannahme. Die Interaktion von Moralischer Identitätsaktivierung mit der Permissivität des Kriegsszenarios war nicht signifikant, unabhängig davon ob die Zusatzannahme berücksichtigt wurde oder nicht. Die Ergebnisse werden in Bezug auf die methodischen Probleme bei der Messung internalisierter moralischer Standards diskutiert. Es fanden vermutlich Prozesse der Moralischen Entkopplung bereits während der Messung dieser Standards statt. Der fehlende Interaktionseffekt kann an der schwachen und unspezifischen Aktivierung der Moralischen Identität liegen, sowie nicht berücksichtigter Mediatoren, wie z.B. die Zentralität von Moralischer Identität.:I. CONTENTS I. Contents 2 II. List of Figures 5 III. List of Tables 6 IV. List of Equations 7 V. Abstract 8 VI. Zusammenfassung 9 1 Introduction 10 2 Theory 12 2.1 Moral Disengagement 12 2.1.1 Socio Cognitive Theory of Morality 12 2.1.2 Mechanisms of Moral Disengagement 13 2.1.3 Moral Disengagement and War 13 2.2 Moral Disengagement and Media 17 2.2.1 Moral Disengagement in Media and Aggressive Behaviour 18 2.2.2 Moral Disengagement in Media and War 19 2.2.3 Operationalization of Scenarios 21 2.3 Moral Identity 22 2.3.1 Moral Identity Centrality 22 2.3.2 Moral Identity Activation and Moral Disengagement 24 2.3.3 Moral Identity Activation in Interaction with Moral Disengagement in Media 26 2.4 Hypotheses 28 2.5 Reasons for this study 30 3 Methods 32 3.1 Experimental Design 32 3.1.1 Participants 33 3.2 Procedures 34 3.2.1 Trait Variables 34 3.2.2 Independent Variables 36 3.2.3 Dependent Variables 38 3.3 Statistical Analysis 39 4 Results 41 4.1 Descriptive Statistics 41 4.1.1 Socio-demographic Characteristics 41 4.1.2 Moral Consideration of War 42 4.1.3 Dependent Variables 42 4.2 Manipulation Check 43 4.3 Hypotheses Tests 44 4.3.1 Hypothesis 1a 47 4.3.2 Hypotheses 2a + 3a 48 4.3.3 Hypothesis 4a 49 4.3.4 Hypotheses 5a + 6a 50 4.3.5 Hypothesis 7a 53 5 Discussion 55 5.1 Moral Disengagement and Support of War 56 5.2 Moral Identity and Support of War 57 5.3 Methodological Problems 58 5.4 Discussion Main Effects without Additional Assumption 60 5.4.1 Moral Disengagement 61 5.4.2 Moral Identity 63 5.5 Limitations 65 5.5.1 Sample 65 5.5.2 Methods 67 5.6 Conclusion 68 6 References 70 7 Annex 76 7.1 Operationalization Permissive Scenario (German) 84 7.2 Operationalization Prohibitive Scenario (German) 86 7.3 Study Description for Participants (German) 88 8 Acknowledgements 89 9 Selbstständigkeitserklärung 90

Doping has serious negative consequences for athletes and the integrity of sports, implying the need for effective prevention programs. Since educating young athletes about doping-related knowledge is deemed to be not sufficiently effective to minimize doping, a focus on values, emotions and morality is seen as a promising approach and previous research indicates which variables exactly could be addressed in anti-doping efforts. These variables are anticipated guilt, empathy, moral disengagement, and collective moral norms, since these constructs have been strongly and consistently linked to doping intention, likelihood, or behavior. Therefore, the purpose of this study was to develop a values-based anti-doping intervention, which targets the aforementioned variables, and to evaluate its effectiveness in producing changes in outcomes in comparison to an information-based intervention and a waiting control group. To evaluate their effectiveness, both interventions, which each consist of six 45-min sessions (one session per week) were implemented in a sample of 321 young elite athletes, aged 13–19 years, from a broad range of team and individual sports. Thirty different teams, training groups or classes were randomly assigned to either the values-based intervention, the information-based intervention, or to a waiting control group. Doping intention, doping susceptibility as well as the above mentioned variables were assessed at pre- and posttest and, for participants of the values- and information-based conditions also at a 3 to 4-month follow up. Within a multilevel modeling framework general linear mixed regression analyses revealed that the values-based intervention, compared to the control group, was able to decrease athletes' moral disengagement and increase their anticipated guilt immediately after the intervention (at posttest), whereas no effects for the information-based intervention emerged. Looking at how the outcomes developed over time (i.e., at the follow up measurement), it could be demonstrated that the reduction in moral disengagement sustained. The increase in anticipated guilt, however, was not sustainable and, surprisingly, decreased from post to follow up. Furthermore, athletes in the values-based intervention reported higher empathy from post to follow up, which could possibly indicate a “delayed” effect. This study provides support that a values-based approach can produce changes in some, yet, not all addressed variables and specific elements from this intervention could potentially be a useful addition to traditional anti-doping education (i.e., information provision).