This thesis explores how different leadership styles affect employee engagement during organizational change. As modern organizations frequently undergo structural, strategic, or cultural transformations, understanding the role of leadership in shaping employee engagement becomes critical for managing transitions effectively and sustaining workforce morale.

The research question is: “How do transformational leadership and coercive leadership impact employee engagement during organizational change?”

The study adopts a qualitative research strategy using a phenomenological approach. Semi-structured interviews were conducted with ten professionals from diverse organizational and cultural backgrounds. A six-phase thematic analysis (Braun & Clarke, 2006) was applied to analyze the data, supported by tools such as Taguette (v1.4.1), Microsoft Excel (v2504), and draw.io (V27.0.9).

The findings demonstrate that leadership style significantly influences employee engagement during change. Transformational leadership marked by empathy, inclusion, and clarity was associated with higher levels of emotional resilience, trust, and motivation. Conversely, coercive leadership characterised by control, poor communication, and emotional detachment led to disengagement, anxiety, and frustration. Importantly, engagement outcomes were driven less by the nature of the change itself, and more by how leaders communicated, involved, and supported their teams.

These findings underscore the importance of emotionally intelligent leadership in navigating organisational change. While transformational behaviours helped foster psychological safety and trust, coercive approaches undermined morale and engagement. Practical implications include the need for leadership development programmes that prioritise empathy, transparency, and participatory communication. Ethical considerations around power dynamics and emotional wellbeing are also highlighted. Limitations include a small, non-generalisable sample and the subjective interpretation of leadership behaviour. Future research could explore leadership responses longitudinally or adopt mixed-method approaches to enhance depth and applicability.

Artificial Intelligence (AI) is reshaping decision-making by enabling automation and innovation, but it also introduces risks such as bias, opacity, and ethical concerns. Organizations recognize the need for effective AI governance to ensure fairness, transparency, and legal compliance. Translating these principles into everyday practice remains challenging. Complexity in regulations, rapid technological change, and conflicting stakeholder priorities compound these difficulties.

This thesis investigates the question: What are the challenges and strategies involved in implementing AI governance?

A systematic literature review using the PRISMA framework analyzed peer-reviewed publications from 2019 to 2025. Studies were sourced from academic databases, including Scopus, Springer Link, Web of Science, IEEE Xplore, and ACM. Thematic analysis was then used to identify recurring patterns, challenges, and strategies across the selected literature.

The results highlight several key challenges organizations encounter when implementing AI governance, including regulatory fragmentation, limited transparency of AI systems, ethical and data governance issues, organizational integration barriers, and gaps in AI literacy among staff. To address these challenges, organizations have adopted strategies such as improving transparency, conducting ethics audits, engaging stakeholders, strengthening data governance, implementing risk management frameworks, and using monitoring tools. These approaches integrate governance into organizational processes, build trust in AI, and support regulatory compliance. In the literature reviewed, regulatory fragmentation emerged as the biggest obstacle to effective AI governance. Among these, the ethical integration of AI principles stood out across sectors as the most consistently emphasized strategy to guide responsible and trustworthy implementation.

This thesis offers a holistic, cross-sector synthesis of AI governance implementation. Specifically, each major challenge - regulatory, technical, organizational, and ethical - is outlined with appropriate strategies for overcoming them, providing practitioners with a ready-to-use roadmap. By addressing the identified barriers through targeted measures, organizations can foster the trustworthy and responsible use of AI technologies. The findings provide a foundation for advancing AI governance in a complex technological environment.

This thesis explores how junior consultants at PwC navigate career progression in a structured, performance driven corporate setting, whereas official promotions systems exist, but the real social game is often played behind the scenes. Although PwC has a clear promotion system and individual performance, the experiences for junior consultants reveals a great complex reality. 

Through a qualitative case study by using semi-structured interviews, the study reveals how networking and relationships shape a junior consultant's career progression. The theoretical frameworks are based on Bourdieu's (1986) concept of social capital, Shein's (2010) model of organizational culture along with the human capital theory (Becker 1962). The findings show that playing the so-called “social game” is perceived as significant, where likability, visibility and cultural fit are often the core characterizations within the PwC corporate environment. This thesis contributes to research centering on early career phases in management consulting firms, which is often overlooked. Behind every promotion lies an untold story, whereas this thesis sheds light on those hidden narratives. 

This study compares AI-generated prototypes with prototypes created by studying computer-interaction students. The prototypes were compared in effectiveness, efficiency and satisfaction. Since Artificial intelligence has become more widespread on the market there has been speculation how some roles will be replaced or supported by AI. A large variation of AI-models are available on the market, but in this study the focus was Visily.ai, a generative AI that creates user interfaces and design. This AI prototyping tool is prevalent in the IT-industry. The prototypes created by the students followed a task description, which the AI-prototype followed. The participants were divided into two groups, using the method A/B-testing. Each group completed tasks for only one of the prototypes. The results were that the participants' evaluation of both prototypes were similar. Also, the total time to complete was similar. However, the human created prototype was better in effectiveness, which means that missclicks were much higher with the AI-prototype, almost double the amount compared to the other prototype. The human created prototype was quite similar to the AI-prototype in all measurement values. Thus, the human created prototype was better on some of the tasks but not overall.

Denna uppsats har undersökt hur det digitala communityt BookTok påverkar användarnas bokval, och hur studiens deltagare upplever den digitala gemenskapen på plattformen. Studien genomfördes genom semistrukturerade intervjuer med sju deltagare i åldern 18–25. Materialet analyserades genom en tematisk analys, och de teoretiska ramverken kring Word-of-Mouth, deltagarkulturer och affinitetsrum applicerades. Studiens resultat visade att deltagarna påverkas av den stora mängd rekommendationer som sprids på plattformen, och att det är mängden rekommendationer snarare än innehållet i dessa som påverkar deras litterära val. Vidare framstod känslomässigt innehåll ha en stor påverkan på deltagarnas bokval. Det framgick att en digital gemenskap finns på plattformen, men att deltagarna inte själva är medvetna om denna. Avslutningsvis framkom en stark kritik mot det litterära innehåll som rekommenderas på plattformen.

Uppsatsen “Den Humoristiska Klimatfrågan” är författad av Raen-Osín Thelander inom ämnesområdet medie- och kommunikationsvetenskap vid institutionen för kommunikation och medier vid Stockholms universitet. Syftet med uppsatsen är att undersöka framställningen av klimat- och miljörelaterade frågor i nyhetssatiren Svenska nyheter. Uppsatsen tar hjälp av teorier rörande satir, miljöjournalistik och framing. Vidare undersöks även hur satiren används för framställningen. Slutligen undersöks även frågan om vad satiren tillför den journalistiska granskningen. Analysen utgår ifrån en tematisk metod där fem teman identifierats utifrån upprepningar inom materialet. Ur analysen framkom resultatet att satiriska verktyg som ironi och hån används i Svenska nyheter främst för att kritisera politiska makthavare genom att synliggöra hyckleri och okunskap i frågan, men även för att kritisera traditionella nyhetsmediers sätt att framställa frågan, samt påvisa att klimat- och miljöfrågan är polariserad. I studiens slutsats fastställs att satiren i Svenska nyheter kan förstås tillföra ytterligare en kritiserande faktor till den traditionella nyhetsjournalistiken.

This thesis explores how male-led feminist content is structured on TikTok and whether their content can serve as a catalyst for greater male participation in feminist efforts. The study first consists of an inductive thematic analysis of 98 educational TikTok videos, followed by a survey with 83 male respondents aged 18–35 to assess the perception and potential of such pro-feminist content.

Findings show that creators reject hegemonic masculinity and promote values aligned with feminism, yet few explicitly reference feminism or address structural patriarchy. Survey results reveal generally positive attitudes toward feminism, especially among self-identifying feminists, though traditional views on masculinity persist. Engagement was largely shaped by pre-existing beliefs, with content resonating most with supportive audiences. While TikTok is seen as a valid space for feminist dialogue, offline conversations are perceived as more valuable. Future research must examine how to better reach those uncertain about their stance on feminism and non-feminists to more effectively challenge patriarchal norms.

Drag queens have become increasingly popular in the news media over the years, even in Sweden, where they have been part of mainstream culture for a long time. This thesis investigates how Swedish media discusses drag queens, to reveal underlying attitudes towards minorities. The topic is explored through qualitative content analysis, through the lenses of Social Representation Theory, Queer Theory and Moral Panic Theory, which offer the possibility to investigate how media discourses influence public opinion, the way drag queens embrace concepts that are central in queer studies, and how these two elements can generate moral panics. The results reveal that the representation of drag is extremely nuanced, presenting both positive and negative attitudes, supportive and excluding narratives. Finally, this thesis contributes to a better understanding of how media have a role in shaping an understanding of drag, and the mechanism they can use to enforce exclusion or inclusion.

This study investigates the underlying ideologies of Finnish and Swedish news media by analyzing news texts about underwater infrastructure disruptions in the Baltic Sea between 2023–2024. Using an inductive, multi-method design—first conducting thematic analysis on news headlines and leads, then applying news frame analysis to a purposive sub-sample across four outlet types (public service TV, commercial TV, broadsheets, tabloids)—five dominant frames emerged: intention, threat, support, negotiation, and economic. The study revealed that both countries foreground intention and threat frames, but Swedish outlets mirror official narratives more strongly and portray events as distant with limited economic fallout, whereas Finnish outlets convey heightened vulnerability and financial concern. In Finland, broadsheets emphasize culprit and conflict narratives and tabloids highlight economic impacts; in Sweden, framing is broadly consistent, though tabloids occasionally invoke a threat frame via Finnish-Russian tensions. These findings reveal how national context and media format shape crisis narratives, underscoring the need for further cross-national, longitudinal, and computational framing research.

This research investigates the impact of paywall strategies on newspaper availability in Sweden and Finland, focusing on young adults aged 18–25, news consumption, perceptions, and attitudes on paid news. Using mixed methods, it analyzes 18 newspapers and interviews 16 young adults, finding Swedish newspapers have more paywalls (59%) than Finnish newspapers (29%), contradicting the media subsidies’ decrease in paywalls. Lifestyle-, Culture & Entertainment-, and Sport-related content are heavily paywalled, while content related to economics and politics is less restricted. Young adults find paywalls irritating and often bypass them through social media, driven by cost-benefit evaluations and loss aversion by Kahneman’s behavioral economics. The Media Day approach demonstrates the integration of news into daily routines, while the Attention Economy emphasizes competition for attention. Paywalls risk information inequalities and news avoidance, undermining civic engagement.

Studien “Om kriskommunikationen kommer” är en kritisk diskursanalys ur ett hegemoniskt perspektiv av MSB:s informationsbroschyr Om krisen eller kriget kommer. Dess syfte är att besvara hur MSB har utformat sin broschyr och vilken effekt det kan få på läsarens tolkning av texten och de fenomen vilka texten berör. Arbetet redovisar för hur broschyren porträtterar kris, krig och invånaren och använder sedan detta material i ett analysschema i fyra delar. Den första delen behandlar avsändaren, kontexten och läsarens förförståelse av dessa. Den andra delen fokuserar istället på den utformning och ton som används och vilka konnotationer det kan orsaka. Den tredje delen granskar hur aktörerna i språket utformas på en mikrostrukturell nivå. Den fjärde och sista delen går djupare in på Gramscis teori om det sunda förnuftet och hur det kan växa fram eller bli påtagligt i broschyren.

Analysen påvisar att MSB:s broschyr fokuserar på krig och invånarens skyldighet i högre mån än andra kriser och vad som är statens skyldighet. Vidare, menar författaren, kan man genom kritisk diskursanalys se spår av hur broschyren försöker omforma vad som är sunt förnuft.

Active ownership has been identified as a key factor in ensuring the long-term success of public e-services, especially within the transformational government context. However, without a clear measurement approach, it is difficult to apply it in practice. This study addresses this research gap by developing a structured measurement toolkit for assessing the amount of active ownership of public e-services. First, thematic analysis was used to derive measurable dimensions of active ownership. Then, a scoping literature review was conducted using the PRISMA-ScR protocol to identify existing measurement methods and units associated with each dimension. Finally, findings were synthesized through thematic analysis into a comprehensive measurement toolkit. As a result, 15 dimensions of active ownership were defined. For each dimension, specific measurement methods and units were determined and formed into a toolkit consisting of 10 assessment instruments. This toolkit provides a systematic and integrated approach to evaluating active ownership, allowing public organizations to identify capability gaps and guide strategic interventions to improve e-service governance. The study offers both theoretical and practical contributions, providing a foundation for future empirical validation and enabling public sector leaders to proactively manage and enhance the capabilities needed for a public e-service to succeed over time.

The fashion industry is increasingly criticised for its lack of transparency and unverifiable sustainability promises. This has led to an erosion of consumer trust and undermined responsible consumption. This thesis aims to design, implement and evaluate a user-centric, blockchain-based smart label system to optimise transparency and trust in sustainable fashion. As part of this study, a prototype was developed using a design science research approach that integrates a Hyperledger Fabric-based blockchain network, QR code-enabled smart labels and a user-oriented web interface. This system enables decentralised, tamper-proof documentation of sustainability-related product information that consumers can access in real time.

The prototype was evaluated using the System Usability Scale (SUS) and qualitative interviews with five participants. The results show excellent user-friendliness (SUS score of 88) and demonstrate that direct access to detailed, verifiable information on the supply chain and sustainability can strengthen trust and positively influence purchasing decisions. Participants appreciated the transparency and reliability of the system and expressed a greater willingness to choose products with accessible, trustworthy sustainability information.

The results demonstrate that an open, modular and user-oriented blockchain solution is technically feasible and effective in promoting trust and transparency in a sustainable fashion. The work concludes with practical recommendations for future implementations and highlights the relevance of interoperable systems in light of new regulatory frameworks such as the European Union's Digital Product Passport.

Artificial intelligence (AI) is changing the very foundation of the education landscape. The purpose of this study is to investigate how the use of AI affects the engagement of IT students and what the primary driving forces behind its possible change are. A qualitative research strategy was adopted by gathering data through semi-structured interviews with seven students from the Department of Computer and System Sciences at Stockholm University. The study used thematic analysis with the purpose of identifying patterns throughout the students’ experiences, reflections and feelings regarding the use of AI within their studies. The results of the study showed that AI could both enhance and undermine the engagement of students. The use of AI affected cognitive engagement negatively through reduced reflection but also positively through improved understanding. Emotional engagement was affected, as shown by the interviews, which shed light on feelings of enthusiasm and concern. Behavioral engagement also changed, in terms of reduced human interaction alongside increased study efficiency. This study aims to contribute to the existing knowledge of how AI influences the engagement of students. This study reveals an increased awareness among IT students of the possible downsides that AI can have, regarding engagement in an educational context, which partly contrasts with previous research.

Background: Artificial Intelligence (AI) holds significant potential to transform mental healthcare. This thesis investigates how mental healthcare professionals perceive and experience the integration of an AI-based “clinical co-pilot” tool designed to support clinical documentation within their daily routines. Understanding how clinicians perceive and integrate AI tools into their workflows is critical for successful implementation and sustainable digital transformation.

Methods: This study employed a sequential explanatory mixed-methods design. A quantitative survey was completed by 93 mental healthcare professionals to assess attitudes toward an AI-based clinical documentation tool. This was followed by semi-structured interviews with 8 unit-level managers, exploring the factors that influence AI adoption and integration in clinical practice. The study drew on multiple theoretical perspectives, including systemic and organizational-level theories, professional and workplace-level theories, individual behavioral theories, and cognitive and psychological theories. This comprehensive, multi-theoretical approach facilitated a deep exploration of the various factors influencing AI adoption.

Results: The findings revealed that healthcare professionals' willingness to recommend the AI tool was strongly associated with perceptions of practical utility, educational support, and preserved clinical autonomy. Significant variability in AI adoption was observed across different professional roles, age, and prior experience, with local leadership, organizational culture, and clinical context emerging as key factors. 

Business Insights: The study provides actionable recommendations for organizations seeking to enhance AI adoption. These include developing comprehensive training programs and maintaining clinicians' professional autonomy, while carefully balancing individual adaptation with organizational cohesion. Empowering leadership and fostering a strong organizational culture at all levels are also essential. In addition,establishing a clear governance structure and ongoing monitoring of key performance indicators (KPIs) are critical for sustained AI adoption.

Conclusion: Successful AI adoption in mental healthcare is not solely a technical challenge but a complex process shaped by human, organizational, and contextual factors. By understanding these dynamics, organizations can develop more effective strategies for integrating AI into clinical workflows, ensuring both improved patient care and optimized operational efficiency.

Denna studie bygger på min kandidatuppsats Rasprofilerad? - En kvalitativ studie om svarta mäns upplevelser av interaktionen med polis (Koné, 2024) om svarta mäns erfarenheter av interaktioner med polisen. Medan min tidigare studie behandlade svarta mäns upplevelser av polisens kontroller generellt, fokuserar föreliggande studie på svarta mäns upplevelser av visitationer i utsatta områden. Eftersom en ny lag om säkerhetszoner (visitationszoner) infördes strax efter min första studie, och hittills endast inrättats i utsatta områden, undersöks även svarta mäns kännedom om och inställning till visitationszoner samt skillnader i erfarenheter för dem som bott i visitationszoner jämfört med tidigare. Trots att visitationszonerna syftar till att förbättra ordning och bekämpa brott, finns en oro för ökad diskriminering, då svarta män ofta stereotypiseras som gärningsmän. Syftet med studien är att undersöka hur visitationer i utsatta områden påverkar och upplevs av svarta män. Syftet är vidare att undersöka svarta mäns uppfattning av och inställning till visitationszoner. Studien består av sex semistrukturerade intervjuer med svarta män mellan 18-30 år, boende eller tidigare boende i utsatta områden och visitationszoner, analyserade med teorier om proceduriell rättvisa, stämpling- och stigma, obehag samt Hirschmans (1970) begrepp om ‘’exit’’ och ‘’voice’’. Resultatet sammanfattas i tre teman: Misstänkliggörandet, strategier och konsekvenser, Upplevelser av makt och kontroll, samt Påverkan på identiteten. Intervjupersonerna rapporterar negativa erfarenheter av visitationer, inklusive hot, våld, att känna sig utpekade och marginaliserade, samt att de använder strategier för att uttrycka missnöje. De berättar även att visitationerna kan leda till maktmissbruk och kan påverka deras självbild.

Upplevd trötthet har såväl fysiska som psykiska effekter och kan exempelvis försämra individens emotionsigenkänning. Syftet med denna studie var att undersöka om det finns ett indom-individuellt samband mellan självskattad trötthet och förmågan att känna igen emotioner hos andra. Forskarna hade tillgång till redan färdiga data som samlats in genom ett bekvämlighetsurval. Studien omfattade fem respondenter (fyra kvinnor och en man), varav en exkluderades från vissa analyser på grund av att Karolinska Sleepiness Scale (KSS) inte besvarades. Deltagarna var i åldern 22-24 år. Enligt Spearman´s Rho och LOWESS-analyserna fanns en möjlig tendens till samband mellan självskattad trötthet och emotionsigenkänning hos två av deltagarna, efter att inlärningseffekten kontrollerats för. Resultaten, som både överensstämmer och avviker från tidigare forskning, diskuteras vidare i uppsatsens diskussionsavsnitt. Resultaten kan utgöra en grund för framtida forskning och bidra till utveckling av arbetslivet.

Rare earth elements (REEs) are critical raw materials with growing demand in green technologies and limited global supply. Iron oxide-apatite (IOA) deposits can host significant REE concentrations, particularly in minerals such as fluorapatite, monazite and allanite. This study investigates the occurrence, distribution and mineral hosts of REEs in drill core BB-14-005 from the Blötberget IOA deposit, Bergslagen, Sweden. A combination of core logging, XCT-XRF scanning (Orexplore GeoCore X10), petrography and SEM-EDS was used to identify REE-bearing phases, assess their distribution in relation to lithological units and evaluate the performance of XCT-XRF scanning as a tool for visialization of REE bearing minerals in IOA deposits. Two distinct mineral assemblages were identified: (1) in magnetite-dominated massive iron ore, REEs are hosted primarily by fluorapatite and allanite; (2) in hematite-dominated banded iron ore, REEs occur in fluorapatite, monazite, allanite and minor xenotime. Elemental data from XCT-XRF scanning correlate well with assay results (R²=0.9-0.97), though the system underestimates P and Ca. 3D tomographic models highlight high-attenuation zones interpreted as REE-bearing phases. While XCT-XRF scanning was helpful in the early stages of the study for visualizing general REE distribution throughout the core, its effectiveness is limited where REEs are hosted in low-concentration phases such as fluorapatite and may significantly underestimate total REE content in such settings.

Data often flows through disconnected systems and departments in large organizations, making it difficult to trace its origin, transformations, and ownership. This fragmentation leads to duplicated efforts, inconsistent reporting, and limited confidence in analytics. These challenges became increasingly apparent within the Purchase-to-Pay (P2P) process of a global industrial firm migrating from legacy platforms to a consolidated data lake architecture. This thesis investigated whether a data mapping approach can capture and maintain reliable data lineage, as well as dismantle data silos in a fragmented data landscape. Guided by the research question, “How can a standardized data mapping approach be employed to capture and maintain reliable data lineage across Atlas Copco’s heterogeneous systems?”, the thesis followed a Design Science Research (DSR) methodology. Through 24 stakeholder interviews, site observations, analysis of organizational documents, and prior literature on similar artifacts, a visual framework was developed to map end-to-end data flows. Triangulating these sources enabled the developed artifact to identify lineage breaks, highlight data silos, clarify ownership, and quality controls. Applied in a real operational setting, the artifact revealed hidden inefficiencies, reduced manual reconciliation, and improved alignment between business processes and technical systems. The thesis concluded that making data lineage and silos visible and actionable in complex organizational settings can improve operational coherence and contextualization. It provides a repeatable blueprint with for large firms pursuing data-centric transformation.

This thesis explores the human ability to assess manipulated media and examines whether IT expertise influences detection performance. While machine-aided detection methods like deep learning have shown promise, human judgment remains understudied in identifying fabricated images, videos, and audio. The study investigates whether Swedish office IT professionals and non-IT professionals differ in their ability to to detect deepfakes and if general awareness or knowledge of generative artificial intelligence affects detection accuracy. A quantitative approach was employed using an online survey, with statistical analyses including T-test and Spearman’s correlation to compare the groups and explore how familiarity affects accuracy. The findings indicate that human ability to detect deepfakes is limited, with respondents correctly identifying 57.6% of the presented content. Although IT professionals reported higher accuracy and greater certainty, the difference was minimal, suggesting that IT expertise alone does not significantly enhance detection. Instead, a higher level of familiarity with relevant concepts was associated with better performance, emphasizing the importance of increasing public awareness and education to counter the challenges posed by manipulated media. Future research could examine deepfake detection in realistic, contextual settings, such as social media, using a collaboration of human and machine-aided approaches.

Introduction

Negation detection is essential in biomedical NLP to correctly interpret medical records. Greek remains underrepresented compared to other languages, due to limited annotated corpora. This thesis develops and evaluates a negation detection system for Greek biomedical text using a deep learning approach, comparing it with rule based methods to improve automated biomedical text processing.

Research Question

This study examines: "How effectively can fine tuned transformer based models detect negation in Greek biomedical text, compared to rule based approaches?" BERT based models, fine tuned on Greek biomedical text, are evaluated against a NegEx adaptation, analyzing strengths and limitations in handling negation cues and scope.

Method

A quantitative experimental approach is applied, utilizing a Greek biomedical negation corpus to fine tune GreekBERT and XLM-RoBERTa for negation detection. Performance is assessed against a rule based NegEx system using precision, recall, and F1-score. Preprocessing includes tokenization, data augmentation, and other NLP techniques.

Results

The results demonstrate that the deep learning based approach offers a clear advantage over the rule based system in handling negation in Greek biomedical text. The transformer models demonstrate high accuracy and reliability, effectively capturing complex negation patterns that challenge traditional rule based methods.

Discussion

The findings confirm that transformer based models can robustly detect negation in Greek biomedical text. While performance is strong, handling complex or implicit negation is challenging, highlighting opportunities for future improvement.

Denna studie syftar till att utveckla och pröva undervisningsmaterial för gymnasiekursen Naturkunskap 1b med fokus på naturvetenskapens karaktär (NOS) och arbetssätt, särskilt tolkning och analys av data. Med utgångspunkt i Vision II av scientific literacy designades en undervisningssekvens där elever arbetade med autentiska dataset och övade på att tolka, analysera och dra slutsatser. Studien bygger på en designbaserad forskningsansats och involverade fyra erfarna lärare som samarbetade i planering, genomförande och gemensam bedömning. Resultaten visar att undervisningen delvis lyckades stödja elevers NOS-förståelse, men också att begreppsliga och kontextuella svårigheter begränsade förståelsen. Vidare visar studien att det kollegiala arbetet – särskilt genom korsbedömning – bidrog till ökad samsyn, reflektion och utveckling av undervisningen. Studien visar hur systematiskt samarbete kan öka likvärdighet och kvalitet i undervisning för SL, men belyser också behovet av mer tid, tydlighet och kontinuitet för att sådana insatser ska få långsiktig effekt.

Toxicity in online games is a prominent issue in the rapidly growing gaming industry. Although titles such as League of Legends and Counter-Strike 2 are often associated with high levels of toxicity, there is a lack of clear understanding of how game design and social interaction within different genres contribute to the emergence and reinforcement of these behaviors, making it difficult for developers to effectively address the problem. This thesis aims to investigate how specific design patterns contribute to a more toxic gaming environment. Through formal analysis, as well as participatory observation in a total of sixty matches, divided between the two earlier mentioned games, key design patterns such as communication channels, downtime, and competition are identified as particularly contributing factors. The results show that it is not individual design patterns themselves that cause toxicity, but how they interact under certain gameplay conditions. By highlighting these connections, the thesis provides valuable insights for future game design aimed at fostering more inclusive and positive gaming environments.

This thesis explores the user experience and attitudes in relation to generative AI-content on the two of the biggest social media platforms in the world: TikTok and Facebook. With the rise of generative AI-tools in the last couple of years the content generated from AI is increasingly becoming integrated into both content creation and the social media atmosphere. Through a qualitative approach with semi-structured interviews and a thematic analysis, the study investigates Swedish people from generation z and how they perceive awareness, relevance, authenticity and emotions when interacting with AI-generated content. The results revealed a predominantly negative attitude against generative AI-content, especially on Facebook. The respondents reported a varied awareness in being able to recognise content derived from AI-tools and how they felt about it. While most of the respondents could appreciate AI-generated content for creative and entertainment purposes, they expressed concerns about manipulation, misinformation, stolen creative intellect and the loss of human touch in the content. This study aims to open up the discussion on how AI-content affects our social media climate and indirectly our user experience and awareness for misleading content. 

The emergence of deepfake technology poses growing challenges to information security, particularly as manipulated media spreads rapidly through social media platforms. Generation Z is one of the groups that spends the most time on social media, and therefore they should be the most exposed to deepfake content or manipulated content. Several studies have investigated the human ability to detect manipulated images and deepfake videos. However, there is currently a limited amount of research specifically focused on Generation Z’s ability to identify such content, as well as how this ability relates to their social media use and awareness of information security. This study addresses the research gap by examining how social media use and information security awareness influence Generation Z’s actual ability to detect deepfakes. The research question posed was “what is the relationship between social media use, information security awareness, and the ability to identify deepfakes among individuals within Generation Z?”. For the thesis, a quantitative study was conducted that involved people aged 15–30 living in Sweden. The participants took part in a survey built around a questionnaire with questions regarding social media use, information security awareness, and a test with 16 examples of audio, images and videos. The examples were then graded on the likelihood of being real or deepfake on a scale from 1–7. The collected data was first analyzed descriptively and inferential statistics were then conducted to strengthen the analysis, including Kruskal-Wallis and Spearman’s rank correlation tests. The results indicate that participants' ability to identify deepfake content was generally limited, and varied depending on media type. Audio clips were the easiest to identify correctly, while video content posed the greatest challenge. These findings suggest a knowledge gap in information security, even among individuals in Generation Z who are highly digitally engaged. No significant correlations were found between social media use or information security awareness and ability to detect deepfakes. The study highlights the need for targeted educational efforts within information security as well as media and information literacy, and may be useful for developing programs that enhance young people's resilience against digital manipulation.

Unnecessary amounts of food are being wasted in today's society. With an increasing global population, the global meat waste is being driven to higher levels as a result of increased consumption and production. The production of red meats has the largest amount of carbon monoxide emissions that negatively affects our climate. In addition, there is a large knowledge gap regarding how well different large language models are able to forecast demand using historical data from a grocery store. This study aims to find out the precision different large language models have during forecasting with the goal of reducing meat waste in grocery stores by answering the following research question: With what accuracy can different AI models - ChatGPT, Mistral, DeepSeek, Gemini - predict the demand for red meat in grocery stores using past data? The chosen method includes the research strategy experiments which used to measure how different variables affect each other. By randomly choosing grocery stores from a list of possible candidates were we able to, with the usage of document analysis of secondary data, gather 2.5 years of weekly historical data of four types of red meats from a grocery store. With the usage of predictive analysis as the data analysis method, together with evaluation methods like MAE, MSE and MAPE, were we able to analyze the precision of the different models, which made it possible to present the results using descriptive analysis. The results show that the ChatGPT-o3 model had the highest precision value at 87.67% while the DeepSeek-R1 model had the lowest precision at 78.49%. The Google Gemini model had the second best precision coming in at 81.22% while the Mistral Le Chat Pro model had a precision of 80.53%. The results also showed that all the models were relatively stable with standard deviations between 4.0 and 8.0 percentage points. Gemini showed the most instability with weekly fluctuations reaching 20 percentage points when it comes to standard deviation. By analysing the results, were we able to establish that the ChatGPT model performed the best, similar to the findings of other related studies. Another interesting find was that most of the models showed weekly fluctuations when it came to precision, which indicates that some forecasts may have been randomly generated. Beyond that the ChatGPT model showed the highest stability and lowest amount of weekly fluctuations. Furthermore, the ChatGPT model also beat all the other models at creating forecasts with higher precision in the weeks where there was a campaign, high season or pay week, while other models showed a bigger weakness with a lower precision. However, the remaining models had a higher precision, inline with ChatGPT, for weeks where such factors did not exist. The results show major opportunities for implementation with other types of products, but also in other contexts where lots of data needs to be analyzed. With these results are we able to come to the conclusion that the ChatGPT model had an average precision of 87.67%, the Gemini model had an average precision of 81.22%, the DeepSeek model had an average precision of 78.49% and that the Mistral model had an average precision of 80.53%. Furthermore, all the models showed a precision that was higher than 70% when it came to forecasting using historical data from a grocery store. Based on the conclusion, we believe that the possibilities of using LLM are great and endless. Different stores can implement low-cost solutions to try to reduce their meat wate using such forecasts.

Climate change is rapidly reshaping marine ecosystems and disrupting fisheries. Coastal fisheries such as Northern Ireland’s Nephrops fishery (known locally as Portavogie prawns and internationally as Norway lobsters) face mounting pressure to adapt or transform in the face of compounding stressors. This thesis assesses the climate resilience of the Northern Ireland Nephrops fishery using an attribute-based framework that evaluates ecological, socio-economic, and governance dimensions. It explores how interactions among resilience attributes shape the system’s response to six climate-related risks. The analysis reveals several key vulnerabilities, including economic precarity, an ageing fleet, governance fragmentation, and ecological constraints. It identifies how hidden risks to long-term viability arise from abundance, where high Nephrops biomass masks ecological simplification and market entrapment. It also highlights an intensifying “spatial squeeze” created by the intersection of immigration policy, territorial access restrictions, and offshore renewable energy development, further constraining the fleet’s adaptive space. More broadly, it demonstrates how systematically examining climate resilience attributes and their interactions across different climate change impacts can reveal critical weaknesses and opportunities.

This study aimed to highlight women’s experiences of working in male-dominated industry and to understand how barriers and opportunities affect their professional roles. This was investigated using semi-structured interviews with nine women, all of whom work at the same male-dominated company, where the data was then analyzed thematically. The results consist of five themes that describe women’s percived characteristics, workplace culture, challenges, opportunities and factors that influence the decision to stay. This included descriptions of soft and hard characteristics, experiences of jargon and collaboration, various barriers and favorable conditions, and motivations for continuing in the industry. The conclusion was that women adapt to a male-dominated enviroment, need to prove their competence, but also see opportunities and use characteristics as strengths. Impotant factors for staying are collegiality, freedom, support from managers and development opportunities. The study shows that women are active change agents and reshape gender norms from within.

Vuxna med adhd upplever ofta betydande utmaningar med att upprätthålla en fungerande balans mellan arbete, fritid och familjeliv, vilket i sin tur kan påverka deras psykisk hälsa och livskvalitet. Trots att tidigare forskning har påvisat att internetbaserade interventioner (IBI) kan vara kostnadseffektiva, kliniskt beprövade och ge goda behandlingsresultat, saknas det etablerat stöd för adhd-patienters egna upplevelser av IBI. Syftet med studien var att undersöka upplevelserna och de upplevda effekterna av två internetbaserade stresshanteringsprogram utvecklade för yrkesverksamma med adhd, i förhållande till respondenternas psykiska hälsa, vardag samt arbetsliv. Åtta respondenter genomförde en ny intervention baserad på kognitiv beteendeterapi (KBT), samt fyra genomförde en intervention med fokus på medveten närvaro. Semistrukturerade intervjuer med tolv yrkesverksamma (nio kvinnor och tre män) genomfördes och analyserade med kvalitativ innehållsanalys. Tre huvudteman identifierades som belyser respondenternas generella upplevelser av programmet, att självständigt arbeta med en IBI samt upplevda stödfunktioner; Uppfattningar om programmets användbarhet i livet med adhd, att jobba med programmet och dess psykologiska påverkan i vardagen och arbetslivet samt upplevelsen av stöd i ett självständigt upplägg, med totalt nio underkategorier. Subkategorierna behandlade positiva och negativa upplevelser, tidsrelaterade svårigheter, förbättringsförslag, användarvänlighet, innehåll och övningar, tanke- och beteendemässiga förändringar, kontakt med forskargruppen, att arbeta självständigt samt stöd från omgivningen. Dessa teman tydliggör respondenternas behov av en IBI med en tydlig struktur, enkel och lättillgänglig, inslag av behandlarstöd samt erbjuder tydliga och tillämpbara strategier anpassade efter individens situation. Framtida studier bör inkludera dels ett större urval dels undersöka långtidseffekter av IBI för adhd-patienter. 

Despite advances in natural language processing, automatic hate speech detection remains challenging, especially for coded and implicit hate speech. While transformer-based models like BERT and ALBERT have improved contextual understanding, they often struggle with evasive language use. The research investigates "To what extent can ALBERT improve the detection of coded and implicit hate speech compared to other NLP models in this domain?" in order to contribute in the ongoing technical advancements in automatic detection of coded and implicit hate speech. To this end, we fine-tuned an ALBERT model on a dataset that underwent targeted augmentation to 10% of the data to better represent real-world patterns. The resulting model, IHSD-BERT as we named it, was evaluated against hateBERT using a similarly augmented tertiary dataset. The results showed that IHSD-BERT outperformed hateBERT in accuracy by 8%, in macro-F1 by 35%, and in weighted-F1 by 23%. Notably, IHSD-BERT demonstrated a 416% improvement in F1-score for true positives, from 6.49% in hateBERT to 33.46% in IHSD-BERT with an increase in recall by 537%. These findings suggest that lightweight models like ALBERT, when combined with targeted augmentation, can improve the accuracy of nuanced classification tasks like coded and implicit hate speech detection.

As organizations face escalating cyber threats amidst rapid digital transformation, effective cybersecurity management has become crucial for ensuring operational resilience and strategic continuity. This study conducts a systematic literature review (SLR) of 100 peer reviewed publications from 2019 to 2025 to explore what organizations can learn from cybersecurity management science. By integrating bibliometric analysis using VOSviewer with thematic analysis, following Braun and Clarke’s six phase framework, and supported by Taguette software, this research provides a comprehensive understanding of the field at both macro and micro levels. The bibliometric analysis reveals dominant research areas including governance, compliance, risk integration, and organizational culture. The thematic analysis identifies five major themes: cybersecurity governance and compliance, risk management and strategic integration, human factors and organizational culture, cybersecurity capability and maturity, and technical foundations supporting management. These findings underscore the shift in cybersecurity from a purely technical domain to a strategic, multidisciplinary concern. The study proposes a layered governance model to translate academic insights into practical guidance, contributing to both scholarly discourse and real world cybersecurity strategy.

Apache Kafka is a high-throughput and distributed messaging system designed to handle large volumes of data feeds with scalability and reliability, serving as an essential part of event-driven architecture. Despite its advantages and popularity, Kafka’s high degree of flexibility reduces the incentive for academic researchers to develop universal resource utilization strategies, making the study of this kind underexplored. A few of very recent studies have initiated research to address this generalizability issue, placing emphasis on responding backpressure from producer-side, but with limited consideration given towards consumer-side. In many IoT scenarios, though, a messaging middleware follows a fire-and-forget model, where message durations are transient, ending once the message processed by its recipient. This study seeks to address this research gap in developing generalized, real-time resource allocation strategies tailored to any Kafka-involved architecture characterized by such IoT cases with minimal storage requirements, by allowing it to learn operational pattern of the consumer-side. This project is initialized with an artificial IoT use case to produce synthetic datasets, ensuring a controlled and repeatable environment with experiments. The result of this study, based on quantitative analysis, demonstrates that a single broker maintains linearity of message lifetime against producing time on a per-batch basis, which reflects Kafka’s append-only design principle as the chosen data structure. Aligning well with this fundamental, the project firstly demonstrates the basis for future implementation of resource optimization with a potential to generalize. This knowledge basis obtained is subsequently applied, upon which a persistent data storage load-shedding engine is developed. Working together, the engine modules coordinate consumers behavior data, make predictions for log retention time setting for topics, and then automate configuration adjustments in real-time for saving persistent storage. Overall, this project serves as a proof of concept for Kafka-based architecture in IoT use cases sharing specific characteristics, with a light-weight local setup. Even though it relies on an artificial use case with synthetic data generated, it fills a gap in high-level Apache Kafka performance and resource utilization research with a possibility of broad-grained optimization around broker layer and creates opportunities to prevent disk failure using unsupervised machine learning algorithms.

This case study explores how coordination and mutual understanding occur in a context of collective dance improvisation. The research interest and purpose of the study is to increase the understanding of the communicative and coordinative activities in collective dance improvisation, thereby contributing to knowledge of collective learning processes. Empirical material was collected through participant observation and semi-structured interviews and analyzed using thematic analysis with an abductive approach. The theoretical framework draws on theories of collective learning in action and concepts from sensemaking. The results show that the dancers’ communicative actions for coordination and mutual understanding constitute of bodily reading, collective bodily reading, collective movement patterns and renegotiations within the ongoing sensemaking collective dance improvisation. The dancers make themselves available to one another by being readable and adaptable in the moment. Through intuitive and embodied responses, mutual understanding and coordination emerge, for which the level of attention is crucial. Embodied hesitation and assumptions hinder the process. The dancers’ mutual understanding is also created in the verbal dialogue, but as often as it creates understanding, it causes confusion. To clarify meanings, verbal interpretations must be tested in practice to establish mutual understanding. By remaining in the dancing activity and playfully renegotiating and challenging the action, it is plausibility, rather than precision, that guides the social learning in the ongoing ambiguous improvisational landscape. By highlighting the potential of a more holistic and embodied approach to collective learning, current dominant theories can be expanded and lead to further knowledge.

Introduction: This thesis aims to address the challenge of accurately predicting customer churn in a highly competitive and digitized retail environment. As customer loss can lead to significant financial consequences, the aim is to support businesses in identifying at-risk customers early, enabling more effective retention strategies and reducing the risk of customers switching to competitors. Specifically, the thesis explores how interpretable machine learning models can be leveraged to uncover behavioral patterns predicting churn, evaluate the effectiveness of predictive models, and estimate customer lifetime value in order to assist in better resource allocation strategies for companies.

Research question: The research questions of this thesis are:

1. What customer behavior patterns are the most significant indicators of churn in a non-contractual retail setting?

2. What machine learning model provides higher predictive accuracy in identifying customer churn in a non-contractual retail setting?

3. How can machine learning models be applied to estimate customer lifetime value in a non-contractual setting?

Method: The research strategy employed in this thesis is experimental in nature, as it aims to systematically explore and evaluate the performance of machine learning models. It takes a quantitative approach, using supervised machine learning models to analyze patterns in customer behavior and predict churn. By applying the quantitative methods of logistic regression, decision trees and random forest, the thesis aims to uncover key predictors of churn, determine the best model and estimate customer lifetime value. The data analyzed in this thesis is the synthetic dataset called Online Retail Customer Churn Dataset and was retrieved from the Kaggle platform.

Results: The results showed that decision trees outperformed logistic regression and random forest in terms of identifying churners. The analysis of feature importance highlighted transactional patterns and engagement behavior as the most important predictors of churn. To estimate customer lifetime value with machine learning a binary variable was needed to enable classification. The strongest predictors of high or low customer lifetime value across all models were spending behavior and marketing response variables, and the random forest model achieved the best F1-score on the better predicted class.

Discussion: The findings suggest that customers’ transactional patterns and engagement behavior are more reliable signals of future churn than who they are demographically. From a business perspective, this insight could be valuable for targeting retention efforts. The decision tree outperformed the other models in detecting churners, which is often the primary focus in churn prediction. Ultimately, the choice of model should be guided by the specific objectives of the churn management strategy and the relative costs of false positives and false negatives. Additionally, the findings suggest that in non-contractual settings, customer lifetime value estimation with the help of machine learning is best approached through modeling with behavioral and demographic features. Some limitations of the study are reproducibility, concerning the code, validity, concerning the historic customer lifetime value approach, and generalizability, due to the relatively small dataset. Further research could explore why transaction amount consistently emerges as a strong predictor in churn, why the number of returns stood out as the most important feature in the decision tree, and study customer lifetime value on a deeper level to predict monetary values and to either corroborate or refute identified predictors.

This thesis investigates how generative AI influences the work processes, creativity and professional identity of web developers. Generative AI tools, such as ChatGPT and GitHub Copilot, have rapidly become integral parts of modern programming and web development. While previous research has mainly focused on quantitative effects like increased productivity, there is a need for deeper qualitative understanding of how developers perceive and experience these tools in their daily work.

The study aims to explore developers’ experiences of using generative AI, with particular focus on how it affects their creative processes, sense of professional identity and the quality of their work. The research question guiding the study is: How do web developers perceive the impact of generative AI on their work processes, creativity, professional identity, and code quality?

A qualitative research method was used, based on ten semi-structured interviews with participants from a web development bootcamp and the data were analyzed using thematic analysis.

The results show that generative AI was used frequently to support learning, problem-solving, and debugging. Participants described how AI tools provided personalized explanations and accelerated their understanding of complex coding issues. Furthermore, AI enhanced creative workflows by reducing cognitive load and enabling faster prototyping. At the same time, many participants expressed concerns about becoming overly dependent on AI and emphasized the importance of maintaining problemsolving skills.

The study also reveals that AI influenced how participants saw themselves as developers, fostering greater confidence and a sense of belonging in the tech field. The ability to craft effective prompts emerged as a valuable new skill. While some voiced concerns about the future job market, most believed that developers will adapt by integrating AI into their workflows rather than being replaced by it.

These findings contribute to the broader understanding of how generative AI reshapes not only the technical but also the human aspects of software development.

The purpose of this thesis is to examine how the Sweden Democrats’ (SD) stance towards the EU has changed between 1992 and 2024. The essay examines how ideological continuity and strategic change can coexist, and how the party’s official positions relate to its practical politics. Our analysis is based on a qualitative method of ideological analysis that focuses on SD:s party platforms and their ideological evolution as well as a quantitative content analysis that focuses on SD:s election manifestos and their evolving political stance on the EU. The theoretical framework draws on concepts such as conservatism, nationalism, and euroscepticism, as well as the relationship between ideological consistency and pragmatic adaptation. The results show that the Sweden Democrats’ ideological core has remained intact but its expression has shifted over time, possibly in response to changing political circumstances. Most notably, the party’s EU policy has moved from a clear call for withdrawal to a more reform-oriented approach, reflecting a broader European trend among like-minded right-wing populist parties. At the same time, a tension emerges between the party’s ideological rhetoric and its concrete political positions. Thus, the thesis ultimately highlights how the Sweden Democrats’ ideological flexibility enables tactical adaptation without fully abandoning the party’s fundamental values.

Under senare decennier har det politiska beslutsfattandet blivit mer inbäddat i en komplex miljö, där regeringen i allt högre grad påverkas av aktörer utanför den statliga sfären. Samverkan och kommunikation sker inte längre på den korporativistiska modellens villkor, där intresseorganisationerna garanterades formell representation i ämbetsstyrelser och statliga kommittéer. Politisk påverkan sker i stället genom lobbning och informella nätverk där möjligheterna till insyn är begränsade. De statliga kommittéerna lever kvar och deras funktion som arenor för kompromisser och samförstånd är särskilt relevant i dagens polariserade politiska klimat. Det finns dock indikationer på att regeringens styrning hämmar konflikter och därigenom begränsar möjligheten till debatt och kompromisser inom de statliga kommittéerna. Syftet med denna uppsats är att undersöka hur regeringens styrning antingen förstärker eller mildrar konflikter inom det statliga kommittéväsendet, med fokus på kommittéer med utredningsuppdrag. Studien består av två delar. Den första innefattar en statistisk analys av 500 kommittéer och i den andra delen undersöks sex av kommittéerna närmare med hjälp av fallstudier. Resultaten visar, å ena sidan, att regeringen kan förstärka konflikter genom att tillsätta fler riksdagsledamöter och intresserepresentanter i kommittéerna, uppmana till samråd med utomstående aktörer, framhäva konflikter i direktiven och ge kommittén frihet att utforma egna förslag. Regeringen kan, å andra sidan, mildra konflikter genom att ställa krav på delredovisning, hänvisa till befintliga regler och beslut i direktiven samt tillsätta fler regeringsrepresentanter i kommittéerna.

This study aimed to examine the process that led to the emergence of the peace agreement New Decade, New Approach (NDNA), trace its development, and analyse it through the lens of historical institutionalism. NDNA is the latest of the eight (8th) agreements in the seemingly perpetual peace process of Northern Ireland and by studying the process on how it was negotiated we may gleam important knowledge about the state of current political actors in the ongoing peace process and their core issues.

The process around NDNA was defined by its lack of progress and the unwillingness of the political actors to compromise, despite significant consequences for the citizens of Northern Ireland. By using theory testing process-tracing methods this study shows that the political actors of Northern Ireland and the surrounding negotiation institutions was influenced by earlier policy decisions and institutional structures in a way that hinders developments. Despite being essentially in agreement, the process took three years due to the inability to change positions once taken. The theories of path dependency and critical junctures explains to a degree why the process developed as it did and is supported by the findings of this study.

Despite the rise in national climate policies, current efforts remain inadequate to prevent the adverse effects of climate change. Evidence shows that only a small fraction of policies enacted over the past two decades have resulted in measurable emission reductions, underscoring the need to better understand what drives climate policy effectiveness. This study explores how political institutions shape the effectiveness of national climate policies, defined as the extent to which policies lead to reduced emissions. Analyzing over 1,000 mitigation policies across 41 countries from 1997 to 2021, the study applies recent advances in causal policy attribution to link institutional features to policy outcomes. Results from fixed and mixed effects regression models reveal that both higher levels of democracy and stronger government power are positively associated with policy effectiveness. While democracy correlates with a greater number of effective policies, it does not increase their proportion. In contrast, stronger government power increases both the number and share of effective policies. No significant effects are found for government ideology or corruption. These findings highlight the critical role of institutional context, especially democracy and government power, in enhancing climate policy outcomes.

Since the year 2002 politicians in Swedish municipal assemblies have had the option to have an item on the assembly agenda returned for reconsideration by minority vote. This essay studies this tool and aims to answer the questions of how such actions are justified by the opposition and whether this tool can be used in a strategy of political obstruction in the context of Swedish municipal politics. The essay achieves this by analyzing the justifications of such actions offered by the opposition in the municipality of Nynäshamn during the year 2023 and comparing those to the opinion pieces written by the same parties in the local news paper. The essay also analyzes the effects this political tool could have on the trilemma and tension between people’s rule, the rule of law and political efficiency.

The essay concludes that the available material is insufficient in order to determine whether any opposition party used their ability to have items on the assembly agenda returned for reconsideration by minority vote in order to delay or obstruct the agenda of the majority. What can be concluded however is that if that was the goal of a minority in a Swedish municipality they could achieve it by acting just like parts of the opposition in the municipality of Nynäshamn in 2023. While the full scope of the effects this tool could have on the before mentioned trilemma it is clear from previous and international research that such tactics do often influence the rules and laws of the parliaments they are employed in. 

The essay suggests that further research delves into the various factors that could contribute to a municipality having more or fewer of these minority induced returns for reconsideration.

This essay contributes to a deeper understanding of how the concept of cultural heritage is implemented, articulated and politicized within the Swedish political sphere. It does so by examining party platforms, ideological programs, and principle statements produced by the political parties represented in the Swedish Parliament in 2025, as well as structured interview responses from 2022 by each party’s cultural policy spokesperson. Through qualitative dokument and discourse analysis, the thesis explores how cultural heritage is conceptualized, which functions it is assigned, and to what extent it is used as a tool for identity formation, symbolic politics, or ideological positioning. The study thus explores similarities and differences in how the political parties employ cultural heritage policy and the concept of cultural heritage itself, as well as whether these are connected to the parties’ ideological positions. The study explains these patterns by analyzing and interpreting Laurajane Smith’s theory of the Authorized Heritage Discourse (AHD), which is used both as an analytical lens and as a critical explanatory model for identifying patterns and ideological differences in how cultural heritage is framed by each party. Preliminary findings suggest that political parties across the ideological spectrum engage with cultural heritage in distinctly different ways, with variations in emphasis on national identity, inclusion, diversity, and cultural preservation. 

This thesis investigates how Donald Trump’s 2025 discourse on Gaza can be understood as a fascist spectacle that legitimizes the US-Israeli domination over the Palestinian people. Drawing on Guy Debord’s theory of the spectacle and Federico Caprotti’s characteristics of fascist spectacle, the study applies John B. Thompson’s method of depth hermeneutics to five press events and one viral video. The analysis identifies four symbolic characteristics -myth of the leader, aestheticization of politics, militarized peace, and depoliticized displacement as central features of Trump’s fascist spectacle.

The results show that Trump constructs himself as a singular redeemer of Gaza, reimagines occupied land as real estate fantasy, recasts military occupation as care, and frames Palestinian displacement as humanitarian relocation. The findings demonstrate that these narratives function ideologically by erasing Palestinian agency and reframing control as moral necessity.

The thesis contributes to situationist critique of Trumpism. It extends Caprotti’s framework beyond its original historical context and shows how Debord’s concept of ideology materialized applies to the spectacle of contemporary foreign policy discourse. The study concludes that Trump’s vision of Gaza is not merely a proposal but a symbolic act of power. A vision where domination is rendered beautiful, inevitable, and unquestionable through the spectacle’s logic.