Publikationer från Stockholms universitet
Not a member yet
31237 research outputs found
Sort by
Lower secondary school students’ conceptions about their effort in mathematics education
No full textPrevious research has shown the importance of affect when seeking knowledge about students’ achievements. However, there are surprisingly few studies looking at students’ expressed effort on a longer time scale. Using data spanning over almost three decades, in this paper, we analyze Swedish lower secondary school students’ responses to items in relation to large-scale assessments in 1992, 2003, and 2019. The two general results were that students express a lower level of goal-directed motivation, whereas the responses regarding students’ motivation as an evaluation of effort, most students state that they do not give up when facing a difficult task, that they would learn more and do better if they put more effort into their studies, and that they are happy with their achievement. Looking at changes between the different years, several items show a significant difference in the Swedish students’ answers between 1992 and 2003 and no difference by 2019. The results might explain the changes in achievement that took place between 1992 and 2003
Incentivizing Health: A Theoretical and Empirical Evaluation of the Effect of Wellness Policy on Health Outcomes
No full textThis thesis explores the effectiveness of wellness policies in promoting healthy behaviors by combining a theoretical model with empirical evidence from a U.S. wellness initiative: the Comprehensive School Physical Activity Program (CSPAP), which is designed to improve adolescent health and combat obesity. The model pre- dicts that only individuals with intermediate preferences are likely to change their behavior in response to policy interventions when the primary mechanism is in- creased accessibility to wellness activities (e.g., physical activity). It also suggests that policies may have a greater impact on health outcomes when accessibility is combined with educational interventions. Empirically, the results reveal a statis- tically significant two-year-lagged reduction in obesity rates, particularly in states with higher adoption intensity. This finding is consistent with the model’s predic- tion: pairing accessibility with education can amplify the policy’s effect on health outcomes. The effect is estimated to have lowered national obesity among high school students by up to 2.7 percentage points in 2022. These findings draw at- tention to the importance of program intensity and sustained implementation in shaping public health outcomes
Nature Revisited : On the construction of the aesthetics of sustainable architecture in early twenty-first century architectural discourse
No full textLooking back on the setting twentieth and dawning twenty-first centuries, the pursuit of environmental sustainability brought about a new landscape for architectural discourse to navigate, and new topics for architectural theory to grapple with. In the early 2010s, the question whether there could be an aesthetics of sustainable architecture, and if so, what it should be like, rose to the fore as an issue demanding attention. In attempting to settle a tension or conflict between the notions of aesthetics and sustainability, some instances of architectural theory turned to the notions of nature and the natural, thus joining a long tradition of Western architectural discourse. This dissertation explores the aesthetics of sustainable architecture as it takes shape as an object of discourse within some instances of architectural theory at this time, more specifically the roles played by the notions of nature and/or the natural in this process. Using a discourse theoretical approach combined with an understanding of nature/society dualism as a way of discursively representing the world, the dissertation combines deeper, text oriented analyses of three enunciations of the discourse on the aesthetics of sustainable architecture from the early 2010s with an ideational exploration of a more comprehensive mass of extended material. The latter forms the basis of the first part of the study, which argues that insofar as the issue of the aesthetics of sustainable architecture falls back upon a conflict between the notions of aesthetics and sustainability in archietctural discourse, this conflict can be understood as a reiteration of a longstanding tension within architectural theory, dating back to the breakthrough of Kantian aesthetics and the idea of disinterestedness. Through closer analysis of three specific texts from the early 2010s, the second part of the study shows how the aesthetics of sustainable architecture taking shape in these texts are both made possible and restrained by what is conceptualised as a dualist discourse on (un)sustainability. Within such a discourse, sustainable architecture is constructed as an architecture involved in a harmonious relation to nature or the natural. The dissertation further shows that the aesthetics of sustainable architecture in question are cast as integral to the establishment and maintenance of such a harmonious relation, while at the same time operating as solutions to the longstanding tension informing the issue of the aesthetics of sustainable architecture. Disintegration and imitation are identified as two possible flexions of a dualist discourse on (un)sustainability, or as two possible means by which to represent architecture and its aesthetics as sustainable. A key argument of the dissertation is that as they play out in the three texts submitted to closer analysis, both imitation and disintegration involve a discursive naturalisation of sustainable architecture through its aesthetics, which goes hand in hand with a renegotiation of the notion of aesthetics as such. This renegotiation involves a decoupling of the notion of aesthetics from the idea of individual authorship, a recasting of the object of aesthetic experience as something other than a visually construed object-thing, and a construction of this object as partly natural.
Feasibility of internet-based multimodal emotion recognition training in adolescents with and without autism : A pilot study
No full textBackground: Research suggests that individuals with autism spectrum disorder (ASD) have difficulties in emotion recognition (ER), which could lead to social difficulties. ER can be enhanced through targeted interventions, but generalization to everyday functioning poses a challenge. Using dynamic multimodal emotional expressions for training may increase similarities to everyday situations. This pilot study investigated the feasibility of internet-based multimodal emotion recognition training (iMERAT) for adolescents with ASD. Method: Eight adolescents with ASD and nine typically developing (TD) adolescents took part in the iMERAT intervention, which included brief online training sessions conducted each weekday during a 3-week period. Training was performed on dynamic facial, vocal and multimodal emotional expressions, with outcome feedback provided after each response. A survey was conducted to explore participants' experiences of the training. ER was measured pre- and post-training using a multimodal ER test. Results: Participants reported that the training was moderately difficult, instructions were relatively easy to understand, and the duration of training was appropriate. Content analysis of open-ended responses suggested further adaptations, such as providing more explanations of emotions and further tailoring content and language for adolescents. ER increased from pre- to post-intervention, with large effect sizes for both ASD and TD adolescents. Conclusion: Results suggest that the iMERAT intervention is feasible for adolescents with ASD. Gains in ER ability were observed, but the small sample size and lack of a control group render these findings tentative. Further research is required to assess the effectiveness of the iMERAT and possible impact on broader social skills
AI i spel: Påverkan på sociala aspekter inom MMORPG-spel
No full textIntroduktion: Denna studie undersöker sambandet mellan introduktionen av AI kompanjoner inom spelgenren Massive Multiplayer Online Role-Playing Game och spelares sociala upplevelse. Att förstå detta samband är viktigt då det påverkar spelares beteenden i ett socialt sammanhang, samt hur det påverkar spelarnas upplevelse av spelen på en social nivå. Forskningsfråga: Studiens mål är att besvara frågan: Hur uppfattar spelare skillnaderna mellan mänskliga interaktioner och AI-drivna roller i sociala spelmiljöer, och hur påverkar AI spelares engagemang och upplevelse av gemenskap? Problemställning: Problemet denna studie försöker lösa är gapet i forskningen om AI:s påverkan på spelares uppfattning av MMORPG:er. Gapet i forskningen beror på att det inte gjorts mycket forskning efter AI:s ökade spridning 2022. Metod: Studien är en fallstudie och kommer att genomföras i två faser, en kvalitativ data fas och en kvantitativ data fas. Fas 1 är utförandet av 10 kvalitativa semistruktuerade intervjuer med erfarna spelare från spelet Final Fantasy XIV. Den data som var utvunnen från dessa intervjuer kommer sedan att gå igenom en tematisk analys. Detta leder in i fas 2, då en kvantitativ enkät ska byggas från informationen framtagen genom den tematiska analysen. Denna enkät kommer skickas ut till forum på Reddit och Discord som är kopplade till spelet Final Fantasy XIV. Den data som utvunnits från dessa enkäter kommer att använda sig av endast likertskalor för att lättare utvinna och presentera information från resultatet. Resultat: Resultatet visar på att AI har gett en plats för spelare att öva och testa sig själva i en miljö där de inte känner sig dömda. AI ger inte spelaren den information som krävs för att utveckla som spelare. Spelare kommer inte lära sig om gruppsynergi och hur de ska spela spelet för att prestera på den nivå som spelet och andra spelare förväntar. AI:s beteende anses vara stelt och omänskligt, men de försöker härma mänskligt beteende som spelare har visat. Spelare som spelar med andra spelare har en större oro över prestanda, men också har det roligare. Spelare uppskattar återkoppling som de får av andra spelare och till stor del gillar att ge återkoppling till andra, vilket är något som AI saknar. När det kommer till upplevelsen av gemenskap skiljer sig detta åt bland spelare. Flera anser att AI har haft lite påverkan på gemenskapen i spelet, medans andra anser att den har haft en stor påverkan. Generellt visar resultatet på att majoriteten av spelare fortfarande föredrar att spela med andra och bevara gemenskapen, men känner att AI kan hota detta i framtiden. Diskussion: Dessa resultat visar på att AI har en påverkan på spelare och spelet och att AI:s implementation i spelet upplevs på olika sätt av spelare. Med detta sagt håller majoriteten av spelare om att AI inte kommer kunna ersätta den sociala aspekt som finns i spelet. Final Fantasy XIV borde fortsätta sin utveckling av olika AI-system så länge det inte hotar dessa sociala aspekter. Studien har varit begränsad i vilka forum som kontaktades för datainsamling, olika tolkningar av ord och fraser kopplade till ämnet, samt AI:s ständiga utveckling kan påverka framtida resultat. Vidare forskning kan utforska hur AI-systemen i spelet påverkar olika typer av spelare inom Final Fantasy XIV. Utforskning av andras MMORPG:s påverkan av AI-system skulle också bidra till djupare diskussion av frågan.Introduction: This study examines the relationship between the introduction of AI companions in the computer game genre Massive Multiplayer Online Role-Playing Game and players' social experience. Understanding this relationship is important when it can affect players' behavior in a social context, as well as how it possibly affects players' experience of playing on a social level. Problem: The problem that this study tries to solve is the gap in research on AI's impact on players' perception of MMORPGs. The gap in research is consistent with the fact that not much research has been done after AI's increased spread in 2022. Research Question: The goal of the study is to answer the question: How do players perceive the differences between human interactions and AI-driven roles in social game environments, and how does AI affect players' engagement and experience of community? Method: The study is a case study and takes place in two phases, a qualitative data phase and a quantitative data phase. Phase 1 consists of 10 qualitative semi-structured interviews with experienced players from the game Final Fantasy XIV. The data extracted from these interviews will then go through a thematic analysis. This leads into phase 2, where a quantitative survey is to be built from the information produced by the thematic analysis. This survey will be sent out to forums on Reddit and Discord related to the game Final Fantasy XIV. The survey will only consists of questions following the Likert scale. This is to more easily extract and present information for the results. Result: The results show that AI has provided a place for players to practice and test themselves in an environment where they do not feel judged. AI does not give players the information they need to develop as players. Players will not learn about party synergy and how to play the game to perform at the level expected by the game and other players. AI's behavior is considered rigid and inhuman, but they try to mimic human behavior that players have shown. Players who play with other players have a greater concern about performance, but also have more fun. Players appreciate the feedback they get from other players and largely like to give feedback to others, something that AI lacks. When it comes to the experience of community, this differs among players. Several believe that AI has had little impact on the community in the game, while others believe that it has had a large impact. Overall, the results show that the majority of players still prefer to play with other players and maintain its community, but feel that AI may threaten this in the future. Discussion: These results show that AI can have an impact on players and the game, and that AI's implementation in the game is experienced in different ways by players. That said, the majority of players agree that AI will not be able to replace the social aspect of the game. Final Fantasy XIV should continue its development of various AI-systems as long as it does not threaten these social aspects. The study has been limited in which forums were contacted for data collection, different interpretations of words and phrases connected to the topic, as well as the constant development of AI may affect future results. Further research could explore how the AI-systems in the game affect different types of players within Final Fantasy XIV. Exploring the impact of AI-systems in other MMORPGs would also contribute to a deeper discussion of the question
Händelsebaserad rymdavbildning och omloppsbanabestämning av satelliter för rymdsituationsmedvetenhet
No full textThe increasing reliance of our society on space-based services such as satellite communication, navigation systems and weather forecasting demands a maintenance of Space Situational Awareness (SSA). SSA efforts aim to ensure the safety and protection of objects in orbit around the Earth. An important aspect of SSA is tracking and determining the orbits of satellites to secure safe operations and avoid collisions. Orbit determination in the optical typically relies on observations with conventional frame-based sensors. However, frame-based imaging tends to carry highly redundant information as each frame stores data from every pixel over a fixed exposure time. This redundancy can be avoided by using an Event-Based Camera (EBC) that asynchronously records binary signals of intensity changes in each pixel. In this thesis, we present two successful observations of satellites utilising one of the latest types of EBCs. These satellites could be identified solely based on the observational data cross-referenced with a comprehensive satellite database. We also present a software framework, designed to effectively filter out event-based data not associated with satellite streaks. Moreover, we have performed coordinate transformations and attempted Initial Orbit Determinations (IODs) as well as orbit updates. The observed satellite streaks had a maximum angular separation of ±0.213°, requiring a corresponding maximum allowable error in azimuth and elevation of ±5e-6° to ensure suitable accuracy in the IODs. However, the estimated error in our angular measurements after converting from pixel coordinates to sky coordinates in the observer frame was estimated to ±0.04° — far exceeding the permissible threshold and thereby preventing any viable orbit determination. We consequently suggest implementing either a high-precision pointing mechanism or a larger field of view, ideally both, to enhance accuracy and data quality. While a complementary frame-based sensor could be beneficial for the coordinate transformations, our results conclude that the current-technology EBCs do not require a complementary optical sensor to successfully identify satellites as well as faint stars on the sky. We also showcase the camera’s potential for determining the shape and orientation of satellites, provided the satellite is large or close enough to be marginally resolved. Lastly, we demonstrate that our EBC reacts to events on a 2 μs timescale — which can have several applications in fast optical astronomy, especially in observing transient and dynamic phenomena such as supernovae, binary stars and stellar pulsations
Aerosol-fog interactions in the Italian Po Valley
No full textAtmospheric aerosol particles, tiny solid or liquid particles suspended in air, play a crucial role in environmental processes by interacting with both radiation and clouds, including fog. These interactions contribute to a range of environmental and health-related issues, both on their own and through complex interactions. One of the most polluted regions in Europe is the Italian Po Valley, frequently covered in fog. With a more than 30-year long history in aerosol-fog measurements, the research station San Pietro Capofiume (SPC), located in a rural area near Bologna, is an ideal site for studying aerosol-fog interactions in a changing atmosphere. Despite the fact that aerosol-fog interactions have been extensively studied worldwide, especially in recent years, many uncertainties and challenges persist. This thesis aims to investigate aerosol-fog interactions in the Po Valley, contributing to a deeper understanding of fog processes such as formation and dissipation, as well as shifts in fog occurrence in the context of a warming climate and cleaner air. These insights are relevant for various fields such as air quality or the transportation sector, but are e.g. also needed to validate fog model results and remote-sensing retrievals of aerosol particles and fog with detailed in-situ measurements. Therefore, during winter and spring 2021/22, we conducted the Fog and Aerosol InteRAction Research Italy (FAIRARI) campaign at the research station SPC. We measured processes on the scale of gas molecules up to droplets and captured not only the wintertime fog period but also the springtime new particle formation period and the transition, which was associated with an increase in the air temperature and global radiation, as well as a decrease in PM2.5 concentration. We found a large variability in the microphysical characteristics of the fog events, which lasted between a few minutes and 13 hours. While the dry aerosol mostly consisted of particles around 100 nm, the droplets grew to diameters around 20 μm. Organics and nitrate dominated the dry mass of the measured submicron chemical composition. In the fog residuals, nitrate and sulfate fractions were enhanced, whereas the organic aerosol fraction was reduced compared to the ambient aerosol during non-fog conditions. Notably, the organics present in the fog residuals showed a higher relative contribution of organic nitrogen, potentially resulting from aqueous-phase fog processing. Moreover, we provide new κ-values for several organic compounds and show that a value of 0.1 reliably estimates their hygroscopicity in both supersaturation-limited (Po Valley) and particle-limited (Arctic) environments. The high amount of nitrate found in the Po Valley led to an overall high hygroscopicity of the particles. The high number concentration of accumulation mode particles in combination with high hygroscopic growth factors make hygroscopically grown - but not activated - particles important for various reasons: These hydrated aerosol particles do not only reduce visibility through scattering but also influence the interpretation of aerosol-fog relationship measurements. For example, the effective diameter of the droplets increases by 81 % when excluding hydrated aerosol particles, while the cloud droplet number concentration decreases by 87 %. The contribution of hydrated aerosol particles to the liquid water content was shown to be dependent on the accumulation mode number concentration and the supersaturation. Moreover, not the total number concentration but the shape of the upper tail of the dry particle number size distribution was shown to be crucial when predicting fog microphysical parameters using Large Eddy Simulations. Additionally, hydrated aerosol particles impact the physical and chemical measurements of interstitial aerosol and fog residuals as those are defined by ambient diameter thresholds which are lower than the largest hydrated aerosol particles. This thesis emphasizes the importance of aerosol chemical composition and particle size in determining aerosol-cloud and aerosol-fog behavior across contrasting environments. It highlights the role of fog and aqueous-phase processing in transforming aerosol properties and affecting visibility, climate-relevant fog parameters, and potentially human health.ERC INTEGRATEH2020 FORCe
Dynamic Adaptation of ML Algorithms for Real-time Resilience (Ensemble Learning Enhanced with LLM-Based Interpretation)
No full textThe growing complexity and volume of cyber threats necessitate adaptive Intrusion Detection Systems (IDS) capable of real-time learning. Traditional IDS models rely on static datasets and offline training, making them ineffective against zero-day attacks and dynamic threats. To address these limitations, this study explores the integration of ensemble learning and online learning techniques augmented by a lightweight retrieval-augmented Large Language Model (LLM) to enhance IDS adaptability, detection accuracy, and interpretability. the research problem this study aims to address is: In resource-constrained settings, online learning algorithms need to be effectively coupled with light-weight LLMs to build an interpretable online intrusion detection system that adapts in real time, improves detection and adaptation of unknown/novel threats, and provides human-understandable explanations while reducing retraining costs. The proposed model combines multiple online classifiers, Adaptive Random Forest, Online k-Nearest Neighbors, and Online Naïve Bayes, within a stacking-based ensemble, while the LLM acts as a second-stage verifier that filters potential false positives and generates concise natural-language explanations. This architecture supports continuous model updates without full retraining, allowing the IDS to adapt dynamically to evolving attack patterns. Evaluated on the CIC-IDS2017 dataset, the system attains 95.99% precision, 99.14% recall, and a 97.54% F1-score for known attacks when the LLM module is enabled. For previously unseen attacks, the IDS achieves 96.29% precision, 92.15% recall, and a 94.17% F1-score after online adaptation. With more features included in RAG, it achieves better performance in recall with 93.81% precision, 95.27% recall and 94.55% F1 score. These results confirm its ability to learn from new threat signals in real time. These findings demonstrate that the proposed LLM-enhanced IDS meets key requirements for real-time learning, high detection performance, and human-readable explanations, offering a practical path toward more resilient cybersecurity defenses in dynamic network environments
Between Independence and Integration : The Pursuit of Self-Determination in Non-Self-Governing Territories
No full textSince the height of the post-Second World War (WWII) decolonisation wave, self-determination has emerged as a cardinal principle in international law. Today, the realisation of self-determination for all peoples is viewed as an unexceptionable goal and an imperative for the full enjoyment of all human rights. At the same time, scholars have identified the very concept of self-determination as particularly vague and ill-defined. While it is a cornerstone of international law, it is paradoxically also one of its most unsettled norms. This thesis argues that self-determination’s unsettled nature hampers peoples’ abilities to exercise or call out violations of it, even where there is an undisputed right. This becomes evident if one considers the peoples of the world’s remaining non-self-governing territories (NSGTs), defined by the United Nations General Assembly (UNGA) as not yet decolonised, which make up the empirical focus of the thesis. NSGT peoples have an unalienable right to self-determination as part of their right to decolonisation. While the vast majority of colonies transferred to independent statehood in the post-WWII decades, NSGTs to this day remain in constitutional relationships with their respective administering (colonising) powers. Studies have found that movements and electoral support for full independence in NSGTs now tend to be limited. Metropolitan states have relatedly objected to the categorisation of their affiliated territories as colonies, seeking to portray the very retention of constitutional ties as a voluntary act – perhaps even an act of self-determination? The thesis argues that such a binary view is overly limited and risks obscuring self-determination violations that still occur within these relationships, as well as the limited opportunities for major constitutional change that have existed in practice. International law on decolonisation stipulates a set of political status options that NSGT peoples may pursue. However, it is less clear regarding what self-determination entails and legitimises in relation to contemporary arrangements for metropolitan affiliation, for as long as they may persist. Using various types of qualitative and field data, including addresses by NSGT representatives to UN decolonisation committees, local news articles and interviews with key actors in two NSGTs and British Overseas Territories – the British Virgin Islands and Anguilla – this thesis inquires into these arrangements. It considers how they are experienced on the territory-level, how they shape local discourses and concepts of self-determination, and how they are in turn thought to promote or undermine it. By analysing this data in light of writings from law and political theory pertaining to Indigenous self-determination and decolonisation, the thesis develops an understanding of self-determination and its requirements within frameworks for metropolitan affiliation. Building on legal scholar James Anaya’s ideas about self-determination as having an ongoing and a constitutive aspect, it argues that this needs to be considered in regards to how the governing institutional order is structured and who hold power in both modifying and upholding it. Identifying shortcomings pertaining to both ongoing and constitutive self-determination in NSGT-metropole relationships, the thesis demonstrates the value of conceptualising this in a more elaborate way, aside from a choice for a permanent political status
Exploring the Role of Adult-Child Preschool Play Pedagogy in the development of Education for Social Sustainability:Perspectives from Icelandic and Swedish Preschool Pedagogues.
No full textThe aim of this research is to explore if and how role and imaginary play in ECEC settings can contribute to fostering education for social sustainability. The research draws on documentation from ECEC educators engaged in Lindqvist’s (1995) play pedagogy during an Erasmus exchange between Icelandic and Swedish preschools. Rooted in Vygotsky’s theories of play, aesthetics, and cultural development, this research builds on Lindqvist’s (1995) early childhood play pedagogy with emphasizes on the role of socio-dramatic play between adults and children in the creation of collaborative idiocultures of play (Fine, 2012; Lecusay and Nilsson, 2024). The study draws on an interpretivist, qualitative methodology (Creswell, 2013). Narrative analyses together with concepts from drama pedagogy (Hallgren, 2018; 2022) are applied to characterize educators' reflections and perspectives on the relationship between children's play and social sustainability (Lindqvist, 1998; Nilsson and Lecusay, 2024). Informed consent was obtained from the participating educators. Preliminary findings indicate that educators perceive character and fantasy play in ECEC settings as contributing to social sustainability. Their reflections highlight how imaginative environments and active educator facilitation shape their understanding of the relationship between play and social learning, as well as their role in fostering relational trust—key components in developing sustainable educational practices. These findings suggest that national ECEC policies should emphasize the value of adult–child fantasy play as part of sustainability education. Practice can be supported through professional development and leadership initiatives that help educators integrate collaborative storytelling and drama into everyday pedagogy to foster inclusion and social learning