Publikationer från Stockholms universitet
Not a member yet
31237 research outputs found
Sort by
How and why upper secondary students use generative AI tools in business administration
No full textDenna uppsats undersöker hur och varför gymnasieelever använder generativa AI-verktyg som hjälpmedel i ämnet företagsekonomi. Undersökningen gjordes med hjälp av en digital enkät i fyra klasser på två olika skolor med totalt 86 enkätsvar, samt en gruppintervju med 7–8 elever på respektive skola. Med hjälp av tematisk analys identifierades olika kategorier bland elevsvaren.Resultaten visar att de vanligaste fördelarna med att använda AI-verktyg som eleverna nämner är att fördjupa eller säkerställa sin förståelse, effektivisera sitt arbete och spara tid samt studiestöd i olika former som exempelvis sammanfattning, planering, förhör och som en extra lärare. Den överlägset vanligaste risken som eleverna ser är att AI kan ge felaktig information. Majoriteten av eleverna tycker inte att det är försvarbart att använda AI för huvuddelen av olika inlämningsuppgifter. De flesta elever försöker antingen hitta en balans mellan att få snabb hjälp av AI och att bygga långsiktiga kunskaper eller prioriterar långsiktigt lärande. Eleverna har knappt fått någon information i skolan om hur man använder AI och de upplever att många lärare är negativa till AI.Baserat på resultaten ser vi implikationer för undervisning såsom att lärare behöver planera för att säkerställa elevers förståelse, undervisa elever mer om AI, förändra inlämningsuppgifter och bedömning som en konsekvens av AI samt få löpande kompetensutveckling då AI:s intåg i skolan förändrar lärarrollen
Stochastic optimal control and stopping, games and time inconsistency
No full textThis thesis contributes to the field of stochastic optimisation considering a version of a dividend problem, a stochastic differential game with incomplete information, as well as game-theoretic approaches to time-inconsistent stopping and control. The content of this thesis is based upon four papers. Paper I considers a game theoretic approach to a time-inconsistent stopping problem, where the time-inconsistency is due to non-exponential discounting. We introduce a novel class of mixed stopping strategies and provide a verification theorem. Furthermore, we consider an example, where there is no equilibrium when using only pure stopping times. In this case we are able to construct an equilibrium in the class of mixed stopping times. Paper II considers a continuous time stochastic controller stopper game with incomplete information. The stopper can be seen as owner of an asset and a controller as the manager who is either effective or non-effective. The manager earns a salary paid by the owner. An effective manager can choose to exert effort at a cost in order to increase the drift of the asset while a non-effective manager cannot act. The owner can choose to stop the game at any time based on observations of the movement of the asset. The owner is not able to observe directly whether the manager is effective or non-effective, making this a game of incomplete information. We derive a Nash equilibrium to this game, given as a threshold solution depending on the conditional probability that the manager is effective. Paper III considers a time-inconsistent singular stochastic control problem, where the time-inconsistency is due to non-exponential discounting. We introduce a new class of "mild" threshold controls, which are given by an exploding rate that generates an inaccessible boundary for the underlying diffusion. These "mild" controls stand in contrast to the "strong" threshold controls that have been considered previously and amount to a Skorokhod reflection at an upper boundary. We provide an appropriate equilibrium condition for these controls as well as a verification theorem. Furthermore we provide an example, where no equilibrium exist if we only consider "strong" threshold strategies. We are, however able to find an equilibrium when considering "mild" threshold control strategies. Paper IV considers a dividend problem with ruin at zero surplus or if the surplus spends too long time below a certain threshold of distress. We completely solve the problem considering three different cases. If the distress threshold is small or large the optimal control results in paying out dividends above a certain threshold. If the distress threshold takes intermediate values, the optimal control results in paying out dividends in two separated regions. Collectively, these results advance the theory for optimal stochastic control and stopping, by enriching the literature with new problems, as well as presenting solution structures that have not been considered previously
Human-in-the-loop learning: Making smarter and safer AI decisions
No full textIntroduction, This thesis explores the benefit of utilizing Human-in-the-loop (HITL) in the Budgeted UCB algorithm. Addressing this limitation is critical, as the budgeted UCB algorithm lacks mechanisms to incorporate human knowledge. This deficiency often leads to unnecessarily costly explorations and terminal constraint violations, which pose a safety risk and reduce the life cycles of IoT devices, a critical factor in resource sensitive systems. Research Question, The study answers the research question “To what extent does human-in-the-loop impact the total performance of a budgeted UCB algorithm, measured in cumulative regret, in a simulated IoT environment?” Method, A controlled and isolated Python-simulation experiment was conducted for evaluating the impact of Human-in-the-loop. The method involved comparing the budgeted UCB baseline against three HITL variants: oracle with perfect knowledge, a noisy expert that does probabilistic errors and an estimate based variant. These were evaluated within a single simulation run, with dynamically changing constraints to replicate a non-stationary environment. Results, The results demonstrate that the HITL-extension reduced cumulative regret by approximately 34% compared to the standard budgeted UCB algorithm. Discussion, These findings suggest that HITL-extended Budgeted UCB algorithms can be of crucial value for IoT-devices and scenarios where constraint violations are harmful. Further research could explore theoretical results and the impact in real world scenarios to determine if the simulated results could transfer to actual IoT environments
När stödet riskerar att brista : Speciallärares erfarenheter av arbetet med elever med läs- och skrivsvårigheter i övergången mellan årskurs 6 och 7
No full textDenna studie syftar till att undersöka hur speciallärare beskriver sitt arbete med att stödja elever med läs- och skrivsvårigheter i övergången mellan årskurs 6 och 7. Studien genomfördes mot bakgrund av tidigare forskning och granskningar som visar på brister i övergångsarbetet. Det finns även ett behov av ökad kunskap om speciallärares roll i denna process. Ett särskilt fokus riktas mot förebyggande insatser och samverkan. Studien tar sin teoretiska utgångspunkt i Bronfenbrenners ekologiska systemteori och i det sociokulturella perspektivet. En kvalitativ metod med semistrukturerade intervjuer användes, där tio speciallärare verksamma på mellanstadiet och på högstadiet deltog. Resultatet indikerar att tidiga och förebyggande insatser, pedagogiska anpassningar och ett nära samarbete mellan professioner är centrala för elevernas utveckling och för att skapa trygghet och kontinuitet. Samverkan mellan speciallärare, lärare, elevhälsa och vårdnadshavare framträder som avgörande. Samtidigt framkommer brister i struktur, dokumentation och informationsöverföring vid stadieövergången. Studien synliggör att ett mer systematiskt och långsiktigt övergångsarbete behövs på organisatorisk nivå. Detta är nödvändigt för att säkerställa en likvärdig och inkluderande utbildning för elever med läs- och skrivsvårigheter
A Monetary Policy Shock and Stock Returns : Evidence from Sweden
No full textMonetary policy announcements can have significant effects on the stock market movement, particularly for sectors with high interest rate sensitivity. The thesis examines how Swedish GICS classified sectors react to the Riksbank's unexpected monetary policy shock on the 20th of September 2022 using an event study methodology across the selected sectors. The analysis documents how sector characteristics and reactions explain differences in abnormal returns. The findings contribute to the understanding of monetary policy impact on the Swedish economy. Sectors like Real Estate, Energy, Communications, Consumer Staples and Utilities experienced large negative cumulative returns. Meanwhile Information Technology, Industrials, Health Care, Materials and Consumer Discretionary sectors displayed statistically insignificant cumulative abnormal returns. The Financials sector was the only sector with a positive cumulative abnormal return on the event date. The results highlight cross-sector heterogeneity in how markets respond to monetary policy shocks.
Boundary Cases : Privacy as a Social Practice in Children’s Everyday Lives in Preschool
No full textWith the overarching aim to create knowledge about children’s privacy in preschool, this doctoral study lies at the intersection of early childhood educational research and privacy studies. Given privacy’s status as a human right and democratic value in Swedish preschool, the study investigates what privacy means to children and how they regulate it in everyday interactions in preschool. Analytically, the study departs from an understanding of privacy as a social practice (Reiman, 1984/2007), and children’s privacy practices are interpreted through Irwin Altman’s (1975, 1976, 1977) privacy regulation theory with concepts such as territory and personal space. Influenced by Erving Goffman (1959/2020, 1961/2014) and Maxine Wolfe (1978; Wolfe & Rivlin, 1987) these practices are interpreted as situated in preschool as an institutional context where children are attributed a social role as preescholers. The analytical point of departure in privacy theory is new to national early childhood educational research and thus contributes new perspectives. Ethnographic fieldwork was conducted in one preschool with children aged 1—6, along with their parents and educators. Building on participant observations, interviews, and documents, the analysis identifies themes of homesickness, physical closeness, seclusion, ownership and documentation. The results show that privacy is a central part of children’s everyday social lives in preschool. Children practice privacy by attempting to control their social boundaries: they regulate social distance, ownership and information sharing based on subjective privacy preferences. Children’s privacy practices often take a “relational logic” as their starting point, one that links close relationships with physical proximity and intimacy, and vice versa. Thus, children’s boundary regulation is about selectively creating closeness and nurturing relationships as well as keeping others at a distance. Furthermore, the results provide understandings of how children’s privacy practices and relation logic conflict with and are conditioned by institutional logics that prioritise the collective over the individual, emphasising adult authority and surveillance that makes privacy difficult, often resulting in forms of “pseudo-seclusion”. Children’s own bodies, which are smaller, make it more difficult for them to assert their boundaries; this is another central condition. In summary, children practice privacy through many different practices and expressions, including ways that span over a wide range of issues and situations. Recognising privacy as part of the complexity of children’s relations is relevant across contexts, private or public, physical or digital. The study thus calls for an understanding (and pedagogy) of privacy attentive to children’s experiences
Multidimensional change point detection using likelihood ratio statistics
No full textThis thesis tackles binary splitting of regression trees through the lens of change-point detection. Consider a dataset with multidimensional features and a one-dimensional response variable. A binary split attempts to form partitions of observations with similar response values. A typical Classification and Regression Tree (CART) lacks an inherent stopping mechanism to avoid over-partitioning which leads to overfitting. CARTs tend to rely on cross-validation to reduce overfitting, but then one loses out on valuable training data. We propose a method that succeeds at generalizing without removing any data from the training set. We model this setup as a change point problem, where the change point is the index of an ordered dataset where the partitions are optimal. A likelihood ratio test is used to determine the significance of each recurring optimal change point. We first study the one-dimensional asymptotic distribution of the split location under the null hypothesis (that there is no change point). Using a likelihood ratio statistic we recover the argmax of a Brownian bridge, which has an arcsine distribution, when the noise has finite variance. In the case where the noise has infinite variance, a stablebridge limit results in an approximate Beta distribution approaching to uniformity as tails thicken. The limiting distribution of the statistic is approximated by a Gumbel distribution that changes by an affine scaling as dimensionality grows. Across Gaussian and t-distributed response variables, this method provides a solid method for partitioning datasets, while avoiding overfitting, and could be useful when regularizing regression trees
Counting Kronor, Losing Trust: Austerity and Populism in Sweden, 1990–2024
No full textAusterity has been central to Swedish economic governance since the 1990s and the political effects of austerity is a well-established research area. While comparative research linkscrisis-driven austerity to rising populism, Sweden presents a paradox: it adopted a permanent fiscal framework during a period of economic stability and high trust but have experienced a strong surge in populist mobilization. This thesis addresses this gap by examining whether different types of austerity (fiscal, welfare, contextual, institutional, and local) have produced distinct forms of populism between 1990 and 2024. Drawing on national time-series survey-based data (SOM, ESS and WVS) and regional panel data from 96 municipalities aggregated into three regions, the study tests five theoretically derived hypotheses using regression analysis, lagged models, random- and fixed-effects estimations, robustness tests, and diagnostic procedures. The results show that fiscal tightening, contextual austerity, and especially institutional austerity increase attitudinal populism, while welfare retrenchment does not systematically drive cultural populism. Locally, cross-sectional analyses show that regions experiencing greater fiscal strain exhibit higher Sweden Democrat vote shares, although this relationship weakens in fixed-effects models due to limited within-region variation. This pattern suggests that local populism in Sweden is shaped partly by structural fiscal pressures, but that demographic factors remain important. Overall, the findings show that in high-trust welfare states, populist reactions emerge primarily from long-term institutionalized austerity while local and welfare-based mechanisms act more unevenly
Skrivstrategiundervisning enligt metoden Self-Regulated Strategy Development - för elever som uppvisar skrivsvårigheter : En systematisk litteraturöversikt
No full textSpråkutvecklande arbetssätt i matematik : En kvalitativ studie om matematiklärares perspektiv på elevers språkutveckling, begreppsförståelse och matematisk litteracitet
No full textSyftet med vår studie är att undersöka hur matematiklärare i grundskolans årskurs 4–9, beskriver sitt språkutvecklande arbete i matematikundervisningen. Den teoretiska utgångspunkten är det sociokulturella perspektivet på lärande och matematisk litteracitet. Studien har genomförts med kvalitativa semistrukturerade intervjuer med tio matematiklärare. Datan analyserades induktivt med en reflexiv tematisk analys. I resultatet framkommer att lärare i matematik använder sig av olika språkutvecklande arbetssätt för att stödja elevernas begreppsförståelse och tolkningsförmåga i det ämnesspecifika språket i matematik. Några av de strategier lärarna beskriver handlar främst om modellering, muntlig interaktion, multimodalt arbete (multimodalitet), repetition och kontextualisering. Dessutom upplever lärarna att elevers språkförmåga har stor betydelse för deras möjligheter att förstå och tolka textbaserade uppgifter i matematik. I studien framkommer även att psykosociala faktorer såsom motivation, självförtroende och goda relationer, är viktiga förutsättningar för elevers lärande och utveckling av sin matematiska litteracitet. Sammanfattningsvis tyder resultaten på att ett språkutvecklande arbetssätt behöver användas som en integrerad del av den språkutvecklande matematikundervisningen. Studien bidrar med kunskap om hur olika strategier och metoder utvecklar elevers matematiska litteracitet och vilka hinder och möjligheter som kan erfaras i detta arbete