Swedish universities have broadened recruitment of underrepresented groups in

higher education since the early 2000s. Previously, students with foreign

backgrounds were underrepresented, but now participate at levels that correspond

to their proportion in the population, which is approximately one-third of students

in Swedish higher education. Students with foreign backgrounds can face

challenges in the study environment, including language barriers and cultural

factors. The aim of this licentiate thesis was to explore the study-related

experiences of nursing and social work students with foreign backgrounds in

higher education from a salutogenic perspective. Study I was a quantitative study

with a repeated cross-sectional, multicenter design aimed at describing and

investigating changes in health-promoting resources and social study factors

between the first and last semesters. In the first semester, there were differences in

health-promoting resources and social study factors, as shown by lower measures

among students with foreign backgrounds than among other students. The

differences had evened out last semester. When assessed in the total sample during

the final semester, social relationships emerged as central for health-promoting

resources among the students. Study II was a qualitative, semi-structured

interview study examining perceptions of support and interaction in higher

education among students with foreign backgrounds. Fourteen students with

diverse migration experiences were interviewed, and findings highlighted

ambivalent feelings about a sense of belonging in relationships within the higher

education context. Conclusion: In Study I, lower measures of health-promoting

resources and social study factors between students with foreign backgrounds and

other students, evened out from the first to the last semester. In the last semester,

there was a correlation between health-promoting resources and social study

factors, and social study factors predicted health-promoting resources in the total

group. In Study II, strengthening students' sense of belonging was a key issue for

students with foreign backgrounds to be fully engaged in higher education.

This thesis presents the design, implementation, and evaluation of a hardware accelerator for the ASCON-CXOF128 extendable-output function. The work is scoped to simulation, synthesis, and post-implementation analysis; no physical hardware deployment was performed. The accelerator was implemented in Verilog using an iterative architecture in which the ASCON permutation logic is reused across twelve active round cycles per p12 invocation, reducing hardware resource usage relative to unrolled designs. Functional verification was performed using behavioral simulation against reference outputs derived from a software implementation following NIST SP 800-232.The design was synthesized and implemented using AMD Vivado 2023.2 targeting two platforms: an AMD/Xilinx Alveo U280 (UltraScale+, 16 nm) as the primary development target, and a Xilinx Artix-7 xc7a100tcsg324-1 (28 nm) as are presentative embedded evaluation target. Both implementations achieved timing closure at 100 MHz with positive worst-case slack.On the Artix-7, resource usage was 1254 look-up tables and 1238 flip-flops, with no block-memory or DSP usage, corresponding to approximately 2.0 % ofavailable LUTs on the device. Total on-chip power on the Artix-7 was estimated at 0.098 W (static: 0.091 W; dynamic: 0.007 W), based on a post-implementation netlist SAIF file with 100 % net coverage (2906/2906) and high Vivado confidence. For the verified case-5 vector, this corresponds to an energy efficiency of approximately 0.72 nJ/bit. For a verified case (customisation string V2X, 32 bytes of message data absorbed across five 64-bit blocks – four data blocks plus one padding-only block,four 64-bit output blocks) the full ASCON-CXOF128 operation – comprising eleven p12 invocations plus FSM and handshake overhead – completed in 189 clock cycles (1.89 µs at 100 MHz), giving an effective throughput of 135.45 Mbit/s for that case. Across the complete 1089-vector KAT sweep with a fixed 512-bit output, thepositive per-vector effective throughput ranged from approximately 3.16 Mbit/s to 105.53 Mbit/s at 100 MHz. Zero-length messages produce 0 Mbit/s by definition. Throughput generally increases as fixed operation overhead is amortised, although padding boundaries make the relationship non-monotonic. The results demonstrate that an iterative ASCON-CXOF128 hardware accelerator can be realized on a mid-range embedded FPGA with a small logic footprint and predictable timing. Whether the design meets the power and area budgets of a specific constrained application requires measurement on target hardware, which is left as future work.

Mining engines operate in highly contaminated environments where dust, mud, moisture, and particulate matter accumulate on engine surfaces and create maintenance challenges. Surface contamination increases cleaning frequency, contributes to corrosion risks, and may reduce equipment reliability and operational efficiency. This thesis investigates contamination sensitivity on mining engine components and evaluates commercially available protective solutions capable of reducing contamination accumulation.The study combines literature review, experimental contamination testing, and practical evaluation of protective coatings and sprays. Several hydrophobic, ceramic, and high-temperature protective products were selected based on industrial relevance, application simplicity, temperature resistance, and contamination-repellent properties. Experimental testing methods included residue weight analysis, digital microscopy, tape lift testing, and profilometry-based particle analysis.Results indicate that hydrophobic and ceramic-based coatings reduce contamination adhesion and improve surface cleanability compared to untreated surfaces. Profilometry analysis demonstrated measurable differences in particle density and surface contamination behaviour between coated and uncoated samples.The findings suggest that commercially available protective solutions can provide practical contamination mitigation for mining engine applications while reducing maintenance effort and downtime. The thesis contributes practical contamination assessment methodologies and industrial recommendations for contamination management in mining environments.

This thesis examines how the failed Arctic expedition led by Salomon August Andrée in 1897 was transformed into a heroic and meaningful national narrative within Swedish national identity construction. The study investigates how the image of Andrée and his expedition was interpreted, renegotiated, and represented from the late nineteenth century to the present day. Drawing on theories of nationalism, collective memory, and discourse, the thesis explores how historical failure can be reinterpreted as a symbolic national success.

The analysis is conducted through a qualitative and systematic discourse analysis, involving close readings of the articles to identify recurring themes, language choices, descriptions, evaluations, and narratives surrounding Andrée. The findings are then compared across the different time periods to examine how his representation has changed or been reproduced over time. The theories of imagined communities and national identity provide the analytical framework for interpreting these discursive patterns and for understanding how media narratives contribute to the construction of national heroes.

The work demonstrates that Swedish newspapers initially presented Andrée as a symbol of scientific progress, technological modernity, and national ambition. Following the discovery of the expedition’s remains, the narrative shifted toward themes of heroism, sacrifice, and national pride, transforming the failed expedition into a powerful national myth. In more recent decades, popular culture and academic research have increasingly challenged these heroic narratives, emphasizing human vulnerability, technological limitations, and the role of nationalism in shaping historical memory. Nevertheless, the expedition continues to function as a significant cultural symbol within Swedish society.

The study concludes that heroism is not an inherent quality but a discursive construction shaped by changing social, cultural, and political contexts. The Andrée expedition illustrates how national identities are continuously produced and renegotiated through historical narratives, collective memory, and cultural representation.

This thesis investigates support from multiple Unmanned Aerial Vehicles (UAVs) for safer autonomous Unmanned Ground Vehicle (UGV) navigation in a Robot Operating System (ROS) 2 and Gazebo simulation environment, with communication conditions evaluated using an objective modular network testbed in C++. The work develops a staged aerial–ground coordination system that starts from a single UAV–UGV baseline and extends toward perception-driven following and many-to-one support-chain coordination. Four simulation conditions are evaluated. Firstly, a direct-following baseline is established using a ground-truth odometry signal as the target reference. Then, this reference is replaced with a You Only Look Once (YOLO)-based visual leader estimate to evaluate direct visual estimate-following. Next, the visual approach is extended into a full perception-to-actuation pipeline, including filtering, target-state estimation, follow-point generation, planning, and visual actuation. Finally, asupport-chain setup is evaluated in which two additional UAVs provide separate support observations that are aggregated at the main UAV and forwardedto the UGV-side interface. The results show that the odometry-driven reference provides the most stable direct-follow behaviour, as expected for a simulation-side reference signal. Replacing this reference with direct visual estimate-following introduces route-dependent sensitivity to detection gaps, estimate freshness, and target visibility. The full visual-follow pipeline improves command timing and visual continuity compared with direct estimate-following, but still depends on perception quality and route geometry. The support-chain evaluation validatesthe implemented information flow from support UAV observations, throughmain UAV aggregation, to UGV-side monitoring and advisory topics. The current system, however, does not perform active UGV replanning, Nav2 costmap updates, obstacle avoidance, or communication-quality-based support repositioning. Overall, the thesis provides a modular simulation framework and a baseline of results for further work on perception-driven aerial support, multi-UAV coordination, and future communication-aware UGV assistance.

Buck converters are essential components in automotive, industrial, embedded, and battery-powered electronic systems, where efficient DC–DC power conversion is increasingly important. However, converter simulations often rely on ideal component models that fail to capture the effects of parasitic elements present in practical hardware, reducing the accuracy of performance prediction.This thesis presents an LTspice simulation study of a 12 V to 5 V synchronous buck converter based on the commercially available model. Two converter models were developed: a baseline model and a parasitic-inclusive model incorporating inductor DC resistance (DCR), capacitor equivalent series resistance (ESR), capacitor equivalent series inductance (ESL), and PCB trace resistance. The models were used to investigate the effects of parasitic components, switching frequency (300 kHz–2200 kHz), and soft-start capacitor values on converter efficiency, output-voltage ripple, startup behaviour, power loss, and transient response.The inclusion of parasitic components increased the total power loss from 0.796 W to 1.548 W and reduced the conversion efficiency from 94.9% to 89.49%, with the inductor DCR contributing the largest loss (0.563 W). The highest efficiency of 89.49% was achieved at 700 kHz, while increasing the switching frequency reduced the output-voltage ripple from 19 mV at 300 kHz to 2.2 mV at 2200 kHz. Soft-start analysis showed that increasing the capacitor value from 10 nF to 100 nF extended the startup rise time from 5 ms to 50 ms while reducing the peak inrush current from 5.6 A to 3.4 A.The results demonstrate that incorporating parasitic effects provides a more realistic prediction of converter performance than ideal models. The study also identifies an appropriate switching frequency and soft-start capacitor selection for balancing efficiency, ripple, and startup characteristics, making the proposed simulation approach a reliable tool for evaluating buck converter designs before hardware implementation.

Ultra-narrow linewidth lasers represent the cornerstone of modern frequency metrology,quantum spectroscopy, and precision timekeeping. They serve as the primary local oscillators that drive high-precision technologies, including optical atomicclocks, optical gyroscopes, and coherent LiDAR systems. Accurately characterizing their frequency noise power spectral density, Sν(f), and extracting their intrinsic lorentzian linewidth is essential for evaluating their coherence properties and designing active stabilization feedback loops. Traditional reference-based heterodyne methods require a secondary optical source of superior stability, which is frequently unavailable. Conversely, passive self-referencing techniques like conventional delayed self-heterodyne (DSH) require an optical fiber delay exceeding the laser’s coherence length (L > Lc). Furthermore, standard single-channel photodetectors introduce thermal and shot noise that masks the laser’s white frequency noise floor (Sν,0).To overcome these limitations, this thesis implements a correlated self-heterodyne(COSH) measurement architecture using 1.6 km optical delay line. By combining high-frequency heterodyne translation via an acousto-optic modulator (AOM) with dual-channel cross-correlation digital signal processing, this platform suppresses independent detector noise down to 0.01 Hz2/Hz without requiring multi-hundred kilometre fiber spools. This work directly advances optical frequency metrology and supports the development of portable, high-stability laser sources for our NMI.

This thesis presents the design and testing of a hardware accelerator for CSI-based positioning using an FPGA. The goal was to investigate if a quantized Convolutional Neural Network (CNN) could estimate a position from Channel State Information (CSI) and run fast enough on FPGA hardware. The model was trainedin software using PyTorch and Brevitas. The trained model was then exported to QONNX and converted to a hardware accelerator using the FINN toolchain. The final hardware test was done on a Zybo Z7-20 FPGA board using PYNQ.

Three antenna scenarios were tested: Distributed, Uniform Linear Array (ULA), and Uniform Rectangular Array (URA). The evaluation used an indoor MaMIMO CSI dataset that covers an area of approximately 2.5 m by 2.5 m. The verified quantized software model showed a mean positioning error of 15.83 mm for Distributed, 34.28 mm for ULA, and 17.16 mm for URA. The hardware performance test on the Zybo Z7-20 showed an average latency of about 4.608 ms and a throughput of about 217 FPS for all three scenarios. The Vivado resource reports showed that the design fit on the selected FPGA board.

The results show that a quantized CNN model for CSI-based positioning can be implemented as an FPGA accelerator. The accelerator reached low latency, while the positioning accuracy was evaluated using the verified quantized software model. The FPGA test focused on deployment, latency, throughput, and resource usage. Future work should validate the end-to-end FPGA positioning output by comparing the FPGA output values directly with the software model andground-truth labels.

This thesis investigated the development of a sensor-based pillbox capable of providing scheduled dose reminders and automatically verifying whether a pill compartment had been emptied and subsequently refilled. The main technical challenge was to achieve reliable classification of compartments as empty or nonempty despite variations in pill position, optical properties, ambient lighting, and sensor response, while maintaining a simple system architecture. An iterative development method was used, in which optical proximity sensing and ambient light sensing were evaluated experimentally under different lighting conditions, pill counts, and pill positions. Measurement series were analysed using statistical measures and observed class overlap, while the compartment geometry was successively modified to improve the separation between empty and non-empty states. The final prototype used one optical proximity sensor per compartment, a microcontroller, a multiplexer, visual indicators, an acoustic alarm, and statebased decision logic. A decision threshold of 6 was selected experimentally, and verification of emptying and refilling required 20 consecutive sensor readings that satisfied the corresponding criterion. In the final stationary tests, all investigated critical pill positions produced values above the decision threshold, while the empty compartment produced values below it; no observed readings were misclassified. The results showed that reliable compartment-state detection was achievable under the investigated stationary conditions through the combined optimisation of sensor selection, mechanical compartment design, and state-based decision logic.

Early-stage green ventures are important for Sweden’s move toward sustainability, but they still struggle with financial challenges even though there is strong institutional support. This thesis looks at how these ventures handle financial sustainability in real situations, with a focus on how they manage financial resource dependence and financial sustainability while pursuing environmental objectives.The study examines how ventures access and manage financial resources while addressing tensions associated with sustainability. Through qualitative research and interviews, the results show that these ventures mix active and passive ways of managing financial resource dependence, relationships with investors, and technology. The analysis shows that founders expressed differing views on how environmental objectives and financial viability relate, though most described them as mutually supportive rather than competing goals.This thesis contributes to the literature by showing how early-stage green ventures navigate financial sustainability through their management of financial resource relationships while pursuing environmental objectives and maintaining financial viability. It further demonstrates that strong institutional support does not necessarily eliminate financial barriers, particularly for highly innovative or inexperienced ventures. The findings provide insights for policymakers, incubators, and investors seeking to support early-stage green innovation in Sweden.

‘Quality’ in education has become a global phenomenon, and is frequently discussed in transnational and national policy. In such discussions, ‘quality’ seems to be both worthwhile and demanded, but at the same time, the concept has also been criticised in previous research as it seems to have attained a ‘common sense’ status. In this context, the role of the Swedish Schools Inspectorate (SSI) is raised, as the SSI is commissioned by the Swedish Government to audit and review ‘quality’ in education. However, as the concept of ‘quality’ is left undefined in policy, the SSI has to enact national policy to review ‘quality’ in its quality audits. Hence, the SSI plays an active part in construing ‘quality’, as it makes it auditable through the indicators and systems it uses when conducting quality audits. Thus, this paper has critically scrutinised underlying assumptions and the discursive effects of how the SSI construes ‘quality’ in different policy proposals. To access and analyse such constructs, this article draws on Bacchi’s (2009) post-structural, Foucault-influenced ‘What’s the problem represented to be?’ (WPR) approach. The results indicate that ‘quality’ is presupposed to be both measurable and elusive at the same time. Hence, the SSI operates in a field of tension, targeting something more elusive than just figures and numbers when it comes to ‘quality’. The discursive effects of these presuppositions have implications for what can be said and thought about ‘quality’ at both national and local levels, as well as for students and school staff who are involved in the educational system. © 2022 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.

This study examines how auditors in practice manage clients with a high risk of moneylaundering, and how they navigate the commercial challenges and conflicts of interest that arisein their role as the state's gatekeepers. The study adopts a qualitative approach with anabductive research design, utilizing semi-structured interviews conducted with practicingauditors and subject matter experts. The results demonstrate that while auditors play a pivotalrole in anti-money laundering efforts, the management of high-risk clients is governed bydefensive screening strategies aimed at protecting firm legitimacy. Challenges in practiceinclude loyalty conflicts within the client relationship, the fear of regulatory sanctions, and anunintended regulatory gap resulting from the abolition of the mandatory audit. Despite astatutory reporting obligation, auditors most frequently choose to quietly terminate the clientrelationship instead of reporting suspicions to the Financial Intelligence Unit. The studycontributes a deeper understanding of how market realities and administrative documentationrequirements in practice reshape and counteract the legislator's intentions regarding thegatekeeper role.

Freshwater is facing a variety of anthropogenically caused stressors, ultimately affecting ecosystems. One persistent stressor is acidification, with impacts that are well studied. To combat these effects liming has continuously been carried out in different waterbodies in Sweden to improve water quality and favor aquatic biodiversity, especially fish. Previous research found that liming of lakes can reduce the diversity of adult Odonata and negatively impact larval numbers. The common four-spotted chaser (Libellula quadrimaculata) hasbeen occurring in fewer limed lakes compared to untreated waterbodies. Moreover, there are indications of the condition of L. quadrimaculata being negatively impacted by lower pH levels. To understand why, this study conducted an experiment, testing if a change in pH is the sole cause by impacting the growth. Two populations of first-year L. quadrimaculata larvae (N =32) were collected from separate locations in Sweden (Halmstad and Uppsala) and divided equally based on pH (5, 6.5 and 8) into a total of six treatment groups,measuring head width and weight for a total of 28 days. Using a linear mixed model, we found no significant difference in growth rate and weight gain between the pH treatments. Indicating that L. quadrimaculata growth is not explicitly impacted by a change in water pH. But rather, the effect of liming on occurrence and the condition of L. quadrimaculata is a result of individual or compounding factors. Furthermore, a significant difference was found between the two locations for both growth rate and weight gain, where the larvae from Halmstad grew faster. We speculate this is due to the experiment setup more closely resembling factors from those found in Halmstad. More research is needed to further understand the impact of liming coupled with other stressors on L. quadrimaculata to understand the decline seen. 

Generative design is Artificial Intelligence software that generates design options based on set constraints. It is commonly used in various fields, but its application in medical device development remains uncertain due to high safety and documentation requirements. Using the Task-Technology Fit (TTF) framework, this study explores how medical device developers perceive generative design's values, limits, and deployment potential. Interviews with seven participants, five from Atos Medical and two external experts, show that its usefulness varies by each development stage. Developers see generative design software as helpful for brainstorming and optimizing designs in the early stages. However, its value goes down later when regulations require proof of a single final design. Different roles have different views: regulatory staff point out the tool's drawbacks, while engineers appreciate its benefits early on. There are also concerns about who owns the intellectual property, limited access to the software, and the difficulty of reviewing a lot of AI-generated results. The results suggest that generative design is better suited to early concept development, indicating that organisations should pilot it at this stage while resolving data security and ownership questions in advance. It also suggests exploring AI applications beyond design itself, such as supporting regulatory documentation.

Many lives are lost due to motorcycle crashes worldwide. Riding a motorcycle is a complex task that riders carry out by using their body to modify the motorcycle dynamics. Safety systems and protective equipment need to be able to estimate rider postures to be as performant as possible which is a challenge in real riding environments. This thesis presents the development and validation of a wearable system for estimating upper body posture using six inertial measurement units (IMUs) mounted on the torso and arms of a riding jacket. The work is a laboratory proof-of-concept: all experiments were performed indoors on a stationary bicycle, not on a motorcycle. The IMUs are connected to an ESP32 microcontroller via SPI and streamed wirelessly to a host computer. The recorded data was processed offline using gyroscope bias calibration, Madgwick sensor fusion, and joint angle computation to estimate the rider’s posture. A dual-camera system based on MediaPipe Pose was developed as a reference and the two measurement systems were compared after time synchronization using the root mean square error and the correlation coefficient. Indoor laboratory experiments were conducted with a participant seated on a stationary bicycle in a riding position. An agreement with the video reference was achieved in the forward bend and lateral lean of the torso to within roughly 4-8 degrees during the best sessions, with an RMSE of 7.3 and 11.8 degrees across all 11 recordings. The R8 target of 5 degrees was achieved in one session only which wasn’t the average. The synchronization target of 20 ms in R7 was not verifiable due to the residual timing error being 0-102 ms in the longer recordings and a resolution limit of 20-30 fps cameras. Deviations for the arm angles were considerably larger. Analysis indicates that a large part of these deviations is consistent with the limitations of the two-dimensional video reference, in particular its sensitivity to camera viewing angle. The results show that the system tracks torso posture in good agreement with the video reference under laboratory conditions, and highlight the need for a 3D reference with a higher frame rate and for validation on a motorcycle in future work.

This qualitative study with an abductive approach examines, from the perspective of preschool teachers, which dilemmas can arise in everyday preschool work with inclusion. The aim is to identify dilemmas preschool teachers experience in situations where they strive to include all children, but where exclusion nonetheless occurs.

Data were collected through semi-structured interviews with five practicing preschool teachers, and the analysis drew on the developed concepts of inclusion, the inclusion ideal, dilemma and exclusion. The results show that preschool teachers inclusion ideal is characterized by children’s emotional experience of belonging and the individual abilities rather than physical presence. Exclusion arises in both children’s own relationships and through preschool teachers decisions, which may be reinterpreted as care or adaptation according to the child’s and group’s needs. Collegial conversations about inclusion and exclusion can be challenging, as the concepts are experienced as emotionally charged. Methods for inclusion centre around organizational adaptations and a preventive approach, but the room for action is limited by resources such as staff-to-child ratios and group size.

The study concludes that inclusion is a complex, context-dependent phenomenon shaped by the tension between the inclusion ideal and the preschool’s available resources. The results should therefore be understood as examples of how the phenomenon can manifest in specifics contexts, rather than as generalizable patterns. 

The study aims to gain a deeper understanding of young women's fear of violence in public settings and the consequences it has on their everyday lives. The empirical data has been collected through a qualitative method where eight semi-structured interviews were conducted digitally via Teams with young women aged 21-25. After being transcribed and coded, the collected material has then undergone a thematic analysis to analyze the empirical data. The study is based on a social constructionist scientific theoretical approach and the theories of emotional work, emotional rules and gender performativity. The results show that the majority of the young women experience fear of violence in public environments, where darkness in particular has a major impact. The participants provide a consistent picture of what is perceived as safe versus unsafe in public and what coping strategies they have. At the same time, the results show that the women's coping strategies are not always sufficient and that their everyday lives are affected in many different ways. It shows many consequences in the lives of women, in the form of obstacles, precautions or the choice not to go out at all after dark.

According to Paulsen (2014, p. 5), empty labor refers to time during the working day spent on non-work-related activities, such as scrolling on the phone or answering private messages. This study examines empty labor within digital customer service and how employees manage demands, control, and workload in an everyday working life increasingly characterised by measurement and performance management.  This study is based on a qualitative method consisting of semi-structured interviews with employees working within digital customer service. The empirical material was analysed through thematic analysis, focusing on how employees describe and manage performance demands, organizational control, and limited room for action in their everyday work. The analysis also explores actions that can be understood as empty labor, and how they can be interpreted in relation to recovery, coping, and informal resistance.The results show how a high work pace, extensive follow-ups, and strong control through monitoring and control systems create a demanding work environment for the participants. To manage the workload and the perceived surveillance, the participants developed different strategies, such as taking micropauses, having private conversations, and consciously adjusting their work pace. This study shows that actions understood as empty labor can function as strategies for recovery and as a way of reclaiming control in a work environment characterised by high demands and surveillance. It also shows how it could take form as informal resistance against the organisations. Employees’ experiences of these demands and forms of control were largely influenced by the degree of support and trust they received from managers and colleagues. Empty labor should therefore be understood as a complex phenomenon that cannot solely be seen as a form of resistance. It can instead hold multiple meanings depending on the situation and work context.

The aim of this study was to explore and analyze upper secondary school students’ experiences of social media and its significance in their everyday lives. Particular attention was given to the students’ experiences of the role of social media in relation to self-image and identity, as well as to how material conditions may shape these experiences. The study adopted a qualitative approach and was based on eight semi-structured interviews with upper secondary school students aged 18 and above. The empirical material was examined through thematic analysis from a hermeneutic perspective. The analysis was informed by Pierre Bourdieu’s theories of capital and social fields and Erving Goffman’s theories of stigma and self-presentation.

The findings indicate that social media constitutes a central part of the participants’ everyday lives. The platforms were described both as arenas for community, inspiration, and self-expression and as environments in which norms related to appearance, achievement, and social status are reproduced. The participants reported that repeated exposure to idealized content may influence their self-image and identity through social comparison and perceived pressure to conform to different ideals. At the same time, the findings suggest that participants’ experiences of this influence varied and were related to factors such as self-esteem, previous experiences, and material conditions. Several participants also described strategies for managing the influence of social media, including critical reflection and an increased awareness that content on social media platforms is often selective and idealized.

Overall, the study indicates that social media constitutes an important arena for identity formation among upper secondary school students, in which individual experiences, social relationships, and material conditions interact. The findings highlight the importance of recognizing the role of social media in social work and of developing interventions that may promote young people’s mental and social well-being.

The energy transition is creating new challenges for district heating (DH) firms. These firms operate capital-intensive infrastructure with long investment cycles and strong requirements for reliability, while at the same time facing pressure from decarbonisation, electrification, digitalisation, new customer demands, and increasing competition from alternative heating solutions such as heat pumps. Historically, innovation in DH has mainly focused on improving existing production and distribution systems. This production-oriented and incremental innovation logic may become insufficient as the energy transition requires DH firms to address increasingly complex problems involving new technologies, customers, and external actors.

Less is known about how incumbent DH firms search for more distant solutions, collaborate with external actors, and reorganise value creation when incremental improvement is no longer sufficient.

The purpose of this doctoral thesis is to understand how innovation in the district heating sector develops under conditions of energy transition. More specifically, the thesis examines how DH firms engage in exploration with external partners, and how such exploration may reconfigure inter-firm relationships and existing value systems.

The thesis comprises a compendium and five papers and uses a predominantly qualitative and exploratory research design. The empirical material includes a structured literature review, workshops, interviews, observations, documentary material, and longitudinal case studies involving DH firms and other actors in several European contexts.

Three major findings emerge from the thesis.

First, innovation in the DH sector remains predominantly incremental and strongly shaped by path dependence. DH firms have developed considerable capabilities for improving production, distribution, fuel use, and operational efficiency within established technological and organisational structures. Long-lived assets, sunk costs, regulatory arrangements, established supplier relationships, and demands for reliability all reinforce this exploitation-oriented logic. These conditions have supported reliable and increasingly low-carbon heat supply, but they also encourage local search around familiar technologies and solutions. Transformation of the DH firms therefore tends to be gradual, uneven, and selective rather than disruptive. 

Second, the empirical cases show that radical exploration in DH takes several distinct forms, depending on technological distance, the actors involved, and how innovation is organised across organisational boundaries. The thesis identifies five patterns of radical exploration, ranging from supplier-led development within the incumbent DH system to customer integration, entrant-led innovation, and emerging ecosystem-based configurations in the production respectively customer sides. Across these patterns, the locus of innovation gradually moves away from the production core of the DH firm toward customers, property owners, technology suppliers, entrants, complementors, and other external actors. The actor leading innovation may therefore also change. Radical exploration is consequently not only a search for new technology; it is also a search for partners, capabilities, and suitable ways of organising collaboration across firms.

Third, the findings show that transformation in heating is increasingly about how value creation is organised, not only about which technology is used. Traditional DH is typically organised as a hierarchy-based value system in which one focal firm controls production, distribution, and the customer interface. Individual heat pumps are more commonly embedded in market-based arrangements, while new combinations of district heating, heat pumps, cooling, storage, digital technologies, and customer-side resources may require more ecosystem-based forms of coordination. Competition in the heating sector can therefore increasingly be understood as competition between alternative value-system configurations. Platform-based ecosystems represent one possible emerging coordination mechanism, although the thesis finds that such configurations are still at an early stage and introduce new challenges related to governance, control, dependency, and value capture.

Overall, the thesis shows that the transformation of district heating is best understood as a gradual reconfiguration of innovation search, actor roles, and value systems. Incumbent DH firms remain strongly rooted in established infrastructure logics, but pressures from electrification, digitalisation, customer integration, and new entrants are expanding the solution landscape. Moving from silos toward more ecosystem-like forms of innovation therefore requires not only new technologies, but also new ways of searching, collaborating, and coordinating value creation across organisational boundaries. 

The findings may also inform understanding of innovation and transformation in other mature infrastructure-based utilities characterised by long-lived assets, path dependence, and increasing technological uncertainty.

This study aims to analyze how organizational structures and competence levels affectaccounting quality in small businesses, with a focus on a digitalized context. The studyaddresses the issue of accounting errors, where previous research has mainly examinedinternal control and competence separately. A qualitative approach was applied, based onsemi-structured interviews with accounting consultants and small business representatives.The empirical data was analyzed using thematic analysis. The findings show that accountingerrors arise from an interaction between organizational factors and individual competence.Weak internal control and limited knowledge increase the risk of errors, while digitalizationboth reduces and creates new types of errors depending on user understanding. The study alsoidentifies economic considerations as an underlying factor, where small businesses oftenhandle accounting internally to reduce costs, which can negatively affect quality. The studycontributes to a deeper understanding of how organizational, individual, and economic factorsinteract in relation to accounting errors in small businesses.

The aim of this study is to examine how perpetrators and victims are represented in district court judgments concerning assault. The study also explores how these representations reproduce or challenge gender norms. 

The study is based on a social constructionist perspective and uses Judith Butler’s theory of gender and performativity, Connell’s theory of masculinity, Christie’s theory of the ideal victim, and Scheff and Retzinger’s theory of shame and aggression. The empirical material consists of 36 assault judgments from a district court in southern Sweden. The material has been analyzed through a thematic analysis based on district court judgements, with the purpose of identifying recurring patterns in the representations of perpetrators and victims.

The results show that perpetrators are represented through five themes: the aggressive perpetrator, the self-defending perpetrator, the remorseful perpetrator, the intoxicated perpetrator, and the controlling man. Victims are represented through four themes: the remorseful woman, the consistent woman, the male victim, and the intoxicated victim. The analysis shows that perpetrators and victims are represented relationally, meaning that the understanding of one role affects the understanding of the other. Violence is therefore not only represented through the physical act itself, but also through issues of responsibility, credibility, relationships, gender, and alcohol intoxication. The study further shows that gender norms are both reproduced and challenged in the material. Men are often represented through aggression, control, self-defense, and physical agency, while female victimhood is frequently connected to credibility, consistency in narration, and relational dependency. At the same time, gender norms are challenged through representations of women as perpetrators and men as victims. However, these norm-breaking representations often need to be made understandable through specific explanations. The results also show that alcohol is given different meanings depending on whether the intoxicated person is represented as a perpetrator or a victim. The study is relevant to social work, as professionals encounter people who have been subjected to violence, people who have used violence, and individuals whose narratives may be shaped by ambivalence, dependency, and complex relational contexts. 

Background Technological advances have lowered the threshold for AI and led to rapid adoption among Swedish small businesses. Depsite this, a lack of digital competence remains one of the primary obstacles to stable integration. There is a need to understand how resource-constrained organizations practically handle these challanges and what drives everyday AI use. 

Purpose The purpose of the study is to explore the perceived barriers and benefits of implementing AI in small businesses, with a particular focus on how adoption takes shape in practice. 

Method To describe organizational and competence related factors affecting small businesses ability to implement AI and to identify perceived barriers and benefits of the technology. 

Results & Discussion AI is informally adopted as an operational complement to offset resource gaps, rather than through formal strategic transformation. A key finding is a paradoxical effect where competence shortage acts as a driver as companies implement AI precisely due to a lack of specialized personell. The main benefits are efficency and the pararell handeling of routine tasks. Conversely technical limitations, the need for manual quality control, uncertainty regarding cost effectiveness and privacy concerns emerge as central barriers to further integration.

Artificial intelligence (AI) is becoming more common in higher education, including in the development of teaching materials. While previous research has examined the pedagogical uses of AI and its environmental implications, less attention has been given to how teachers use AI in developing teaching materials and how they understand its environmental consequences. This study explores these issues through semi-structured interviews with 12 higher education teachers from different disciplinary and geographical contexts. The findings show that teachers used AI for lesson planning, assessment design, generating ideas, classroom activities, and organizing teaching materials. AI was valued for saving time and supporting idea generation, but teachers continued to rely on their professional judgment to review, adapt, and evaluate AI-generated content. Their AI literacy developed mainly through practical experience and self-directed learning, with limited formal training. Environmental sustainability was less familiar to teachers, with most reporting limited awareness of AI’s environmental impacts. Some teachers had nevertheless begun to avoid unnecessary interactions, reuse existing materials, and use AI when it served a clear teaching purpose. Teachers also expressed a need for greater knowledge and institutional support to understand these issues. The study shows that teachers’ professional judgment remains central to AI-supported teaching material development, while environmental sustainability is still developing as a consideration in the use of AI.

Lean manufacturing is associated with operational improvement, yet similar practices can generate different results across firms. This study investigates the organisational conditions that shape the influence of lean manufacturing practices on operational efficiency. It applies an interpretivist qualitative design and analyses eleven interviews using thematic analysis. Respondents represented production, quality, lean and industrial engineering, process improvement, manufacturing support, inspection, and management roles. The analysis identified improvements linked to 5S, standardised work, visual management, built-in quality, TPM, JIT, Kaizen, problem-solving, and daily management systems. Six themes emerged: leadership and review discipline, training and capability, employee involvement, coordination and resources, supplier reliability, and production complexity. Leadership was the most pervasive condition, although its effect depended on interaction with capability, ownership, measurement, and resources. Supplier and complexity conditions produced practice constraints. The study concludes that lean effectiveness is configurational and contingent rather than universal, requiring conditions aligned with each practice and context. 

I uppsatsen undersöker jag hur dagstidningarna Svenska Dagbladet, Dagens Nyheter och Arbetet skrev om kvinnor som begått brottet barnamord och vad som skrevs om straffen och lösningarna på brottet under perioden 1890-1930. Utifrån en komparativ textanalys kommer 60 olika nyhetsartiklar och insändare analyseras utifrån specifika frågor för att undersöka hur kvinnorna blev beskriva och benämnda samt undersöka vad tidningarna skrev om straffen och lösningarna.

Studien syftar till att undersöka hur de tre dagstidningarna skrev om barnamord och kvinnorna under en period där det skedde många förändringar både i synen på kvinnor som begått brottet och i lagändringar kring barnamorden. Utifrån en period på 40 år undersöks det också om man kan se en förändring över tid, både i hur man skrev om kvinnorna, straffen och lösningarna på problemet barnamord.

Resultatet visar på att kvinnorna beskrevs och benämndes på olika sätt över tid i de tre dagstidningarna under den undersökta perioden. Uppsatsen bidrar till att visa på hur dagstidningar skrev och gestaltade brottet barnamord, kvinnorna som begått brottet samt vilka straff och lösningar som de olika tidningarna skrev om under perioden 1890-1930.