Publikationer från Linköpings universitet
Not a member yet
58266 research outputs found
Sort by
Contagious Secularization : Social Influence and Membership Dynamics in Religious Organizations
No full textIn recent decades, most Western countries have experienced a religious decline, reflected in shrinking membership in religious organizations. Explanations of this decline have primarily emphasized individual-level predictors of joining or leaving religious organizations or broad macro-level processes such as modernization. Missing from this picture is the meso-level: the social context and networks in which individuals are embedded in. Attending to this level is crucial to understand not only how individuals make decisions in response to their social environment, but also how these interdependent behaviors aggregate into broader patterns of organizational growth and decline dynamics. To address this gap, this dissertation examines how joining and leaving a religious organization are shaped by the behavior of others, that is, by social influence, and how these micro-level processes generate organizational growth or decline. This is investigated in three essays, all using the Church of Sweden as the empirical case and analyzing full-population Swedish register data by combining conventional statistical approaches with machine-learning techniques and computational simulation methods. Essay I, co-authored with Peter Hedstrom, analyzes how exposure to leavers shapes membership dynamics in the Church of Sweden. Using Swedish full-population register data from 2002 to 2023 and a dynamic matched-sample design, the analysis shows that exposure to leavers increases the likelihood of leaving the Church of Sweden. However, this effect is moderated by how deeply socially embedded an individual is in the Church, being particularly strong among individuals with few social ties to other members. To explore how these dynamics shape collective membership trajectories, an empirically calibrated agent-based simulation model is used. The simulation results suggest that social influence has amplified the overall decline in membership of the Church of Sweden, but also that the decline would have been even steeper in the absence of religious embeddedness. The essay demonstrates how, and under what conditions, social influence shapes individual decisions to leave and what implications this has for the broader organizational decline of the Church of Sweden. Essay II analyzes how exposure to religious pluralism through social ties shapes parents’ decisions to baptize their children in the Church of Sweden. Using Swedish full-population register data on all parents with children born between 2002 and 2022, network-based measures of religious pluralism are constructed, and an inverse probability weighting design is used to estimate the effect of exposure to religious pluralism. The results show that exposure to religious pluralism decreases the likelihood of baptism, with particularly strong effects under religious cross-pressure (when parents differ in religious affiliation) and especially when the father has another affiliation. The essays highlight how social networks shape the intergenerational transmission of religious behavior, providing a meso-level perspective on the link between religious pluralism and religious participation. Essay III examines how individuals’ social ties to other church members shape their susceptibility to the Church of Sweden’s membership fee. Using full-population register data from 2001 to 2023, the analysis shows that higher fees increase the risk of leaving the Church of Sweden, but that this effect is weaker for individuals with many social ties to other church members. To explore how these micro-level dynamics shape collective membership trajectories, an empirically calibrated agent-based simulation model is used. The simulations show that while church fees contributed to membership decline, religious embeddedness dampens this effect overall while simultaneously reinforcing spatial differences in affiliation. Together, these findings demonstrate that social networks shape when and for whom costs affect religious participation, and that this micro-level heterogeneity has large consequences for the collective patterns of religious participation. Together, the three essays show that social influence is crucial for understanding the membership dynamics of religious organizations. In doing so, the dissertation demonstrates how micro-level interactions and their social interdependencies aggregate into the collective patterns of religious decline and secularization.
Making Sense of Smart Electricity Meters : A Socio-Technical Perspective
No full textSmart electricity meters are widely promoted as enablers of efficient, flexible, and sustainable energy systems, yet their real-world outcomes often diverge from these ambitions. This thesis examines smart meters through a socio-technical perspective, showing that smartness is not an inherent technical feature but an effect emerging from interactions among technology, governance, institutions, and everyday practice. The thesis, guided by Socio-Technical Systems theory, brings together five studies that were conducted separately and later analysed together. Each addresses a different layer of the socio-technical system, moving from conceptualisation to practice and governance. The first and third papers establish the theoretical foundation by examining how smart meters and the idea of “smartness” are represented in research and by developing a multidimensional framework for evaluating smartness as a relational and evolving quality. The second paper explores how Swedish households understand and experience smart meters introduced through a nationwide rollout, revealing limited engagement and the normalisation of the technology as background infrastructure. The fourth paper critiques Information Systems research, showing how smart meters are often “black-boxed” and detached from their institutional and infrastructural contexts. The fifth paper examines Sri Lanka’s emerging smart grid, demonstrating how infrastructural fragility and fragmented governance create socio-technical misalignments that constrain functionality and legitimacy. Together, the studies show that smartness is relational, situated, and continuously negotiated. Misalignments appear at informational, institutional, infrastructural, and discursive levels, revealing the gap between policy ambitions and lived experience. Achieving alignment requires governance reflexivity - the capacity to iteratively adjust technical systems, policy frameworks, and social practices over time. Conceptually, the thesis reframes smartness as a socio-technical effect that arises from alignment across multiple layers of practice. Empirically, it demonstrates how different contexts shape the enactment and consequences of smart meter adoption. Practically, it highlights the importance of reflexive and context-sensitive governance that anticipates misalignments, fosters interpretive engagement, and aligns technological ambition with institutional capacity and infrastructural realities. Smarta elmätare framhålls ofta som verktyg för att skapa effektiva, flexibla och hållbara energisystem, men deras faktiska utfall motsvarar inte alltid dessa ambitioner. I denna avhandling analyseras smarta elmätare ur ett sociotekniskt perspektiv och visar att ”smarthet” inte är en inneboende teknisk egenskap utan uppstår genom samspelet mellan teknik, styrning, institutioner och vardaglig användning. Avhandlingen, som utgår från socioteknisk systemteori, sammanför fem enskilda studier som analyseras gemensamt. Varje studie belyser olika nivåer i det sociotekniska systemet - från konceptualisering till praktik och styrning. Den första och tredje artikeln utgör den teoretiska grunden genom att analysera hur smarta elmätare och ”smarthet” beskrivs i forskning samt genom att utveckla ett ramverk för att förstå smarthet som relationell och föränderlig. Den andra artikeln undersöker hur svenska hushåll uppfattar och använder de nationellt införda mätarna, och visar på begränsat engagemang och hur tekniken normaliserats som infrastruktur. Den fjärde artikeln kritiserar informationssystemforskningens tendens att ”svartlåda” tekniken och koppla bort den från organisatoriska och infrastrukturella sammanhang. Den femte artikeln analyserar Sri Lankas framväxande smarta elnät och visar hur bristande infrastruktur och fragmenterad styrning skapar sociotekniska missanpassningar som begränsar funktion och legitimitet. Tillsammans visar studierna att smarthet är relationell, situerad och förhandlad över tid. Missanpassningar uppstår på informativa, institutionella, infrastrukturella och diskursiva nivåer och tydliggör gapet mellan politiska ambitioner och vardagliga erfarenheter. En reflexiv styrning som successivt anpassar tekniska system, policyer och sociala praktiker krävs för att nå bättre samordning. Avhandlingen omdefinierar därmed smarthet som en socioteknisk effekt som uppstår genom samspel mellan flera nivåer av praktik, visar hur olika nationella och institutionella sammanhang formar införande och användning, och betonar vikten av en kontextkänslig och reflexiv styrning som förutser missanpassningar, stärker tolkande engagemang och anpassar tekniska ambitioner till institutionella och infrastrukturella förutsättningar.In the printed version two ISSN numbers were incorrectly listed, these are removed in the digital version. The name of the series was incorrect on the cover and on the title page. The correct official name is "FiF-avhandling, Filosofiska fakulteten, Linköpings universitet". This is changed in the digital version.</p
Deep networks for system identification: A survey
No full textDeep learning is a topic of considerable current interest. The availability of massive data collections and powerful software resources has led to an impressive amount of results in many application areas that reveal essential but hidden properties of the observations. System identification learns mathematical descriptions of dynamic systems from input-output data and can thus benefit from the advances of deep neural networks to enrich the possible range of models to choose from. For this reason, we provide a survey of deep learning from a system identification perspective. We cover a wide spectrum of topics to enable researchers to understand the methods, providing rigorous practical and theoretical insights into the benefits and challenges of using them. The main aim of the identified model is to predict new data from previous observations. This can be achieved with different deep learning-based modelling techniques and we discuss architectures commonly adopted in the literature, like feedforward, convolutional, and recurrent networks. Their parameters have to be estimated from past data to optimize the prediction performance. For this purpose, we discuss a specific set of first-order optimization tools that have emerged as efficient. The survey then draws connections to the well-studied area of kernel-based methods. They control the data fit by regularization terms that penalize models not in line with prior assumptions. We illustrate how to cast them in deep architectures to obtain deep kernel-based methods. The success of deep learning also resulted in surprising empirical observations, like the counter-intuitive behaviour of models with many parameters. We discuss the role of overparameterized models, including their connection to kernels, as well as implicit regularization mechanisms which affect generalization, specifically the interesting phenomena of benign overfitting and double-descent. Finally, we highlight numerical, computational and software aspects in the area with the help of applied examples. (c) 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)
More knee reinjuries after ACL reconstruction compared to nonsurgical treatment of the ACL. A 2-year follow-up of the NACOX prospective cohort study
No full textPurposeTo describe knee reinjuries and surgeries within 2 years after an acute anterior cruciate ligament (ACL) injury, in patients treated with or without ACL reconstruction (ACLR).MethodsTwo years follow-up of 275 patients (52% females, mean age 25.2 [SD 7.0] years) about knee reinjuries and surgeries were analysed from the prospective NACOX cohort study, aiming to describe recovery after an acute ACL injury treated according to clinical routine. Knee reinjury was defined as increase or new symptoms due to new trauma. At 2 years after injury, 169 (61%) had undergone an ACLR. Results are presented with descriptive statistics and risk ratios (RR).ResultsThirty-two patients sustained reinjuries within 2 years; 6 in the non-ACLR group, and 26 in the ACLR group (7 before and 19 after ACLR). Twelve patients in the ACLR group sustained a graft rupture and three did an ACL revision. Patients with non-ACLR, aged 21-25 years, had a 5.9-fold higher risk for reinjury than those aged 15-20 years (RR 5.9 [1.3-26.9]; p = 0.012). Twenty-four patients had surgery in the non-ACLR group and 36 patients in the ACLR group (excluding primary ACLR), where six were before ACLR and 30 were after ACLR.ConclusionTwelve percent sustained a knee reinjury and 21% underwent knee surgery within 2 years after the index ACL injury. There was a higher reinjury incidence in the ACLR group, but no difference in incidence of surgeries. Thus, ACLR did not reduce the risk of traumatic reinjuries or surgeries, which is important for treatment decision considerations.Level of EvidenceLevel I.Funding Agencies|Swedish Medical Research Council [VR2018-02563]; Swedish Research Council for Sport Science [CIFP2022-0058]; Medical Research Council of Southeast Sweden [FORSS856741]; ALF Grants Region Ostergotland [LIO-934538]</p
Association Between Maternal Infections in Pregnancy and the Risk of Inflammatory Bowel Disease in the Offspring: Findings From Two Scandinavian Birth Cohorts
No full textBackground The association of infections and antibiotic use in pregnancy and the risk of inflammatory bowel disease (IBD) development in the offspring have been scarcely investigated. We examined infection and antibiotic use in pregnancy and the risk of IBD in offspring.Methods We followed participants from the All Babies in Southeast Sweden (ABIS) and the Norwegian mother father and child cohort (MoBa) from birth (1997-2009) until 2020-2021. IBD diagnosis was classified as &gt;= 2 records in national registers. Information on infections (any, gastrointestinal, and respiratory), their timing (early or late in pregnancy), and antibiotic use in pregnancy were collected from questionnaires. Cox proportional-hazard regression and meta-analytic methods were used to estimate pooled adjusted hazard ratios (aHRs) for IBD and its subtypes, adjusted for parental IBD, maternal smoking, and education. Sensitivity analyses accounted for exposure to antibiotics and infections 0-12 months of age.Results We followed 117 493 children for 2 024 299 person-years (follow-up 22.3 years in ABIS and 16.4 years in MoBa), including 451 IBD cases. The aHRs for any infection and respiratory infections in pregnancy and offspring IBD were close to one (aHR = 0.99 [95% CI = 0.73-1.33] and aHR = 1.00 [95% CI = 0.81-1.23], respectively). However, any versus no infection in early pregnancy was associated with IBD development (aHR = 1.26 [95% CI = 1.02-1.55]), particularly Crohn's disease (CD; aHR = 1.40 [95% CI = 1.01-1.93]). Any versus no gastrointestinal infection in late pregnancy was associated with offspring CD (aHR = 1.95 [95% CI = 1.34-2.84]). Antibiotic use in pregnancy was not associated with IBD in the child (aHR = 1.15 [95% CI = 0.93-1.44]).Conclusions In this binational birth cohort study, the risk of offspring IBD varied by infection type and timing but not with maternal antibiotic use in pregnancy. In this study, with over 117 000 participants from 2 Scandinavian birth cohorts, the type and timing of maternal infections in pregnancy were associated with the offspring's risk of inflammatory bowel disease.Funding Agencies|Swedish Society for Medical Research [S20-0007]; Swedish Research Council [2020-01980, K2005-72X-11242-11A, K2008-69X-20826-01-4]; ALF [ALFGBG-915661]; Southeast Sweden study was supported by Barndiabetesfonden (Swedish Child Diabetes Foundation) [0000]; Swedish Council for Working Life and Social Research [FAS2004-1775]; Medical Research Council of Southeast Sweden (FORSS) [0000]; JDRF Wallenberg Foundation [K 98-99D-12813-01A]; Region OEstergoetland and Linkoeping University, Sweden [0000]; Joanna Cocozza Foundation [0000]; Norwegian Ministry of Health and Care Services; Ministry of Education and Research</p
Shaping reflective political leaders
No full textUnlike many other countries, in Sweden there are more significant paths to political leadership than the university degree. Most Swedish political parties offer leadership training programmes that include practical and ideological content and several have a design rooted in the popular education movement. This article explores how party political training for leaders happens in practice, using data from interviews, surveys, text files and ethnographic field notes. Using reflection and ideology as theoretical concepts, the study shows how reflection is used for different aims, and how it can shape participants ideologically. Ideology, in terms of describing what is good, is communicated from the top down in these courses in a general way. In terms of showing what is possible and desirable to either legitimise or challenge, these programmes leave that open to negotiation among the participants, which implies a confidence in their competence by virtue of being elected. Allowing different interpretations risks undermining the party's ideological core; however, since popular education methods are not negotiated, participants are required to reflect deliberately and listen to each others views on the issues under discussion. This can be understood as a safety line as it most likely leads to self-sanitation from the party's perspective
Optimal Test Design for Estimation of Mean Ability Growth
No full textThe design of an achievement test is crucial for many reasons. This article focuses on a population's ability growth between school grades. We define design as the allocating of test items concerning the difficulties. The objective is to present an optimal test design method for estimating the mean and percentile ability growth with good precision. We use the asymptotic expression of the variance in terms of the test information. With that criterion for optimization, we propose to use particle swarm optimization to find the optimal design. The results show that the allocation of the item difficulties depends on item discrimination and the magnitude of the ability growth. The optimization function depends on the examinees' abilities, hence, the value of the unknown mean ability growth. Therefore, we will also use an optimum in-average design and conclude that it is robust to uncertainty in the mean ability growth. A test is, in practice, assembled from items stored in an item pool with calibrated item parameters. Hence, we also perform a discrete optimization using simulated annealing and compare the results to the particle swarm optimization
Human Milk Oligosaccharides in Breast Milk at Two Weeks of Age in Relation to Neurodevelopment in 2-Year-Old Children Born Extremely Preterm : An Explorative Trial
No full textBackground: Preventing neurodevelopmental impairment after extremely preterm birth remains challenging. While breast milk feeding is linked to better neurodevelopment, the underlying mechanisms are unclear. This study explored the association between individual human milk oligosaccharides (HMO) and neurodevelopment at two years of age in extremely preterm children. Methods: Milk samples from mothers of 76 extremely preterm infants collected at two weeks after birth were analyzed for 15 dominant HMOs. Register data from examination and Bayley-III neurodevelopmental assessment at two years' corrected age was retrieved and categorized into levels of impairment. An exploratory analysis examined associations between the HMO composition and neurodevelopment. Results: Bioinformatic volcano plots revealed associations between specific HMOs and outcomes: 3FL with less neurodevelopmental impairment, LSTb with higher Bayley-III cognitive scores, and LSTa with worse neurodevelopmental impairment outcomes. Spearman correlations indicated LSTa was linked to more neurodevelopmental impairment (p = 0.018), lower language (p = 0.009), and motor (p = 0.02) scores, whereas 3FL correlated with less neurodevelopmental impairment (p = 0.02). Dichotomized analysis showed LSTa was associated with more neurodevelopmental impairment and lower language scores (p &lt; 0.05), 3FL with milder neurodevelopmental impairment (p &lt; 0.05), and LSTb with better cognitive (p &lt; 0.01) and language (p &lt; 0.05) scores. No significant associations were found for HMO diversity, total sialic acid content, or secretor/Lewis patterns. Conclusions: In this explorative hypothesis-generating study, certain HMOs appeared to be associated with both potentially beneficial and adverse neurodevelopmental outcomes in extremely preterm infants. However, these findings should be interpreted with caution, as they do not constitute evidence but rather serve as a preliminary foundation for future hypothesis-driven research.Funding Agencies|Swedish Research Council [921.2014-7060]; Swedish Society for Medical Research; Swedish Society of Medicine; Research Council for the South-East Sweden; ALF Grants; Ekhaga Foundation; BioGaia AB; Region Ostergotland</p
Psychological Distress, Skipped Meals, and Insufficient Sleep, and the Occurrence of Back Pain in Adolescent Female Soccer Players: The Karolinska Football Injury Cohort Study
No full textBackground: There is growing awareness that back pain in adolescent soccer (European football) players is not always related to local tissue damage. Approaches taking into consideration lifestyle factors are needed. This study aimed to investigate the association between psychological distress, skipped meals, and insufficient sleep, and the occurrence of back pain among adolescent female soccer players.Hypothesis: Psychological distress, skipped meals, and insufficient sleep are associated with the occurrence of back pain.Study Design: Cohort study.Level of Evidence: Level 3.Methods: Swedish adolescent female soccer players aged 12 to 17 years, without back pain at baseline, were included. The exposures psychological distress, skipped meals, and insufficient sleep were assessed at baseline. The players were followed for 1 year with a weekly web-based survey, where back pain intensity was measured on a numeric rating scale (NRS) ranging from 0 to 10. The outcome back pain was defined as reporting low back pain and/or upper back/neck pain intensity of &gt;= 4 on the NRS. Multivariable Cox proportional hazard regression analyses were performed to estimate the hazard rate ratio (HRR) with 95% CI for the association between each exposure and the outcome back pain, adjusted for age at baseline and parents' education.Results: In total, 351 players were included, and 141 players reported back pain at some point during the 1-year follow-up. The adjusted HRR for back pain was 1.79 (95% CI, 1.11-2.91) for insufficient sleep, 1.45 (95% CI, 0.97-2.17) for psychological distress, and 0.98 (95% CI, 0.61-1.59) for skipped meals.Conclusion: Insufficient sleep and psychological distress are associated with the occurrence of back pain among female adolescent soccer players.Clinical Relevance: Our results contribute to the understanding of what may influence the occurrence of back pain in adolescence, offering potential strategies for the prevention of back pain in adolescent female soccer players
Early-Onset Colorectal Cancer : Risk Factors and Outcomes
No full textColorectal cancer is one of the most common and lethal forms of cancer, and the Western world is experiencing an increase in the incidence among adults aged under 50 years. Up to one third of these patients, referred to as ´early-onset colorectal cancer´ (EOCRC) have a family history of the disease. Lifestyle-related risk factors and a personal history of inflammatory bowel disease (IBD), such as ulcerative colitis and Crohn’s disease, are other important risk factors. The aim of this thesis was to explore hereditary and non-hereditary risk factors associated with early-onset colorectal cancer (EOCRC) and to study the outcomes for patients with IBD coinciding with colorectal cancer. Study I was a population-based observational study that evaluated the awareness of hereditary colorectal cancer, measured by the number of patients referred for genetic counselling. Patients diagnosed with EOCRC in the South-Eastern Healthcare Region of Sweden from 2008 to 2017 were included, and all electronic medical records were reviewed. Of the 278 patients, 116 (42%) had been referred to genetic counselling, among whom 74 (64%) completed their genetic investigation. In 13 (18%) patients, a germline mutation was found. Studies II, III, and IV used the Swedish Colorectal Cancer Database as a data source, which provided multi-linkage between several national health registers. Study II was a case– control study of non-hereditary risk factors, such as metabolic and autoimmune disorders, and their association with EOCRC. The study examined data from 2,626 patients with EOCRC who were diagnosed in Sweden between 2007 and 2016, along with 15,759 comparators matched for age, sex, and county of residence. Metabolic disease nearly doubled the incidence of EOCRC, and IBD was associated with a sixfold increase in the incidence of EOCRC. Presence of non-inflammatory bowel disease was not associated with an increased incidence of EOCRC. Study III compared the recurrence-free survival (RFS) rates for patients who underwent surgical resection for rectal cancer, diagnosed between 1997 and 2021, with or without IBD. Flexible parametric survival models were used and were adjusted for time since surgery, year of diagnosis, sex, age at diagnosis, and Charlson comorbidity index score. Of the 22,082 patients included, 323 (1.5%) had IBD. Neoadjuvant radiotherapy and chemoradiotherapy were administered to 55% and 63% of patients, respectively. The overall RFS was similar between the groups (HR 1.05, 95% CI 0.87–1.26); however, a poorer RFS was observed during the first year after surgery (1-year HR 1.36, 95% CI 1.06–1.73) in patients with IBD. Study IV compared the overall survival, RFS, and cancer-specific survival rates for patients with and without IBDs, who were diagnosed with colon cancer from 2007 to 2021 and surgically resected with curative intent. The Cox proportional hazards model was used, adjusted for sex, age, date of colon cancer surgery, Charlson comorbidity index score, and primary sclerosing cholangitis. Among the 35,640 patients studied, 675 Futher funding from:Lions Forskningsfond</p