40639 research outputs found
Sort by
Environmental dynamics in a Middle Pleistocene refugium: palynological and charcoal evidence from the Megalopolis Basin (Southern Greece)
No full textA fundamental climate transformation commenced at the onset of the Quaternary period, where the recurring glacial and interglacial cycles became progressively extreme, leading to a major build-up of global ice volume towards the Middle Pleistocene. This major climate shift had inescapable consequences on biota. Glacial refugia in southern Europe played a crucial role in the survival and evolution of the temperate biota, including hominins, during the Middle Pleistocene glaciations. This dissertation focuses on the Lower Palaeolithic site Marathousa 1, located in the Megalopolis Basin (Peloponnese, southern Greece), providing a new three-step analysis of its palaeoenvironmental record. The studied profile is dated between 485 and 410 thousand years before present, corresponding to the severe glaciation of the Marine Isotope Stage (MIS) 12. By performing palaeoecological analytical methods, the aim is to explore the environmental and climatic context of a putative hominin refugium in the Megalopolis Basin, at the southernmost tip of the Balkans, through the contribution of palynological and charcoal evidence. A particular focus is placed on the reconstruction of the vegetation, both wetland and terrestrial, providing new insights into environmental and climatic conditions at the site local to hominins, as well as the ecosystem responses to changing climate. Analysis of algae and fungi, including coprophilous fungal spores, has allowed further conclusions about the highly local (wetland) environment,
where Marathousa 1 was situated. For the first time, a multi-species record of coprophilous fungal spores from a Middle Pleistocene site in Greece is available to track the presence of large herbivores and deepen our understanding of the available sources of subsistence for Middle Pleistocene hominins. Another central point is the analysis of the local fire history, using pollen-slide and sedimentary charcoal and charcoal morphotypes, which allows further insights into the Megalopolis ecosystem. Finally, these new palaeoecological records from Marathousa 1 are compared with other proxy records to highlight spatial and temporal patterns of climate variability in southern Europe during the MIS 12 glaciation and the MIS 12/11 transition. In summary, a variety of new insights have been gained, thereby contributing to our understanding of hominin evolution and dispersal in Europe during the Middle Pleistocene.Die Dissertation ist gesperrt bis zum 15. Juli 2027
From Protein Adsorption to Crystallization
No full textProteins are essential to all forms of life. They are involved in almost all biological processes, acting as enzymes, structural components, transporting molecules, involved in signaling, and in immune responses. In addition, proteins are important in medicine, biotechnology, and industry. Many therapeutic drugs are based in proteins, and understanding protein function is essential for developing new treatments. Each protein has a specific function, which is determined by its unique three-dimensional structure.
Because the function of the protein is directly connected to its three-dimensional conformation, solving protein structure is essential. X-ray crystallography remains one of the most used methods for structural determination, but it requires high-quality protein crystals. Despite advances in protein crystallization research, growing suitable protein crystals is still a challenge. Protein crystallization is a complex process that depends on subtle balances between attractive and repulsive interactions. The crystallization mechanism for every known protein is still not understood. As a consequence, many proteins have not yet been crystallized, and the strategy is often based on trial and error, which is time-consuming.
This is why understanding the underlying mechanisms of protein crystallization is essential, and it is the main focus of this dissertation. Specifically, protein crystallization at a solid surface is often neglected, even though it is usually energetically favored compared to crystallization in the bulk solution. In this work, protein adsorption and crystallization at the interface are investigated using both surface-sensitive and bulk-sensitive techniques.
For this purpose, a suitable protein system was first identified. Two negatively charged globular proteins were screened under several conditions. beta-lactoglobulin (BLG) and human serum albumin (HSA) in the presence of three different trivalent salts: YCl3, LaCl3 and CeCl3. The main technique used in this work is a real-time surface-sensitive method called quartz crystal microbalance with dissipation monitoring (QCM-D), which is applied here for the first time to study protein crystallization at the surface. In the first results chapter, HSA in the presence of LaCl3 was chosen as the model system to study protein crystallization with QCM-D.
In the second results chapter, evidence for surface-assisted crystallization is presented. The crystallization of HSA in the presence of a trivalent salt was monitored in real-time. The experimental conditions were chosen near the phase boundary where protein-protein attractions are enhanced but amorphous aggregation and liquid-liquid phase separation (LLPS) are not present. A soft protein multilayer forms at the surface, which acts as a reservoir and environment for heterogeneous nucleation. As a result, protein crystals were only growing at the surface.
In the third results chapter, the same protein system was investigated with two different solvents. Heavy water (D2O) was compared to normal water (H2O). This substitution enabled a fine-tuning of protein interactions, providing a deeper understanding of the mechanisms behind protein crystallization at the surface. The results reinforce the connection between bulk properties and interfacial behavior, highlighting that small changes in the solution environment directly influence surface-induced crystallization.Dissertation ist gesperrt bis 31. Juli 2027
Molecular Interactions Governing Phage Specificity in Staphylococcus aureus
No full textDissertation gesperrt bis zum 10.07.2027!Staphylococcus aureus ist eines der häufigsten bakteriellen Pathogene weltweit. Während viele Antibiotika in der Vergangenheit erfolgreich genutzt wurden, um bakterielle Infektionen zu bekämpfen, steigen Antibiotikaresistenzraten weltweit. Aus diesem Grund sind neue Ansätze erforderlich. Einer dieser Ansätze ist die Nutzung von Bakteriophagen, Viren, die Bakterien infizieren. Da Phagen nur ihr Wirtsbakterium angreifen, üben sie minimalen Druck auf den Rest des Mikrobioms aus. Der Grund für diese Selektivität war bisher jedoch unklar. Es war zwar bekannt, dass S. aureus Phagen an die zellwandgebundene Wandteichonsäure (WTA) bindet, welche in S. aureus aus Ribitolphosphatketten (RboP) mit verschiedenen Glykosylierungen besteht, das genaue WTA-glykosylierungsabhängige Bindemuster der meisten Phagen war jedoch unbekannt. In dieser Arbeit entdeckten wir das Bindemuster der meisten S. aureus-Phagen mithilfe computergestützter Identifikation und Clusterung von Phagen-Rezeptorbindeproteinen (RBPs) und Analyse des Bindeverhaltens an S. aureus-WTA-Mutanten mithilfe fluoreszent markierter RBPs. Hierdurch fanden wir heraus, dass RBPs in Gruppen mit spezifischen Bindemustern abgegrenzt werden können, welche mit dem Bindeverhalten des intakten Phagen übereinstimmt. Zusätzlich fanden wir heraus, dass viele S. aureus-Phagen zwei separate RBPs besitzen, welche beide die WTA ihres Wirts binden. Als Nächstes entdeckten wir, dass S. aureus sein WTA-Glykosylierungsmuster mithilfe von Autoinducerpeptiden (AIPs) des Quorum-Sensing-Systems agr modifizieren kann. Dieser Prozess kann außerdem von anderen Staphylokokken ausgenutzt werden, indem diese alternative AIPs mit inhibitorischer oder stimulatorischer Aktivität ausschütten, wodurch der Metabolismus und die WTA-Zusammensetzung verändert werden. Als Letztes entdeckten wir, dass Staphylococcus epidermidis E73 eine neuartige Glykosyltransferase (TarM(Se)) nutzt, um Glukose an seine RboP-WTA anzubringen, statt der N-Acetylglukosamin-Modifizierung, die S. aureus TarM nutzt. Unsere Studie ist die erste Beschreibung von RboP-WTA mit Glukose, und die Struktur und Funktion von TarM(Se) wurde aufgeklärt. Insgesamt waren wir in der Lage, viele Aspekte der Phagen-Wirts-Interaktion von S. aureus aufzudecken, jedoch ist weitere Forschung notwendig, um die Interaktionen noch besser zu verstehen und die Behandlung von S. aureus mit Bakteriophagen weiter zu optimieren.Staphylococcus aureus is one of the most prevalent bacterial pathogens worldwide. While many antibiotics were successfully used to treat infections in the past, antibiotic resistance rates are rising, and new approaches are required. Among these approaches, viruses of bacteria, called bacteriophages, are especially interesting. Since phages are highly selective for their host bacteria, they exert minimal pressure on the overall microbiota. However, the exact interactions, and the cause for this selectivity, remained unclear. It was known that most S. aureus phages bind to the cell wall bound wall teichoic acid (WTA) glycopolymer, which, in S. aureus, consists of ribitol-phosphate (RboP) repeats with different glycosylations, but the exact WTA-glycosylation dependent binding pattern of most phages was unknown. In this work, we elucidated the binding pattern of S. aureus phages by combining computational identification and clustering of phage receptor-binding proteins (RBPs) and analysis of binding to S. aureus WTA mutant panels via fluorescently labelled RBPs. This way, we discovered that S. aureus phage RBPs cluster in distinct groups, each with its own distinct binding pattern that matches the binding behavior of the intact phage. Additionally, we discovered that many S. aureus phages carry two separate RBPs, both binding to the WTA of their host. Next, we discovered that S. aureus can modulate its WTA glycosylation pattern via autoinducer peptides (AIPs) of the quorum-sensing agr system. This process can also be exploited by other staphylococci by using alternative AIPs with inhibitory or stimulatory activity, thus altering the metabolism and WTA composition of S. aureus. Lastly, we discovered that Staphylococcus epidermidis E73 uses a novel glycosyltransferase, TarM(Se), to attach glucose to the RboP-WTA instead of the N-acetylglucosamine modification conferred by S. aureus TarM. Our study is the first description of RboP-WTA decorated with glucose, and the structure and function of TarM(Se) was elucidated. Overall, we were able to uncover many aspects of the phage-host interaction in S. aureus, but further research is necessary to better understand and optimize the treatment of S. aureus with bacteriophages
Genetic aspects of insect cuticle function
No full textDie Dissertation ist gesperrt bis zum 27. Juli 2027 !As the outermost organ of insects, the cuticle has diverse functions conferring
strong adaptability to various environments: 1) as a solid exoskeleton, it offers
support and muscle attachment sites needed for locomotion; 2) as a barrier, it
not only prevents inner water evaporation under extreme conditions, but also
protects against penetration of xenobiotics and pathogenic microorganisms; 3)
it may also serve as a vector to transport attached pollens, microorganisms and
even diseases, thereby contributing to ecological communication.
To fulfill its ecological impact, the insect cuticle is composed of mainly three
components: the polysaccharide chitin, proteins and lipids. Thickness mainly
depends on the co-assembly of chitin and cuticular proteins; while the barrier
function counts on the presence and composition of cuticular lipids.
In this study, two insect species (Drosophila melanogaster and Locusta
migratoria) were chosen to address different aspects of the function of the
cuticle in legs. Using RNA interference (RNAi), I studied the function of cuticular
lipids in D. melanogaster with respect to their importance as xenobiotics barrier
and as a possible substrate for the cuticle microbiome. When silencing key
genes (desat1, fasn2, cyp4g1) of the lipid synthesis pathway, unsaturated,
methyl branched and total CHC production were reduced, respectively.
Xenobiotics penetration (insecticides and dye) and surface microorganism
analyses showed that especially unsaturated CHCs are important for both
aspects. Moreover, data presented here indicate that the tarsal cuticle actively
contribute to the interaction of the insect with its proximal environment and its
modification at least partly through the surface lipids.
The laminate structure of cuticle in locust tibia has been reported. It is formed
following a circadian rhythm. Indeed, two different chitin orientation modes
during day and night occur. The main difference between day and night
organization is possibly the involvement of specific proteins. Here, two proteins,
LmObst-B and LmObst-C, were identified by proteomics in “night” cuticlesamples. Antibody staining was performed to study the localization of Obst- B
and Obst-C during the day. I show that Obst- B and Obst-C activity does not
follow the circadian rhythm of their expression but rather of the subcellular
localization from the cytoplasm (night) to the membrane (day). Functional
studies were conducted by RNAi; after co-injection of dsLmObst-B and
dsLmObst-C, locomotion (walking and jumping) deficiency was observed. In
addition, mechanical analyses indicate a weaker cuticle with a lower Young's
modulus.
In summary, the functions of different components of the insect cuticle were
described in this dissertation, focusing on the legs. This work contributes, hence,
to the advancement of our understanding of the function of the insect leg cuticle
Functional diversity in wild-type and degenerated mouse retinae
No full textRetinitis Pigmentosa (RP) is a hereditary disease characterised by the progressive degeneration of rod and cone photoreceptors, ultimately leading to blindness. Despite extensive research, effective therapies remain scarce. Understanding how photoreceptor degeneration affects downstream retinal processing is crucial for developing targeted interventions. In this thesis, I investigate the functional consequences of RP on the retinal ganglion cells (RGCs), the final output neurons of the retina, using the Pde6brd10 (rd10, Chang et al. 2002) mouse model. This model closely mimics human RP, with rod degeneration beginning around postnatal day (P) 16, followed by secondary cone loss from P30 onwards. By P180, nearly all photoreceptors are lost, leading to significant
changes in retinal circuitry (e.g., Puthussery et al. 2009).
To study these effects, I performed two-photon Ca2+ imaging of RGCs in ex vivo retinal whole mounts from both rd10 and wild-type mice at four key time points: P30, P45, P90, and P180. Using a semi-automated cell classification approach, I assessed how photoreceptor degeneration influences RGC light responses, functional diversity, and signal reliability. I first confirmed that RGC density remained stable throughout degeneration, ruling out major cell loss as a confounding factor. However, the proportion of light-responsive RGCs declined significantly, with a 21% reduction between P30 and P45 and another 14% between P45 and P90. By P180, nearly all light responses were lost. To examine how the composition of functional RGC types changed over time,
I applied an established classification framework (Baden et al. 2016; Qiu et al.
2023; Gonschorek et al. 2024) to cluster RGC responses. Even at P90, most
wild-type RGC types were still present in rd10 retinae, suggesting that earlystage
degeneration does not immediately disrupt functional diversity. However, when investigating the fractions of broad response types (i.e. functional supergroups),
RGC subtypes with ‘Off’ response components (‘Off’ and ‘On-Off’ RGCs) exhibited greater vulnerability than ‘On’-RGCs, with ‘Fast-On’ cells showing the highest resilience. Additionally, direction-selective RGCs declined earlier than orientation-selective types, indicating differential sensitivity to retinal
rewiring.
These findings demonstrate that RP progression leads to differential disruptions
in RGC function, even in the absence of cell loss. This suggests that
degeneration-induced remodelling in the retinal network alters visual signal processing
at the level of the retinal output. This supports the idea that (yet
undiscovered) changes in the presynaptic retinal circuitry and/or other alterations
affect functional diversity among RGCs during degeneration, potentially
contributing to the observed differences in RGC response stability. One possible
explanation is an emerging imbalance between ‘On’- and ‘Off’-pathways, which
may differentially impact specific RGC types and their ability to maintain stable
responses. Understanding these circuit-level changes is essential for future
research on synaptic plasticity and retinal reorganisation in RP and may help
inform therapeutic strategies aimed at preserving functional vision in patients
with photoreceptor degenerations.Die Dissertation ist gesperrt bis zum 05. Juni 2027
Building a Computational Model of Goal Setting, Performance, and Emotions: Toward a Mechanistic Understanding of Dynamic Self-Regulated Learning
No full textDie Dissertation ist gesperrt bis zum 24. Februar 2027 !Selbstreguliertes Lernen (SRL) bezieht sich auf die Fähigkeit von Lernenden, kognitive,
behaviorale, emotionale, motivationale, metakognitive und umgebungsbezogene Aspekte ihres
Lernens zu steuern und zu kontrollieren (Panadero, 2017; Zimmerman, 1989). Lernende
müssen Selbstregulationsstrategien entwickeln und effektiv anwenden, um bei ihren
Lernaufgaben erfolgreich zu sein. Es wurde beobachtet, dass Selbstregulationsstrategien nicht
nur mit besseren akademischen Leistungen korrelieren (Dignath & Büttner, 2008; Dignath et
al., 2008), sondern dass die Verbesserung von SRL-Fähigkeiten auch notwendig ist, um
zukünftige Herausforderungen bei der Arbeit zu bewältigen (Zimmerman, 2002). Um zu
verstehen, wie und wann bestimmte Aspekte des SRL auftreten, haben Forscher diese Prozesse
in dynamische Modelle organisiert, die häufig in drei wiederkehrende Phasen unterteilt werden.
In Anlehnung an die Begriffe einer aktuellen Übersichtsarbeit (Panadero, 2017) durchlaufen
Lernende während des SRL typischerweise die folgenden Phasen. Zuerst durchlaufen sie eine
Planungsphase, in der Ziele gesetzt und Strategien zur Erreichung dieser Ziele ausgewählt
werden. Auch Selbstüberzeugungen, einschließlich Selbstwirksamkeitsüberzeugungen (Tolli &
Schmidt, 2008) und Kontrollbewertungen (Converse et al., 2010), können sich hier entwickeln
und diese Prozesse beeinflussen. Anschließend betreten Lernende die Leitungsphase, in der sie
die Lern- oder Studienaufgabe ausführen. In dieser Phase überwachen sie ihren Fortschritt, und
die in der Planungsphase gewählten Strategien können eingesetzt werden, um die Motivation
zur Erreichung ihrer Ziele aufrechtzuerhalten. Schließlich beenden Lernende ihre Lernsitzung
in der Bewertungsphase, in der sie ihre Leistung im Verhältnis zu ihren Zielen evaluieren.
Selbstbewertungen entstehen, die zu emotionalen Reaktionen führen. Obwohl bereits viele
Studien durchgeführt wurden, um zu verstehen, wie diese Phasen miteinander interagieren,
beschränkt sich ein Großteil der Forschung auf lineare paarweise Vergleiche der zugehörigen
Variablen. Es wurde jedoch wenig unternommen, um alle drei Phasen gemeinsam in einem
formalen Modell zu untersuchen, das die gesamte dynamische Natur des SRL erfasst.
Diese Dissertation zielt darauf ab, dynamisches SRL auf eine ganzheitliche Weise zu
verstehen. Drei Schlüsselvariablen, eine für jede der drei Phasen des SRL, wurden identifiziert:
Zielsetzung als Reflexion der Planungsphase, Aufgabenleistung für die Leistungsphase und
Emotionen als Konsequenzen der Bewertungsphase. Diese Variablen wurden aus folgenden
Gründen gewählt. Erstens wurde die Beziehung zwischen Zielen und Leistung intensiv
untersucht, wobei Diskrepanzen zwischen ihnen als zentral in mehreren Theorien (e.g.,
Bandura, 1986; Carver & Scheier, 1998) und empirischen Studien (e.g., Ilies & Judge, 2005;
Ilies et al., 2010; Theobald et al., 2021; Tolli & Schmidt, 2008) in der Literatur zur
Selbstregulation hervorgehoben wurden. Diese Diskrepanzen können auch die Entstehung von
Leistungsemotionen beeinflussen (Pekrun et al., 2002), die wiederum nachfolgende
Zielsetzungen (e.g., Ahrens, 1987; Cervone et al., 1994; Ilies & Judge, 2005; Ilies et al., 2010;
Scott & Cervone, 2002) und die Leistung (e.g., Carver, 2003; Fulford et al., 2010; Seo & Patall,
2021; Thürmer et al., 2019) beeinflussen. Um Theorie und empirische Befunde miteinander zu
verbinden, wird ein computationales Modell zur Zielsetzung, Leistung und Emotionen (Ballard
et al., 2021; Farrell & Lewandowsky, 2010) entwickelt, getestet und in drei verschiedenen
Datensätzen angewendet, um zu zeigen, wie dieses ganzheitliche Verständnis von
dynamischem SRL erreicht werden kann.
Es gibt mehrere Motivationen für die Durchführung der in dieser Dissertation
vorgestellten Forschung. Erstens wird in Kapitel 2 (The Need for Computational Modeling in
Educational Psychology) argumentiert, warum eine solche Modellierung in der
Bildungspsychologie erforderlich ist. Während ein Großteil der Literatur in der
Bildungspsychologie auf die Untersuchung linearer paarweiser Beziehungen zwischen
Variablen setzt, wird computationale Modellierung flexibel eingesetzt, um das Verhalten
dynamischer Systeme als Ganzes zu modellieren, die mehrere Interaktionen berücksichtigen
(e.g., Ballard et al., 2016; Gee et al., 2018; Hardy et al., 2019). Die Vorteile der computationalen
Modellierung werden in der Übersichtsarbeit erörtert, einschließlich ihrer Flexibilität in der
Spezifikation, der Möglichkeit, Daten zu simulieren, um theoretische Annahmen zu verstehen,
und des Beitrags zum Verständnis und zur Weiterentwicklung von Theorien. Die
Übersichtsarbeit zielt darauf ab, die Bedenken einiger Psychologen hinsichtlich des aktuellen
Stands des Fachgebiets (e.g., Greene, 2022; Murayama & von Keyserlingk, 2024; Pekrun et al.,
2024) zu zerstreuen und Argumente dafür zu liefern, warum computationale Modellierung, die
in verwandten Disziplinen wie der Organisationspsychologie eingesetzt wurde, auch in der
Bildungspsychologie verstärkt angewendet werden sollte.
Zweitens haben Untersuchungen zur Rolle von Emotionen bei der Zielsetzung und
Leistung in der Literatur uneinheitliche Ergebnisse geliefert. In Kapitel 3 (Study 1: The Role of
Emotions on Goal Setting and Performance in Dynamic Self-Regulated Learning: A
Computational Modeling Approach) wird ein computationales Modell zur Zielsetzung,
Leistung und Emotionen vorgestellt, welches ein Werkzeug zur Bewertung der Auswirkungen
von Emotionen bietet. Diese Studie verwendet zwei Datensätze. Der erste Datensatz (N = 357)
wurde in einer Online-Mathematikaufgabe erhoben, bei der die Teilnehmenden eine Serie von
20 Blöcken mathematischer Gleichungen bearbeiteten, während Ziele und Emotionen erfasst
wurden. Dieser Datensatz wurde genutzt, um das computationale Modell zu validieren. Die
Ergebnisse zeigten, dass das Modell in der Lage war, die Effekte der Diskrepanz zwischen Ziel
und Leistung (Ziel-Leistung-Diskrepanz, GPD) und Emotionen abzubilden. Dies deutet darauf
hin, dass Emotionen zusätzliche Informationen darüber liefern können, wie Individuen Ziele
setzen und Aufgaben bearbeiten, die über die bloßen Effekte von Diskrepanzen hinausgehen.
Der zweite Datensatz (N = 305) wurde in einer Studie zur Zielsetzungsverhalten von
Medizinstudierenden erhoben, die sich auf das zweite Staatsexamen in Deutschland
vorbereiteten, welches als entscheidend für die Fortsetzung ihrer Berufsausbildung gilt.
Mithilfe einer Online-App bearbeiteten die Studierenden frühere Prüfungsfragen, setzten
tägliche Lernziele und berichteten über ihre Emotionen vor und nach jeder Lernsitzung. In
diesem ökologisch validen Kontext zeigten die Ergebnisse, dass positive Emotionen vor der
Lerneinheit einen positiven Einfluss auf die Leistung hatten, während keine direkten Effekte
auf die Zielsetzung beobachtet wurden (welche möglicherweise durch die Wichtigkeit der
Prüfung maskiert sind). Dies deutet darauf hin, dass positive Emotionen den Einsatz von
Anstrengungen während des Lernens fördern könnten. Diese Studie trägt zur Literatur bei,
indem sie eine Anwendungsfallstudie der computationalen Modellierung präsentiert und zeigt,
wie ein solches Modell entwickelt und getestet werden kann, um die Rolle von Emotionen im
dynamischen SRL zu verstehen.
Drittens wurde das Setzen ambitionierter Ziele bereits in mehreren Studien beobachtet
(Campion & Lord, 1982; Donovan, 2009; Lin et al., 2024) und untersucht (Chevance et al.,
2021; Ritzema, 2016). Um die Auswirkungen dieser Strategie im Bildungskontext mithilfe
eines dynamischen SRL-Rahmens zu verstehen, beschreibt Kapitel 4 (Study 2 – Pushing the
Boundaries: The Effects of High Goal Setting in Dynamic Self-Regulated Learning) eine Studie,
in der Studierende entweder reguläre Ziele (Kontrollgruppe, N = 77) oder hohe, ambitionierte
Ziele (Experimentalgruppe, N = 69) setzen sollten, während sie für den Leseverständnisteil des
Test of English as a Foreign Language (TOEFL) lernten. Das am besten passende
computationale Modell aus Kapitel 3 wurde auf diesen Datensatz angewendet, wobei ein
hierarchischer Bayes-Ansatz genutzt wurde, um die Parameter der beiden Gruppen zu schätzen.
Die Ergebnisse dieser Studie zeigten, dass computationale Modellierung Unterschiede
zwischen den Gruppen in den zugrunde liegenden Mechanismen der Zielsetzungsverhalten
nach Zielversagen aufzeigen konnte. Darüber hinaus wurde ein positiver Effekt von Emotionen
vor dem Lernen auf die Zielsetzung beobachtet, was darauf hindeutet, dass Emotionen die
Wahrnehmung der eigenen Kompetenz fördern können. Diese Studie unterstützt weiter die
Nützlichkeit der computationalen Modellierung in der empirischen Bildungsforschung, indem
sie zusätzliche Belege liefert, dass diese Technik verwendet werden kann, um zugrunde
liegende psychologische Prozesse zu beleuchten.
Durch die Anwendung einer neuartigen Analysemethodik in der empirischen
Bildungsforschung liefert diese Dissertation eine erfolgreiche Darstellung der computationalen
Modellierung und zeigt, wie sie verwendet werden kann, um zugrunde liegende Prozesse im
Bereich der Zielsetzung, Aufgabenleistung und Emotionen zu identifizieren. Dabei leistet sie
einen theoretischen Beitrag zum Verständnis bestimmter Prozesse, einschließlich der
Auswirkungen von Zielversagen auf Zielsetzungsverhalten sowie der zeitlichen Bedeutung von
Emotionen für die Zielsetzung und die zielgerichtete Handlung. Durch die Übersichtsarbeit, die
für die Einführung der computationalen Modellierung plädiert, sowie die beiden empirischen
Studien, die Ihren Wert bei der Untersuchung des dynamischen SRL demonstrieren, zielt diese
Dissertation darauf ab, eine neuartige Methodik für den Einsatz in der empirischen
Bildungsforschung voranzutreiben.Self-regulated learning (SRL) refers to students’ ability to manage and control the cognitive, behavioral, emotional, motivational, metacognitive, and environmental aspects of their learning (Panadero, 2017; Zimmerman, 1989). Students must develop and effectively apply self-regulation strategies to succeed in their learning tasks. Doing so has not only been observed to correlate with better academic outcomes (Dignath, et al., 2008; Dignath & Büttner, 2008), but honing SRL skills is also necessary for overcoming future challenges in the workplace (Zimmerman, 2002). To understand how and when certain aspects of SRL occur, researchers have organized its processes into dynamic models, most notably in three reoccurring phases. Borrowing the terms from a recent review paper (Panadero, 2017), students typically transgress through the following phases during SRL. First, students go through a planning phase in which goals are set and strategies to achieve these goals are chosen. Self-beliefs may also take shape here and influence these processes, including self-efficacy beliefs (Tolli & Schmidt, 2008) and control appraisals (Converse et al., 2010). Next, students enter the performance phase and where they complete the learning or studying task. Here, students typically monitor their progress and employ the strategies chosen in the initial planning phase may be used here to keep up their motivation to reach their goals. Finally, students finish their learning session in the appraisal phase in which they evaluate their performance in relation to their goals. Self-evaluations arise, leading to emotional reactions. While much work has been done to understand how these phases interact with each other, much of the research has only investigated linear pairwise comparisons of associated variables. Little has been done to succinctly investigate all three phases together in a formal model that captures the whole
dynamic nature of SRL.
To this end, this dissertation seeks to understand dynamic SRL in a holistic manner. Three key variables, one for each of the three phases of SRL, are identified: goal setting as a reflection of the planning phase, task performance for the performance phase, and emotions as consequences of the appraisal phase. These variables were chosen for the following reasons. First, the relationship between goals and performance has been heavily researched, where discrepancies between them have been found to play a major role, as noted in several theories (e.g., Bandura, 1986; Carver &Scheier, 1998) and empirical studies (e.g., Ilies et al., 2010; Ilies & Judge, 2005; Theobald et al., 2021; Tolli & Schmidt, 2008) in the self-regulation literature. Furthermore, these discrepancies can be implicated in the feeling of achievement emotions (Pekrun et al., 2002), which then have subsequent effects on further goal setting (e.g., Ahrens, 1987; Cervone et al., 1994; Ilies & Judge, 2005; Ilies et al., 2010; Scott & Cervone, 2002) and performance (e.g., Carver, 2003; Fulford et al., 2010; Seo & Patall, 2021; Thürmer et al., 2019). To incorporate theory and the empirical findings together, a computational model of goal setting, performance, and emotions (Ballard et al., 2021; Farrell & Lewandowsky, 2010) is developed, tested, and applied in three separate datasets to show how this holistic understanding of dynamic SRL can be achieved.
There are several motivations for conducting the research presented in this dissertation. First, as detailed in the Chapter 2 (The Need for Computational Modeling in Educational Psychology), the dissertation provides arguments as to why such modeling is necessary in educational psychology. While much of the literature in educational psychology relies on investigating linear pairwise relationships between variables, computational modeling can and has been flexibly used to model the behavior of dynamic systems as a whole that account for multiple interactions (e.g., Ballard et al., 2016; Gee et al., 2018; Hardy et al., 2019). Benefits of computational modeling are discussed in the review paper, including its flexibility in its specification, the ability to simulate data to understand theoretical assumptions, and its benefit to understanding and further developing theory. The review paper hopes to alleviate some psychologists’ concerns for the current state of the field (e.g., Greene, 2022; Murayama and von Keyserlingk, 2024; Pekrun et al., 2024) and to provide arguments as to why computational modeling, which has been used in other associated fields such as organizational psychology, should be employed more in educational psychology research. Second, investigations into the role of emotions on goal setting and performance have yielded inconsistent findings in the literature. In Chapter 3 (Study 1: The Role of Emotions on Goal Setting and Performance in Dynamic Self-Regulated Learning: A Computational Modeling Approach), a computational model of goal setting, performance, and emotions is presented that provides a tool for the assessment of the effects of emotions. This study uses two datasets. The first dataset (N = 357) was collected in an online math task in which participants completed an online math task where goals and emotions were assessed as students completed 20 blocks of math equations. This dataset was used to validate the computational model and showed that the model was able to capture effects of the goal-performance discrepancy (GPD) and emotions, indicating that emotions could provide extra information as to how individuals set goals and performed on tasks that went above and beyond the effects of discrepancies. The second dataset (N = 305) was collected from a study investigating goal setting behavior in medical students preparing for the second state exam in Germany, which had high stakes as it determined whether students would be able to continue with their career training. Using an online app, students reviewed previous exam questions while setting daily goals and reporting their emotions before and after each study session. In this more ecologically valid setting, the results revealed a positive effect of pre-learning emotions on performance while no effects were observed in goal setting (with a possible masking of effects due to the high-stakes nature of the exam), suggesting that positive emotions encouraged more effort to be exerted while studying. This study contributes to the literature by presenting a use case of computational modeling by showing how such a model can be developed and tested in an attempt to understand the role of emotions in dynamic SRL.
Third, high goal setting has also been observed (Campion & Lord, 1982; Donovan, 2009; Lin et al., 2024) and investigated (Chevance et al., 2021; Ritzema, 2016). To understand their effects in the educational setting using a dynamic SRL framework, Chapter 4 (Study 2 – Pushing the Boundaries: The Effects of High Goal Setting in Dynamic Self-Regulated Learning) describes a study in which students were asked to set normal goals (control group, N = 77) or high, ambitious goals (experimental group, N = 69) while studying for the reading comprehension section of the Test of English as a Foreign Language. The best-fitting computational model ascertained from the work presented in Chapter 3 was also applied to this dataset using a hierarchical Bayesian framework that estimated parameters for the two groups. The findings in this study provided evidence that computational modeling could elucidate differences between groups in underlying mechanisms governing goal setting behavior after goal failure. Furthermore, a positive effect of pre-learning emotions on goal setting were observed, suggesting that emotions may enhance how students feel about their competencies. This study further promotes the usefulness of computational modeling in educational psychology by providing more evidence that it can be used to highlight underlying psychological processes.
By using a novel analysis technique in educational psychology, this dissertation provides a successful account of computational modeling and how it could be used to identify underlying processes governing goal setting behavior, task performance, and the role of emotions. In doing so, it advances our theoretical understanding of certain processes, including how goal failure may affect goal setting behaviors as well as how timing of emotions may be crucial to how students set goals for themselves and how goal-directed behavior is impacted. Through the review paper arguing for the implementation of computational modeling as well as the two empirical studies that shows its value in investigating dynamic SRL, this dissertation aims to introduce a novel technique for use in educational psychology. untranslate
Iron and carbon cycling in permafrost thaw ponds: Implications for greenhouse gas emissions
No full textDie Dissertation ist gesperrt bis zum 26. Februar 2027 !Rapid warming in the Arctic is driving permafrost thaw, which in turn alters hydrology, vegetation, carbon accumulation, and microbial communities. All of these changes significantly influence greenhouse gas (GHG) emissions, including carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O). A key factor controlling these emissions is the association between organic carbon (OC) and iron (Fe) minerals, as up to 50% of OC in permafrost soils is stabilized by Fe(III) minerals. As the climate warms, these Fe(III)-OC associations become destabilized through the reductive dissolution of Fe(III) minerals, a process driven by Fe(III)-reducing bacteria in waterlogged and oxygen-limited conditions. This process releases dissolved organic carbon (DOC), which can then be metabolized, leading to greater GHG emissions (primarily CO2 and CH4).
As permafrost continues to thaw, aqueous iron (Fe2+) and associated OC are released in both dissolved and particulate forms and are transported into small thaw ponds where redox conditions fluctuate. Seasonal runoff and extended waterlogging can shift redox conditions from oxic to anoxic, affecting the balance between Fe(II) oxidation and Fe(III) reduction. However, the impact of these redox shifts on Fe-OC interactions, and consequently on carbon mobilization and GHG production, remains unclear.
To address this knowledge gap, I first characterized Fe and OC phases in partially thawed ‘bog’ and fully thawed ‘fen’ ponds at Stordalen Mire near Abisko, Sweden. I collected samples and analyzed them for Fe and OC concentrations, Fe redox states, mineralogy, OC functional groups, and the spatial correlation between Fe and OC using both bulk and nanoscale techniques. I then incubated bog pond samples in the laboratory under different redox conditions (anoxic, oxic, and fluctuating anoxic-oxic), and monitored Fe(II) and total Fe concentrations, OC levels, CO2 and CH4 fluxes, Fe mineral speciation, and particle charge interactions.
The results showed that bog thaw ponds had higher OC concentrations (50 to 352 mg/L), whereas fen thaw ponds contained more iron (1.5 to 212 mg/L). Iron occurred in both Fe(II) and Fe(III) redox states, either associated with OC (8 to 80%) or precipitated as poorly crystalline Fe(III) oxyhydroxides. Fe-OC associations were present in dissolved, small particulate, and large particulate size fractions. While Fe(II) dominated the large particulate fraction, Fe(III) was detected in the dissolved fraction, possibly as Fe(III)-OC associations or ferrihydrite, and remained stable under the anoxic conditions typical of thaw ponds.
Subjecting pond water to anoxic incubation indicated that microbial Fe(III) reduction released OC into the dissolved phase of thaw ponds. Under oxic conditions, however, OC re-adsorbed onto Fe(III) minerals. It can be concluded that microbial Fe(III) reduction not only drove iron cycling but also controlled OC availability and transformation, particularly when redox conditions changed. With continuous thawing, intensified redox fluctuations may speed up OC breakdown, increasing GHG emissions and accelerating climate change. Ultimately, mineral-associated OC in thaw ponds was susceptible to microbial decomposition and shifting redox conditions promoted its mobilization and the release of GHGs.
In addition to environmental samples, I examined an iron-organic complex called pulcherrimin which is produced by the aerobic bacterium Staphylococcus epidermidis. I observed that iron concentration in the growth medium influenced pulcherrimin production. This concentration affected the distribution of iron between the supernatant and pellet, with approximately 70% of the iron remaining in the supernatant alongside extracellular pulcherrimin. This served as another example of an iron-organic complex with potential applications, such as managing biofilm-related infections through iron sequestration and complexation
Intratumoral Senescence Drives Cancer Metastasis
No full textCellular senescence is a stress-induced cell cycle arrest that can be triggered by various stimuli such as telomere shortening, oncogene activation, radiation or chemotherapy. Through their senescence-associated secretory phenotype (SASP), senescent cells shape their microenvironment and have both beneficial and detrimental effects, affecting processes such as tissue repair, inflammation and tumorigenesis. While extensive data exists on the role of therapy-induced senescent tumor and stromal cells, little is known about senescent cancer cells under therapy-naive conditions.
The aim of this work was to characterize the role of spontaneously senescent tumor cells (SSTC) under therapy-naïve conditions. SSTC were found in therapy-naive murine and human colorectal carcinoma (CRC), cholangiocarcinoma (CCA) and pancreatic ductal adenocarcinoma (PDAC) as well as in 27 different human and murine tumor lines. Interestingly, the percentage of SSTCs within the lines correlated with their in vitro invasive capacity. Analysis of primary human tumor material also showed an increased frequency of SSTC in patients with regional lymph node metastasis compared to patients with localized disease. Pharmacological senolysis with navitoclax or genetic ablation of SSTC significantly reduced tumor cell invasion in vitro and in vivo, resulting in significantly reduced metastasis in the passive splenic seeding model and in two syngenic mouse models of CRC and CCA. Spontaneous senescence and tumor invasion/metastasis were shown to be functionally linked, with selection of highly invasive and metastatic cells leading to increased SSTC numbers.
Collected data suggest a model in which SSTC are derived from a subset of cancer cells with increased resistance to apoptosis, and functional studies confirmed the induction of SSTC in such cells upon exposure to microenvironmental factors such as acidosis or hypoxia. In turn, several SASP factors from SSTC were found to act on neighboring cancer cells to increase their metastatic potential.
Overall this thesis pinpoints a crucial role of SSTC for initiation of metastasis. The data harbor important translational implications, as, despite intensified screening and early detection efforts, almost half all human tumors are still diagnosed at advanced stages. Pharmacological senolysis in patients with increased cancer risk might allow to detect more cancers at earlier curative stages and thus greatly increase the prognosis of cancer patients.Dissertation ist gesperrt bis 10. Februar 2027
Following the Traces of Curious Minds: Intellectual Curiosity and its Behavioral Manifestations Across Learning Contexts
No full textIntellectual curiosity is a desire for information and knowledge that drives individuals to seek intellectually stimulating activities and environments (Litman & Spielberger, 2003; Mussel, 2013a; von Stumm & Ackerman, 2013). It plays a central role in how humans learn and develop, both in everyday life and formal education (Grossnickle, 2016; von Stumm et al., 2011). Curiosity-driven learning has been widely studied, from basic cognitive mechanisms to its function in classrooms. However, research across disciplines often focuses on different aspects, making it difficult to form a comprehensive understanding of the curiosity-driven learning process (Hassin & Shohamy, 2020). Moreover, although information-seeking and learning behaviors are seen as central expressions of curiosity, the behavioral mechanisms underlying knowledge acquisition have received limited attention so far.
To address these gaps, this dissertation investigated curiosity-driven learning as a multi-level process shaped by state-level dynamics, person characteristics, and the learning context. Each of these levels play an important role but they also interact to influence how we seek and acquire knowledge. Two key objectives guided this dissertation: (1) to integrate disciplinary perspectives by examining intellectual curiosity across states, traits, and learning contexts; and (2) to enrich our understanding of intellectual curiosity by focusing on its behavioral expressions. To this end, the dissertation introduced behavioral trace data as a promising method for capturing curiosity-driven learning behaviors across varied settings. Three empirical studies were conducted to address these objectives, each exploring different facets of curiosity. These studies investigated intellectual curiosity at different degrees of granularity, moving from cognitive mechanisms captured by basic metrics of information seeking to linking students’ trait curiosity to their real-world learning behaviors. Collectively, this dissertation contributes to a richer and behaviorally grounded account of intellectual curiosity, deepening our understanding of how intellectual curiosity supports human learning.Die Dissertation ist gesperrt bis zum 19. September 2027
Moral Ordering. A Study of Neighborly Relations
No full textDie Dissertation ist gesperrt bis zum 19. Dezember 2027 !Die Studie analysiert den Zusammenhang von Nachbarschaftsbeziehungen, Moral und Bedrohungen. Empirisch baut die Arbeit auf einer ethnografischen Fallstudie einer Nachbarschaftskonstellation in Johannesburg auf, bestehend aus der Mittelschichts-Suburb Mindalore mit weißer und schwarzer Bevölkerung und der informellen Siedlung Soul City mit einer marginalisierten schwarzen Bevölkerung. Der entwickelte figurationssoziologische Blick fokussiert auf die alltäglichen Praktiken der Nachbar*innen, in denen sie sich wechselseitig bewerten, in Gruppen einteilen und hierarchisieren. Diese Praktiken des nachbarschaftlichen Ordnens sind inhärent moralisch, also operieren entlang der Unterscheidung gut/richtig – falsch/schlecht/böse. Die Arbeit zeigt in drei Kapiteln anhand verschiedener Bedrohungen auf, wie diese das figurational-moralische Ordnen anstoßen und die Nachbarn sich jeweils unterschiedlich zueinander in Beziehung setzen. Die Bedrohung durch Kriminalität moralisiert die class Differenzierungen und vereinigt so die Bewohner*innen der Mittelschichtsnachbarschaft (unabhängig von deren race Differenzierungen) gegen die Bewohner*innen der informellen Siedlung als vermeintliche Täter. Die Umwelt- und Gesundheitsbedrohung durch giftige Überreste des lokalen Goldabbaus führte zu einem Aktivismus, der Nachbar*innen aus beiden Vierteln zusammenbrachte und zusammenschweißte gegen die Verantwortlichen aus Wirtschaft und Politik. In der informellen Siedlung Soul City führte die Bedrohung durch Bandengewalt, die mit dem informellen Goldabbau vor Ort einhergeht, zu einer Spaltung der lokalen Bevölkerung entlang von xenophoben Unterscheidungen und Verdächtigungen der Kooperation mit den Banden. In der Mittelschichtsnachbarschaft Mindalore wurde der Wandel der race Komposition nach Ende der Apartheid von der alteingesessenen etablierten weißen Bevölkerung als bedrohlich wahrgenommen, wobei ihre daraus resultierenden rassistischen Zurückweisungen der neuen schwarzen Nachbarn von diesen wiederum als Bedrohung des friedlichen (nachbarschaftlichen) Zusammenlebens des Post-Apartheid Versprechens angesehen wurden mit dem Resultat der figurativen Spaltung der Nachbarschaft entlang von race Differenzierungen. Die verschiedenen Kapitel verdeutlichen die unterschiedlichen koexistierenden Modi des nachbarschaftlichen Sich-in-Beziehung-setzens, welche in und zwischen den beiden Vierteln Soul City und Mindalore vorherrschen.This study analyzes the relationship between neighborhood relations, morality, and threats. Empirically, the work is based on an ethnographic case study of a neighborhood constellation in Johannesburg, consisting of the middle-class suburb of Mindalore with a white and black population and the informal settlement of Soul City with a marginalized black population. The figurational perspective developed focuses on the everyday practices of neighbors, in which they evaluate each other, divide themselves into groups, and establish hierarchies. These practices of neighborhood organization are inherently moral, i.e., they operate along the distinction between good/right and wrong/bad/evil. In three chapters, the study uses different threats to show how these trigger figurational-moral ordering and how neighbors relate to each other in differing ways. The threat of crime moralizes class differentiations and thus unites the residents of the middle-class neighborhood (regardless of their racialized differences) against the residents of the informal settlement as alleged perpetrators. The environmental and health threat posed by toxic residues from local gold mining led to activism that brought neighbors from both neighborhoods together and united them against those responsible in business and politics. In the informal settlement of Soul City, the threat of gang violence associated with informal gold mining in the area led to a division of the local population along xenophobic lines and suspicions of cooperation with the gangs. In the middle-class neighborhood of Mindalore, the change in racial composition after the end of apartheid was perceived as threatening by the established white population, whose resulting racist rejection of their new black neighbors was in turn seen by the latter as a threat to peaceful (neighborly) coexistence promised by post-apartheid, resulting in the figurative division of the neighborhood along racial lines. As a whole, this work illustrates the different coexisting modes of neighborly relations that prevail in and between the two neighborhoods of Soul City and Mindalore