40639 research outputs found
Sort by
Diagrammverständnis von Wirtschaftsstudierenden: Eine Analyse der fachlichen Elaboration und Sicherheit beim Sprechen über Diagramme
No full textDiagramme prägen die Vermittlung von Wissen und Daten in Alltag und Wissenschaft und nehmen in der ökonomischen Bildung eine wichtige Stellung ein (Davies & Mangan, 2009; Franke, 2024). Allerdings beziehen sich nur wenige Studien speziell auf Diagramme und die Diagrammkompetenz der Lernenden im ökonomischen Bereich. Die vorliegende Masterarbeit untersucht das Diagrammverständnis von Wirtschaftsstudierenden anhand qualitativer Interviews, die auf der Methode des lauten Denkens basieren. Im Mittelpunkt stehen zwei Fragen: (1) Wie elaboriert und sicher sprechen die Studierenden über die Diagramme? (2) Welche Fehler und Unsicherheiten treten dabei auf? Die Ergebnisse zeigen, dass viele Studierende wenig elaboriert über die Diagramme sprechen und Schwierigkeiten beim Interpretieren sowie Verknüpfen von Inhalten aufweisen. Fehler und Fehlkonzepte lassen sich auf inhaltlicher, prozessbezogener und mathematischer Ebene identifizieren. Trotz dieser Defizite zeigen sich die Studierenden beim Sprechen über die Diagramme insgesamt eher sicher. Allgemein bestätigen die Ergebnisse bisherige Befunde, die auf ein geringes Diagrammverständnis von Wirtschaftsstudierenden sowie Schwierigkeiten im Umgang mit (ökonomischen) Diagrammen hinweisen (u.a. Cohn et al., 2001, 2004; Franke, 2024; Klein et al., 2019; Marire, 2018; Ring & Oberrauch, 2024)
Platelet stiffness correlates inversely with platelet volume during expansion and compression
No full textMusikbox Spezial: Ten Years After live at the Fillmore East, 1970
No full textUnser Format Musikbox Spezial ist immer einem besonderen Konzert aus der Musikgeschichte gewidmet. Heute bringen wir die britische Bluesrockband Ten Years After mit ihrer denkwürdigen Show im New Yorker Fillmore East. Dort gastierten sie am 27. und 28. Februar 1970. Sie gaben vier Konzerte vor jeweils 3.534 Zuschauer*innen. Ten Years After, das waren Chuck Churchill an den Keybords, Rick Lee am Schlagzeug, Leo Lyons am Bass und Alvin Lee Gesang und an der Gitarre. Als Ten Years After spielten die Vier seit 1967 und im August 1969 hatten sie einen legendären Auftritt in Woodstock mit dem zehnminütigen Stück "I'm Going Home,", das fortan meist am Ende ihrer Shows zu hören war.
Das Fillmore East an der Second Avenue in Manhattan existierte seit 1968 lediglich drei Jahre lang und gehörte dem US-amerikanischen Konzertveranstalter Bill Graham, der auch das Fillmore West in San Francisco betrieb. Die Umstände für die Aufzeichnung waren schwierig. Den Auftrag hatte Tonmeister Eddie Kramer, der mit seinem Achtspurbandgerät ein halbes Jahr zuvor auch das Woodstock-Festival aufgezeichnet hatte
Linking Neural and Behavioral Data: Discriminative and Generative Models
No full textA central goal in neuroscience is to understand how neural activity gives rise to behavior. Encoding and decoding models formalize this relationship by estimating conditional distributions of neural activity given behavior or other covariates, and vice versa. Advances in neural and behavioral recording techniques now enable recordings of thousands of neurons during continuous behaviors beyond rigid trial structures. These developments pose a challenge to classical encoding and decoding methods, which often struggle with high dimensionality, variability, and non-linear relationships. Thus, the overarching aim of this thesis was to develop statistical approaches linking high-dimensional neural activity with continuous behavior.
The first contribution addresses whether the midbrain, in particular superior colliculus (SC), differentiates visual motion resulting from self-movement or object-movement. We tested this in virtual reality, where mice experienced identical visual loom for these two contexts. Because neural activity drives, but is also shaped by behavior, we had to account for behavioral differences when decoding context from neural activity. Using a multivariate discriminative decoding framework, we found that SC activity, particularly in intermediate layers, differs between contexts, even after controlling for behavior.
The second contribution directly accounts for this neural-behavioral bidirectionality, hidden in distinct encoding and decoding models. We developed a probabilistic latent variable model based on masked variational autoencoders (VAEs) to jointly model conditional distributions of neural activity and behavior. This framework allowed us to model and sample from the distribution over continuous behaviors given neural activity and to generate neural activity conditioned on unseen behavior.
Masked VAEs provided calibrated uncertainty estimates, indicating higher uncertainty when predictions were likely wrong—an advance increasingly important for highly variable data and only achievable with probabilistic approaches.
In the third contribution, we extended these approaches to diffusion-based probabilistic models that enhance sampling fidelity and conditioning flexibility. To preserve low-dimensional neural representations and account for the discrete nature of neural spikes, we introduced Latent Diffusion for Neural Spiking Data (LDNS). LDNS enabled the extraction of behaviorally meaningful neural latents, and training a diffusion model directly on these latents generated realistic spiking data for various tasks. Flexible conditioning on scalars or entire time-series renders LDNS a highly powerful encoding model enabling scalable hypothesis generation.
Collectively, these works develop complementary statistical approaches for continuous neural-behavioral datasets. By integrating classical encoding and decoding with probabilistic deep generative models, this work scales classical analyses to large-scale datasets and highlights the importance of modeling variability and uncertainty
Enganging with differences, challenging perspectives : Kontroversität im Englischunterricht
No full text'Pass the Mountains, follow the Megaliths and there you are!' Landscape, Resources and Society in Western Toledo (Spain) in the Context of Iberian Late Prehistory
No full textOur knowledge of the Central Iberian prehistoric communities is still worse than that of those in Andalucía or southern Portugal. Recent work on the southern Meseta uncovered site types previously unknown in the central area: On the territory of the modern
village of Azután, a large double-ditched enclosure (El Prado) and several other ditched systems were detected on the terraces of the Tagus River. El Prado became the key topic of this study, as it is the only large enclosure of the Mesetas with a chronology
contradicting the idea of a diffusion of such sites from the coastal areas.
New investigations revealed outstanding results about the Chalcolithic communities of the southern Meseta. The territorial study of the whole province, including the mapping of the known late prehistoric sites, painted a new picture of the late Neolithic/
Chalcolithic of the Tagus middle basin. On-site research at El Prado – taking into account scientific analyses as well – provided new data on chronology and late prehistoric archaeological structures comparable on the southern Meseta only to the
results of Primitiva Bueno and her team in Huecas.
The unexpected results from El Prado surpassed the initial expectations by far, proving that the sedentary communities of the interior were in close contact and constant exchange with the coastal areas of the peninsula
Characterization of stiffness as a biomarker in osteoarthritis and in bone- and soft tissue cancers
No full textMechanobiology encompasses all of the mechanical processes that influence the cell and
its environment. Thereby, the major actors are the cell´s structural components (i.e. cell
membrane, cytoskeleton), the components of the extracellular matrix (ECM), and the
interface between both (i.e. mechanosensors, adhesion molecules). It is well established
that biomechanics plays a critical role in the homeostasis of living tissue, but the concrete
mechanisms by which mechanical cues are assimilated and regulate the balance between
health and disease are not yet fully understood. In recent decades, atomic force
microscopy (AFM) has emerged as the gold standard to measure the foremost mechanical
marker of a tissue/cell, namely the stiffness (measured as the Young´s modulus). This
parameter provides important information about the state of a living entity, and can be
used to investigate a diseased/degenerative condition. In the present work, AFM was
employed to characterize the biomechanical alterations occurring during two major
disorders from the orthopedic field: osteoarthritis (OA) and bone/soft-tissue cancers.
OA is defined as a degenerative joint condition, characterized by articular cartilage
destruction. With progression of the disease the chondrocyte organization pattern is
evolving, from single string (SS) in healthy cartilage to double string (DS), small cluster
(SC), big cluster (BC) and finally diffuse pattern in the most advanced arthritic state.
Cartilage discs of 1mm thickness and 4mm diameter were generated from condyles of
patient undergoing total knee arthroplasty and were sorted according to the
aforementioned cellular patterns. Subsequently the stiffness of the discs was measured
with AFM. A stepwise stiffness decrease was observed with an increasing degeneration
level, except for the group of the diffuse pattern, which displayed a large data spread with
a relatively high stiffness on average. The discs used for those experiments have the
advantage of being of appreciable thickness, they exhibit the native cartilage surface at
their upmost side and the cells within the tissue are kept alive. They hence represent a
suitable model to further investigate articular cartilage degenerative processes in situ.
Bone and soft-tissue sarcomas represent a group of diverse cancer entities from
mesenchymal origin, with overall poor prognosis. The stiffness of malignant cells of five
different bone and soft-tissue sarcoma cell lines were investigated using AFM and were
compared to their respective healthy control cells. Osteosarcoma, Ewing sarcoma,
fibrosarcoma and rhabdomyosarcoma cells were significantly less stiff than osteoblasts,
66
Mesenchymal stem cells (MSCs), fibroblasts and skeletal muscle cells (SKMCs)
respectively, reflecting a typical feature observed in most of the neoplastic
transformation. However, chondrosarcoma cells appeared to be stiffer than chondrocytes
and the other malignant cell lines. Further cytoskeleton examination using fluorescence
microscopy, ELISA and qPCR revealed that if most of the malignant cells displayed a
disorganized internal scaffold and a decreased F-actin content, chondrosarcoma
maintained a cytoskeletal structure relatively close to the one of the healthy cells, with
the microtubule being the most affected components (increase in the β-tubulin gene
expression and protein content). These results define the biomechanical fingerprints of
the investigated cancer cell lines and expose the related molecular mechanisms.
Overall, the work presented in this thesis contributes to the effort of mechanical
characterization of the biological systems respectively addressed in the disclosed studies.
Understanding the processes underlying the biomechanical properties of cells and tissue
is of upmost importance for the development of novel diagnostic and therapy approaches