Efforts to induce customers to buy groceries through the Internet have existed for around twenty years. Early on, the market structures of the digital grocery trade were still strongly fragmented and poorly coordinated. Due to the technological advancement in the past decade, the digital purchase of groceries has become more attractive. The adoption rate of these services varies greatly between different regions. In Germany in particular, the digital grocery trade is stagnating at a comparatively low level. In this regard, this dissertation analyzes both the retail-side market structures and the expectations and obstacles of German consumers. The year 2020 connotes a turning point for the online grocery trade, as daily routines such as grocery shopping were subject to strict regulations imparted at a governmental level in order to reduce COVID-19 infections. At the same time, despite this opportunity, the digital grocery trade has not yet established itself nationwide in Germany. This can be attributed to a lack of investments, but also to inadequate digitization measures. A stronger synchronization between the digital and stationary offer, better integration of digital food services at a regional level as well as adapted, target group-appropriate digital solutions for the efficient breakdown of usage barriers will benefit service usage. The importance of stable food chains and distribution channels was illustrated by the COVID-19 pandemic. Further research should help to develop the digital food trade into a stable and sustainable supplementation of the stationary store.

Today’s agriculture heavily relies on pesticides to manage diverse pests and maximise crop yields. Despite elaborate regulation of pesticide use based on a complex environmental risk assessment (ERA) scheme, the widespread use of these biologically active compounds has been shown to be a threat to the environment. For surface waters, pesticide exposure has been observed to exceed safe concentration levels and negatively impact stream ecology leading to the question whether current ERA schemes ensure a sustainable use of pesticides. To answer this, the large-scale “Kleingewässer-Monitoring” (KgM) assessed the occurrence of pesticides and related effects in 124 streams throughout Germany, Central Europe, in 2018 and 2019. Based on five scientific publications originating from the KgM, this thesis evaluated pesticide exposure in streams, ecological effects and the regulatory implications. More than 1,000 water samples were analysed for over 100 pesticide analytes to characterise occurrence patterns (publication 1). Measured concentrations and effects were used to validate the exposure and effect concentrations predicted in the ERA (publication 2). By jointly analysing real-world pesticide application data and measured pesticide mixtures in streams, the disregard of environmental pesticide mixtures in the ERA was evaluated (publication 3). The toxic potential of mixtures in stream water was additionally investigated using suspect screening for 395 chemicals and a battery of in-vitro bioassays (publication 4). Finally, the results from the KgM stream monitoring were used to assess the capability to identify pesticide risks in governmental monitoring programmes (publication 5). The results of this thesis reveal the widespread occurrence of pesticides in non-target stream ecosystems. The water samples contained a variety of pesticides occurring in complex mixtures predominantly in short-term peaks after rainfall events (publications 1 & 4). Respective pesticide concentration maxima were linked to declines in vulnerable invertebrate species and exceeded regulatory acceptable concentrations in about 80% of agricultural streams, while these thresholds were still estimated partly insufficient to protect the invertebrate community (publication 2). The co-occurrence of pesticides in streams led to a risk underestimated in the single substance-oriented ERA by a factor of about 3.2 in realistic worst-case scenarios, which is further exacerbated by a high frequency at which non-target organism are exposed to pesticides (publication 3). Stream water samples taken after rainfall caused distinct effects in bioassays which were only explainable to a minor extent by the many analytes, indicating the relevance of unknown chemical or biological mixture components (publication 4). Finally, the regulatory monitoring of surface waters under the Water Framework Directive (WFD) was found to significantly underestimate pesticide risks, as about three quarters of critical pesticides and more than half of streams at risk were overlooked (publication 5). Essentially, this thesis involves a new level of validation of the ERA of pesticides in aquatic ecosystems by assessing pesticide occurrence and environmental impacts at a scale so far unique. The overall results demonstrate that the current agricultural use of pesticides leads to significant impacts on stream ecology that go beyond the level tolerated under the ERA. This thesis identified the underestimation of pesticide exposure, the potential insufficiency of regulatory thresholds and the general inertia of the authorisation process as the main causes why the ERA fails to meet its objectives. To achieve a sustainable use of pesticides, the thesis proposes substantial refinements of the ERA. Adequate monitoring programmes such as the KgM, which go beyond current government monitoring efforts, will continue to be needed to keep pesticide regulators constantly informed of the validity of their prospective ERA, which will always be subject to uncertainty.

Social media provides a powerful way for people to share opinions and sentiments about a specific topic, allowing others to benefit from these thoughts and feelings. This procedure generates a huge amount of unstructured data, such as texts, images, and references that are constantly increasing through daily comments to related discussions. However, the vast amount of unstructured data presents risks to the information-extraction process, and so decision making becomes highly challenging. This is because data overload may cause the loss of useful data due to its inappropriate presentation and its accumulation. To this extent, this thesis contributed to the field of analyzing and detecting feelings in images and texts. And that by extracting the feelings and opinions hidden in a huge collection of image data and texts on social networks After that, these feelings are classified into positive, negative, or neutral, according to the features of the classified data. The process of extracting these feelings greatly helps in decision-making processes on various topics as will be explained in the first chapter of the thesis. A system has been built that can classify the feelings inherent in the images and texts on social media sites, such as people’s opinions about products and companies, personal posts, and general messages. This thesis begins by introducing a new method of reducing the dimension of text data based on data-mining approaches and then examines the sentiment based on neural and deep neural network classification algorithms. Subsequently, in contrast to sentiment analysis research in text datasets, we examine sentiment expression and polarity classification within and across image datasets by building deep neural networks based on the attention mechanism.

Augmented reality (AR) applications typically extend the user's view of the real world with virtual objects. In recent years, AR has gained increasing popularity and attention, which has led to improvements in the required technologies. AR has become available to almost everyone. Researchers have made great progress towards the goal of believable AR, in which the real and virtual worlds are combined seamlessly. They mainly focus on issues like tracking, display technologies and user interaction, and give little attention to visual and physical coherence when real and virtual objects are combined. For example, virtual objects should not only respond to the user's input; they should also interact with real objects. Generally, AR becomes more believable and realistic if virtual objects appear fixed or anchored in the real scene, appear indistinguishable from the real scene, and response to any changes within it. This thesis examines on three challenges in the field of computer vision to meet the goal of a believable combined world in which virtual objects appear and behave like real objects. Firstly, the thesis concentrates on the well-known tracking and registration problem. The tracking and registration challenge is discussed and an approach is presented to estimate the position and viewpoint of the user so that virtual objects appear fixed in the real world. Appearance-based line models, which keep only relevant edges for tracking purposes, enable absolute registration in the real world and provide robust tracking. On the one hand, there is no need to spend much time creating suitable models manually. On the other hand, the tracking can deal with changes within the object or the scene to be tracked. Experiments have shown that the use of appearance-based line models improves the robustness, accuracy and re-initialization speed of the tracking process. Secondly, the thesis deals with the subject of reconstructing the surface of a real environment and presents an algorithm to optimize an ongoing surface reconstruction. A complete 3D surface reconstruction of the target scene offers new possibilities for creating more realistic AR applications. Several interactions between real and virtual objects, such as collision and occlusions, can be handled with physical correctness. Whereas previous methods focused on improving surface reconstructions offline after a capturing step, the presented method de-noises, extends and fills holes during the capturing process. Thus, users can explore an unknown environment without any preparation tasks such as moving around and scanning the scene, and without having to deal with the underlying technology in advance. In experiments, the approach provided realistic results where known surfaces were extended and filled in plausibly for different surface types. Finally, the thesis focuses on handling occlusions between the real and virtual worlds more realistically, by re-interpreting the occlusion challenge as an alpha matting problem. The presented method overcomes limitations in state-of-the-art methods by estimating a blending coefficient per pixel of the rendered virtual scene, instead of calculating only their visibility. In several experiments and comparisons with other methods, occlusion handling through alpha matting worked robustly and overcame limitations of low-cost sensor data; it also outperformed previous work in terms of quality, realism and practical applicability. The method can deal with noisy depth data and yields realistic results in regions where foreground and background are not strictly separable (e.g. caused by fuzzy objects or motion blur).

Wild bees are essential for the pollination of wild and cultivated plants. However, within the last decades, the increasing intensification of modern agriculture has led to both a reduction and fragmentation as well as a degradation of the habitats wild bees need. The resulting loss of pollinators and their pollination poses an immense challenge to global food production. To support wild bees, the availability of flowering resources is essential. However, the flowering period of each resource is temporally limited and has different effects on pollinators and their pollination, depending on the time of their flowering. Therefore, to efficiently promote and manage wild bee pollinators in agricultural landscapes, we identified species-specific key floral resources of three selected wild bee species and their spatial and temporal availability (CHAPTERS 2, 3 & 4). We examined, which habitat types predominantly provide these resources (CHAPTERS 3 & 4). We also investigated whether floral resource maps based on the use of these key resources and their spatial and temporal availability explain the abundance and development of the selected wild bees (CHAPTERS 3 & 4) and pollination (CHAPTER 5) better than habitat maps, that only indirectly account for the availability of floral resources. For each of the species studied, we were able to identify different key pollen sources, predominantly woody plants in the early season (April/May) and increasingly herbaceous plants in the later season (June/July; CHAPTERS 2, 3 & 4). The open woody semi-natural habitats of our agricultural landscapes provided about 75% of the floral resources for the buff-tailed bumblebees, 60% for the red mason bees, and 55% for the horned mason bees studied, although they accounted for only 3% of the area (CHAPTERS 3 & 4). In addition, fruit orchards provided about 35% of the floral resources for the horned mason bees on 4% of the landscape area (CHAPTER 3). We showed that both mason bee species benefited from the resource availability in the surrounding landscapes (CHAPTER 3). Yet this was not the case for the bumblebees (CHAPTER 4). Instead, the weight gain of their colonies, the number of developed queen cells and their colony survival were higher with increasing proximity to forests. The proximity to forests also had a positive effect on the mason bees studied (CHAPTER 3). In addition, the red mason bees benefited from herbaceous semi-natural habitats. The proportion of built-up areas had a negative effect on the horned mason bees, and the proportion of arable land on the red mason bees. The habitat maps explained horned mason bee abundances equally well as the floral resource maps, but red mason bee abundances were distinctly better explained by key floral resources. The pollination of field bean increased with higher proportions of early floral resources, whereas synchronous floral resources showed no measurable reduction in their pollination (CHAPTER 5). Habitat maps also explained field bean pollination better than floral resource maps. Here, pollination increased with increasing proportions of built-up areas in the landscapes and decreased with increasing proportions of arable land. Our results highlight the importance of the spatio-temporal availability of certain key species as resource plants of wild bees in agricultural landscapes. They show that habitat maps are ahead of, or at least equal to, spatio-temporally resolved floral resource maps in predicting wild bee development and pollination. Nevertheless, floral resource maps allow us to draw more accurate conclusions between key floral resources and the organisms studied. The proximity to forest edges had a positive effect on each of the three wild bee species studied. However, besides pure food availability, other factors seem to co-determine the occurrence of wild bees in agricultural landscapes.

Efficient Cochlear Implant (CI) surgery requires prior knowledge of the cochlea’s size and its characteristics. This information helps to select suitable implants for different patients. Registered and fused images helps doctors by providing more informative image that takes advantages of different modalities. The cochlea’s small size and complex structure, in addition to the different resolutions and head positions during imaging, reveals a big challenge for the automated registration of the different image modalities. To obtain an automatic measurement of the cochlea length and the volume size, a segmentation method of cochlea medical images is needed. The goal of this dissertation is to introduce new practical and automatic algorithms for the human cochlea multi-modal 3D image registration, fusion, segmentation and analysis. Two novel methods for automatic cochlea image registration (ACIR) and automatic cochlea analysis (ACA) are introduced. The proposed methods crop the input images to the cochlea part and then align the cropped images to obtain the optimal transformation. After that, this transformation is used to align the original images. ACIR and ACA use Mattes mutual information as similarity metric, the adaptive stochastic gradient descent (ASGD) or the stochastic limited memory Broyden–Fletcher–Goldfarb–Shanno (s-LBFGS) optimizer to estimate the parameters of 3D rigid transform. The second stage of nonrigid registration estimates B-spline coefficients that are used in an atlas-model-based segmentation to extract cochlea scalae and the relative measurements of the input image. The image which has segmentation is aligned to the input image to obtain the non-rigid transformation. After that the segmentation of the first image, in addition to point-models are transformed to the input image. The detailed transformed segmentation provides the scala volume size. Using the transformed point-models, the A-value, the central scala lengths, the lateral and the organ of corti scala tympani lengths are computed. The methods have been tested using clinical 3D images of total 67 patients: from Germany (41 patients) and Egypt (26 patients). The atients are of different ages and gender. The number of images used in the experiments is 217, which are multi-modal 3D clinical images from CT, CBCT, and MRI scanners. The proposed methods are compared to the state of the arts ptimizers related medical image registration methods e.g. fast adaptive stochastic gradient descent (FASGD) and efficient preconditioned tochastic gradient descent (EPSGD). The comparison used the root mean squared distance (RMSE) between the ground truth landmarks and the resulted landmarks. The landmarks are located manually by two experts to represent the round window and the top of the cochlea. After obtaining the transformation using ACIR, the landmarks of the moving image are transformed using the resulted transformation and RMSE of the transformed landmarks, and at the same time the fixed image landmarks are computed. I also used the active length of the cochlea implant electrodes to compute the error aroused by the image artifact, and I found out an error ranged from 0.5 mm to 1.12 mm. ACIR method’s RMSE average was 0.36 mm with a standard deviation (SD) of 0.17 mm. The total time average required for registration of an image pair using ACIR was 4.62 seconds with SD of 1.19 seconds. All experiments are repeated 3 times for justifications. Comparing the RMSE of ACIR2017 and ACIR2020 using paired T-test shows no significant difference (p-value = 0.17). The total RMSE average of ACA method was 0.61 mm with a SD of 0.22 mm. The total time average required for analysing an image was 5.21 seconds with SD of 0.93 seconds. The statistical tests show that there is no difference between the results from automatic A-value method and the manual A-value method (p-value = 0.42). There is no difference also between length’s measurements of the left and the right ear sides (p-value > 0.16). Comparing the results from German and Egypt dataset shows there is no difference when using manual or automatic A-value methods (p-value > 0.20). However, there is a significant difference when using ACA2000 method between the German and the Egyptian results (p-value < 0.001). The average time to obtain the segmentation and all measurements was 5.21 second per image. The cochlea scala tympani volume size ranged from 38.98 mm3 to 57.67 mm3 . The combined scala media and scala vestibuli volume size ranged from 34.98 mm 3 to 49.3 mm 3 . The overall volume size of the cochlea should range from 73.96 mm 3 to 106.97 mm 3 . The lateral wall length of scala tympani ranged from 42.93 mm to 47.19 mm. The organ-of-Corti length of scala tympani ranged from 31.11 mm to 34.08 mm. Using the A-value method, the lateral length of scala tympani ranged from 36.69 mm to 45.91 mm. The organ-of-Corti length of scala tympani ranged from 29.12 mm to 39.05 mm. The length from ACA2020 method can be visualised and has a well-defined endpoints. The ACA2020 method works on different modalities and different images despite the noise level or the resolution. In the other hand, the A-value method works neither on MRI nor noisy images. Hence, ACA2020 method may provide more reliable and accurate measurement than the A-value method. The source-code and the datasets are made publicly available to help reproduction and validation of my result.

Over the past few decades, Single-Particle Analysis (SPA), in combination with cryo-transmission electron microscopy, has evolved into one of the leading technologies for structural analysis of biological macromolecules. It allows the investigation of biological structures in a close to native state at the molecular level. Within the last five years the achievable resolution of SPA surpassed 2°A and is now approaching atomic resolution, which so far has only been possible with Xray crystallography in a far from native environment. One remaining problem of Cryo-Electron Microscopy (cryo-EM) is the weak image contrast. Since the introduction of cryo-EM in the 1980s phase plates have been investigated as a potential tool to overcome these contrast limitations. Until now, technical problems and instrumental deficiencies have made the use of phase plates difficult; an automated workflow, crucial for the acquisition of 1000s of micrographs needed for SPA, was not possible. In this thesis, a new Zernike-type Phase Plate (PP) was developed and investigated. Freestanding metal films were used as a PP material to overcome the ageing and contamination problems of standard carbon-based PPs. Several experiments, evaluating and testing various metals, ended with iridium as the best-suited material. A thorough investigation of the properties of iridium PP followed in the second part of this thesis. One key outcome is a new operation mode, the rocking PP. By using this rocking-mode, fringing artifacts, another obstacle of Zernike PPs, could be solved. In the last part of this work, acquisition and reconstruction of SPA data of apoferritin was performed using the iridium PP in rocking-mode. A special semi-automated workflow for the acquisition of PP data was developed and tested. The recorded PP data was compared to an additional reference dataset without a PP, acquired following a conventional workflow.

The protected areas of Rwanda are facing various challenges resulting from the anthropogenic activities of the surrounding communities especially in the adjacent area to Cyamudongo isolated rain forest, which results in climate change, soil degradation, and loss of biodiversity. Therefore, this study aims to broaden current knowledge on the impact of sustainable Agroforestry (AF) on the Carbon (C) stock and Biodiversity conservation on the surroundings of Cyamudongo isolated rain forest and Ruhande Arboretum. To understand this, the permanent sample plots (PSPs) were established mainly in the designed four transects of four km long originating on the boundary of the Cyamudongo isolated rain forest following the slope gradient ranging from 1286 to 2015 m asl. A total number of 73 PSPs were established in the Cyamudongo study area while 3 PSPs were established in the Ruhande AF plot. The Arc Map GIS 10.4 was used to design and map the sampling areas while GPS was used for localization of collected items. Statistical significance was analyzed through the R-software especially for wood and soil variables while for biodiversity indicator species, MVSP Software 3.0 was used to determine the Shannon Diversity indices and similarities among species. In this study, I have obtained comprehensive results demonstrating that in all study areas, the various AF tree species contribute differently to C stock and C sequestration and the amount of C stored and removed from the atmosphere depends on different factors such as tree species, plantation density, growth stage, or the age of establishment, applied management practices, wood specific density (WSD), wood C concentration, and climatic conditions. The estimated quantity of sequestrated C for 2 years and 34 years AF species were 13.11 t C ha -1 yr-1 (equivalent to 48 t CO2 ha -1 yr-1) and 6.85 t ha-1 yr-1 (equivalent to 25.1 t CO2 ha -1 yr-1) in Cyamudongo and Ruhande respectively. The estimated quantity of C stored by the Ruhande AF plot is 232.94 t ha-1. In Cyamudongo, the overall C stored by the AF systems was 823 t ha-1 by both young tree species established by the Cyamudongo Project (35.84 t ha-1) and C stored by existed AF species before the existence of the Project (787.12 t ha-1). In all study areas, the Grevillea robusta was found to contribute more to overall stored C compared to other species under this study. The tests revealed differences in terms of nutrient contents (C, N, C: N ratio, K, Na, Ca, and Mg) for various AF tree species of Cyamudongo and Ruhande study areas. The differences in terms of correlation for various variables of AF tree species in different study areas varied with tree species, age, stage of growth, and tree shape. By comparing the correlation coefficients for various tree variables for young and mature AF tree species, the results showed a high correlation variability for young species than mature or old species recorded in different environmental conditions of Cyamudongo and Ruhande study areas. The recorded soil pH mean value across in Cyamudongo study area is 4.2, which is very strongly acidic. The tests revealed that the soil pH, C, C: N ratio, OM, NH4+, NO3-+NO2-, PO43-, and CEC were significantly (P < 0.05) different in various soil depths whereas the N was not statistically significant. The pH, N, C: N ratio, CEC, NH4+, PO43-, and Al3+ showed a significant difference across land uses whereas the C and NO3-+NO2- did not show any statistical difference. All tested chemical elements showed a statistical difference as far as altitude ranges are concerned. The only NH4+, PO43-, and CEC showed significant differences with time whereas all other remaining chemical elements did not show any statistical significance. The bulk density of soil was statistically different across land uses and altitude ranges. The soil pH was very strongly correlated with CEC, Mg, and Ca in cropland (CL) whereas it was strongly correlated in both AF and natural forest (NF) except for Mg, which was moderately correlated in AF. Furthermore, its correlation with K was strong in CL, moderate in AF while it was weak in NF. Finally, the pH correlation with Na was weak in both AF and CL whereas it was negligible in NF. The overall estimated soil C stock of the study area was 16848 t ha -1. The sustainable AF practices changed significantly the frequency of reptiles, amphibians, and flowering plants while there was no statistical change observed on ferns with time. In terms of species richness, 16 flowering plants, 14 ferns, 5 amphibians, and 3 reptiles were recorded and monitored. These findings add to a growing body of literature on the impact of AF on the C stock, soil improvement, and Biodiversity. It is recommended that further researches should be undertaken for the contribution of other AF tree species to the C stock found in the agricultural landscape around all protected areas of Rwanda and the impact on them on the soil and biodiversity.

Die bisherige Forschung zur Diagnosekompetenz hat trotz langjähriger Forschungstraditionen für die Praxis kaum relevante Befunde erbracht. Dies wurde in der vorliegenden Arbeit als Anlass für den Vorschlag genommen, diagnostische Kompetenz als reine Urteilsgenauigkeit, gemäß der Komponentenanalyse nach Cronbach (1955), zu erweitern. Mit der Einführung der Beobachtungskompetenz als einer weiteren Facette diagnostischer Kompetenz knüpft die vorliegende Studie zudem an der Forschung zu formativem Assessment an. Die zentralen Fragen der Arbeit beziehen sich darauf, wie die beiden Facetten der diagnostischen Kompetenz Urteilsgenauigkeit und Beobachtungskompetenz von pädagogischen Fachkräften im Elementarbereich ausgeprägt sind, ob diese Facetten zusammenhängen und wie sie durch bereichsspezifische Fortbildungen und explizit durch verschiedene Fortbildungsansätze (Lehrgangs- vs. Lernwegsorientierung) verändert werden können. Untersucht wurden insgesamt 103 Erzieherinnen aus 21 Einrichtungen mit eigens für die Arbeit entwickelten Instrumenten (Fragebögen, Filmvignetten). Zur Beantwortung der ersten Forschungsfrage, wie die bereichsspezifische diagnostische Kompetenz von Erzieherinnen hinsichtlich der Facetten Urteilsgenauigkeit und Beobachtungskompetenz ausgeprägt ist, zeigt sich, dass Erzieherinnen das Niveau ihrer Gruppe im Vergleich der Urteilsgenauigkeits-Komponenten am besten einschätzen können (Niveaukomponente), mit Tendenz zur Überschätzung. Auch die Wahrnehmung von Unterschieden zwischen den Kindern ihrer Gruppe gelingt ihnen gut (Personenkomponente), dabei neigen sie eher zu einer Überschätzung der Heterogenität. Wesentlich schwerer fällt es ihnen, spezifische Aufgabenschwierigkeiten im mathematischen Bereich korrekt einzuschätzen (Merkmalskomponente) und Unterschiede zwischen den Kindern im Hinblick auf einzelne Aufgaben (Wechselwirkungskomponente) zu beurteilen. Die Erzieherinnen sind jedoch sehr gut in der Lage, die Kinder entsprechend ihrer Leistungen in eine Rangreihenfolge zu bringen, wie es die Korrelationskomponente der Personenkomponente zeigt. Die Beobachtungskompetenz von Erzieherinnen zeigt sich in der Anzahl erkannter Fähigkeiten von Kindern während verschiedener Tätigkeiten mit mathematischem Bezug. Dabei wurde der Schwierigkeitsgrad der Beobachtung mittels Expertinneneinschätzung berücksichtigt. Die Erzieherinnen erkennen durchschnittlich die Hälfte der mathematischen Fähigkeiten von Kindern in verschiedenen komplexen Alltagssituationen bei mathematischen Tätigkeiten. Die zweite Forschungsfrage, ob es einen Zusammenhang zwischen der Urteilsgenauigkeit und der Beobachtungskompetenz gibt, kann klar beantwortet werden. Erwartungswidrig zeigt die vorliegende Untersuchung, dass die Facetten Urteilsgenauigkeit und Beobachtungskompetenz voneinander unabhängig sind. Damit konnte die Annahme, dass auf der einen Seite eine genaue summative Beurteilung auf einer Vielzahl kompetenter Einzelbeobachtungen beruht, auf der anderen Seite aber häufige gezielte Einzelbeobachtungen in ihrer Summe zu einem genaueren Gesamturteil führen, nicht bestätigt werden. Die dritte Forschungsfrage, ob eine bereichsspezifische Fortbildung von Erzieherinnen zur Verbesserung ihrer Urteilsgenauigkeit führt, muss verneint werden. Dagegen beeinflusst der Faktor Zeit (neun Monate zwischen Prä- und Posttest) die Fähigkeit der Erzieherinnen, Unterschiede zwischen den Kindern signifikant besser wahrzunehmen (Personenkomponente) und spezifische Aufgabenschwierigkeiten signifikant besser einzuschätzen (Merkmalskomponente). Damit verbessert sich auch die globale Urteilsgenauigkeit über die Zeit signifikant. Auf die Einschätzung des Leistungsniveaus der Gruppe (Niveaukomponente) sowie von Unterschieden zwischen den Kindern für einzelne Items (Wechselwirkungskomponente) hat jedoch auch der Faktor Zeit keinen signifikanten Einfluss. Die vierte Forschungsfrage, ob die Unterschiede in der Veränderung der Urteilsgenauigkeit von Erzieherinnen davon abhängig sind, nach welchem Ansatz sie Kinder bereichsspezifisch fördern, lässt sich zusammenfassend verneinen. Die Art und Weise der bereichsspezifischen Förderung spielt keine Rolle für die Variabilität der Urteilsgenauigkeit von Erzieherinnen. Auch die Unterschiedlichkeit in den einzelnen Urteilsgenauigkeitskomponenten lässt sich nicht auf die Art der Fortbildung zurückführen. Die fünfte Forschungsfrage, ob sich die Beobachtungskompetenz von Erzieherinnen durch eine bereichsspezifische Fortbildung beeinflussen lässt, kann aufgrund einer signifikanten Zunahme der Beobachtungskompetenz nach nur zwei Fortbildungstagen bejaht werden. Damit wird die These gestützt, dass eine Fortbildung zur bereichsspezifischen Förderung für eine gezielte und damit fokussiertere Wahrnehmung früher bereichsspezifischer Kompetenzen bedeutsam ist, selbst wenn sie nicht explizit auf diagnostische Kompetenz, sondern allgemein auf Förderung ausgerichtet ist Die sechste Forschungsfrage zielt auf Veränderungen der Beobachtungskompetenz von Erzieherinnen in Abhängigkeit von Fortbildungen in verschiedenen bereichsspezifischen Förderansätzen ab. Für eine Zunahme der Beobachtungskompetenz, so zeigt sich, spielt der bereichsspezifische Ansatz, welcher der Fortbildung zugrunde gelegt ist, keine Rolle. Die vorliegende Studie ist die erste Arbeit, die sowohl Beobachtungskompetenz (Forschung zu formativem Assessment) als auch Urteilsgenauigkeit (Accuracy-Forschung) anhand derselben Stichprobe untersucht. Trotz des im Ergebnis nicht nachgewiesenen Zusammenhangs zwischen Urteilsgenauigkeit und Beobachtungskompetenz stellt die Studie für die zukünftige Forschung bedeutsame Hinweise für die weitere Untersuchung der Facetten diagnostischer Kompetenz bereit. Im Hinblick auf den frühpädagogischen Kontext zeigt die vorliegende Arbeit zum einen, dass ein ökologisch valides Treatment (Fortbildung) in einem Umfang von zwei Tagen implizit zu einer Verbesserung diagnostischer Kompetenzen führen kann. Zum anderen wird hier erstmalig im deutschsprachigen Raum die Messung der Urteilsgenauigkeit unter vollumfänglicher Erfüllung eines auf Hoge und Coladarci (1989) beruhenden Anforderungsprofils durchgeführt. Ein weiteres Alleinstellungsmerkmal dieser Studie ist die erstmals mit einer Mindestanzahl von vier Einschätzungen pro Erzieherin unter Berücksichtigung der Kontaktzeiten zwischen Erzieherinnen und Bezugskindern erfolgte Erhebung der Urteilsgenauigkeit von Erzieherinnen.

Content and Language Integrated Learning (CLIL) has experienced growing importance in the last decades and an increasing number of schools have already implemented CLIL programmes or are planning to do so. Though the potentials of CLIL programmes are widely praised, first research results also raise doubts if CLIL students can live up to these high expectations. Both Fehling (2005) as well as Rumlich (2013; 2016), for example, found that CLIL programmes not inevitably show the expected results but that the CLIL students’ success might also be at least partially explained by other influences, such as the selection process of future CLIL students. Hence, CLIL students apparently fall short of the high expectations that are usually connected to the respective CLIL programmes and as this is mainly based on the unsatisfactory quality of these programmes, Rumlich concludes that “it is now high time to focus on the quality of CLIL provision” (Rumlich 2016: 452). He continues to explain that “the promises of CLIL do not materialise automatically owing to the fact that another language is used for learning in a non-language subject” (Rumlich 2016: 452). It must be assumed that the success of CLIL teaching also highly depends on the quality of the CLIL teachers. In contrast to the continuously growing number of CLIL schools, however, the number of specifically trained CLIL teachers is comparably small. In Germany, CLIL teachers are not (yet) required to attend any special training in order to teach in a CLIL programme. Notwithstanding, is it sufficient for a CLIL teacher only to be trained in the content subject and the foreign language? Or does CLIL teaching require more than the sum of these two components? Do CLIL teachers need additional teaching competences to the ones of a content and a language teacher? In the light of the recent findings of CLIL programmes falling short of the high expectations, the answer to these questions must clearly be “Yes”. Hence, in order to appropriately train (future) CLIL teachers, special training programmes need to be developed which consider the teachers’ individual educational backgrounds, i.e. their qualifications as language and/or as content teachers and build up on these competences through adding the CLIL-specific teaching competences. Therefore, this thesis aims at developing a German Framework for CLIL Teacher Education, which considers both the already published, theory-based standards of CLIL teacher education as well as the practical perspective of experienced CLIL teachers in Germany. This German Framework for CLIL Teacher Education classifies the different teaching competences, which are derived from integrating the theoretical and the practical perspective on CLIL teacher education, with regard to the three different competence areas, i.e. the general teaching competence, the language teaching competence and the subject teaching competence and is hence adaptable to different CLIL settings and educational backgrounds. In addition to developing this German Framework for CLIL Teacher Education, which provides the content of future CLIL teacher education programmes, this thesis discusses different forms of structurally implementing CLIL teacher education programmes in the existing structures of teacher education in Germany. This is achieved through analysing the current state of the art of CLIL teacher education at German universities and systematising the different forms of implementing these training programmes in the prevailing educational structures. Building on these first two steps, in the third and final step, this thesis develops a CLIL teacher education programme at a German university that is based on the results and elements of the German Framework for CLIL Teacher Education as well as the state of the art of CLIL teacher education in Germany. Thus, this thesis is allocated at the intersection between foreign language teaching as well as teacher education and is structured in eleven chapters.