German Data Research Articles

On the Respective Roles of National Libraries, National Archives and Research Data Centers in the Preservation of and Access to Research Data

A meeting was held in Berlin on the 22nd of June, 2010, under the aegis of the UK Data Forum and the German Data Forum (RatSWD). The purpose of this meeting was to consider the role of libraries and archives for the acquisition, management and preservation of research data in the behavioural, social and economic sciences.For the purposes of this document we define research data broadly as “the primary inputs into research, as well as the first-order results of that research.” Our conclusion from the meeting was that it would be useful to formulate a memorandum on the respective roles of national libraries, national archives and research data centres. In our capacities as the chairpersons of the UK and German Data Forums we have written the following short overview and memorandum.

Strategic Obfuscation and Retail Pricing

Consumer-product manufacturers—and retailers that sell their products—often sell slightly differentiated items for reasons other than appealing to heterogeneous tastes—different sizes of a popular brand, or different flavors in a common product line for instance. We argue that this practice is a form of strategic obfuscation, which is intended to make price-comparison more difficult, and thereby raise margins on non-comparable products. We test our hypothesis with the use of examples from consumer-packaged good categories in German and French retail scanner data. We find that—after controlling for other explanations for how margins can vary with package size and type—we cannot rule out strategic obfuscation as a feature of our retail sales data.

Secure Linking of Data from Population-Based Cancer Registries with Healthcare Data to Evaluate Screening Programs.

The evaluation of population-based screening programs, like the German Mammography Screening Program (MSP), requires collection and linking data from population-based cancer registries and other sources of the healthcare system on a case- specific level. To link such sensitive data, we developed a method that is compliant with German data protection regulations and does not require written individual consent. Our method combines a probabilistic record linkage on encrypted identifying data with 'blinded anonymisation'. It ensures that all data either are encrypted or have a defined and measurable degree of anonymity. The data sources use a software to transform plain-text identifying data into a set of irreversibly encrypted person cryptograms, while the evaluation attributes are aggregated in multiple stages and are reversibly encrypted. A pseudonymisation service encrypts the person cryptograms into record assignment numbers and a downstream data-collecting centre uses them to perform the probabilistic record linkage. The blinded anonymisation solves the problem of quasi-identifiers within the evaluation data. It allows selecting a specific set of the encrypted aggregations to produce data export with ensured k-anonymity, without any plain-text information. These data are finally transferred to an evaluation centre where they are decrypted and analysed. Our approach allows creating several such generalisations, with different resulting suppression rates allowing dynamic balance information depth with privacy protection and also highlights how this affects data analysability. German data protection authorities approved our concept for the evaluation of the impact of the German MSP on breast cancer mortality. We implemented a prototype and tested it with 1.5 million simulated records, containing realistically distributed identifying data, calculated different generalisations and the respective suppression rates. Here, we also discuss limitations for large data sets in the cancer registry domain, as well as approaches for further improvements like l-diversity and how to reduce the amount of manual post-processing. Our approach enables secure linking of data from population-based cancer registries and other sources of the healthcare system. Despite some limitations, it enables evaluation of the German MSP program and can be generalised to be applicable to other projects.

Short and Long-Term Effectiveness of Ustekinumab in Patients with Crohn's Disease: Real-World Data from a German IBD Cohort.

Background and Aims: The IL-12/23 inhibitor ustekinumab (UST) opened up new treatment options for patients with Crohn’s disease (CD). Due to the recent approval, real-world German data on long-term efficacy and safety are lacking. This study aimed to assess the clinical course of CD patients under UST therapy and to identify potential predictive markers. Methods: Patients with CD receiving UST treatment in three hospitals and two outpatient centers were included and retrospectively analyzed. Rates for short- and long-term remission and response were analyzed with the help of clinical (Harvey–Bradshaw Index (HBI)) and biochemical (C-reactive protein (CRP), Fecal calprotectin (fCal)) parameters for disease activity. Results: Data from 180 patients were evaluated. One-hundred-and-six patients had a follow-up of at least eight weeks and were included. 96.2% of the patients were pre-exposed to anti- TNFα agents and 34.4% to both anti-TNFα and anti-integrin antibodies. The median follow-up was 49.1 weeks (95% CI 42.03-56.25). At week 8, 51 patients (54.8%) showed response to UST, and 24 (24.7%) were in remission. At week 48, 48 (51.6%) responded to UST, and 25 patients (26.9%) were in remission. Steroid-free response and remission at week eight was achieved by 30.1% and 19.3% of patients, respectively. At week 48, 37.6% showed steroid-free response to UST, and 20.4% of the initial patient population was in steroid-free remission. Conclusion: Our study confirms short- and long-term UST effectiveness and tolerability in a cohort of multi-treatment-exposed patients.

The Productivity Effects of Worker Mobility Between Heterogeneous Firms

Abstract We analyze the effect of worker inflows on establishments’ productivity, using German data. Previous studies for other countries have found positive effects of hiring workers from superior (more productive or higher paying) firms. Ranking establishments by their median wage, we find that inflows from inferior establishments seem to increase hiring establishments’ productivity. Further empirical analyses suggest our findings are due to a positive selection of such inflows from their sending establishments. These workers might have to find a better job match in order to advance their careers, an interpretation supported by the finding that the effect is driven by workers with short tenure at their previous employer. Our findings reflect the increasingly assortative pattern of worker mobility in Germany found in a related strand of literature.

Epstein–Zin Utility, Asset Prices, and the Business Cycle Revisited

Abstract This paper discusses the advantages of Epstein and Zin (1989) (EZ) preferences when building dynamic stochastic general equilibrium (DSGE) models that are consistent with well-known stylized facts of both the business cycle and asset markets. To this end, we combine EZ preferences with several building blocks from the DSGE literature that has tried to solve the equity premium puzzle and to replicate characteristic statistics of the labor market. Our goal is to guide researchers in this area to useful modeling devices and to discuss EZ preferences vis-a-vis the standard time-additive expected utility function. EZ preferences separate the attitude toward risk from the attitude toward intertemporal substitution. We demonstrate that this additional degree of freedom allows us to closely match the empirical facts already in a frictionless production economy with endogenous labor supply. Our study follows Heer and Maußner (2013). We examine models that consider adjustment costs of capital accumulation, consumption habits, and frictions in the allocation of labor. Our empirical targets are estimated from German data.

Ereigniskontrollierte Adjunkte im Deutschen

AbstractThe present paper investigates small clause adjuncts displaying the phenomenon referred to as ‘event control’ in literature. Many languages, including German, employ non-finite clauses (besides finite clauses) as propositional adjuncts, for instance infinitival, participial and small clause adjuncts. The subject of these adjunct clauses is left unexpressed and must generally be interpreted co-referentially with the subject or object of the matrix clause (subject or object control), but the matrix event itself can also be interpreted as the controller. Adjuncts involving event control have, to my knowledge, never been examined jointly or particularly thoroughly. The aim of this paper is therefore to provide insight into German data involving event control in different kinds of non-finite propositional adjunct clauses, by examining common and diverging syntactic and semantic properties. The data comprises nominative DPs (Germ.Satzappositionen), adverbial infinitives headed byum(Engl.in order to), adverbial present and past participle constructions, and adverbial small clauses headed by the particleals. Furthermore, I discuss briefly how these data could be captured theoretically, by analyzing them as adjuncts in different syntactic-semantic domains and as obligatorily controlled (OC) adjuncts according to the OC-properties described by Landau (2013).

Diversity and uniformity of German data protection oversight

Angesichts des gegenwartigen Reformprozesses der Datenschutzpolitik in der EU und Deutschland untersucht dieser Beitrag, in welchem Mase die deutsche Datenschutzaufsichtspraxis von foderaler Vielfalt gepragt ist. Denn wahrend Datenschutzrecht vorrangig auf der EU-Ebene gesetzt wird, findet die Rechtsdurchsetzung weiterhin auf nationaler, oder, wie im Bundesdeutschen Fall, auf der Landesebene statt. Eine qualitative Analyse der Aufsichtspraxis in den 18 deutschen Datenschutzaufsichtsbehorden zeigt, dass diese formal betrachtet uber ahnliche Handlungsspielraume verfugen, doch in ihrer Aufsichtspraxis Unterschiede aufweisen. Vielfalt findet sich mit Blick auf ihre Handlungsorientierung, ihr Handeln, ihr regulierungsrelevantes Wissen und ihre offentliche Kommunikation. Eine wichtige Unterscheidungsdimension ist dabei das Spannungsfeld zwischen Pravention und Sanktion. Daruber hinaus haben die Aufsichtsbehorden unterschiedliche Expertisen und Kommunikationsformen entwickelt. Welche Faktoren das beobachtete Muster der Einheitlichkeit und Vielfalt erklaren konnen, stand nicht im Fokus der Analyse, die nur erste Hypothesen generiert hat. Es finden sich Hinweise darauf, dass soziookonomische Faktoren, die politische Kultur und politische Unternehmer einen Einfluss darauf haben konnten, dass eine Aufsichtsbehorde starker praventiv oder starker sanktionierend agiert. Dafur, dass die deutsche Teilungsgeschichte, regulativer Wettbewerb oder die parteipolitische Farbung der Landesregierung eine Rolle spielen konnten, gab es keine Anzeichen. Der Beitrag zeigt somit, welche Strategien und Masnahmen Datenschutzbehorden in Deutschland zur Erreichung eines hohen Datenschutzniveaus derzeitig einsetzen. Daraus lassen sich Schlusse fur die Gestaltung der Datenschutzaufsicht auf europaischer Ebene ziehen, etwa dazu, wie regulativer Wettbewerb in der europaischen Datenschutzpolitik verhindert werden kann.

Self-employment as a source of income inequality

It is well known that the self-employed are over-represented at the bottom as well as the top of the income distribution. This paper shifts the focus from the income situation of the self-employed to the distributive effects of a change in self-employment rates. With representative German data and unconditional quantile regression analysis we show that an increase in the proportion of self-employed individuals in the labor force increases income polarization by tearing down floors at the bottom and allowing higher income potentials at the very top of the hourly income distribution. Recentered influence function regression of inequality measures corroborate that self-employment is a source of income inequality in the labor market.

The Relationship Between Dictionary Look-up Frequency and Corpus Frequency Revisited: A Log-File Analysis of a Decade of User Interaction with a Swahili-English Dictionary

In an earlier publication it was claimed that there is no useful relationship between Swahili-English dictionary look-up frequencies and the occurrence frequencies for the same wordforms in Swahili-English corpora, at least not beyond the top few thousand wordforms. This result was challenged using data for German by a different team of researchers using an improved methodology. In the present article the original Swahili-English data is revisited, using ten years’ worth of it rather than just two, and using the improved methodology. We conclude that there is indeed a positive relationship. In addition, we show that online dictionary look-up behaviour is remarkably similar across languages, even when, as in our case, one is dealing with languages from very dissimilar language families. Furthermore, online dictionaries turn out to have minimum look-up success rates, below which they simply cannot go. These minima are language-sensitive and vary depending on the regularity of the searched-for entries, but are otherwise constant no matter the size of randomly sampled dictionaries. Corpus-informed sampling always improves on any random method. Lastly, from the point of view of the graphical user interface, we argue that the average user of an online bilingual dictionary is better served with a single search box, rather than separate search boxes for each dictionary side.

Polarity particles revisited

Claus et al. (2017) present novel experimental evidence concerning polarity particle responses in German, and discuss the challenges these findings raise for three approaches to such responses, namely the saliency approach of Krifka (20130, the feature model of Roelofsen and Farkas (2015), and the ellipsis approaches of Kramer and Rawlins (2012), and Holmberg (2016). The authors then sketch a way to account for the data within the feature model, as well as a revised version of the saliency account. The first goal of this paper is to work out in full detail an account of the new German data in the feature model. In the process, we will clarify and better articulate those aspects of the feature model that are responsible for how preference patterns are to be explained. Our second goal is to deepen the comparison between this model and the saliency approach by taking a wider cross-linguistic perspective. EARLY ACCESS

INNV-15. CLINICAL DATA THAT MATTERS: A DISTILLATION OF NEURO-ONCOLOGY CLINICAL TRIAL INCLUSION CRITERIA USING MACHINE LEARNING

Abstract INTRODUCTION Neuro-oncologic conditions have dismal outcomes, ineffective treatments, poor access to clinical trials, and variability in care. Clinical trials do not capture a patient’s complete journey and are restricted to select populations. ‘Real-world-evidence’ (RWE) attempts to inform point of care decisions through routine collection of data with a clinical-trial-like rigor. RWE complements existing knowledge through broad patient participation, collection throughout disease course, and creation of large multidimensional datasets “knowledge network of disease” 1,2. RWE implementation is hindered by unstructured data, uncertainty of relevant features, and semantic heterogeneity. Clinical attributes were selected from trial inclusion criteria and prioritized for structuring in clinic notes for abstraction. METHOD We queried Clinicaltrials.gov from 1/1/2018-12/31/2018, refined to North America, recruiting, interventional, and adult. Meningioma, pituitary, glioblastoma, astrocytoma, oligodendroglioma, and ependymoma were chosen based on incidence3. Lymphoma and nerve sheath tumors were omitted. “Brain tumor” and “glioma” were added. ‘K-nearest-neighbor’ tokenization parsed inclusion criteria4. Document term matrix (n-gram) converted text to vectors5. A generative probabilistic model using ‘Latent Dirichlet Allocation’ plotted words into 10 clusters6. Hierarchal clustering was used to compare histology with terms. RESULTS 401 trials parsed into 3676 statements and 4008 keywords. 10 clusters of terms were similarly distributed amongst histologies, suggesting generalizability across tumor types. Cluster revealed 8 categories: 1) Time: enrollment; 2) Performance status: KPS; 3) Testing: mutations, upper limit of normal, routine hematologic laboratory assays; 4) Imaging: extent of surgery; 5) Pregnancy/childbearing; 6) Tumor grade; 7) Treatment history: recurrence, chemotherapy, radiation, time; 8) Informed consent CONCLUSIONS Dissecting the compendium of clinical trials using machine learning can identify general parameters for trial enrollment to guide RWE clinical collection. Using practical definitions of the most germane trial data, specific information can be sought after and defined to improve research quality, maximize research yields and improve patient care whilst minimizing wasted research and clinical endeavors.

One Year of Newborn Screening for SMA - Results of a German Pilot Project.

Objective:Spinal muscular atrophy (SMA) is the most common neurodegenerative disease in childhood. The study was conducted to assess the impact of early detection of SMA by newborn screening (NBS) on the clinical course of the disease.Methods:Screening was performed in two federal states of Germany, Bavaria and North Rhine Westphalia, between January 2018 and February 2019. The incidence in the screening population was calculated as number of detected patients with a homozygous deletion in the SMN1-gene per number of screened patients. To get an idea about the incidence of newly diagnosed SMA in the year prior to screening a survey covering all neuropediatric centers in the state of Bavaria was conducted, identifying all SMA-cases in 2017 and 2018. Following positive NBS and confirmatory diagnostic test, treatment was advised according to the recommendations of the “American SMA NBS Multidisciplinary Working Group”. Immediate treatment with Nusinersen was recommended in children with 2 and 3 SMN2 copies and a conservative strict follow-up strategy in children with ≥4 copies. All children underwent regular standardized neuropediatric examination, CHOP INTEND and HINE-2 testing as well as electrophysiological exams every 2-3 months.Results:165,525 children were screened. 22 cases of SMA were identified, meaning an incidence rate of 1:7524. SMN2 copy number analysis showed 2 SMN2 copies in 45% of patients, 3 SMN2 copies in 19 % and 4 SMN2 copies in 36%. These findings are confirmed in the most recent statistical data-cut from 31st August 2019 (incidence 1:7089, 2 SMN2 copies in 44%, 3 in 15% and 4 in 38%). Comparison with up-to-date German data on SMA incidence and the Bavarian survey give evidence that NBS did not lead to a relevant increase in incidence. 10 patients with 2 or 3 SMN2 copies were treated with Nusinersen, starting between 15– 39 days after birth, in 7/10 patients before onset of symptoms. Presymptomatically treated patients (age at last examination: 1– 12 months, median 8 months) showed no muscle weakness by the age of one month to one year. One child with 4 SMN2 copies became symptomatic at the age of 8 months.Conclusions:Newborn screening, resulting in presymptomatic treatment, improves outcome in children with genetically proven SMA. Newborn screening for SMA should be introduced in all countries where therapy is available. An immediate therapy in cases with 4 SMN2 copies should be considered.

The Frequency and Timing of Recurrent Stroke: An Analysis of Routine Health Insurance Data.

Stroke is among the leading causes of death in Germany and one of the main reasons for handicap and disability. The risk factors for stroke include hypertension, metabolic disorders, and diabetes. An estimated 250 000 persons sustain a stroke in Germany each year. German data on the occurrence of stroke, and of recurrent stroke in particular, are still inadequate. The main objective of this study was to determine the frequency and timing of recurrent stroke. The analyses were based on routine data of the AOK statutory health insurance fund in the German federal state of Lower Saxony. Inpatient billing and diagnosis data were used to identify initial and recurrent strokes (ICD-10 I60-I64). The derived incidence and prevalence were standardized for age and sex. The risks of recurrence and mortality were estimated with Kaplan-Meier analyses and Cox regressions. In 2010/2011, the standardized incidence of stroke was 292 per 100 000 persons per year, and the standardized prevalence was 336 per 100 000 persons. The risk of recurrence was 1.2% in the first 30 days, 3.4% within 90 days, 7.4% within 1 year, and 19.4% within 5 years. The mortality after an initial stroke was 6.8% in the first 30 days, 9.4% within 90 days, 17.0% within 1 year, and 45% within 5 years. Patients with hemorrhagic strokes had higher mortality. Treatment in a stroke unit was associated with lower mortality. Patients with acute stroke should be treated in a specialized center whenever possible. After hospital discharge, stroke patients should be followed up at short intervals and any risk factors for stroke should be treated.

Refining animal research: The Animal Study Registry.

The Animal Study Registry (ASR; www.animalstudyregistry.org) was launched in January 2019 for preregistration of animal studies in order to increase transparency and reproducibility of bioscience research and to promote animal welfare. The registry is free of charge and is designed for exploratory and confirmatory studies within applied science as well as basic and preclinical research. The registration form helps scientists plan their study thoroughly by asking detailed questions concerning study design, methods, and statistics. With registration, the study automatically receives a digital object identifier (DOI) that marks it as intellectual property of the researcher. To accommodate the researchers concerns about theft of ideas, users can restrict the visibility of their registered studies for up to 5 years. The full content of the study becomes publicly accessible at the end of the embargo period. Because the platform is embedded in the infrastructure of the German Federal Government, continuity and data security are provided. By registering a study in the ASR, researchers can show their commitment to transparency and data quality to reviewers and editors, to third-party donors, and to the general public.

Risk aversion and the teaching profession: an analysis including different forms of risk aversion, different control groups, selection and socialization effects

ABSTRACTRisk aversion might affect current and potential teachers’ reaction to reforms, such as payment reforms. However, evidence on teachers’ risk aversion in comparison to other occupations is limited. The present study is based on twelve waves of a representative German data set (N = 18,381) and shows that teaching relates positively to risk aversion, especially to risk aversion with respect to occupational career. Teachers score higher in risk aversion even than other civil servants. Our results suggest that risk-averse individuals are attracted to teaching, while we find no evidence for a socialization effect during the career.

EFAS Score – Multilingual Development and Validation of a Patient-Reported Outcome Measure (PROM) by the Score Committee of the European Foot and Ankle Society (EFAS)

Category:Basic Sciences/BiologicsIntroduction/Purpose:A scientifically sound validated foot and ankle specific outcome measure for different European languages is still missing. Indeed, language-specific cross cultural validation in other languages than English is largely absent. Some outcome measures were validated for specific pathologies such as hallux valgus, ankle arthritis or flatfoot. The European Foot and Ankle Society (EFAS) established in 2013 a Score Committee to develop, validate, and publish a new score, the “EFAS Score”, which is not specific for single pathologies for different European languages. The principal aim of this project was to develop and validate the EFAS Score simultaneously for different European languages.Methods:The EFAS Score was developed and validated in three stages: 1) item (question) identification, 2) item reduction and scale exploration, 3) confirmatory analyses and responsiveness. The following score specifications were chosen: scale/subscale (Likert 0-4), questionnaire based, outcome measure, patient related outcome measurement. For stage 3, data were collected pre- operatively and post-operatively at a minimum follow-up of 3 months and mean follow-up of 6 months. Item reduction, scale exploration, confirmatory analyses and responsiveness were executed using analyses from classical test theory and item response theory.Results:Stage 1 resulted in 31 general and 7 sports related questions. In Stage 2, a 6-item general EFAS Score was constructed using English, German, French and Swedish language data. In Stage 3, internal consistency of the scale was confirmed in seven languages: the original four languages, plus Dutch, Italian and Polish (Cronbach’s Alpha >0.86 in all language versions).Responsiveness was good, with moderate to large effect sizes in all languages, and significant positive association between the EFAS Score and patient-reported improvement.No sound EFAS Sports Score could be constructed.Conclusion:The multi-language EFAS Score has been successfully validated for orthopaedic foot and ankle surgery populations incorporating a wide variety of foot and ankle pathologies, including language-specific validation in seven languages so far (English, German, French, Swedish, Dutch, Italian, Polish). Validation for other languages (Danish, Finnish, Norwegian, Portuguese, Spanish, Turkish) is in progress. All validated score versions are freely available at www.efas.co.

Gestural Cosuppositions within the Transparency Theory

Gestural Cosuppositions within the Transparency Theory

The insect apocalypse, and why it matters.

The majority of conservation efforts and public attention are focused on large, charismatic mammals and birds such as tigers, pandas and penguins, yet the bulk of animal life, whether measured by biomass, numerical abundance or numbers of species, consists of invertebrates such as insects. Arguably, these innumerable little creatures are far more important for the functioning of ecosystems than their furry or feathered brethren, but until recently we had few long-term data on their population trends. Recent studies from Germany and Puerto Rico suggest that insects may be in a state of catastrophic population collapse: the German data describe a 76% decline in biomass over 26 years, while the Puerto Rican study estimates a decline of between 75% and 98% over 35 years. Corroborative evidence, for example from butterflies in Europe and California (which both show slightly less dramatic reductions in abundance), suggest that these declines are not isolated. The causes are much debated, but almost certainly include habitat loss, chronic exposure to pesticides, and climate change. The consequences are clear; insects are integral to every terrestrial food web, being food for numerous birds, bats, reptiles, amphibians and fish, and performing vital roles such as pollination, pest control and nutrient recycling. Terrestrial and freshwater ecosystems will collapse without insects. These studies are a warning that we may have failed to appreciate the full scale and pace of environmental degradation caused by human activities in the Anthropocene.

The Success of Business Forecasting: Comparisons Across Industries

Purpose: This study measures and assesses the accuracy of forecasts by industry branches. Such an investigation provides a view on the relative benefits of forecasting in different industries. Design/methodology/approach: Accuracy is studied here by investigating survey data covering manufacturing firms in the U.S. and Germany. The two data sets are conceptually different and cover different time periods: the U.S. data mostly cover the 1980s while the German data have been continuously available since the 1990s. Findings: As it turns out we can identify industries that are among the most (e.g., electric machinery) and least accurate forecasters (e.g., the food industry). These findings can be of help for management in different industries and countries regarding whether resources should be allocated to forecasting. Yet, the printing industry – at the top of the U.S. ranking and towards the bottom in Germany – reminds us of the structural changes affecting the production processes and the need for forecasting. Practical implications: The findings can help management to make decisions regarding the allocation of resources to forecasting. Originality/value: This study presents a widely applicable approach to measuring and comparing forecasting performance across industries.