Düsterhus, André: Statistical Quality Assurance and Peer Review in Primary Data Publication. - Bonn, 2013. - Dissertation, Rheinische Friedrich-Wilhelms-Universität Bonn.
Online-Ausgabe in bonndoc: https://nbn-resolving.org/urn:nbn:de:hbz:5n-31001
@phdthesis{handle:20.500.11811/5610,
urn: https://nbn-resolving.org/urn:nbn:de:hbz:5n-31001,
author = {{André Düsterhus}},
title = {Statistical Quality Assurance and Peer Review in Primary Data Publication},
school = {Rheinische Friedrich-Wilhelms-Universität Bonn},
year = 2013,
month = jan,

note = {Publication of results is an elementary part of scientific work. Started in in 17th century with the traditional scientific publications several new forms emerged in recent years, triggered by the digital revolution and focused on increasing the data availability. These changes offer new chances for scientists and science in general, but also imply risks, which might compromise this established institution in science.
Data can be seen as an essential foundation of science. Therefore, it is important to think about new ways to distribute scientific data between the scientists, which are generated by the advent of the world wide web. One way, the primary data publications, aims at the publication of raw data and their metadata and seeks to be comparable to the traditional forms of publishing manuscripts.
This thesis will present the possible ways to publish data, which are typical in meteorological and climatological sciences. Additionally, it shows, how a publication process of the primary data itself can be included into the traditional scientific working scheme. Thereby, it will especially focus on the development of an effective quality assurance of these publications. A fundamental part of this will be general quality tests, which are needed to obtain estimations on the quality of datasets. These quality checks are characterised by its parameter-driven flexibility and can be applied on a lot of different types of datasets.
Some of these checks, like a newly developed histogram test and a bayesian change point detection, are described and undergo some sensitivity tests. In addition, the possible automatisation and application to meteorological and climatological datasets of the tests will be investigated. All these developments will be discussed under the aspects of a usability in data publications, an effective quality assurance system and the possibility to generate a peer review procedure on data publications.},

url = {https://hdl.handle.net/20.500.11811/5610}
}

Die folgenden Nutzungsbestimmungen sind mit dieser Ressource verbunden:

InCopyright