Previous Page Table of Contents Next Page


8. Data Quality


8.1 Policy

All GTOS datasets will be provided with adequate metadata, enabling users to assess quality in light of the intended use.

8.2 Discussion

Judgements of quality of scientific data are traditionally done through a peer review process. Researchers and users with knowledge in the relevant fields will examine the methods of data collection, analysis techniques used, and the manner in which the results have been interpreted. Detailed documentation of the steps taken and the techniques used to ensure and preserve quality at all stages is required to enable an assessment of quality to be made. Given the number of different types of datasets potentially required, it would be impossible for GTOS to establish an effective quality assessment process to deliver meaningful results in a useful time frame.

GTOS can however ensure that there is sufficient documentation associated with any data and information to allow an assessment to be made by the user community. This builds on the premise that data quality is best defined as "fitness for use" and should be accurately documented to allow an assessment to be made on that basis i.e. taking into account the use which is to be made of the data. Just as the objective in collecting data may influence the collection method, so the prospective use to be made of data has a bearing on their suitability. Thus a dataset judged to be of acceptable quality for one use may be unacceptable for another.

The requirement for this type of documentation at all stages is an essential element of the end-to-end data management concept. Any "good-quality" dataset must carry such information as part of the metadata associated with it.

Specifically, in establishing a dataset as a GTOS product (see Appendix 1) it should be reviewed to ensure that metadata are complete, i.e. that users may be able to make a quality assessment. Any deficiencies should be identified to the custodian. The proposed operation of the pilot Information Centre (GOSIC) aims to assist this process through carrying out a review as G3OS datasets are "registered " with them. However consideration also needs to be given to ensure that there is a formal recognition of datasets as GTOS products through the Secretariat.

8.3 Actions


Previous Page Top of Page Next Page