Research data

Research data include information and facts that have been collected, generated or analysed in the course of research. They serve as a basis for investigating new phenomena, testing hypotheses, formulating conclusions, producing publications, developing new technologies, or sharing and replicating results among researchers. Research data take many forms depending on the field of research and the methods of collection.

Examples of research data: 

  • Numerical data: measurements, tables, statistics (e.g., experimental results, surveys, physical measurements); 
  • Textual data: interview records, field notes, documents, laboratory journals; 
  • Image data: photographs, microscopic images, maps, video recordings; 
  • Audio data: recordings of conversations, audio signals, music; 
  • Codes and models: algorithms, simulation models, software codes; 
  • Experimental samples: biological samples, chemical substances, archaeological artifacts. 

 

FAIR data 

According to the Open Science principles, research data should satisfy FAIR principles: 

  • Findable – the data are described by sufficiently detailed metadata and are assigned a persistent identifier, e.g. DOI; 
  • Accessible – the data or at least the metadata are freely accessible, preferably stored in a trusted repository; 
  • Interoperable – the data are described by standardized expressions and can be integrated with other datasets. 
  • Reusable – the data are described sufficiently and shared under the least restrictive license to make it clear how they were created, what they describe, and how other users can use them. 

 f.a.i.r.
Open data 

Research data should be open but this is not always possible due to a number of specificities: 

  • Immediate access (right of first use) cannot be requested; 
  • Data cannot always be opened because they contain sensitive personal or commercial data; 
  • Data are often large in size (thousands to millions of files in one dataset, rapid growth over time); 
  • Data have a large variety of formats and forms (often non-textual); 
  • There are big differences between disciplines – different standards; 
  • There are different categories of data – raw data, processed data, analysed data (which category to share?); 
  • Making data available to someone else involves a lot of work (organisation, description, transfer, access control); 
  • Etc. 

 

Therefore, the following rule applies for research data opening: 

"As Open as Possible, as Closed as Necessary." 

 Data
 
 

Section navigation: Open Science Office