The data interview

The researcher wants your help in writing a data paragraph or DMP. Or he wants to share his research data through a public data portal like 3TU.Datacentrum. At this moment in the advisory process, it's important to learn about the story of the data. In the first stage you can ask broad questions, like those in the "Conducting a data interview"1 handout van Purdue University Libraries.

                      

The questions are:  

  • What is the story of the data? 
    • Let the researcher talk freely about his research. Give attention, listen, wrap up. In this way you have the opportunity to take the researcher's perspective thereby understanding the context in which the data were produced 
  • In what form are the data (or will the data be) available? 
    • what is the data format? what hard- or software is needed to be able to read the data? 
  • In what way will others be able to build upon the data? 
    • Reuse? Other purposes? What is the intended audience? 
  • For how long should the dataset be preserved? 
  • How large is the dataset en how fast does it grow? 
  • Who owns the data? 
  • Does the dataset contain sensitive informatie (privacy, confidentiality?) 
  • What publications or discoveries have emerged from these data? 
  • Are metadata being allocated? If not, is this still possible?  
  • How should the data be made available?

 

Depending on the stage the research finds itself in, you ask questions about data that are to be expected and/or data that are already present. 

By getting global information about a dataset first, you can signal in an early stage whether certain aspects might prevent the dataset from being submitted to 3TU.Datacentrum. 

If the data meet the selection criteria you can gather more in depth information about the dataset when needed. An instrument to gather this kind of detailed information is the so called Data Curation Profile (DCP). Purdue University Libraries have developed a corresponding Data Curation Profiles Toolkit. The DCP is the document which describes in a structured way what the background of the dataset is and how one should deal with it. 

While a DMP (data management plan) is a documenten serving the researcher/research group, a DCP is a document for the institution which manages the data portal. A data interview is the conversation through which facts about a dataset can be gathered. 

Here3 you can look at an example of a DCP. This DCP belongs to the "Atmospheric Collection Cabauw". A manual for holding data interviews was developed in the project CARDS. Here you see4 the resulting document. 

 

1. Purdue University Libraries. Conducting a Data Interview. Retrieved 10-12-2011 from
http://www.vprgs.msu.edu/files_vprgs/data_interview_HANDOUT.pdf
2. Witt, M., Carlson, J., Scott Brandt, D. (2009). Constructing Data Curation Profiles. International Journal of Digital Curation. 3 (4), 93-103. Retrieved 15-12-2011 from ijdc.net/index.php/ijdc/article/view/137/165 
3. Rombouts, J., Gramsbergen, E. van. (2010). Data Curation Profile - IDRA. 
Retrieved 10-12-2011 from 
https://docs.google.com/document/d/1pLtthOz7DGlgHXQIV9yKWTdZ_Lof_oGUdi8n5UJAsS4/edit
4. Data Management Plan - interview. Project CARDS. (2010). Retrieved 10-12-2011 from https://docs.google.com/open?id=0B9Ccd1URnalvZTZkNWQ2MmYtZDE4NS00MjBlLTk0NTYtNzNiODUzMDQ4ZmNh 

Twitter
Loading..