Research Data Management Strategy: Documenting and Providing Context for Data

High level information about the research should include: 

• Purpose (project history, aims, objectives, investigators, funders, publications)
• Context (kinds of data, structure of files, number of cases/records/files, relationship between)
• Collection Methods (methodology for collection, sampling method, instruments used, transcription methods, secondary data sources)
• Collection Details (who collected it, where, when)
• Processing (tools, instruments, procedures, to edit, clean, code, or classify data)
• Modifications (anonymization for example)
• Quality Assurance Procedures (any data validation, calibration procedures, etc)
• Access (where the data set is available, persistent identifier, access and use conditions, copyright, confidentiality, citation)

Constructing documentation during analysis can support the analytic process and results in richer contextual information when data are shared. 

• Create classifications for persons, such as interviewees, data sources, etc. 
• Classifications can contain demographic information, dates/times of interviews, etc.
• Classification sheets can be imported into NVivo12
• Documentation files such as the description of the methodology, research plan, interview questions, consent form templates can be imported into the NVivo12 project file and stored in a "documentation" folder in the "Memos" folder 
• Descriptions can be added to all objects created in, or imported into the project file including data, documents, memos, nodes and classifications. 
• All textual documentation compiled during the project cycle can later be exported as spreadsheets to document preserved data collections. 
• Summary information about the project can be exported via project summary extract reports as text, MS Excel or XML file

Data level documentation provides information about files, for example interview transcripts or pictures. Such documentation also includes elements within the files, such as describing the variables or code descriptions. 

Please note that coded data can reduce the usefulness of the data for future use as other researchers may want to recode original data for different analyses. 

1. Data List: Allows researchers to identify and locate relevant items. Example: interview identifier, age, gender, occupation, location, place of interview, date, transcript file name.

2. Metadata: Research Data Alliance Metadata Working Group provides a comprehensive open directory of metadata standards for documenting research data, regardless of discipline. The metadata schema promoted by DataCite is also widely used for data catalogues. DataCite (2017) identifies three fundamental goals: establish better access to scientific research data on the internet, increase the acceptance of research data as legitimate and citable contributions, and to support data management and preservation for future use.