At the point in a data-based undertaking when the work of the data collectors and originating project staff is nearing its end, the process of preparing data for deposit in an archive reaches a critical phase. This might be a good time to have a final look at the Checklist for Data Management which has been developed by the UK Data Archive:
- Have you gained written consent from your respondents?
- Are you sure about who owns the IPR of your data?
- Have you used standardised and consistent measures and procedures to collect and check and verify data?
- Have you provided enough documentation to inform a secondary user about project methods, data collection and data preparation?
- Are your data and files labelled sufficiently? Do all variables have a set of exhaustive mutually-exclusive codes that are documented in the codebook?
- Are you using up-to-date recommended data formats?
- Do you need to anonymise data? Does the data contain information that could identify an individual?
- Are your copies of data - digital and non-digital - held in a safe location?
- Are your files backed up sufficiently?
- Do you know which the master version of your data is?
- If the dataset contains more than one file, does the dataset contain the requisite link variables with full documentation to allow separate files to be linked?
- Are all derived variables fully documented?
- Have you included analytic information too, for example, derived variables or interview summaries?
If data have been managed throughout the data collection effort, the work involved in preparing the data for sharing and archiving will be minimal and straightforward. There are, however, a few issues which will still require attention before the data can be sent to the data archive. These relate particularly to the responsibility of the data collector to preserve and protect the confidentiality of all respondents. There are a variety of methods that are employed to create data which does not violate respondent confidentiality and can therefore be shared openly with other researchers. These are described below in general terms only.