Best Practices for Research Organization
by Craig Emerson - Vice President ? Editorial Operations, ProQuest
- Document absolutely everything, review, and then fill in the gaps! It is crucial that all aspects of data collection and analysis be fully documented. Assume (1) that you'll forget everything about your data within one month, and that (2) there will be a pressing need to use those data. Without documented context, the data are useless.
- Your institution or funding agency may require you to submit your data to a repository. You may also wish to store your data at a central data center. Review submission guidelines for these repositories and organize your data accordingly
- Anticipate the type and volume of data you'll be collecting. This will help predict the type and extent of data storage you'll need (computer memory, hardcopy notebooks, etc.), and the most likely types of analytical software you'll require. Some software used to manipulate/analyze data requires storage in a specific data format.
- Do not discard any data! Annotate questionable data with a precise description of the possible problems, but do not delete. The same goes for hardbound notebooks for laboratory or field research. Do not destroy or alter any notations.
- When saving data in electronic files (e.g. spreadsheets etc.), create as few files as possible, and consider saving data in formats that can be easily shared.
- File names for data sets should be short and meaningful:
- Avoid unnecessary repetition and redundancy in file names and file paths, and avoid using non-alphanumeric characters.
- If using a date in the file name always use the format "YYYYMMDD"
- Order the elements in a file name in the most appropriate way to retrieve the record.
- Perform consistent backups of electronic files, and store backups in a separate location. Keep an inventory of all data sets files, including backups. Convert old formats into contemporary machine-readable formats as necessary.
