• Backup your data at regular intervals
• After changing anything
• Upon completion of data collection
• Follow the 3, 2, 1 rule
• Store 3 copies of your data (2 copies in addition to your original data file)
• Store the data on 2 different media types (hard drive, external drive, central server etc.)
• Have 1 copy at an off-site location (central server, cloud storage, etc.)
• Document where your data is located and how often it is backed up
• Include data documentation with your backups
• Have a contingency plan for restoring lost data

Use Metadata to record important aspects of your data. Create a document entitled readme.txt that accompanies your data and includes metadata as well as any other pertinent information others would need to use or interpret your data.

See Cornell's template for a readme.txt document 

Metadata elements include (but are not limited to) things such as:

• title
• date of data collection,
• method of data collection,
• settings of instruments used for data collection,
• investigators,
• file types,
• information on variables/variable names
• format
• subject
• unique identifier
• rights
• funding agencies

Metadata standards exist for different disciplines to help you record the important information. Some examples are: 

• General: Dublin Core | MODS
• Social Science: DDI
• Humanities: TEI | VRA
• Sciences: Darwin Core | ITIS 

For a list of metadata standards, see the UK's Digital Curation Centre (DCC) inventory of discipline-specific metadata standards or ask at the library for help choosing a standard. 

Data Management Plans:

• Even if you do not need it for funding requirements, it is a good idea to create a data management plan (DMP) prior to starting your research project. See the Data management Plans page for help on creating DMPs. 

Naming Conventions: 

• File names should be descriptive enough to be meaningful. They can include sections such as project name or acronym, investigator, study title, location, version, date of data collection, and data type.
• Do not include spaces or special characters. Instead use underscore (_) or dashes (-). 
• Ensure everyone on your project team understands and adheres to the naming convention
• Apply the naming convention to folders and individual files
• Check for database limitations on filename length or use of certain characters. 
• Agree and stick to a date convention. Ensure everyone creating files sticks to this. (ex. dd-mm-yyyy or yyyymmdd)
• Example of a good naming convention: VI_Surveydata_01022015.csv (name of study = Value and Impact, data type was a survey data, Feb 1, 2015 was date of data collection, .csv is file format)

Formats:

• To prevent incompatibility, store data in file types and on hardware that are open, not proprietary. 

Before you share your research data, use your data management plan to address the following questions: 

• Is the data anonymous? Remove or redact sensitive information 
• Who will have access to the data? Ensure appropriate access restrictions are in place (dark archives, embargo periods, etc.)
• What are the intellectual property rights? Ensure that copyright, ownership, and licenses for the use of the data are clear
• Have participants given informed consent? Consent forms can include language that allows data sharing. ICPSR has a great page with suggested language that allows data sharing 

Be sure to check Memorial University's Ethics of Research Involving Human Participants policy. 

For more information on Best Practices, see: