There is an important difference between anonymous data collection and confidential data collection.
Anonymous data has no identifiers recorded with no way for the investigator to link the data back to the individual participant. This means that no one not even the investigator is able to identify participants identities by their data. Anonymous just not mean no name other data can be used to identify the participant. Online surveys are generally the easiest way to design an anonymous study.
Confidential data can potentially be identified or linked to a specific participant. If any identifying information is collected, even if it is coded, it is confidential rather than anonymous. Special protections must be in place to prevent a data breech when conducting confidential research. Data must be kept in secure locations with codes or keys identifying specific participants kept separately from the data. Examples of secure locations are locked file cabinets in a private office, electronic data secured on a password protected computer, or data stored on an encrypted cloud. Any research assistants, investigators, or staff members that have access to the data must be properly educated on how to handle confidential data (which can be done through completing the CITI program, as required by the Fordham University IRB). High risk data can be protected by a certificate of confidentiality to ensure that the data cannot be subpoenaed. More information about obtaining a certificate of confidentiality can be found at grants.nih.gov/grants/policy/coc/index.htm.
A researcher can de-identify confidential data through aggregating data to report means and standard deviations rather than individual responses. If there are enough individuals in each demographic group, individual data can be reported without a risk of identification. Planning to de-identify your data at a later date does not mean that your initial data collection procedures are anonymous.
Examples of data collection procedures that are NOT anonymous:
- If you collect too much data from a small group of individuals (for example, ethnicity, age, gender, department, and number of years working at a company), your data collection procedures may not be anonymous.
- If you have a list of codes connecting the participant to their data, it is not anonymous.
- Face-to-face interviews are not anonymous.
- Online surveys that collect IP (Internet Protocol) addresses are not anonymous.
Please ensure that your consent forms make it clear whether your data collection procedures are anonymous or confidential.