Anonymity and Confidentiality

While both anonymity and confidentiality aim to protect participant privacy in survey research, a key distinction lies within the degree of information being collected and the level of assurance that is provided to study participants. The choice between an anonymous or confidential approach depends on the research objectives being studied, ethical considerations that are relevant to the study, and the nature of the information being collected from participants. In either case, it is essential that researchers communicate clearly with participants about how their privacy will be protected to establish trust and ensure ethical research practices.

Anonymity in survey research refers to the state in which the identity of individual responses is completely unknown and cannot be linked to their survey responses. In an anonymous survey, researchers deliberately avoid collecting any identifying information including names, contact details, IP address, or other personal identifiers. The purpose of doing so is to create a setting where participants feel safe to provide honest and candid responses without fear of repercussion. Anonymity ensures that even the researchers are unable to connect specific responses to specific individuals. This approach is particularly important when dealing with subjects that respondents may feel are sensitive or that otherwise make them feel uncomfortable about openly sharing their opinions or experiences.

Confidentiality, on the other hand, involves a commitment to safeguarding the identity of survey participants. While researchers collect identifiable information from participants, such as their names, contact details, or demographic information, they assure respondents that this information will be kept confidential and will not be disclosed to any unauthorized individuals. Confidentiality implies that the researcher(s) have access to the participant’s identity and individual responses, but no personally identifying information is shared or reported in any way that could expose the participant’s identity. This includes ensuring that no combination of reported factors such as race, gender, or department can be used to identify a single individual. Confidentiality is often essential in studies where personalized reminders, follow-up, or additional information may be needed, or participants need to be traced for validation or further inquiry.

Keep in mind that even with a survey instrument that is designed to be confidential or anonymous, these safeguards, the steps you take in analysis and reporting may compromise the privacy of individual responses. This most commonly occurs when cross-tabulating data or reporting on open-ended survey responses. 

• Cross-Tabulations. Surveys often ask respondents to provide personal characteristics so that we can compare the responses of different subpopulations. When we cross-tabulate groups that have small numbers of members (e.g., reporting on the responses of international faculty in a program where there are only one or two international faculty members), it can become obvious who responded to a survey or how. For this reason, it is important to consider establishing data-suppression thresholds (e.g., calculations that specify the minimum number of given responses reported).  

• Open-ended Questions. Open-ended responses sometimes refer to individuals by name or position or include details that identify the respondent. When open-ended questions are included in a survey, you should develop a plan for how you will address responses like these. For instance, depending on the survey’s purpose, a sanitized summary of the open-ended responses may be appropriate. Alternatively, the open-ended responses may be de-identified or cleaned of identifying content.