Objectivity, Confidentiality and anonymity in Research

Confidentiality means that disclosed financial interests marked “Confidential” shall be kept in a locked cabinet accessible only to the Director of Research Integrity and Compliance and /or as encrypted files on a password protected computer. To the extent permitted by law, all records of financial interest submitted by an Investigator shall be made available only to the Conflict-of-Interest Committee and to others on a “need to know” basis with clear understanding of the confidentiality of the information. Records will be destroyed once the record retention period expires.

Objectivity is difficult because experience, our main source of information, is inherently subjective. Generally speaking, we eliminate subjectivity by comparing notes about what we experience with other people who might have been in a position to experience the same thing.

Objective are the persistent patterns of agreement about aspects of experience. The things that everyone agrees one, at least in theory. In practice we sometimes stipulate that any observer will see the same thing after many checks.

For example, anyone who does a thorough job of observing the universe will eventually discover atoms. Atoms are objective, despite not being apparent to the naked senses, because, at least in principle, any suitably skilled observer will observe the evidence that makes atoms seem real.

Thus, objectivity is ensured by comparing notes or having other people look at your data, your methods, and your conclusions. If they make the same inferences then the likelihood of objectivity is higher, if they do not then it is lower. Objectivity is thus never an absolute. We seek it, we approach it, but we never personally attain it. Communities of researchers working together attain much higher degrees of objectivity than individuals, though we may also fall into biases and fallacies that result in collective delusion. Hence a network of communities is ideal so that no one individual or one community dominates the field.

Confidentiality in Research

Confidentiality refers to a condition in which the researcher knows the identity of a research subject, but takes steps to protect that identity from being discovered by others. Most human subjects research requires the collection of a signed consent agreement from participants, and the collection of other personally identifiable data, and thus researchers are aware of the identity of their subjects. In such cases, maintaining confidentiality is a key measure to ensure the protection of private information.

Researchers employ various methods to keep their subjects’ identity confidential.  Foremost, they keep their records secure through the use of password protected files, encryption when sending information over the internet, and even old-fashioned locked doors and drawers. They frequently do not record information in a way that links subject responses with identifying information (usually by use of a code known only to them). And because subjects may not be identified by names alone, but by other identifiers or by combinations of information about subjects, researchers will often only report aggregate findings, not individual-level data, to the public.

Anonymity in Research

Anonymity is a condition in which the identity of individual subjects is not known to researchers. Because most human subjects research requires signed documentation of consent, subject anonymity is not as common in human subjects’ research.  Federal law does allow an IRB to waive the requirement for signed consent documents in cases where the collection of that document is the only identifying information linking the subject to the project.  Such documentation is most often waived for projects such as online survey that present no more than minimal risk to subjects.

Anonymity means that there is no way for anyone (including the researcher) to personally identify participants in the study. This means that no personally-identifying information can be collected in an anonymous study. Personally-identifying information includes, but is not limited to, names, addresses, e-mail addresses, phone numbers, government-issued ID numbers (e.g., social security numbers), photographs, and IP addresses. This also means that any study conducted face-to-face or over the phone cannot be considered anonymous; this rules out virtually all qualitative research that involves interviews.

As you develop your human subjects review application, please be certain you understand the distinction between confidentiality and anonymity, and that you use the appropriate terms in your project description and consent documents.