Dr. Smith's experiment raises three difficulties for confidentiality:
he conducted his study in the workplace, he had few subjects, and he included in his report information that could be used to identify individuals (i.e., the number of known programming languages). In workplace studies, such as Dr. Smith's, it is often difficult to maintain anonymity, since co-workers can often witness the interactions between the researchers and subjects. Workplace studies also increase the likelihood that subjects will be identified from reports of individual subjects'
characteristics. Because co-workers know each other, they will be more successful at identifying the subjects than strangers would. For example, Dr. Smith reported the number of programming languages each subject knew, making it easier for co-workers to determine who participated in the study. Once the subjects have been identified, it then becomes possible to link their identity to portions of the data. In the above example, it was possible for co-workers to determine each subject’s success at training. It is preferable to report aggregated data (e.g. cross-subject averages) instead of raw data, as this makes it much more difficult to identify individual subjects and their data. Unfortunately, such aggregation is less effective at preserving confidentiality and anonymity when there are few subjects, as is common in ESSE. Because
ESSE studies often have features (few subjects, occur in the workplace) that make the preservation of anonymity and confidentiality more difficult, ESSE researchers should disclose the limits of confidentiality and the implications
thereof to the subjects as part of the informed consent procedure.