Confidentiality

Confidentiality is often a difficult problem to resolve. Firstly, there is an issue around protecting the anonymity of questionnaire / interview respondents if this is felt to be necessary or desirable. Sometimes it won’t be. For instance, it could be that after a large-sample questionnaire the evaluators may want to go back to some of the respondents that have raised particular issues for a follow up interview. There are choices to be made – either the questionnaire can be completely anonymous, or it can be identified by a code known only to a limited number of people or it can be readily identified by name.
Whatever the circumstances this must be made clear to the respondents. It is not sufficient to assure confidentiality simply by telling people they do not have to put their names on the questionnaire. With small groups it is often easy to identify individuals by other means (nature of the response, handwriting etc.).
Similarly, comments made during “confidential” interviews may be written in evaluation products as reported speech or as direct quotations.  Again, it may be easy to identify the authorship because of the subject matter or the particular perspective from which their observations are made.  It should be clarified with each interviewee the limits they want to impose, if any.
Evaluators may also have access to project or institutional documents which have a restricted circulation and the use of material from these documents in evaluation reports should be clarified.
Ownership and confidentiality of the raw data (e.g. completed questionnaires, frequency counts, interview notes etc) can also be an issue.  Evaluators will frequently want to discuss points arising from the interview notes with clients, particularly the project staff themselves, to check the significance of particular responses, to check for accuracy or to provide early feedback.  Do the project staff have a right to ask for the source?  Whether this data is `owned’ by the evaluator or the contractor should be agreed.
A second issue is whether either the evaluator or the project staff who provide the data can re-use it at a later date in another context and what the confidentiality limits are? For example, an evaluator recently undertook an evaluation of a project that was concerned with social exclusion issues in mainly Black communities in Cardiff.  One of the many data sets provided information about the levels of domestic violence.  In a subsequent research report by the evaluator on domestic violence, this data was re-used and indicated that the level of domestic violence in that particular Black Community was higher than in adjacent White areas.  The clients for the first evaluation complained because they felt that a negative use was being put to the data and argued that as it was `their’ data, it should not have been re-used without permission as their report was confidential.  The evaluator argued that as the report had not been mentioned, nor in fact the project, the raw data he had collected belonged to him.
These examples are fortunately not common but, nevertheless, need to be resolved.
Handout 5 Confidentiality Checklist