Research Tips

Respondent Anonymity

We sometimes see surveys that ask for respondents’ names or are administered in a way in which respondents’ identities would be obvious. While there are some circumstances in which knowing who said what is essential, not providing anonymity to respondents may influence results. For example, some surveys we’ve seen have been evaluations of various types, and when someone completing such an evaluation knows that their response will be individually identifiable they’re likely to provide a more positive evaluation and less critical feedback than they would if the evaluation was anonymous. That’s particularly true if the individual or organization conducting the evaluation has any power over them — whether that’s as an instructor, a regulator, or a business associate. If the respondent has any sort of social relationship with the evaluator, then social desirability bias is also possible. One way to avoid these problems when knowing who said what is important is using a third party to collect and analyse the information so that identifying information can be used to interpret responses, but is then stripped to preserve anonymity before the results are turned over to the party using the evaluation.