I have a client that is concerned about his constituents potentially skewing a survey by adding multiple survey responses from the same user. Is there a good way to limit a survey response to one submission, say per IP address (when using the web survey)?
As you point out, a cookie can be easily worked around by the use of multiple browsers or clearing cookie settings. IP restrictions cannot work because it is very common for an internal network (say from one company) to be aggregated into a single public IP; this would restrict users from submitting if another person in their organization submitted previously. We are researching this topic to consider other methodologies of creating unique survey entries.
Thank you for your reply and apologies for not replying back to you sooner. I did see your reply in email when you posted it.
I think that a cookie option would at least be a relatively simple start. IMHO, it's creates at least an obstacle that people would have to go around in order to skew results, and doing something is better than doing nothing at all. Survey Monkey and others implement the cookie strategy to prevent this problem, so why not do the same? I would assume most respondents would not care to take the time to respond multiple times, and for those who might want to their level of technological intellect may not enable them to expose the workaround.
We are considering implanting this method of submission prevention. Part of our concern before implementing is ensuring that the solution, while not perfect, would be acceptable to users and that when it's released, we can appropriately describe the scenarios in which it will and won't work.